Aryl to aryl palladium migration in the Heck and Suzuki coupling of o-halobiaryls

Article abstract of DOI:10.1021/ja069238z

A novel 1,4-palladium migration between the o- and o′-positions of biaryls has been observed in organopalladium intermediates derived from o-halobiaryls. The organopalladium intermediates generated by this migration have been trapped either by a Heck reaction employing ethyl acrylate or by Suzuki cross-coupling using arylboronic acids. This palladium migration can be activated or deactivated by choosing the appropriate reaction conditions. Chemical and computational evidence supports the presence of an equilibrium that correlates with the C-H acidity of the available arene positions.

Full text of DOI:10.1021/ja069238z

Published on Web 04/24/2007
Aryl to Aryl Palladium Migration in the Heck and Suzuki
Coupling of o-Halobiaryls
Marino A. Campo, Haiming Zhang, Tuanli Yao, Abdellatif Ibdah, Ryan D. McCulla,
Qinhua Huang, Jian Zhao, William S. Jenks,* and Richard C. Larock*
Contribution from the Department of Chemistry, Iowa State UniVersity, Ames, Iowa 50011
Received December 22, 2006; E-mail: wsjenks@iastate.edu; larock@iastate.edu
Abstract: A novel 1,4-palladium migration between the o- and o′-positions of biaryls has been observed
in organopalladium intermediates derived from o-halobiaryls. The organopalladium intermediates generated
by this migration have been trapped either by a Heck reaction employing ethyl acrylate or by Suzuki cross-
coupling using arylboronic acids. This palladium migration can be activated or deactivated by choosing the
appropriate reaction conditions. Chemical and computational evidence supports the presence of an
equilibrium that correlates with the C-H acidity of the available arene positions.
Scheme 1
Introduction
The Heck1-7 and Suzuki8-11 cross-coupling reactions are
probably the two most important Pd-catalyzed C-C bond
forming methodologies, having found numerous applications
in synthetic organic chemistry. The key organopalladium
intermediates are generally prepared in situ by the oxidative
addition of organic halides or triflates to Pd(0), and subsequent
C-C bond formation usually occurs at the position originally
occupied by the halide or triflate. For instance, Heck reactions
employing o-bromobiaryls12-14 and Suzuki cross-couplings with
o-halobiphenyls15 have been reported to give only the corre-
sponding ortho substitution products. Thus, it was a great
surprise to find that, upon carrying out Heck16 reactions and
Suzuki cross-couplings with o-halobiaryls under various reaction
conditions, we obtained mixtures of the expected o- as well as
the unexpected o′-derived Heck and Suzuki products. These
observations suggest the presence of a 1,4-rearrangement of the
palladium moiety between the o- and o′-positions of these
biaryls. Scheme 1 illustrates how the palladium migration
generates two different arylpalladium intermediates from a single
o-halobiaryl substrate. Subsequent trapping of these two inter-
mediates with either ethyl acrylate or arylboronic acids gives
(1) Guiry, P. J.; Kiely, D. Curr. Org. Chem. 2004, 8, 781-794.
(2) Shibasaki, M.; Boden, C. D. J.; Kojima, A. Tetrahedron 1997, 53, 7371-
7395.
(3) De Meijere, A.; Meyer, F. E. Angew. Chem. Int. Ed. Engl. 1994, 33, 2379-
2411.
(4) Gibson, S. E.; Middleton, R. J. Contemp. Org. Synth. 1996, 3, 447-471.
(5) Overman, L. E. Pure Appl. Chem. 1994, 66, 1423-1430.
(6) Shibasaki, M.; Vogl, E. M.; Oschima, T. Compr. Asymmetric Catal., Suppl.
2004, 1, 73-81.
(7) Cabri, W.; Candiani, I. Acc. Chem. Res. 1995, 28, 2-7.
(8) Kotha, S.; Lahiri, K.; Kashinath, D. Tetrahedron 2002, 58, 9633-9695.
(9) Hassan, J.; Sevignon, M.; Gozzi, C.; Schulz, E.; Lemaire, M. Chem. ReV.
2002, 102, 1359-1470.
mixtures of o- and o′-derived Heck and Suzuki products,
respectively.
We have previously communicated the 1,4-palladium migra-
tion in organopalladium intermediates derived from o-halobiaryls
under Heck reaction conditions16 and now wish to disclose the
full details of our investigation on this fascinating process.
(10) Suzuki, A. J. Organomet. Chem. 1999, 576, 147-168.
(11) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457-2483.
(12) Feuerstein, M.; Doucet, H.; Santelli, M. Synlett 2001, 1980-1982.
(13) Diaz-Ortiz, A.; Prieto, P.; Vazquez, E. Synlett 1997, 269-270.
(14) Heck, R. F. Org. React. 1982, 27, 345-390.
(15) Kothare´, M. A.; Okhanda, J.; Lockman, J. W.; Quian, Y.; Blaskovich, M.
A.; Sebti, S.; Hamilton, A. D. Tetrahedron 2000, 56, 9833-9841.
(16) Campo, M. A.; Larock, R. C. J. Am. Chem. Soc. 2002, 124, 14326-14327.
9
6298
J. AM. CHEM. SOC. 2007, 129, 6298-6307
10.1021/ja069238z CCC: $37.00 © 2007 American Chemical Society

Aryl to Aryl Palladium Migration of o-Halobiaryls
A R T I C L E S
Table 1. Pd-Catalyzed Reaction of 2-Iodo-4′-methylbiphenyl (1a)
Table 2. Heck Reaction of Substituted o-Halobiphenyls with Ethyl
Acrylate
and Ethyl Acrylate (EA)a
mole ratio
2a:3ab
mole ratio
2:3b
entry
equiv EA
conditions
time (d)
% yield
entry
substrate
X
procedurea
time (d)
% yield
c
1
2
3
4
5
6
7
8
9
4
4
1
1
1
1
1
1
1
1
1
TBAC, NaHCO3
TBAC, NaHCO3
TBAC, NaHCO3
TBAC, CsPiv
CsPiv
1.0
1.0
1.0
1.5
1.5
1.5
1.5
1.5
1.0
1.0
1.0
100:0
95:5
100
100
92
89
93
88
87
90
75
94
90
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1a
Me
A
C
A
C
A
C
A
C
A
C
A
C
A
D
A
D
B
D
B
D
1.0
1.5
1.0
1.5
1.0
1.0
1.0
1.5
2.0
2.0
2.0
2.0
2.0
1.0
1.0
1.0
1.0
2.5
1.0
2.5
100
100:0
50:50
0:100
49:51
100:0
55:45
0:100
49:51
100:0
52:48
0:100
48:52
100:0
48:52
0:100
42:58
100:0
39:61
0:100
33:67
88
93
86
80
90
100
93
100
93
99
92
72
86
99
83
85
77:23
55:45
54:46
50:50
50:50
55:45
69:31
75:25
100:0
4a
1b
4b
1c
4c
1d
4d
1e
4e
Me
NMe2
NMe2
OMe
OMe
CO2Et
CO2Et
NO2
5% dppm, CsPiv
10% PPh3, CsPiv
5% dppm, CsOAc
5% dppm, NaOAc
5% dppm, CsPivd
5% dppm, Et3N
10
11
a The reaction was run using 0.25 mmol of the iodobiaryl, ethyl acrylate
(EA), 2 equiv of an appropriate base, and 1 equiv of n-Bu4NCl (TBAC)
where indicated in 4 mL of DMF at 100 °C unless otherwise indicated.
b The mole ratio was determined by 1H NMR spectroscopic analysis. c DMF
d
(1 mL) as solvent. Toluene (4 mL) as solvent.
