Controlling site selectivity in Pd-catalyzed oxidative cross-coupling reactions

Article abstract of DOI:10.1021/ja1097918

This paper presents a detailed investigation of the factors controlling site selectivity in the Pd-mediated oxidative coupling of 1,3-disubstituted and 1,2,3-trisubstituted arenes (aryl-H) with cyclometalating substrates (L～C-H). The influence of both the concentration and the steric/electronic properties of the quinone promoter are studied in detail. In addition, the effect of steric/electronic modulation of the carboxylate ligand is discussed. Finally, we demonstrate that substitution of the carboxylate for a carbonate X-type ligand leads to a complete reversal in site selectivity for many arene substrates. The origins of these trends in site selectivity are discussed in the context of the mechanism of Pd-catalyzed oxidative cross-coupling.

Full text of DOI:10.1021/ja1097918

ARTICLE
pubs.acs.org/JACS
Controlling Site Selectivity in Pd-Catalyzed Oxidative Cross-Coupling
Reactions
Thomas W. Lyons, Kami L. Hull, and Melanie S. Sanford
Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109, United States
S Supporting Information
b
ABSTRACT: This paper presents a detailed investigation of the
factors controlling site selectivity in the Pd-mediated oxidative
coupling of 1,3-disubstituted and 1,2,3-trisubstituted arenes (aryl-
H) with cyclometalating substrates (L∼C-H). The influence of
both the concentration and the steric/electronic properties of the
quinone promoter are studied in detail. In addition, the effect of
steric/electronic modulation of the carboxylate ligand is discussed.
Finally, we demonstrate that substitution of the carboxylate for a carbonate X-type ligand leads to a complete reversal in site selectivity for
many arene substrates. The origins of these trends in site selectivity are discussed in the context of the mechanism of Pd-catalyzed oxidative
cross-coupling.
’ INTRODUCTION
transformations can be widely practiced in organic synthesis.
Currently, the most significant limitation is the inability to pre-
dictably control and modify site selectivity. This paper describes
studies that address this crucial issue in the oxidative cross-coupling
between cyclometalating substrates (L∼C-H) and aryl-H deri-
vatives.
The biaryl linkage is an important structural motif that is
prevalent in numerous organic compounds, including natural
products,1a-1e pharmaceuticals,1f,1g agrochemicals,1h and conju-
gated materials.1j Aryl-aryl bonds are most commonly formed
via Pd-, Ni-, or Cu-catalyzed cross-coupling between aryl halide
electrophiles and [M]-aryl nucleophiles (eq 1, i).2 While such
reactions are extremely important and widely used synthetic
methods, they remain fundamentally limited by the requirement
for two prefunctionalized coupling partners.3 C-H oxidative coupl-
ing reactions represent a highly attractive alternative strategy for
constructing aryl-aryl bonds (eq 1, ii).3m,3n,4 In these transforma-
tions, the biaryl linkage is generated via sequential C-H activation
of two aromatic substrates (aryl-H and aryl0-H) followed by C-
C bond formation. As such, this approach precludes the need for
prefunctionalization of either aromatic starting material.
Our group has recently communicated a Pd-catalyzed reaction
for the oxidative cross-coupling of benzo[h]quinoline (bzq) with
arenes (aryl-H) promoted by benzoquinone (BQ) in the
presence of a Ag1 oxidant (eq 2a).8b Extensive studies implicated
a mechanism involving cyclopalladation of bzq to generate
[(bzq)Pd(OAc)]2, reversible activation of aryl-H, BQ-bind-
ing/BQ-promoted C-C bond-forming reductive elimination,9
Several recent reports have described exciting progress in the
development of Pd-catalyzed oxidative coupling reactions.5-8
For example, our group8b and others5,6,8 have demonstrated the
cross-coupling of benzene with arenes, such as benzo[h]quinoline
(eq 2a),8b benzofuran (eq 2b),6c and anisole (eq 2c).5a Despite
these important advances, many challenges remain before these
Received: October 31, 2010
Published: March 07, 2011
r
2011 American Chemical Society
4455
dx.doi.org/10.1021/ja1097918 J. Am. Chem. Soc. 2011, 133, 4455–4464
|

Journal of the American Chemical Society
ARTICLE
Scheme 1. Proposed Mechanism for Oxidative Coupling
Between bzq and aryl-H
these oxidative coupling reactions. We demonstrate that the
X-type ligands have a particularly dramatic influence on selec-
tivity and discuss the origin of these ligand effects in the context
of the mechanistic framework in Scheme 1.
’ RESULTS AND DISCUSSION
Selection of the Arene Substrate for Initial Studies. 1,3-
Dimethoxybenzene (DMB) was selected as a test substrate for
probing site selectivity. DMB contains three different aromatic
hydrogens (HA, HB, and HC) with dramatically different steric and
electronic properties. HB and HC are attached to the most nucleo-
philic carbons in the molecule. HA is located at the most sterically
accessible position. HC is at the most sterically crowded site on the
arene. Under our original conditions [10 mol % of Pd(OAc)2, 2
equiv of Ag2CO3, 4 equiv of DMSO], DMB underwent oxidative
coupling with benzo[h]quinoline to afford a 3:1 mixture of isomers
A/B (derived from replacing HA and HB, respectively) (eq 3).11
This result suggests that there is relatively little inherent selectivity in
this system, which renders it an ideal case for studying catalyst
control.
Selection of the Reaction to Investigate Site Selectivity.
