Role of the base and control of selectivity in the suzuki-miyaura cross-coupling reaction

Article abstract of DOI:10.1002/cctc.201301080

The outcome of the Suzuki-Miyaura cross-coupling for the direct competition reaction between two boronic acids was evaluated under routine synthesis conditions. The reaction selectivity was found to depend on the amount of the base used, with fewer bases

Full text of DOI:10.1002/cctc.201301080

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201301080
Role of the Base and Control of Selectivity in the Suzuki–
Miyaura Cross-Coupling Reaction
[
a, b]
[a]
[b]
[b]
Carlos F. R. A. C. Lima,*[
Luꢀs M. N. B. F. Santos*
Ana S. M. C. Rodrigues, Vera L. M. Silva, Artur M. S. Silva, and
a]
The outcome of the Suzuki–Miyaura cross-coupling for the
direct competition reaction between two boronic acids was
evaluated under routine synthesis conditions. The reaction se-
lectivity was found to depend on the amount of the base
used, with fewer bases favoring the reactivity of the boronic
catalytic cycle. Moreover, the results indicate that under these
specific conditions, the most reactive organoboron species
ꢀ
toward transmetalation is the borate anion RꢀB(OH) instead
3
of the neutral boronic acid RꢀB(OH) . Hence, the main role of
2
the base in the reaction mechanism is to increase the reactivity
of the boronic acid toward the Pd–halide complex by convert-
ing it into the respective organoborate. In addition, boric acid,
an important reaction byproduct, affects the selectivity in the
Suzuki reaction because its gradual formation in the reaction
medium disturbs the acid–base equilibrium.
acid with lower pK (stronger acid). The dependence of the re-
a
action selectivity on base stoichiometry was found to increase
with the increase in the difference in the pK values of the
a
competing boronic acids. These results confirm a relationship
between acid–base chemistry and the Suzuki–Miyaura reaction
Introduction
[
1–3,17]
Since its first appearance in the scientific community in 1979,
the Suzuki–Miyaura cross-coupling reaction has received much
attention, chiefly owing to its great versatility in the formation
smoothly and with satisfactory yields.
Also, the addition
of NaOH or other bases greatly increases the rate of transmeta-
[18–20]
lation between organoboron reagents and metal halides.
[
1–3]
of CꢀC bonds.
The reaction is generally green, relatively
The two main pathways that have been proposed to account
for the role of the base in the Suzuki reaction (the base is con-
easy to perform, and usually gives good yields and selectivities
without much synthetic effort. The use of organoboron com-
pounds is itself an attractive factor because these compounds
are prone to react through cross-coupling, are thermally stable
and inert to water and oxygen, and tolerate various functional
ꢀ
sidered to be HO but other bases are possible) are illustrated
[3,21]
in Scheme 1.
The organoboron species, Rꢀ[B], is typically
a boronic acid, [B]=B(OH) , or a borate ester, [B]=B(OR) . Path
2
2
A consists of the exchange of X (a halogen or triflate) in the
[
4,5]
II
groups. This reaction has become important to the scientific
community with A. Suzuki receiving the joint Nobel Prize in
Chemistry in 2010 “for palladium-catalyzed cross-couplings in
Pd complex formed after oxidative addition with a nucleophilic
species, which can be the basic anion of the base used in the
ꢀ
Suzuki reaction or the HO anion present in the solution, form-
[
6]
organic synthesis”.
ing the corresponding intermediate shown in Scheme 1. The
anion must have a significant nucleophilic power to displace
the X group acting as the leaving group from the metal center.
This intermediate then reacts with Rꢀ[B] through transmetala-
Although the catalytic cycle of the Suzuki–Miyaura reaction
is well established, the transmetalation step is still not com-
pletely understood and is currently a subject of controver-
[
3,7–14]
II
sy,
probably because in most cross-coupling reactions, the
tion and yields the R’ꢀPd ꢀR complex. In path B, the base at-
ꢀ
transmetalation step strongly depends on the reaction condi-
tacks Rꢀ[B], forming the corresponding borate, Rꢀ[B](OH) ,
[
15–17]
tions.
It is in this phase of the catalytic cycle that the base
which subsequently exchanges the R group with the X group
II
plays its main role in the Suzuki coupling. Stoichiometric quan-
tities of the base are necessary to conduct the reaction
in the Pd complex in a typical transmetalation step. In this
path, the quaternization of boron in Rꢀ[B] by the base increas-
II
es its reactivity toward the R’ꢀPd ꢀX complex. Several studies
II
[
a] Dr. C. F. R. A. C. Lima, A. S. M. C. Rodrigues, Prof. L. M. N. B. F. Santos
CIQ, Departamento de Quꢀmica e Bioquꢀmica
Faculdade de CiÞncias da Universidade do Porto
P-4169-007 Porto (Portugal)
support path A, that is, the facile formation of Pd ꢀOR species
[7,8,11,14,22–24]
and their reactivity toward transmetalation.
For ex-
ample, a halogen ligand can be displaced readily from the or-
ganopalladium halide by alkoxy, hydroxyl, or acetoxy anions,
E-mail: carlos.chemistry@gmail.com
II
[22]
lbsantos@fc.up.pt
which leads to the formation of reactive Pd ꢀOR complexes.
[
b] Dr. C. F. R. A. C. Lima, Dr. V. L. M. Silva, Prof. A. M. S. Silva
Department of Chemistry & QOPNA
University of Aveiro
In addition, the cross-coupling with alkenylborates occurs even
II
in the absence of bases if Pd ꢀOR intermediates are generated
[23,24]
through oxidative addition.
In contrast, some studies sup-
P-3810-193 Aveiro (Portugal)
[
10,12,25–32]
ꢀ
port path B.
Quaternary boron species, R ꢀ[B] , have
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201301080.
4
been associated with faster ligand exchange in transmetalation
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1291