46 (40)c
NO2
89
Similar palladium migration side products were reported
simultaneously by Gallagher during his study of the Heck
olefination of halophenylpyridines.17 In order to better under-
stand the parameters governing palladium migration, we have
also carried out a series of experiments under Suzuki cross-
coupling reaction conditions, which have provided useful
information on the basic reactivity of these biarylpalladium
intermediates, as well as the overall kinetics of the system.
Furthermore, we have used computational methods to model
the migration tendencies of the palladium moiety within the
biaryl and obtained results that are in good agreement with
isolated product ratios.
37 (50)c
a Procedure A: 0.25 mmol of the o-iodobiphenyl, 5 mol % Pd(OAc)2,
1.0 equiv of n-Bu4NCl, 2.0 equiv of NaHCO3, and 4.0 equiv of ethyl
acrylate in 1 mL of DMF at 100 °C. Procedure B: same as procedure A,
except using 2.0 equiv of Et3N as the base. Procedure C: 0.25 mmol
of the o-iodobiphenyl, 1.0 equiv of ethyl acrylate, 5 mol % Pd(OAc)2,
5 mol % dppm, and 2.0 equiv of CsPiv in 4 mL of DMF at 100 °C.
Procedure D: same as procedure C, except using DMF (3.8 mL) and
H2O (0.2 mL) as the solvent. bThe mole ratio was determined by
1H NMR spectroscopic analysis. cThe yield in parentheses is for the reduced
biphenyl.
observation or suppression might be effected by manipulation
of the kinetics through adjustment of the reaction conditions.
For example, replacing the NaHCO3 with CsO2CCMe3
(CsPiv), we observed almost equal amounts of the direct
olefination product 2a and the rearranged product 3a (entry 4).
Furthermore, the addition of n-Bu4NCl (TBAC) was unnecessary
to achieve virtually the same results (entry 5). The use of
phosphine ligands, such as 1,1-bis(diphenylphosphino)methane
(dppm) and PPh3, further changed the isomer distribution to
50:50 (entries 6 and 7).
The choice of carboxylate bases in the reaction mixture plays
a critical role in the isomer distribution. Solubility may play an
important role, but Fagnou has recently pointed out that
pivalate is superior to acetate, methylpropionate, and adaman-
tanecarboxylate under conditions where all of them were soluble
in a closely related reaction.20 Computational modeling by
Fagnou demonstrated that pivalate lowered the transition state
for Pd-catalyzed arylation of benzene over bicarbonate by 1.3
kcal/mol. In our experiments, under otherwise identical
conditions, CsPiv, CsOAc, and NaOAc (which have decreasing
solubilities in DMF) gave decreasing amounts of isomer 3a, as
shown in entries 6, 8, and 9. We also carried out the
reaction under our standard migration conditions but used a
less polar solvent, toluene, in which CsPiv has a much lower
solubility than in DMF, and only one-fourth of the Heck product
mixture was 3a (94% overall, entry 10 versus entry 6). Indicating
that solubility is certainly not the only consideration, the use
of Et3N as a base completely suppressed formation of the
migration product. Using Et3N instead of CsPiv under conditions
otherwise identical to those of entry 6 (procedure B), we
Results and Discussion
Palladium Migration in Heck Reactions. In order to obtain
a clear picture of how the reaction variables affect the palladium
biaryl migration, we studied the behavior of 2-iodo-4′-methyl-
biphenyl (1a, X ) CH3) and ethyl acrylate under various
reaction conditions (Table 1). Under the classical reaction
conditions described by Jeffrey18 (procedure A), the expected
ethyl E-3-(4′-methylbiphen-2-yl)acrylate (2a, X ) CH3) was
obtained exclusively and in a quantitative yield (Table 1, entry
1). By diluting the reaction mixture 4-fold, we observed
formation of small amounts of the migration product, ethyl E-3-
(4-methylbiphen-2-yl)acrylate (3a, X ) CH3, entry 2), retaining
the overall quantitative yield.
Reducing the number of equivalents of ethyl acrylate from 4
to 1 reduces the overall yield of Heck product to about 92%
(entry 3) but gives a 77:23 mixture of 2a:3a.19 These observa-
tions suggest an underlying palladium migration reaction whose
(17) Karig, G.; Moon, M. T.; Gallagher, T. Org. Lett. 2002, 4, 3115-3118.
(18) Jeffrey, T. J. Chem. Soc., Chem. Commun. 1984, 1287-1289.
(19) All stated yields are relative to the total amount of starting material, with
reactions run until the starting material was completely consumed, unless
stated otherwise.
(20) Lafrance, M.; Fagnou, K. J. Am. Chem. Soc. 2006, 128, 16496-16497.
9
J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007 6299

A R T I C L E S
Scheme 2
Campo et al.
obtained isomer 2a exclusively in 90% yield (entry 11). The
nature of carboxylates as bidentate ligands/bases may also play
a mechanistic role.20 However, beyond the phenomenological
optimizations discussed here, we have not conducted further
experiments to determine why pivalate appears to be optimal
for migration.
It is important to note that by manipulating the reaction
conditions, we can switch the palladium migration “on” or “off”
in this biphenyl system.21 Relating again to the recent work of
Fagnou,20 a reasonable hypothesis is that the pivalate is an
optimum ligand and/or proton shuttle for enhancing the rate of
the palladium migration. With other bases, the rate of migration
is much slower than the rate at which the intermediate in the
Heck reaction is captured. Beyond this association, though, our
work remains simply phenomenological on this point.
Our optimal migration conditions for observing palladium
migration are those described in entry 6 of Table 1 (procedure
C), and the conditions employed to prevent palladium migration
are those described in entry 1 (procedure A). Furthermore, it
turned out that ethyl acrylate was superior to other olefins as a
Heck trap. For instance, the reaction of 1a with either methyl
vinyl ketone or styrene using procedure C gave complex
mixtures, and none of the desired Heck products were isolated.