The studies described throughout this paper involve the reaction
of DMB with [(bzq)PdX]2 to afford mixtures of isomers A, B,
and C (eq 4). This stoichiometric transformation allows us to
directly probe DMB functionalization without interference from
the cyclometalation or oxidation steps of the catalytic cycle. Most
importantly, this approach excludes any influence of oxidant-
derived X/L-type ligands on reactivity or selectivity. The results
obtained in regime 1 (1 equiv of quinone relative to [Pd]) and
regime 2 (20 equiv of quinone relative to [Pd])12 are discussed
below as a function of three variables: (1) quinone, (2) carboxy-
late, and (3) other X-type ligands.13
Figure 1. Initial rate versus equivalents of BQ in the coupling of com-
pound 1 with 1,2-dimethoxybenzene.
and finally, reoxidation of Pd0 by AgI to regenerate the PdII
catalyst (Scheme 1).10
Intriguingly, the rate-determining step of this sequence is
dependent upon the concentration of the quinone promoter. For
example, in the reaction of [(bzq)Pd(OAc)]2 with 1,2-dimethox-
ybenzene, a first-order dependence upon [BQ] was observed at
low concentrations of BQ, suggesting that quinone complexation
(step ii) is rate-determining.9,10 In contrast, saturation was
observed at high [BQ], implicating rate-limiting C-H activation
(step i) under these conditions (Figure 1).10 These two kinetic
regimes will be referred to as regime 1 (low [BQ], 1 equiv) and
regime 2 (high [BQ], 20 equiv) throughout this paper.
With this mechanistic framework in hand, we turned our
attention toward understanding and controlling the site selec-
tivity of aryl-H functionalization. We hypothesized that the
change in rate-determining step between regimes 1 and 2 might
also lead to a change in site selectivity. Furthermore, we reasoned
that systematic modification of the ancillary ligands at Pd inter-
mediates 1-3 might enable catalyst control. This paper describes
a detailed investigation of the factors dictating site selectivity in
Role of Quinone. The reaction of DMB with [(bzq)Pd-
(OAc)]2 was first studied as a function of the concentration of
BQ. The site selectivity showed a clear dependence upon [BQ],
with the ratio of isomers A/B ranging from 11:1 at 0.005 M (0.2
equiv) to 1.1:1 at 0.5 M (20 equiv). (Less than 5% of isomer C was
observed in any of these reactions.) A plot of A/Bversus equiv of BQ
(Figure 2) showed that the ratio levels out at approximately 1:1 at
0.20 M (8 equiv). This is the same concentration at which saturation
is observed in Figure 1, consistent with a change in both rate- and
selectivity-determining steps between regimes 1 and 2.
4456
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464

Journal of the American Chemical Society
ARTICLE
This reaction was next studied using a series of alkyl-substituted
quinone promoters. In regime 1 (0.025 M, 1 equiv of quinone),12
quinone substitution had a dramatic effect on selectivity, with A/B
increasing from 5:1 with BQ to 14:1 with 2,3,5,6-tetramethylbenzo-
quinone (Table 1).14 Notably, the improved selectivity was generally
accompanied by a decrease in the overall yield. In regime 2 (0.5 M, 20
equiv of quinone),12 the A/B ratio was significantly lower with all of
the quinone derivatives. This ratio dropped even further as [quinone]
was increased to 1.0 M (40 equiv).
Selectivity was also investigated with a series of para-substi-
tuted 2,5-diarylquinones. This study was conducted in regime
1, because this is where the largest effect was expected. In
general, the A/B ratio was higher with more electron-donating
X substituents (Table 2). A Hammett plot of (log[A/(A þ B)X/
A/(A þ B)]H) versus σpara (Figure S1 in the Supporting
Information) showed a reasonable correlation (R2 = 0.86), with a
F value of -0.8 ( 0.2.
Role of Carboxylate. The reaction of DMB with [(bzq)Pd-
(O2CR)]2 was next studied using carboxylate ligands with R =
Me, Et, iPr, or tBu. In regime 2, larger R groups provided enhanced
preference for isomer A; for example, A/B increased from 1:1.5 to
1.8:1 upon moving from R = CH3 to tBu (entries 5-8 in Table 3).
Interestingly, a similar trend was observed in regime 1, with A/B
changing from 5:1 to 9:1 for R = CH3 and tBu, respectively (entries
1-4 in Table 3).
We also added 3 equiv of RCO2H to the reactions in regime 1
to facilitate complete equilibration of the C-H activation step
(step i in Scheme 1). Under these conditions, A/B increased
from 5:1 to 16:1. Furthermore, this ratio remained constant over
the entire range of carboxylate derivatives (Table 4).
The electronic influence of the carboxylate ligand was assessed
using [(bzq)Pd(O2C(p-XC6H4))]2. In regime 1, the para substitu-
ent X had a relatively minimal impact on selectivity, particularly upon
the addition of 3 equiv of ArCO2H (entries 1-8 in Table 5). In
regime 2, small electronic effects were observed, with A/B ranging
from 1:1.7 to 1:2.9. However, there was no clear correlation between
A/B and Hammett σ values in this system.
Discussion of Ligand Effects with [(bzq)Pd(O2CR)]2. The
results described above provide a comprehensive picture of the in-
fluence of the quinone promoter and carboxylate ligand on selectivity.
A detailed mechanism that is consistent with all of this data is shown
in Scheme 2, and it implicates two limiting mechanistic scenarios.
In the first (limit 1), intermediates PdA and PdB are in a fast and
reversible equilibrium, where k1[1][DMB] and k-1A/B[PdA/B]-
[RCO H] . k2A/B[PdA/B][quinone]. This is a classic Curtin-
2
Hammett situation, where the rate and selectivity will be dictated by
the relative magnitudes of ΔGq for quinone complexation with PdA
versus PdB, respectively. In the second limiting possibility (limit 2),
k2A/B[PdA/B][quinone] . k1[1][DMB]. Under these conditions,
the rate and selectivity will be determined by the relative magnitude of
ΔGq for C-H activation to form PdA versus PdB, respectively.
The experiments described above use 1 equiv of quinone
(regime 1) and 20 equiv of quinone (regime 2) to approximate
Figure 2. A/B versus equiv of BQ for the reaction of compound 1 with
DMB.