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
in the laboratory and industry
are reproduced as closely as
possible.
Results and Discussion
This study was conducted by
using two main synthetic meth-
odologies. The first one was the
Scheme 1. The two main pathways proposed for explaining the role of the base in the transmetalation step of
the Suzuki–Miyaura reaction.
direct competition reaction be-
[
3,7–14,21]
tween bromobenzene and two
substituted phenylboronic acids
[
25,26]
reactions.
Other studies show that the cross-coupling reac-
in an aqueous organic solvent, H O/DMF (1:1) or H O/acetone
2
2
tion between arylboronic acids and aryl halides at pH 7–8.5 is
(1:1), with Pd(OAc) or Pd(PPh ) as the catalyst. The reactions
2
3 4
[
27]
retarded relative to the reaction at pH 9.5–11. The pK value
were performed at different temperatures and reaction times
by varying the amount of the base used, either a weak (K CO )
a
of phenylboronic acid is 8.8, which suggests that higher con-
2
3
centrations of RꢀB(OH)3ꢀ accelerate the transmetalation step.
or a strong (KOH) base. In the second methodology, one of the
II
Studies by using ESIMS have detected the presence of R’ꢀPd ꢀ boronic acids or both were replaced by the respective potassi-
X and borate intermediates in reaction mixtures; however, the
um borates and the reactions were performed as discussed
earlier. After the completion of the reaction, the final mixture
was extracted with CH Cl , washed with water and aqueous
[
30–32]
PdꢀOH complex could not be detected by using ESIMS.
II
Hence, it is not yet clear how the conversion of R’ꢀPd ꢀX to
2
2
II
R’ꢀPd ꢀR actually occurs. This conversion can be accomplished
KOH, and analyzed by using GC to determine the molar ratio
of the products formed (the GC signal was calibrated with the
standard solutions of the known pure samples). There are usu-
ally three main reaction pathways in Suzuki–Miyaura cross-cou-
pling: the main Suzuki reaction (S) and two important side re-
through path A, path B, or a combination of both in varying
extents, as well as depending on the conditions used in a par-
ꢀ
ticular synthesis. The alternative paths in which Rꢀ[B](OH) at-
II
tacks PdꢀOH or Rꢀ[B] attacks R’ꢀPd ꢀX can, in principle, be
discarded because these species have already been shown to
actions—homocoupling (HC) and hydrolytic deboronation (DB)
[
8,10,14]
[3–5,9]
be virtually unreactive toward each other.
of the two competing boronic acids (Scheme 2).
All the
Herein, the outcome of the Suzuki–Miyaura reaction, per-
formed under routine laboratory conditions, was evaluated as
a function of base stoichiometry. The reactions were conduct-
possible products of these three reactions were observed by
using GC and proven to be the main species present in the
final reaction mixture. Unreacted bromobenzene was seldom
ed in a competition fashion between two boronic acids, Rꢀ observed in significant amounts (generally <1 mol%). Unreact-
ꢀ
B(OH) , and/or respective borates, RꢀB(OH) (the respective
ed boronic acids were never observed. These observations and
the fact that the isolated reaction yield of the two Suzuki prod-
ucts was higher than 80% in most cases indicate that these re-
actions occurred until completion (complete conversion of bro-
2
3
conjugate bases), with different pK values. On adjusting the
a
amount of base used in the synthesis, the equilibrium distribu-
tion of boronic acids and borates in the solution changes ac-
cording to the pK values of the
a
two competing reactants, which
affects the ratio of species that
subsequently react by transme-
talation to yield the final cou-
pling products. The evaluation
of the reaction selectivity as
a function of base stoichiometry
then enables us to determine
which pair of species was more
actively competing for transme-
talation: the two boronic acids,
RꢀB(OH) , or the two respective
2
ꢀ
borates, RꢀB(OH) , present in
3
the initial reaction mixture. In
this experimental approach, the
reaction medium remains unper-
turbed in the course of the reac-
tion, which thus ensures that the
mechanistic study is performed
Scheme 2. The Suzuki–Miyaura competition reaction between two boronic acids considering the conditions used
if the conditions routinely used herein and the three main reaction pathways.
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1292