In principle, the palladium migration ought also to occur with
less reactive o-bromobiphenyls, assuming that the initial pal-
ladium-bearing intermediate can be generated and that the
substitution of bromide for iodide as a possible ligand on the
Pd has little effect on the reactivity. Thus, we attempted the
Heck reaction of 2-bromo-4′-methylbiphenyl (1a′) and ethyl
acrylate using procedure C but failed to observe any significant
amount of Heck products after 3 d. When the same reaction
was carried out using wet DMF (5% H2O by volume) as the
reaction solvent, compound 1a′ produced a 50:50 mixture of
2a and 3a in 49% yield, along with 45% of 4-methylbiphenyl,
the product of reduction of the C-Br bond. The iodo-bearing
analogue, 2-iodo-4′-methylbiphenyl (1a), and ethyl acrylate also
reacted under the latter reaction conditions to produce a 50:50
mixture of 2a and 3a in 85% yield, verifying that 1a and 1a′
generate identical mixtures of 2a and 3a under the same reaction
conditions. Thus, two different palladium migration procedures
have been developed, namely procedure C, which is described
in entry 6 of Table 1, and procedure D, where 5% of the DMF
is replaced by water.
a 49:51 mixture of isomers 2a and 3a in 86% yield using our
optimized palladium migration procedure C (entry 4, Table 2).
This result seems to indicate that, under our optimal palladium
migration reaction conditions, the arylpalladium intermediates,
generated from either 1a or 4a, undergo apparent equilibration
prior to olefin trapping and, thus, generate essentially identical
mixtures of 2a and 3a. (Compare entries 2 and 4 of Table 2.)
As with 1a, we were able to suppress the palladium migration
by carrying the reaction out under the conditions described by
Jeffrey (procedure A). Thus, 4a produced 3a exclusively, in
93% yield, under the Jeffrey conditions (entry 3, Table 2).
To explore the relationship between electronic effects in the
o-iodobiphenyls and the ratio of Heck products, we studied the
reaction of various substituted o-iodobiphenyls with ethyl
acrylate under our standard palladium migration conditions; the
data are presented in Table 2. Two groups classically thought
of as π-electron donating were chosen: 4′-methoxy and 4′-
dimethylamino. The use of Jeffrey’s conditions led to formation
of only the “expected” Heck product; upon reaction with ethyl
acrylate, 1b and 4b produced 2b and 3b, respectively, in 80
and 100% yields (entries 5 and 7). Correspondingly, under
Jeffrey’s conditions, 1c and 4c produced 2c and 3c in quantita-
tive and 99% yields, respectively (entries 9 and 11).
Reactions of the methoxy- and dimethylamino- substrates
using procedure C produced near, but apparently not quite
complete, equilibration of the Pd before trapping. In each
instance, there was a slight preference for trapping the initial
Pd intermediate. Thus, reaction of 1b produced a 55:45 mixture
of isomers 2b and 3b in 90% overall yield, while a 49:51
mixture of 2b and 3b was obtained in 93% yield from 4b (entries
6 and 8). Compound 1c produced a 52:48 mixture of 2c and 3c
in 93% total yield (entry 10), and compound 4c gave a 48:52
distribution in 92% yield (entry 12). This observation of “near-
equilibration” is important mechanistically, because it indicates
that the product distribution is not likely dominated by dif-
ferential trapping rates after full and relatively rapid equilibra-
tion. Instead, it implies that the equilibration and trapping occur
on comparable timescales and that the product distributions are
at least qualitative indicators of the equilibrium distribution of
the Pd among the available sites.
Next, we switched our attention to the Heck reactions of
biphenyls bearing electron-withdrawing substituents, choosing
CO2Et and NO2 as representative groups. Although the Pd-
catalyzed reactions of 1d and 4d with ethyl acrylate were
unsuccessful using procedure C, the reaction proceeded smoothly
using procedure D. Thus, 1d produced a 48:52 mixture of 2d
and 3d in 86% yield (entry 14), while 4d produced a 42:58
mixture of 2d and 3d in 83% yield (entry 16). Again, the small
difference in distribution is attributed to near-equilibration of
the corresponding arylpalladium intermediates. The average ratio
of 2d to 3d is 45:55, which indicates that there is a slight
Having observed these 50:50 product distributions, an obvious
question is whether, under such conditions, 2-iodo-4-methyl-
biphenyl (4a) and ethyl acrylate would generate the same
distribution of isomers 2a and 3a as previously obtained from
1a and ethyl acrylate (Scheme 2). Indeed, substrate 4a generated
(21) To be more precise, we can switch on or off the formation of the migration
product. The present evidence does not directly speak to whether the
migration is slowed or whether the coupling reactions are accelerated.
9
6300 J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007

Aryl to Aryl Palladium Migration of o-Halobiaryls
A R T I C L E S
Scheme 3
Figure 1. Computed enthalpy differences for isomerization of W and Y.
tendency for these iodobiphenyls to form isomer 3d. The Heck
reactions of the strong electron-withdrawing nitro-substituted
biphenyls 1e and 4e exhibit a pattern like that observed
previously with the ester-substituted biphenyls. Using procedure
D, 1e generated a 39:61 mixture of 2e and 3e respectively in
46% yield, plus a 40% yield of reduction product (entry 18).
Under the same conditions, aryl iodide 4e produced a 33:67
mixture of 2e and 3e in 37% yield, plus a 50% yield of the
reduction product (entry 20). The average ratio of 2e to 3e is
36:64, again showing a preference for forming the isomer
in which the Heck coupling occurs on the ring that bears the
electron-withdrawing substituent. Entries 13, 15, 17, and 19
show the controls, in which each starting material is con-
verted exclusively to the conventional Heck product. Use of
Et3N was necessary in the case of compounds 1e and 4e in
order to eliminate small amounts of migration products that
were observed under the usual Jeffrey NaHCO3 conditions
(procedure A).
Computational Evidence of Equilibrium Distributions. In
order to test the hypothesis that the equilibrium mixtures of the
key arylpalladium intermediates are reflected in the product
mixtures obtained under equilibrating conditions, calculations
were carried out on two selected systems, one with a strong
electron-donating substituent and the other with a strong
electron-withdrawing group. Thus, the energies22 of the four
species W-Z were calculated at both the MP2 and B3LYP
levels of theory with basis sets chosen to be as accurate as
reasonable without making the calculation unreasonably time-
and resource-consuming. As shown in Figure 1, the energy
differences between the 2- and 2′-palladated isomers of both
4-(dimethylamino)biphenyl and 4-nitrobiphenyl are small and
correspond to the direction of the apparent equilibria derived
from the products.
consistent with the near 50:50 mixtures of Heck products
obtained under equilibrating conditions. Moreover, compound
Z corresponds to the precursor to 3e (entries 17-20, Table 2),
and the slight preference for the 2-isomer is correctly predicted;
the trend slightly in the other direction for electron-donating
substituents (e.g., 2a-c vs 3a-c) is also predicted.