Table 1. Site Selectivity as a Function of Quinone Substitutiona
a Yields were determined by gas chromatography (GC) analysis of the crude reaction mixtures versus an internal standard and represent an average of 2 runs.
4457
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464

Journal of the American Chemical Society
ARTICLE
Table 2. Site Selectivity as a Function of Quinone Substitu-
tion (Electronic Effects)
Table 4. Site Selectivity as a Function of Carboxylate Sub-
stitution in the Presence of Exogenous Acid
entry
conditions
R
yield (%)a
A/B
entry
X
yield (%)a
A/B
1
2
3
4
1 equiv of BQ, 3 equiv of MeCO2H
1 equiv of BQ, 3 equiv of EtCO2H
1 equiv of BQ, 3 of equiv iPrCO2H
1 equiv of BQ, 3 equiv of tBuCO2H
Me
Et
iPr
tBu
94
95
91
64
16:1
16:1
16:1
16:1
1
2
3
4
5
CF3
F
62
59
65
49
68
1.1:1
2.1:1
3:1
H
Me
OMe
2.6:1
a Yields were determined by GC analysis of the crude reaction mixtures
versus an internal standard and represent an average of 2 runs.
5.7:1
a Yields were determined by GC analysis of the crude reaction mixtures
versus an internal standard and represent an average of two runs.
Table 5. Site Selectivity as a Function of Benzoate Substitu-
tion (Electronic Effects)
Table 3. Site Selectivity as a Function of Carboxylate Sub-
stitution
entry
conditions
R
yield (%)a
A/B
yield
entry
conditions
1 equiv of BQ
p-XC6H4 (Ar) (%)a A/B
1
2
3
4
5
6
7
8
1 equiv of BQ
1 equiv of BQ
1 equiv of BQ
1 equiv of BQ
20 equiv of BQ
20 equiv of BQ
20 equiv of BQ
20 equiv of BQ
Me
Et
iPr
tBu
Me
Et
76
78
79
84
61
62
82
65
5:1
6:1
1
2
3
4
5
6
7
8
9
p-MeOC6H4
C6H5
58 14:1
62 10:1
45 11:1
39 11:1
36 17:1
35 16:1
23 19:1
7:1
1 equiv of BQ
1 equiv of BQ
1 equiv of BQ
9:1
p-CF3C6H4
p-NO2C6H4
1:1.5
1:1.2
1.5:1
1.8:1
1 equiv of BQ/3 equiv of ArCO2H p-MeOC6H4
1 equiv of BQ/3 equiv of ArCO2H C6H5
iPr
tBu
1 equiv of BQ/3 equiv of ArCO2H p-CF3C6H4
1 equiv of BQ/3 equiv of ArCO2H p-NO2C6H4
a Yields were determined by GC analysis of the crude reaction mixtures
versus an internal standard and represent an average of 2 runs.
9
36
40
38
38
20:1
1:1.7
20 (regime 2)
p-MeOC6H4
C6H5
these two limiting scenarios. However, changes to the quinone
and the carboxylate ligands can clearly influence both the relative
and absolute values of the six key rate constants in this system
10 20 (regime 2)
11 20 (regime 2)
12 20 (regime 2)
1:2
p-CF3C6H4
p-NO2C6H4
1:2.9
1:2.5
A
B
A
B
(k1 , k1 , k-1A, k-1B, k2 , and k2 ). As such, the experimental
results in regimes 1 and 2 approach but do not necessarily reach the
limiting scenarios outlined above. In other words, while quinone
complexation always dominates the selectivity in regime 1, the C-H
activation step can also provide a small contribution under some
conditions. Similarly, the C-H activation step always dominates the
selectivity in regime 2, but quinone complexation can also have some
influence depending upon the ligands present and relative rate
constants. A detailed discussion of the effects of quinone and car-
boxylate on regimes 1 and 2 is outlined below.
a Yields were determined by GC analysis of the crude reaction mixtures
versus an internal standard and represent an average of two runs.
PdA because of the ortho-substitued σ-aryl ligand. As a result, we
hypothesize that quinone association (and subsequent irreversible
bzq-DMB coupling) is slower at PdB than PdA, leading to selec-
tive formation of isomer A.15
In regime 1, more electron-deficient quinones provided lower
A/B ratios (Table 2), and there are likely two factors that
contribute to this effect. First, electron-deficient BQ derivatives are
more reactive π-acids and are thus expected to be less selective in
distinguishing between PdA and PdB. Second, the higher reactivity of
Discussion of the Quinone Promoter. In regime 1, we
propose that the rate and selectivity are primarily dictated by
quinone complexation through the dz2 orbital of one of two
interconverting σ-aryl Pd intermediates (PdA and PdB). The dz2
orbital of PdB is significantly more sterically congested than that of
electron-deficient quinones should increase k2A and k2 , leading to a
B
larger contribution of the C-H activation step (which shows less
kinetic preference for isomer A) to selectivity.16
4458
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464

Journal of the American Chemical Society
ARTICLE
Scheme 2. Detailed Mechanism for Coupling between [(bzq)Pd(O2CR)]2 and DMB
In regime 1, alkyl substituents on the quinone resulted in en-
hanced selectivity for A(Table 1). This is likely due to a combination
of electronic and steric effects. Alkyl groups are moderately electron-
donating, which, as discussed above, favors the formation of A. Alkyl
substitution also increases the size of the quinone, which should
exacerbate unfavorable steric interactions in its reaction with PdB.
In regime 2, the rate and selectivity are dominated by the
C-H activation step. Interestingly, under these conditions, the reac-
tion showed almost no selectivity for isomer A, providing an approxi-
mately 1:1 ratio of A/B (20 equiv of BQ). This result indicates that,
as conditions approach limit 2, there is much less kinetic preference
for activation of HA versus HB at this PdII center. Notably, most other
C-H activation reactions at PdII are strongly affected by steric/
electronic factors;3,6,8,17 therefore, the apparently low inherent site
selectivity of this step is an unusual feature of the current transforma-
tion, which remains the subject of detailed exploration.