Mechanistic Considerations. Taking all these data into
account, the migration of Pd in intermediates represented by
W and X seems well, if indirectly, established. However, the
precise mechanism by which migration occurs is not. In Scheme
3, we consider potential pathways for the palladium migration.
(For simplicity, not all ligands on the Pd are shown.) After
oxidative addition of the aryl halide to Pd(0) to generate
intermediate i, three mechanistic possibilities are imagined. The
initial intermediate may undergo oxidative addition of a
neighboring C-H bond to produce hydridopallada(IV)cycle (ii),
followed by reductive elimination of C-H to generate either
iii or i. Alternatively, i, ii, and iii, might interconvert via pallada-
(II)cycle iv.
Finally, iv might be formed by an electrophilic reaction step
via v, as illustrated for i.23,24-27 The current data in Table 2
and elsewhere are inconsistent with the selectivity expected with
such a mechanism. Moreover, selectivity studies on intramo-
lecular arylation, similar to the chemistry of Fagnou, by Maseras
and Echavarren,28 and others29 do not support an electrophilic
Given the uncertainties involved in computational chemistry,
including the slight simplification of the molecules and the lack
of solvent treatment, the energy differences cannot be taken
literally. However, the fact that they are very small is quite
(23) For leading mechanistic studies suggesting electrophilic species in intramo-
lecular C-H activation by palladium, see refs 24-27.
(24) Martin-Matute, B.; Mateo, C.; Cardenas, D. J.; Echavarren, A. M. Chem.s
Eur. J. 2001, 7, 2341-2348.
(25) Catellani, M.; Chiusoli, G. P. J. Organomet. Chem. 1992, 425, 151-4.
(26) Park, C.-H.; Ryabova, V.; Seregin, I. V.; Sromek, A. W.; Gevorgyan, V.
Org. Lett. 2004, 6, 1159-1162.
(22) Energies are for gas phase calculations, including zero-point energies. Basis
sets are as follows: C and H, 6-31G(d); N and O, 6-311+G(2d); P,
6-311+G(3df); I, LANL2DZ + s and p diffuse functions and d and f
polarization functions; Pd, LANL2DZ with f polarization functions.
(27) Campeau, L.-C.; Parisien, M.; Jean, A.; Fagnou, K. J. Am. Chem. Soc.
2006, 128, 581-590.
(28) Garcia-Cuadrado, D.; Braga, A. A. C.; Maseras, F.; Echavarren, A. M. J.
Am. Chem. Soc. 2006, 128, 1066-1067.
9
J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007 6301

A R T I C L E S
Scheme 4
Campo et al.
mechanism. Simply put, the pattern of reactivity for electron-
donating and -withdrawing groups is inconsistent with formation
of v, and we will not consider an electrophilic palladation
mechanism further.
Nonetheless, as they are directly consistent with both possibili-
ties, they merit discussion.
The first measurements involve the incorporation of
deuterium into the product when employing an overwhelming
excess of D2O. In the presence of D2O, deuterium incorporation
in the ortho positions would be expected if the migration
proceeds through formation of iv. If equilibration of the Pd
substituent among the ortho positions is substantially faster
than the coupling step, then virtually complete D incorporation
would be expected in the three available positions, as shown
in Scheme 4. Lack of deuterium incorporation would, of
course, imply that no intermediate with exchangeable H was
involved and, thus, eliminate iv as an intermediate. The rate of
hydrogen exchange in ii is unknown. However, if it were slow
enough, migration could occur without incorporation of deu-
terium.
Although recent computational evidence in a related reaction
favored an analogue of iv over the analogue of ii,30 we wished
to consider the intermediacy of ii for two reasons. First, it seems
improbable that intermediate iv could react with HI under the
basic reaction conditions we have employed. Second, Catellani
and Chiusoli have suggested that pallada(II)cycles analogous
to intermediate iv easily undergo oxidative addition of aryl and
alkyl halides to generate palladium(IV) intermediates generating
characteristic polycyclic compounds,31-37 which have not been
observed under our reaction conditions.
Organopalladium(IV) species are well-known,38 although no
such hydride-containing species have ever been isolated.39-41
As a result, we sought evidence for the reversible interconver-
sion between i and iii via hydridopallada(IV)cycle ii. We report
two sets of experiments carried out to test whether the
intermediacy of ii could be disproved or supported. Neither has
allowed us to distinguish between ii and iv definitively.
We first treated 2-iodobiphenyl and 4 equiv of ethyl acrylate
with 5 mol % Pd(OAc)2, 5 mol % dppm, 1 equiv of TBAC, 2
equiv of NaHCO3 in 1 mL of DMF, and 0.05 mL of D2O
(procedure A), conditions under which migration is not ob-
served. No deuterium incorporation in the ortho positions of
the coupled product is expected, because only the original
position of the I-substituent is ever activated. This expectation
(29) Campeau, L.-C.; Fagnou, K. Chem. Commun. 2006, 1253-1264.
(30) Mota, A. J.; Dedieu, A.; Bour, C.; Suffert, J. J. Am. Chem. Soc. 2005, 127,
7171-7182.
1
was met for the ester product, as analyzed by H NMR and
(31) Catellani, M.; Motti, E.; Paterlini, L. J. Organomet. Chem. 2000, 593-
594, 240-244.
GC-MS.
(32) Catellani, M.; Motti, E.; Minari, M. Chem. Commun. 2000, 157-158.
(33) Catellani, M.; Cugini, F. Tetrahedron 1999, 55, 6595-6602.
(34) Catellani, M.; Frignani, F.; Rangoni, A. Angew. Chem., Int. Ed. Engl. 1997,
36, 119-122.
However, for this reaction, when conducted using procedure
D, but employing 0.2 mL of D2O, instead of 0.2 mL of H2O,
an overall average of approximately two hydrogens were
substituted by deuterium, as indicated by the 1H NMR spectrum
(35) Catellani, M.; Chiusoli, G. P. J. Organomet. Chem. 1985, 286, C13-C16.
(36) Catellani, M.; Fagnola, M. C. Angew. Chem., Int. Ed. Engl. 1994, 33, 2421-
2423.
2
(37) Canty, A. J. Acc. Chem. Res. 1992, 25, 83-90.
(38) Canty, A. J.; van Koten, G. Acc. Chem. Res. 1995, 28, 406-13.
(39) For a Pt(IV) hydride, see refs 40 and 41.
(40) Lo, H. C.; Haskel, A.; Kapon, M.; Kelnan, E. J. Am. Chem. Soc. 2002,
124, 3226-3228.