In regime 2, changes to the quinone structure are expected to
have minimal impact on selectivity, because quinone is not involved
in the C-H activation step. As illustrated in Table 1, increasing the
[quinone] resulted in a decrease of the A/B ratio in all cases.
However, even at 1.0 M (40 equiv), the differentially substituted
quinones provided somewhat different selectivities. These data
suggest that many of these reactions have not completely reached
limit 2, and thus, the quinone complexation step still provides some
contribution to selectivity.
protonation of PdA/PdB (k-1[RCO2H][PdA/B]) compared to that
of BQ complexation (k2[BQ][PdA/B]). As such, it should promote
more complete equilibration of the C-H activation step, such that
selectivity is dictated solely by ΔΔGq for quinone complexation.
In regime 2, the carboxylate ligand is expected to be coordi-
nated to Pd during the rate/selectivity-determining C-H activa-
tion step. Consistent with this, the A/B ratio in regime 2 was
sensitive to both steric and electronic perturbation of R. Sterically
large carboxylate ligands provided enhanced selectivity for
functionalization at the least hindered site of DMB (to form
A), implicating unfavorable steric interactions in the transition
state for C-H activation. Small electronic effects were also
observed with p-substituted benzoate derivatives; however, the
modest magnititude and the lack of a clear trend preclude
definitive interpretation of these results at this time.
Influence of Varying the X-Type Ligand. On the basis of the
studies described above, we reasoned that substituting carbox-
ylates with alternative X-type ligands might further modulate
selectivity in these systems. Thus, we examined the reaction of
DMB with [(bzq)PdCl]2 (5) in the presence of a variety of AgX
salts. Importantly, complex 5 is unreactive under these condi-
tions without added AgX (entry 1 in Table 6), presumably
because of its low solubility. Therefore, any observed reactivity
implicates at least partial substitution of Cl with X.
We first examined the reaction of compound 5 with AgOAc
and AgOPiv. The A/B selectivity in these systems was nearly
identical to that obtained using the preformed carboxylate-
bridged dimers (compare entries 1 and 4 in Table 3 to entries
2 and 3 in Table 6). This suggests that the reaction between
compound 5 and AgO2CR generates [(bzq)Pd(O2CR)] in situ.
We next surveyed Ag salts containing weakly coordinating
Discussion of the Effect of Carboxylate Ligands. In regime 1,
sterically larger carboxylates provided enhanced selectivity for A
(entries 1-4 in Table 3). We initially considered that the carboxylic
acid might serve as the ligand L in PdA/B (Scheme 2). However, this
proposal is inconsistent with the inverse first-order kinetics in AcOH
observed during this transformation,10 which implicate AcOH dis-
sociation prior to the rate-determining step.
-
-
counterions (e.g., CF3CO2 , NO3 , BF4-, and PF6-). In all cases,
modest to poor yields and low levels of site selectivity were observed
(entries 4-7 in Table 6). In striking contrast, Ag2CO3 provided an
excellent chemical yield (92%) with a complete reversal of selectivity
(1:6 ratio of A/B; entry 8 in Table 6).18 In addition, for the first time,
traces (∼5%) of isomer C (derived from functionalization at the
most hindered site on DMB) were observed.
Instead, we propose that C-H activation is not fully reversible
in the presence of 1 equiv of BQ. As a result, a Curtin-Hammett
situation (involving fast pre-equilibration between PdA and PdB
followed by rate- and selectivity-determining BQ complexation) is
not fully established. The different relative rates of C-H activation/
B
protonation (k1A/k-1A versus k1 /k-1B) for each carboxylate com-
Alkali metal carbonate salts provided similar results to Ag2CO3
(Table 7), and both reactivity and selectivity generally improved
with increasing size of the cation. For example, 1 equiv of
Na2CO3 afforded 51% yield of product as a 1:3 ratio of isomers
plex thus contribute to the observed selectivities.
Consistent with this proposal, the A/B selectivity became
independent of R upon the addition of 3 equiv of RCO2H to these
reactions. The added acid should increase the relative rate of
4459
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464

Journal of the American Chemical Society
ARTICLE
Table 6. Site Selectivity as a Function of AgXa
entry
AgX
none
A/B
yield (%)b
1
2
3
4
5
6
7
8
nrc
98
79
13
38
nrc
nrc
92
AgOAc
5:1
9:1
1:1
2:1
nrc
nrc
1:6d
AgOPiv
AgO2CCF3
AgNO3
Ag2SO4
AgBF4
Ag2CO3
a A total of 1 mol of Ag was used per mole of Pd (e.g., 1 mol of AgOAc/
0.5 mol of compound 5). b Yields were determined by GC analysis of the
crude reaction mixtures versus an internal standard and represent an
average of 2 runs. c nr = no reaction observed. d Trace amounts (∼5%) of
C were also observed.
Figure 3. Site selectivity (B/A) as a function of the equivalent of BQ for
the coupling of compound 5/Cs2CO3 with DMB.
selectivity as with DMB. With X = AcO-, A/B was 6:1, and this
could be increased to 15:1 by adding 3 equiv of AcOH (entry 1 in
Table 8). Moving from BQ to more electron-rich and sterically
hindered quinones, such as 2,3,5,6-tetramethylbenzoquinone,
also led to significant improvements in selectivity (A/B increased
to 13:1); however, this was accompanied by a decrease in
chemical yield (49%; Table S2 in the Supporting Information).
Site selectivity could be reversed through the use of [(bzq)-
PdCl]2/Cs2CO3, which provided the C-C coupled product as a
1:6 ratio of A/B in excellent (85%) yield (entry 2).
Table 7. Selectivity, Yield, and Function of M2CO3
1,2,3-Trisubstituted arenes also reacted with compound 1 in the
presence of 1 equiv of BQ to afford high selectivity for isomer A (A/
B was typically >50:1) (entries 3, 5, 7, and 9 in Table 8). With
complex 5/Cs2CO3, the site selectivity changed dramatically in all
cases; however, the magnitude of this change was subject to the
electronic character of the 2-substituent. For example, with a 2-nitro
group, A/B = 1:5 (entry 4), while with a methoxy substituent at the
2 position, A/B = 1:1 (entry 8). Finally, 1,2-dimethoxybenzene
(11) showed analogous trends in selectivity. This arene reacted with
compound 1 to give high selectivity for isomer A(A/B = 41:1; entry
11 in Table 8). In contrast, with compound 5/Cs2CO3, the A/B
ratio was 1:1 (entry 12 in Table 8).
entry
M2CO3
A/Ba
yield (%)b
1
2
3
4
Na2CO3
K2CO3
1:3
51
25
1:2
Rb2CO3
Cs2CO3
1:10
1:11
99
100
a Less than 5% of isomer C was observed in these reactions. b Yields were
determined by GC analysis of the crude reaction mixtures versus an
internal standard and represent an average of 2 runs.
A/B, while 1 equiv of Cs2CO3 provided quantitative yield of a
1:11 mixture of A/B (entries 1 and 4 in Table 7, respectively).
We also evaluated A/B selectivity in the carbonate system as a
function of the quinone concentration (Figure 3). Intriguingly,
the A/B ratio changed only a small amount from 0.1 to 20 equiv
of BQ (A/B moved from 1:7 to 1:11). This is in marked con-
trast to the much larger effect observed in the acetate system
(where A/B decreased from 11:1 to 1.1:1 over the same
concentration range).
In summary, metal carbonate salts are uniquely effective in
reversing the selectivity of oxidative coupling between com-
pound 5 and DMB. The generality of this effect with respect to
other aromatic substrates as well as preliminary investigations of
its mechanistic origin are discussed below.
Similar trends were observed in complexes with other cyclo-
metalated ligands (Table 9). For example, [(phpy)PdX]2 (phpy =
2-phenylpyridine) reacted with DMB to afford an 8:1 ratio of A/B
when X = AcO- in the presence of AcOH and a 1:6 ratio of A/B
2-
when X = CO3 (entries 3 and 4 in Table 9). Similarly, the
8-methylquinoline (mq) complex [(mq)PdX]2 provided a 5:1 ratio
of A/B when X = AcO- in the presence of AcOH and a 1:4 ratio of
A/B when X = CO32- (entries 5 and 6 in Table 9). The results in
Tables 8 and 9 clearly demonstrate that the selectivity trends derived
for DMB are applicable across a range of different aromatic
substrates and cyclometalating ligands.
Studies To Understand Selectivity in Carbonate Systems.
Finally, we sought mechanistic insight into the reversal of site
selectivity between the carboxylate and carbonate systems. In the
carboxylate case, steric effects appear to dominate selectivity. Parti-
cularly, in regime 1, the least sterically hindered site (C-HA) was
functionalized with >5:1 (and often much higher) selectivity in all
Application to Diverse Substrates. The reaction of 1,3-
diisopropoxybenzene with [(bzq)PdX] in the presence of 1
equiv of BQ showed nearly identical trends in reactivity and
4460
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464

Journal of the American Chemical Society
ARTICLE
Table 8. Arene Scope with Bzq
Table 9. Scope of Cyclometalated Ligands for Oxidative
Coupling with DMBa
a Conditions: 1 equiv of [Pd], 1 equiv of BQ, 4 equiv of DMSO, 1 or 0
2-
equiv of Cs2CO3, 0.8 mL of DMB, 150 °C, and 15 h. b X = CO3
generated from [(bzq)PdCl]2 (0.5 equiv) and Cs2CO3 (1 equiv).
c Yields were determined by GC analysis of the crude reaction mixtures
versus an internal standard and represent an average of 2 runs. d Traces
of isomer C (<5%) were observed in these reactions.
0
0
Table 10. Calculated Values of EA - EB for Different Arene
Substrates
a DFT calculations performed using Spartan ’08 using B3LYP 6-31G**
basis set.
substrates examined. The dominance of steric effects is fully
consistent with the mechanism outlined in Scheme 2.
In marked contrast, the combination of complex 5/Cs2CO3
provided modest to high selectivity for functionalization of more
sterically hindered C-HB in DMB to afford B. Because C-HB is
more nucleophilic than C-HA (by virtue of being ortho and para
to MeO substituents), we first hypothesized that the observed
selectivity might arise from an electrophilic aromatic substitution
(SEAr) type mechanism for C-H cleavage. To test this possibi-
lity, we conducted density functional theory (DFT) calculations
to assess the relative energies of SEAr intermediates A0 and B0
a X = CO3 generated from [(bzq)PdCl]2 (0.5 equiv) and Cs2CO3
2-
(1 equiv). b Yields were determined by GC analysis of the crude reaction
mixtures versus an internal standard and represent an average of 2 runs.
4461
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464

Journal of the American Chemical Society
ARTICLE
These studies provide evidence against two limiting possibi-
lities for selectivity in the compound 5/Cs2CO3 reactions: (i)
selectivity-determining palladation via an SEAr pathway and (ii)
selectivity-determining thermodynamically controlled deproto-
nation. The origin of selectivity in this system remains the subject
of ongoing investigations.24 We are currently considering at least
two alternative possibilities: (1) the selectivity-determining step
changes as a function of the arene substrate (for example,
involving C-H cleavage via electrophilic palladation with certain
arenes and concerted metalation-deprotonation with others),
and/or (2) this transformation involves a mechanistically unique
selectivity-determining step that remains to be elucidated.
Figure 4. Plot of B/A ratio (experimental from reactions of compound
5/Cs2CO3) versus EA - EB (DFT).
’ CONCLUSION
0
0
By examining the factors controlling site selectivity under the
two regimes outlined above, we have gained insight into the
mechanism of Pd-catalyzed aryl-aryl0 oxidative cross-coupling.