(41) Haskel, A.; Kelnan, E. Organometallics 1999, 18, 4677-4680.
of the ester product obtained. A broad peak in the H NMR
spectrum at 7.4 ppm was also observed, consistent with
deuterium incorporation occurring at more than one carbon
atom. Mass spectral data indicated that comparable amounts of
9
6302 J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007

Aryl to Aryl Palladium Migration of o-Halobiaryls
A R T I C L E S
the nondeuterated (m/z 252), monodeuterated (m/z 253), dideu-
terated (m/z 254), and trideuterated (m/z 255) esters were
observed.
biphenylene C-C bond. Indeed, Gallagher has demonstrated
directly a kinetic preference for Pd insertion into aryl bromides
over biphenylene.44 Alternatively, it could be that iv is reversibly
formed but unreactive under the conditions (and thus excluded
mechanistically from the palladium migration chemistry).
However, without any other evidence for the formation of iv,
such a conclusion cannot be drawn.
The observed incomplete deuteration indicates that our
“equilibrating” conditions are not such that migration of the Pd
species is orders of magnitude faster than the coupling step,
just as the product distributions from 1 and 4 also indictate.
However, this result does not clearly distinguish between the
intermediacy of ii and iv, because, while it is obvious that H
exchange would occur with formation of iv, it is also reasonable
that H exchange could occur with ii as the key intermediate,
even if iv were never formed.
Other workers have considered this question as well. The
computational results of Dedieu30 and Fagnou45 on couplings
To explore this further, an equivalent experiment using the
substituted starting material 1d in the presence of excess D2O
was run. For 1d to yield 3d, it is required that palladium
migration occurs as part of the sequence. Thus, if iv is an
intermediate, then at least one deuterium must be incorporated.
However, if iv is not an intermediate and Pd-H exchange from
intermediates like ii is sufficiently slow, some 3d might be
formed without any deuteria. In the event, all of the isolated
3d contained at least one deuterium, as determined by mass
spectrometry. Again, this result is necessary, but not sufficient,
to prove the involvement of iv.
A second experiment involved the formation of iv via an
alternate synthetic pathway. Biphenylene has been reported to
react under some conditions with Pd(0) to generate iv (X )
H),42 which can also undergo Heck and Suzuki couplings.43 If
this were to occur under our optimal equilibration reaction
conditions, the same ester products should be observed from
biphenylene as from 2-iodobiphenyl. Thus, use of a substituted
biphenylene derivative should give very similar product ratios
to the other precursors generated under equilibrating conditions
(see Table 2) and might give a distinct product distribution under
our nonequilibrating conditions.
However, biphenylene was not an effective precursor to the
Heck product under our conditions. When biphenylene was
allowed to react with 1 equiv of ethyl acrylate (0.25 mmol) in
the presence of 5 mol % Pd(OAc)2, 5 mol % dppm, 2 equiv of
CsPiv in 3.8 mL of DMF, and 0.2 mL of D2O, GC-MS spectral
analysis indicated that, after reaction for 1 d, none of the
anticipated ester product was obtained and only the starting
biphenylene was present. Since 1 equiv of HI acid is usually
generated in our Heck palladium migration reactions, this
reaction was repeated in the presence of 1 equiv of DCl. Again,
none of the anticipated Heck product was obtained. This reaction
was also conducted using 1 equiv of Pd(OAc)2. After reaction
for 1 d, only biphenylene was evident by GC-MS spectral
analysis.
closely related to the chemistry described here are consistent
with a concerted four-centered transition state for activation of
the arene. Dedieu argues that a Pd-hydride species analogous
to ii is several kcal/mol above the low-energy pathway, in which
a Pd(II) intermediate is formed directly. Fagnou’s computations
show a low-energy pathway that directly produces a Pd(II)
intermediate and HBr (by proton transfer to a Br ligand on the
catalyst.) We recognize that the analogous intermediate in both
cases is iv, rather than ii, and we have no direct evidence
supporting the formation of ii, although the previously outlined
chemical arguments for ii remain.
Effects of Other Substituents. For a more general under-
standing of the scope and generality of the Pd migration in the
Heck reaction, several other substituted biaryls have been
investigated, as illustrated in Scheme 5. The reaction of 2-iodo-
3-phenylbenzofuran (5) and ethyl acrylate using procedure C
gives exclusively ethyl E-3-(3-phenylbenzofuran-2-yl)acrylate
(6) in 85% yield in 24 h. This result, showing no apparent 1,4-
Pd shift, suggested that palladium has a strong preference for
the 2-position of the benzofuran moiety. To test this hypothesis,
3-(2-iodophenyl)benzofuran (8) was used as a substrate under
the same conditions. Acrylate 6 was produced in 24 h in 78%
yield, alongside only ∼5% of isomer 7, indicating a strong
preference for palladium to migrate from the phenyl to the
benzofuran ring.46,47 As before, however, from the product
distributions alone, we cannot specifically rule out the possibility
that very different reactivities of the two arylpalladium inter-
mediates toward ethyl acrylate could be favoring the formation
of 6, although this seems unlikely, given the preponderance of
evidence. Once again, as a control experiment, compound 7 was
prepared as the sole product from 8 in 75% yield in 24 h by
carrying out the Heck reaction with ethyl acrylate under Jeffrey’s
reaction conditions (procedure A).
Similarly, the N-methylindole analogues of 5 and 8 were
reacted with ethyl acrylate (Scheme 5). Using procedure C, 9
produced exclusively E-3-(1-methyl-3-phenylindol-2-yl)acrylate
(10) in 94% yield in 24 h; a mixture of 77% 10 and 13% 11
This result is, again, mechanistically ambiguous regarding
the Pd migration. The most likely cause of the problem may be
that the conditions were not conducive to Pd insertion into the
(44) Masselot, D.; Charmant, J. P. H.; Gallagher, T. J. Am. Chem. Soc. 2006,
128, 694-695.
(45) Lafrance, M.; Rowley, C. N.; Woo, T. K.; Fagnou, K. J. Am. Chem. Soc.
2006, 128, 8754-8756.
(46) The 2-position of benzofuran can readily be palladated catalytically via
C-H activation. See ref 47 for a leading reference.
(47) Jia, C.; Lu, W.; Kitamura, T.; Fujiwara, Y. Org. Lett. 1999, 1, 2097-
2100.
(42) Edelbach, B. L.; Lachicotte, R. J.; Jones, W. D. J. Am. Chem. Soc. 1998,
120, 2843-2853.
(43) Satoh, T.; Jones, W. D. Organometallics 2001, 20, 2916-2919.
9
J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007 6303

A R T I C L E S
Scheme 5
Campo et al.
was obtained from 3-(2-iodophenyl)-1-methylindole (12). Using
procedure A as the nonmigration control again, an 85% yield
of 11 was obtained from 12.
results in the formation of a 65:35 mixture of isomers 17 and
18 (in 91% overall yield), within experimental error of the
statistical 2:1 ratio.
We believed that the 13 mol % of indole acrylate 11,
generated from 12 through procedure C, is a result of slow
palladium equilibration, which might be due to unfavorable
steric interactions imposed on the palladium when migrating
to the relatively more hindered 2-position of the N-methylindole.
Therefore, we postulated that steric hindrance might be a
significant factor in this palladium migration chemistry disfavor-
ing sterically congested arylpalladium intermediates.
In order to study the effect of steric hindrance on the
palladium migration, we prepared 2-iodo-3′,5′-dimethylbiphenyl
(13). The methyl substituent is not likely to chelate palladium,
and it appears to be essentially electronically neutral in the
palladium migration, based on the data reported in Table 2.
Using procedure A as the control, 14 was obtained in a 97%
yield, without formation of 15, as expected. Using the equili-
brating procedure C, 14 was still the exclusive Heck product,
formed in 94% yield. We thus conclude that the steric bulk of
an adjacent methyl group is sufficient to strongly inhibit
palladium migration. Unfortunately, we have been unable to
prepare 2-iodo-3,5-dimethylbiphenyl to examine its palladium
migration chemistry.
The multiply deuterated products discussed previously (Scheme
4) clearly indicate that multiple migrations can occur before
the key coupling step. This can be dramatically demonstrated
in another way by employing a substituent in the meta position,
as illustrated in Scheme 6. Thus 2-iodo-3-methoxybiphenyl (19)
produces isomers 20-22 (53:38:9 respectively, in 97% overall
yield) using equilibrating procedure C. The presence of the key
intermediate 25, generated after a minimum of two 1,4-Pd shifts
between the 2-, 2′-, and 6-positions of the biphenyl, explains
the formation of isomer 22.
Interestingly, the reaction of 2-iodo-3′-methoxybiphenyl (26)
and ethyl acrylate using procedure C produced a 25:62:13
mixture of compounds 20, 21, and 22, respectively, in 87%
overall yield (Scheme 6). The wide discrepancy in the Heck
product distributions obtained from the reaction of 19 versus
26 indicates that palladium is unable to achieve equilibrium
between the three different positions of the biphenyl prior to
olefin trapping under the reaction conditions employed. This
result is consistent with the deuterium exchange results and some
of the entries in Table 2, both of which imply that a true
equilibrium is not quite achieved before the carbon-carbon
coupling step.
Another factor that may add to the complexity of this
particular system is the possibility of intramolecular oxygen
chelation in the arylpalladium intermediate 23. Such chelation
would obviously affect both the rates and equilibrium, presum-
ably in a manner opposite to that expected due to the steric
hindrance of the methoxy group.
A related example of palladium migration involves the Heck
reaction of 2-iodo-2′-methylbiphenyl (16). Steric congestion is
not reduced by migration of the Pd because the methyl and iodo/
Pd substituents are not in the same plane. (We presume the two
phenyl planes are nearly perpendicular in the low energy
conformation.) Thus, while procedure A results in the formation
of 17 in a 92% yield as the only Heck product, procedure C
Palladium Migration in Suzuki Reactions. After exploring
the Heck reaction, we proceeded to examine the Suzuki cross-
coupling2 of the same o-halobiaryls with arylboronic acids.
Because the conditions for the Suzuki and Heck reactions differ,
we cannot assume that the rate of palladium migration will be
identical in both processes. However, understanding the relative
rates of migration vs coupling in the Suzuki reaction and the
9
6304 J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007

Aryl to Aryl Palladium Migration of o-Halobiaryls
A R T I C L E S
Scheme 6
Table 3. Pd-Catalyzed Reaction of 2-Iodo-4′-methylbiphenyl (1a)
and 4-(Methoxycarbonyl)phenylboronic Acid (27a)
an attempt to promote palladium migration, the coupling reaction
of 1a and 27a was then carried out under migration conditions
similar to those employed in procedure C for the Heck reaction
(entry 2). However, only trace amounts of coupling products
were observed. It has been reported that the addition of water
facilitates Suzuki coupling reactions.14 Therefore, we carried
out the coupling reaction of 1a and 27a in the presence of 20
equiv of H2O and obtained a 57:43 mixture of Suzuki product
28a and palladium migration product 29a in 55% overall yield
(entry 3). Based on previous experience with the Heck reaction,
it was assumed that the true “equilibrium” distribution of the
two products should be approximately 1:1, so conditions were
sought which would approach that ratio.
23a,
mole ratio
28a:29ab
entry
equiv
procedurea
time, h
% yield
1
2
3
4
5
1.2
1.0
1.0
1.0
1.4
E
F
8
72
8
24
4
100:0
62
trace
55
57
78
F + 20 H2O
57:43
50:50
51:49
F + 20 H2O + 2 PAc
F + 20 H2O + 2 PAc
a All reactions were run using 0.25 mmol of 1a, 5 mol % Pd(OAc)2,
and appropriate equivalents of boronic acid 27a and base. Procedure E:
10 mol % PPh3, 2.2 equiv of CsF in 1 mL of DME at 90 °C. Procedure F:
5 mol % (Ph2P)2CH2 (dppm) and 2.0 equiv of CsPiv in 4 mL of DMF at
100 °C. b The mole ratio was determined by 1H NMR spectroscopic analysis.
c PA ) pivalic acid (Me3CCO2H).
Realizing that the presence of base probably activates the
arylboronic acid toward direct Suzuki coupling prior to pal-
ladium migration, we buffered the reaction mixture using a
combination of 2 equiv of cesium pivalate (CsPiv) and 2 equiv
of pivalic acid (PA). To our satisfaction, the reaction of 1a and
27a under these buffered conditions produced a 50:50 mixture
of 28a and 29a in 57% overall yield (entry 4). An even higher
overall yield (78%) of a 51:49 mixture of 28a and 29a could
be obtained by employing 1.4 equiv of the arylboronic acid 27a
(entry 5, also entry 1 in Table 4).
compatibility of the Suzuki reaction with palladium migration
for synthetic purposes was of interest. To establish what reaction
conditions best promote palladium migration, we first studied
the Suzuki coupling of 2-iodo-4′-methylbiphenyl (1a) and
4-(methoxycarbonyl)phenylboronic acid (27a) under various
reaction conditions (Table 3).
An identical distribution of Suzuki products, in 83% overall
yield, was obtained from 4a (X ) Me, Scheme 7). Thus, we
chose this latter set of reaction conditions, described in entry
5, Table 3, as our standard palladium migration conditions
for the Suzuki coupling of o-iodobiphenyls with other
arylboronic acids (Scheme 7, Table 4). The iodobiaryls 1a and
4a gave similar distributions of Suzuki products 28a and 29a
when coupled with either phenylboronic acid (27b) or 4-
methoxyphenylboronic acid (27c) using our standard Suzuki
migration procedure (Table 4, entries 3-6). The insensitivity
of the product ratio to the electron demand of the boronic acids
implies that the product distribution is determined by the
equilibrium position of the palladated intermediate before
transmetallation.