We have shown that the steric and electronic environment at
[Pd] can be tuned to control the site selectivity of arene C-H
functionalization. In the process, we have also discovered new
ways of altering the preferred regioisomer formed in these
transformations. In general, isomer A is formed selectively with
carboxylate X-type ligands in the presence of e1 equiv of
quinone, with added RCO2H and with alkyl-substituted qui-
nones. Switching to carbonate as the X-type ligand at [Pd]
reverses the selectivity to favor isomer B. The ability to achieve
this type of catalyst-based control over site selectivity is a
significant advance, and we anticipate that the observed effects
may be more broadly applicable to other C-H arylation reac-
tions. Efforts to investigate the origin of the reversal in site
2-
selectivity with X = CO3 as well as to apply the principles
derived herein to diverse catalytic transformations are ongoing
Figure 5. B/A ratio (experimental from reactions of compound
5/Cs2CO3) versus pKa(HA) - pKA(HB).
and will be reported in due course.25
(Table 10) for a variety of 1,2- and 1,2,3-substituted aromatics.
’ ASSOCIATED CONTENT
0
The difference in energy between these intermediates (EA
EB ) was then used to approximate the energetic preference for
-
0
S
Supporting Information. Experimental details and spec-
b
SEAr at C-HB versus at C-HA. As shown in Figure 4, the
troscopic and analytical data for new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
0
0
calculated EA - EB values showed a poor correlation with the
experimental B/A selectivities. Even in 1,3-dimethyl-substituted
arenes 7-10 (where the C-HB bonds are sterically identical),
the experimental isomer ratios did not track well with those pre-
dicted for SEAr.19 As such, we conclude that B/A selectivity in the
[(bzq)PdCl]/Cs2CO3 system is unlikely to be due to selectivity-
determining electrophilic palladation.
’ AUTHOR INFORMATION
Corresponding Author
mssanfor@umich.edu
Elegant studies by Fagnou/Gorelsky and coworkers20 as well
as Echavarren and coworkers21 have shown that C-H acidity can
play a large role in both reactivity and selectivity in Pd-catalyzed
C-H functionalization reactions. We hypothesized that, if an
analogous effect were responsible for selectivity with compound
5/Cs2CO3, then there should be a correlation between ΔpKa
[pKa(HA) - pKA(HB)] and experimental B/A ratios. DFT
calculations were used to calculate the deprotonation energies
of HA and HB in a series of 1,3- and 1,2,3-substiuted aromatics to
approximate the relative pKa of HA and HB, and the data are
reported in Table S7 in the Supporting Information.22 A plot of
[pKa(HA) - pKA(HB)] versus experimental B/A ratios showed
no clear relationship between these values (Figure 5). Even in the
sterically similar arenes 7-10, B/A ratios were poorly correlated
with pKa(HA) - pKA(HB). These data suggest that a thermodyna-
mically controlled deprotonation is not the selectivity-determining
step of this transformation.23
’ ACKNOWLEDGMENT
This material is based on work supported by the U.S. Depart-
ment of Energy (Office of Basic Energy Sciences, DE-FG02-
08ER 15997). K.L.H. gratefully acknowledges the ACS Division
of Organic Chemistry and Novartis for graduate fellowships. We
thank Prof. Adam Matzger, as well as J. Brannon Gary and
Matthew S. Remy for assistance with DFT calculations. We also
thank Dr. Bala Ramanathan for performing preliminary Ham-
mett studies.
’ REFERENCES
(1) (a) Torsell, K. B. G. Natural Product Chemistry; Wiley:
Chichester, U.K., 1983. (b) Thomson, R. H. The Chemistry of Natural
Products; Blackie and Son: Glasgow, U.K., 1985. (c) Baudoin, O.;
Guꢀeritte, F. Stud. Nat. Prod. Chem. 2003, 29, 335–417.(d) Herz, W.;
Kirby, G. W.; Falk, H.; Moore, R. E. Progress in the Chemistry of Organic
Natural Products; Springer-Verlag: Vienna, Austria, 2001; Vol 82.
4462
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464

Journal of the American Chemical Society
ARTICLE
(e) Kozlowski, M. C.; Morgan, B. J.; Linton, E. C. Chem. Soc. Rev. 2009,
38, 3193 and references therein. (f) Rao, A. V. R.; Gurjar, M. K.; Reddy,
K. L.; Rao, A. S. Chem. Rev. 1995, 95, 2135. (g) Nicolaou, K. C.; Boddy,
C. N. C.; Br€ase, S.; Winssinger, N. Angew. Chem., Int. Ed. 1999, 38, 2096.
(h) Cepanec, I. Synthesis of Biaryls; Elsevier, Inc.: San Diego, CA, 2004.
(i) Roncali, J. Chem. Rev. 1992, 92, 711. (j) Mei, X.; Wolf, C. J. Am. Chem.
Soc. 2006, 128, 13326.
(2) (a) Stille, J. K. Angew. Chem., Int. Ed. 1986, 25, 508. (b) Farina,
V.; Krishnamurthy, V.; Scott, W. J. Org. React. 1997, 50, 1. (c) Miyaura,
N.; Suzuki, A. Chem. Rev. 1995, 95, 2457. (d) Suzuki, A. J. Organomet.
Chem. 1999, 576, 147. (e) Beletskaya, I. P.; Cheprakov, A. Chem. Rev.
2000, 100, 3009. (f) Kotha, S.; Lahiri, K.; Kashinath, D. Tetrahedron
2002, 58, 9633. (g) Denmark, S. E.; Sweis, R. F. Acc. Chem. Res. 2002, 35,
835. (h) Muci, A. R.; Buchwald, S. L. Top. Curr. Chem. 2002, 219, 131.
(i) Hassan, J.; Sꢀevignon, M.; Gozzi, C.; Schulz, E.; Lemaire, M. Chem.