The Pd-catalyzed Suzuki coupling of 1a and 27a under
conditions described by Wright et al.48 [0.25 mmol 1a, 5 mol
% Pd(OAc)2, 10 mol % PPh3, 1.2 equiv of boronic acid and
2.2 equiv of CsF in 1 mL of DME at 90 °C, procedure E]
produced methyl 4′-methyl-o-terphenyl-4-carboxylate (28a) in
62% yield (entry 1, Table 3), with no other Suzuki product. In
(48) Wright, S. W.; Hageman, D. L.; McClure, L. D. J. Org. Chem. 1994, 59,
6095-6097.
9
J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007 6305

A R T I C L E S
Campo et al.
Table 4. Pd-Catalyzed Reaction of o-Iodobiphenyls (1 and 4) and
Arylboronic Acids (27) Using the Standard Suzuki Migration
Procedurea
standard Suzuki palladium migration conditions gave only 32
in 67% yield in 3 h (Scheme 9). When 3-(2-iodophenyl)-1-
methylindole (12) was allowed to react with 27a, the migration
product 32 was produced in a 57% yield in 6 h, along with a
trace amount of isomer 33, which clearly indicates a preference
for palladium migration from the phenyl ring to the indole ring.15
Adduct 33 can be obtained exclusively in 79% yield when 12
is allowed to react with boronic acid 27a under the nonmigration
conditions described by Wright et al.48
Basis for Selectivity. It is our working hypothesis that the
mixtures obtained under “equilibrating” conditions qualitatively
reflect the preferred position of the Pd, rather than differing
rates of reactivity of the arylpalladium complexes. Clearly, the
reactivity patterns are not consistent with the pi system of the
benzene ring acting as a nucleophile, as in electrophilic aromatic
substitution. Rather, as noted, there appears to be a correlation
between inductive electron withdrawal or C-H acidity and the
favored position.49
biphenyl,
ArB(OH)2,
Ar
mole ratio
28:29b
entry
X
)
)
time (h)
% yield
1
2
3
4
5
6
7
8
9
1a, Me
4a, Me
1a, Me
4a, Me
1a, Me
4a, Me
1b, MeO
4b, MeO
1c, CO2Et (27a)
4c, CO2Et (27a)
1d, NO2
4d, NO2
p-MeO2CC6H4 (27a)
(27a)
Ph (27b)
(27b)
p-MeOC6H4 (27c)
(27c)
(27a)
4
6
3
10
3
6
4
9
10
10
25
10
51:49
49:51
52:48
50:50
52:48
49:51
42:58
39:61
40:60
34:66
23:77
16:84
78
83
79
69
93
90
85
75
84
68
61
75
(27a)
10
11
12
(27a)
(27a)
a All reactions were run using 0.25 mmol of the o-iodobiphenyl, 1.4
equiv of arylboronic acid, 5 mol % Pd(OAc)2, 5 mol % dppm, 2.0 equiv of
CsPiv, 2.0 equiv of PA, and 20 equiv of H2O in 4 mL of DMF at 100 °C.
1
b The mole ratio was determined by H NMR spectroscopic analysis.
Fagnou recently showed that haloarenes can be coupled to
various fluoroarenes with Pd catalysts by C-H activation of
the fluoroarene.45 Selectivity studies showed a clear correlation
between C-H acidity and regiochemistry of C-H activation/
coupling in the Fagnou chemistry, indicating at least a kinetic
selectivity, and presumably a thermodynamic one as well. The
current palladium migration chemistry might be viewed as
entropically activated C-H activation, the lower requirements
of inherent C-H acidity being overcome by the preorganization
aspect of the intramolecular reaction. In fact, intramolecular
coupling to form biaryls using arenes bearing no electron-
withdrawing groups (and even methoxy groups) have been
reported by Fagnou as well.50
Scheme 7
Interestingly, the Pd-catalyzed reactions of the methoxy,
carbethoxy, and nitro-substituted o-iodobiphenyls 1b-d and
4b-d (XdMeO, CO2Et, NO2) with boronic acid 27a using our
standard Suzuki migration procedure show a consistent trend
in the distribution of Suzuki products (entries 7-12), slightly
favoring coupling on the substituted phenyl ring. (Control
reactions under the Wright conditions provide 4a-d as the only
Suzuki products in 62-98% yields.) If these product distribu-
tions, along with estimates based on averages from the corre-
sponding Heck reactions, can be taken as qualitative indicators
of the equilibrium position of the Pd in the substituted biphenyls,
the most sensible correlation consistent with the order NO2 >
CO2Et > OCH3 > H is that of inductiVe electron withdrawal
by the substituent, i.e., through the sigma framework, rather
than through the pi system.
Similar to previous observations on the Heck reaction, an
even more marked effect on the product distribution was
observed in the reaction of 2-iodo-3-phenylbenzofuran (5) with
4-(methoxycarbonyl)phenylboronic acid (27a), which under our
Suzuki migration conditions gave exclusively 2-[4-(methoxy-
carbonyl)phenyl]-3-phenylbenzofuran (30) in 79% yield after
3 h (Scheme 8). Obviously, no palladium migration has occurred
in this biaryl system. On the other hand, when 3-(2-iodophenyl)-
benzofuran (8) was allowed to react with boronic acid 27a under
our standard Suzuki palladium migration conditions, 30 was
produced in 78% yield in 3 h, along with only ∼5% of isomer
31. These results again clearly show a preference for palladium
migration from the phenyl ring to the benzofuran ring.47
Furthermore, a 75% yield of 31, along with only ∼7% of
benzofuran 30, was formed from 8 when employing the reaction
conditions described by Wright et al.48
Conclusions
We have observed a number of palladium-catalyzed reactions
that proceed by 1,4-palladium migration in organopalladium
intermediates derived from o-halobiaryls. These arylpalladium
intermediates have been generated under relatively mild reaction
conditions compatible with the Heck and Suzuki reactions. We
have developed standard migration reaction conditions, which
allow arylpalladium intermediates to approach equilibrium prior
to the trapping step. The palladium migration can be activated
or suppressed at will by simple manipulation of the reaction
conditions, and the selectivities for the equilibrated position
appear to correlate most strongly with anticipated C-H acidity.
Experimental Section
General Procedure for the Heck Coupling of o-Iodobiaryls with
Ethyl Acrylate under the Conditions Described by Jeffrey et al.
(Procedures A and B).18 2-Iodo-4′-methylbiphenyl (73.5 mg, 0.25
mmol), Pd(OAc)2 (2.8 mg, 0.0125 mmol), n-Bu4NCl (0.0694 g, 0.25
mmol), NaHCO3 (0.042 g, 0.5 mmol), and ethyl acrylate (0.10 g, 1.0
mmol) in DMF (1 mL) under Ar at 100 °C were stirred for 1 d
(procedure A). The reaction mixture was then cooled to room
temperature, diluted with diethyl ether (35 mL), and washed with brine
(30 mL). The aqueous layer was reextracted with diethyl ether (15 mL).