Rev. 2002, 102, 1359.(j) de Meijere, A.; Diederich, F. Metal-Catalyzed
Cross-Coupling Reactions; Wiley-VCH; Weinheim, Germany, 2004. (k)
Martin, R.; Buchwald, S. L. Acc. Chem. Res. 2008, 41, 1461.
(3) For reviews that discuss C-H arylation reactions, see: (a)
Campeau, L. C.; Fagnou, K. Chem. Commun. 2006, 1253. (b) Daugulis,
O.; Zaitsev, V. G.; Shabashov, D.; Pham, Q. N.; Lazareva, A. Synlett 2006,
3382. (c) Yu, J. Q.; Giri, R.; Chen, X. Org. Biomol. Chem. 2006, 4, 4041.
(d) Godula, K.; Sames, D. Science 2006, 312, 67. (e) Alberico, D.; Scott,
M. E.; Lautens, M. Chem. Rev. 2007, 107, 174. (f) Beccalli, E. M.;
Broggini, G.; Martinelli, M.; Sottocornola, S. Chem. Rev. 2007, 107,
5318. (g) Ackermann, L. Synlett 2007, 507. (h) Fairlamb, I. J. S. Chem.
Soc. Rev. 2007, 36, 1036. (i) Campeau, L. C.; Fagnou, K. Chem. Soc. Rev.
2007, 36, 1058. (j) Catellani, M.; Motti, E.; Della Ca’, N. Acc. Chem. Res.
2008, 41, 1512. (k) Kakiuchi, F.; Kochi, T. Synthesis 2008, 3013. (l) Li,
B. J.; Yang, S. D.; Shi, Z. J. Synlett 2008, 949. (m) McGlacken, G. P.;
Bateman, L. M. Chem. Soc. Rev. 2009, 38, 2447. (n) Chen, X.; Engle,
K. M.; Wang, D. H.; Yu, J. Q. Angew. Chem., Int. Ed. 2009, 48, 5094. (o)
Lyons, T. W.; Sanford, M. S. Chem. Rev. 2010, 110, 1147.
(13) The presence/absence of DMSO did not influence the selec-
tivity in this system. It was used as an additive in our previously reported
catalytic L∼C-H/aryl-H cross-coupling reactions and was maintained
throughout the current investigations for consistency.
(14) The addition of 3 equiv of AcOH to regime 1 in this series of
sterically differentiated quinones altered the magnitude of selectivity;
however, the overall trend remained unchanged. See page S18 of the
Supporting Information for details.
(15) For other examples of BQ-promoted reductive elimination, see:
(a) Temple, J. S.; Riediker, M.; Schwartz, J. J. Am. Chem. Soc. 1982, 104, 1310.
(b) Backvall, J. E.; Bystrom, S. E.; Nordberg, R. E. J. Org. Chem. 1984, 49,
4619. (c) B€ackvall, J. E.; Nordberg, R. E.; Wilhelm, D. J. Am. Chem. Soc. 1985,
107, 6892. (d) Szabꢀo, K. Organometallics 1998, 17, 1677. (e) Albꢀeniz, A. C.;
Espinet, P.; Martn-Ruiz, B. Chem.—Eur. J. 2001, 7, 2481. (f) Chen, M. S.;
Prabagaran, N.; Labenz, N. A.; White, M. C. J. Am. Chem. Soc. 2005, 127,
6970. (g) Chen, X.; Li, J. J.; Hao, X. S.; Goodhue, C. E.; Yu, J. Q. J. Am. Chem.
Soc. 2006, 128, 78. (h) Pꢀerez-Rodríguez, M.; Braga, A. A. C.; Garcia-Mechor,
M.; Pꢁerez-Temprano, M. H.; Casares, J. A.; Ujaque, G.; de Lera, A. R.; lvarez,
R.; Maseras, F.; Espinet, P. J. Am. Chem. Soc. 2009, 131, 3650. (i) Lanci, M. P.;
Remy, M. S.; Kaminsky, W.; Mayer, J. M.; Sanford, M. S. J. Am. Chem. Soc.
2009, 131, 15618. (j) Yin, G.; Wu, Y.; Liu, G. J. Am. Chem. Soc. 2010, 132,
11978. (k) Ishikawa, A.; Nakao, Y.; Sato, H.; Sakaki, S. Dalton Trans. 2010, 39,
3279.
(16) Experiments performed at lower than 1 equiv of quinone provided
further support for this hypothesis. For example, the use of 0.1 equiv of 2,5-
diarylquinone with X = CF3 afforded a 7:1 ratio of A/B, consistent with a
reduced contribution from k2A/B[PdA/B][quinone] to selectivity.
(17) Liꢀegault, B.; Petrov, I.; Gorelsky, S. I.; Fagnou, K. J. Org. Chem.
2010, 75, 1047 and references therein.
(18) This reversal is consistent with the A/B selectivity of 3:1 in the
catalytic reactions, where both AcO- [from the Pd(OAc)2 catalyst] and
CO32- (from the Ag2CO3 oxidant) are present. This is an intermediate
to the A/B ratio, with X = AcO- (5:1) and CO32- (1:6).
(19) The lack of correlation is even more pronounced if transition-
state theory (TST) is used to estimate the expected isomer distribution.
Using the difference in intermediate energies (EA - EB ) as a crude
approximation for the relative transition-state energies, the expected B/
A selectivities were calculated and are shown in Table S6 in the
Supporting Information.
(20) (a) Lafrance, M.; Rowley, C. N.; Woo, T. K.; Fagnou, K. J. Am.
Chem. Soc. 2006, 128, 8754. (b) Gorelsky, S. I.; Lapointe, D.; Fagnou, K.