The organic layers were combined, dried (MgSO4), and filtered, and
(49) Computations on the enthalpies of the corresponding naked anions of model
biaryls are qualitatively consistent with the product distributions, although
the energy differences are much larger than our observed product
distributions justify. See the Supporting Information for these data and a
brief discussion.
(50) Campeau, L.-C.; Parisien, M.; Leblanc, M.; Fagnou, K. J. Am. Chem. Soc.
2004, 126, 9186-9187.
Again, the indole analogue was examined. Reaction of 2-iodo-
1-methyl-3-phenylindole (9) and boronic acid 27a under our
9
6306 J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007

Aryl to Aryl Palladium Migration of o-Halobiaryls
A R T I C L E S
Scheme 8
Scheme 9
the solvent was removed under reduced pressure. The reaction mixture
was chromatographed using 7:1 hexanes/ethyl acetate to afford 66.6
mg (100%) of ethyl E-3-(4′-methylbiphen-2-yl)acrylate (2a) as a clear
oil.16 In procedure B, the NaHCO3 was replaced by Et3N (0.051 g, 0.5
mmol).
optimization, Hessian, and energy runs were performed with the
GAMESS suite of programs.52 Results were visualized using MacMol-
Plt.53 All structures were found to have all positive definite vibrational
frequencies, verifying them as minima.
For Pd-containing compounds, a two-stage procedure was used. After
a semiempirical conformational search was used to establish the lowest
energy conformation and starting geometry, optimizations were carried
out at the RHF and B3LYP54-57 levels of theory using a mixed group
of basis sets. The 6-31G(d) basis set was used for carbon, hydrogen,
nitrogen, and phosphorus. For iodine, the LANL2DV effective core
potential, supplemented with p-diffuse functions and d-polarization
functions, was used. For palladium, the LANL2DV effective core
potential was supplemented with f functions.
Single-point energies with larger basis sets were then obtained at
the MP2 and B3LYP levels of theory, using the RHF and B3LYP
optimized geometries, respectively. For C and H, the basis set was
6-31G(d). For N and O, the basis set was 6-311+G(2d). For P,
6-311+G(3df) was used. For I, the LANL2DZ effective core potential
was supplemented with s and p diffuse functions and d and f polarization
functions. For Pd, LANL2DZ was supplemented with f polarization
functions. Computations on biphenyl anions and their derivatives are
described in the Supporting Information.
General Procedure for the Heck Coupling of o-Iodobiaryls with
Ethyl Acrylate Using Palladium Migration Procedures C and D.
2-Iodo-4′-methylbiphenyl (73.5 mg, 0.25 mmol), Pd(OAc)2 (2.8 mg,
0.0125 mmol), (Ph2P)2CH2 (dppm, 4.8 mg, 0.0125 mmol), CsO2CCMe3
(CsPiv) (0.117 g, 0.5 mmol), and ethyl acrylate (0.025 g, 0.25 mmol)
in DMF (4.0 mL) under Ar at 100 °C were stirred for 1.5 d. The reaction
mixture was then cooled to room temperature, diluted with diethyl ether
(35 mL), and washed with brine (30 mL). The aqueous layer was
reextracted with diethyl ether (15 mL). The organic layers were
combined, dried (MgSO4), and filtered, and the solvent was removed
under reduced pressure. The reaction mixture was chromatographed
using 7:1 hexanes/ethyl acetate to afford 58.6 mg (88%) of ethyl E-3-
(4′-methylbiphen-2-yl)acrylate (2a) and ethyl E-3-(4-methylbiphen-2-
yl)acrylate (3a) as a clear oil in a 50:50 molar ratio as determined by
1H NMR spectroscopic analysis.16 In procedure D, the DMF solvent
was replaced by a 95:5 DMF/H2O mixture.
General Procedure for the Suzuki Coupling of o-Iodobiaryls with
Arylboronic Acids under Palladium Migration Conditions. To a
2-dram vial was added the o-iodobiaryl (0.25 mmol), the arylboronic
acid (0.35 mmol, 1.4 equiv), Pd(OAc)2 (2.8 mg, 5 mol %), (Ph2P)2-
CH2 (dppm, 4.8 mg, 5 mol %), CsO2CCMe3 (CsPiv, 117 mg, 2.0 equiv),
Me3CCO2H (PA, 51 mg, 2 equiv), H2O (90 mg, 20 equiv), and DMF
(4.0 mL). The mixture was flushed with Ar and stirred at 100 °C in an
oil bath. The completion of the reaction was determined by the
observation of palladium black. The reaction mixture was cooled,
diluted with ether, and washed with satd aq NaHCO3. The organic layer
was dried (Na2SO4), filtered, and evaporated. The residue was purified
by column chromatography on a silica gel column, and the molar ratio
Acknowledgment. The authors thank the National Science
Foundation for financial support of this research. Thanks are
also extended to Johnson Matthey, Inc. and Kawaken Fine
Chemicals Co. for donating Pd(OAc)2 and PPh3, and Frontier
Scientific, Inc. for donating the arylboronic acids. The authors
also thank the NCSA for a generous granting of computational
time.
Supporting Information Available: Preparation and charac-
terization of the starting materials and all new compounds, mass
spectral data for the deuteration experiments, and structures and
energies for all computational structures. This material is
available free of charge via the Internet at http://pubs.acs.org.
1
of the products was determined by H NMR spectroscopic analysis.
General Procedure for the Suzuki Coupling of o-Iodobiaryls with
Arylboronic Acids under the Conditions Described by Wright et
al. (Nonmigration Conditions).48 To a 2-dram vial was added the
o-iodobiaryl (0.25 mmol), the arylboronic acid (0.30 mmol, 1.2 equiv),
Pd(OAc)2 (2.8 mg, 5 mol %), PPh3 (6.5 mg, 10 mol %), CsF (84.0
mg, 2.2 equiv), and DME (1.0 mL). The mixture was flushed with Ar
and stirred at 90 °C in an oil bath. The completion of the reaction was
determined by the observation of palladium black. The reaction mixture
was cooled and purified directly by column chromatography on a silica
gel column.
JA069238Z
(51) Hehre, W. Spartan, v. 3.1; Wavefunction, Inc: 18401 Karman Ave. Irvine,
CA, 2002.
(52) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M.
S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, N.; Su, S. J.; Windus,
T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem. 1993, 14, 1347-
1363.
(53) Bode, B. M.; Gordon, M. S. J. Mol. Graphics Modell. 1998, 16, 133-138.
(54) Becke, A. D. Phys. ReV. A 1988, 38, 3098-3100.
(55) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(56) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys.
Chem. 1994, 98, 11623-11627.
Computational Methods. To find the lowest energy conformations
for subsequent ab initio optimizations, semiempirical conformational
searches were performed, as implemented in MacSpartan.51 All other
(57) Hertwig, R. H.; Koch, W. Chem. Phys. Lett. 1997, 268, 345-351.
9
J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007 6307