J. Am. Chem. Soc. 2008, 130, 10848.
(21) Garcia-Cuadrado, D.; Braga, A. A. C.; Maseras, F.; Echaverren,
A. M. J. Am. Chem. Soc. 2006, 128, 1066.
(22) Literature studies have shown that gas-phase calculations with
diffuse basis sets are appropriate for calculating the deprotonation
energies of simple molecules. For example, see: (a) Hehre, W. J.; Radom,
L.; Schleyer, P. R.; Pople, J. A. Ab Initio Molecular Orbital Theory; John
Wiley and Sons: New York, 1986. (b) Maksic, Z. B.; Kovacek, D.; Eckert-
Maksic, M.; Zrinski, I. J. Org. Chem. 1996, 61, 6717.
(4) Ashenhurst, J. A. Chem. Soc. Rev. 2010, 39, 540.
(5) (a) Li, R.; Jiang, L.; Lu, W. Organometallics 2006, 25, 5973. (b)
Rong, Y.; Li, R.; Lu, W. Organometallics 2007, 26, 4376. (c) You, S.; Xia,
J. Top. Curr. Chem. 2010, 292, 165–195.
0
0
(6) (a) Stuart, D. R.; Fagnou, K. Science 2007, 316, 1172. (b) Stuart,
D. R.; Villemure, E.; Fagnou, K. J. Am. Chem. Soc. 2007, 129, 12072. (c)
Dwight, T. A.; Rue, N. R.; Charyk, D.; Josselyn, R.; DeBoef, B. Org. Lett. 2007,
9, 3137. (d) Potavathri, S.; Dumas, A. S.; Dwight, T. A.; Naumiec, G. R.;
Hammann, J. M.; DeBoef, B. Tetrahedron Lett. 2008, 49, 4050. (e) Xi, P.;
Yang, F.; Qin, S.; Zhao, D.; Lan, J.; Gao, G.; Hu, C.; You, J. J. Am. Chem. Soc.
2010, 132, 1822. (f) Potavathri, S.; Pereira, K. C.; Gorelsky, S. I.; Pike, A.;
LeBris, A. P.; DeBoef, B. J. Am. Chem. Soc. 2010, 132, 14676.
(7) For examples of Pd-catalyzed oxidative Ar-H homocoupling,
see: (a) Hull, K. L.; Lanni, E. L.; Sanford, M. S. J. Am. Chem. Soc. 2006,
128, 14047. (b) Mori, A.; Sugie, A. Bull. Chem. Soc. Jpn. 2008, 81, 548.(c)
Ref 5b. (e) Liang, Z.; Zhao, J.; Zang, Y. J. Org. Chem. 2010, 75, 170.
(8) (a) Xia, J. B.; You, S. L. Organometallics 2007, 26, 4869. (b) Hull,
K. L.; Sanford, M. S. J. Am. Chem. Soc. 2007, 129, 11904. (c) Tian, S. L.;
Fang, Z.; Shi, Z. J. Angew. Chem., Int. Ed. 2008, 47, 1115. (d) Brasche, G.;
García-Fortanet, J.; Buchwald, S. L. Org. Lett. 2008, 47, 2207. (e) Zhao,
X.; Yeung, C. S.; Dong, V. M. J. Am. Chem. Soc. 2010, 132, 5837. (f)
Yeung, C. S.; Zhao, X.; Borduas, N.; Dong, V. M. Chem. Sci. 2010, 1, 331.
(g) Scheuermann, C. J. Chem.—Asian J. 2010, 5, 436.
(9) The exact mechanism of BQ complexation and BQ-promoted
reductive elimination remains to be elucidated. For example, the initial
BQ binding could occur via a 5-coordinate intermediate, such as
compound 3 in Scheme 1, or via an associative ligand substitution
reaction (where compound 3 is a transition state for the ligand
substitution event).
(10) Hull, K. L.; Sanford, M. S. J. Am. Chem. Soc. 2009, 131, 9651.
(11) Functionalization at HC (to form isomer C) was not detected
under these conditions.
(23) A recent report by Potavathri et al.6f suggested that the relative
reactivity of arene C-H bonds toward Pd-catalyzed C-H functionali-
zation via a concerted metalation/deprotonaton mechanism can be
predicted by their ground-state bond lengths. As shown in Table S9 in
the Supporting Information, the C-HA and C-HB bond lengths for the
1,3- and 1,2,3-substituted arenes in Table 10 show no correlation with
B/A isomer ratios from the compound 5/Cs2CO3 reactions.
(24) Another possibility to account for the reversal in selectivity with
compound 5/Cs2CO3 is that the presence of a carbonate base renders the
C-H activation step of Scheme 2 irreversible. In this scenario, equilibration
between PdA and PdB would be inhibited and the reversal in selectivity would
be due to a kinetic preference for C-H cleavage to form PdB over PdA
(analogous to the situation in limiting regime 1). Several pieces of evidence
suggest that this is not the complete explanation for the observed effect with
complex 5/Cs2CO3. First, Figure 2 shows that the A/B selectivity levels off at
close to 1:1 in regime 1, which is considerably different from that obtained
with complex 5/Cs2CO3. Second, the reaction of acetate complex 1 with
DMB in the presence of 1 equiv of BQ and 1 equiv of Cs2CO3 afforded
(12) The use of 1 and 20 equiv of BQ were examined as representa-
tive points in regimes 1 and 2. For consistency, we have compared all of
the reactions presented herein under these conditions.
4463
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464

Journal of the American Chemical Society
ARTICLE
an A/B ratio of 1:1. This is much lower than the 1:11 ratio obtained with
compound 5/Cs2CO3, suggesting that the carbonate is doing more than
simply serving as a base to push the reaction toward regime 1 (otherwise
comparable selectivities would be expected in these two experiments).
(25) Our initial efforts to apply the findings with [(bzq)PdCl]2/
Cs2CO3 to a catalytic process have thus far failed to find a suitable
oxidant to reoxidize Pd0. Significant efforts are underway to address this
challenge.
4464
dx.doi.org/10.1021/ja1097918 |J. Am. Chem. Soc. 2011, 133, 4455–4464