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
ꢀ
mobenzene and both boronic acids). In all cases, the desired
Suzuki reaction occurred to the greatest extent, followed by
HC and usually less extensive DB, from 0 to 19 mol% [not
much considering the stoichiometry of bromobenzene and
boronic acids (1:2)]. Throughout this study, the stoichiometry
of all reagents is considered relative to bromobenzene.
sidering that in 2-bPhB(OH)3 , the ortho-phenyl ring partially
ꢀ
shields the ꢀB(OH)3 group from the aqueous solvent, which
thus prevents the efficient stabilization of the negative charge
ꢀ
ꢀ
through solvation. In both 3-bPhB(OH)3 and 4-bPhB(OH)3
,
the borate group is completely exposed to the solvent and
conjugation effects are not strong enough to justify significant
[
34,35]
difference in acid strength.
The reported pK values also in-
a
dicate that ortho-substituted phenylboronic acids are generally
Acid–base equilibrium
[5,33]
less acidic.
The relative acidity of all boronic acids used
The boronic acids used herein are presented in Figure 1 along
herein was evaluated computationally in the gas phase at the
B3LYP/6-311+ +G(d,p) and MP2/cc-pVDZ levels of theory. This
evaluation was accomplished through the use of a homodes-
with the order of their acid strength. The pK values of 4-NO -,
a
2
4
-CHO-, 4-CH -, and 4-OCH -phenylboronic acids in water were
3
3
[
5,33]
ꢀ
taken from the literature.
The relative acid strength of the
motic reaction, in which HO is exchanged between phenylbo-
three biphenylboronic acids—2-bPhB(OH) , 3-bPhB(OH) , and
rate and another arylboronic acid. Because the contributions
from entropy and thermal enthalpy corrections can be greatly
reduced by the use of such homodesmotic reaction schemes,
the difference in electronic energies between products and re-
2
2
actants at 0 K (DE (0 K)) is a good evaluation of relative acid
el
strength. These results are summarized in Table 1.
ꢀ
Table 1. Gas-phase HO exchange reaction used to compute the relative
acidities of the boronic acids used herein.
ꢀ
1
R
DEel(0 K) [kJmol
]
[
a]
B3LYP/6-311+ +G(d,p) MP2/cc-pVDZ SCS-MP2/cc-pVDZ
Figure 1. The boronic acids used herein in order of acid strength by pK
a
2
3
4
4
4
4
4
-Ph
-Ph
-Ph
-OCH
-CHO
ꢀ13.5
ꢀ11.2
ꢀ10.5
7.2
ꢀ39.6
4.5
ꢀ25.7
ꢀ14.3
ꢀ11.6
3.2
ꢀ31.8
0.2
ꢀ23.7
ꢀ13.8
ꢀ11.2
3.2
ꢀ31.6
0.2
value. The relative acid strength of the biphenyl isomers was evaluated
1
herein by using H NMR spectroscopy, and the pK
a
values shown refer to
[
5,33]
pure water and were taken from the literature.
3
-CH
-NO
3
2
ꢀ57.4
ꢀ44.4
ꢀ45.1
4
-bPhB(OH) —was evaluated in D O/[D ]acetone (1:1) and
2
2
6
[36]
1
[a] Spin-component-scaled second-order Møller–Plesset approach.
D O/[D ]DMSO (1:2) by using H NMR spectroscopy; the spec-
2
6
ꢀ
,
3
tra of equimolar mixtures of 2-bPhB(OH) /4-bPhB(OH)
2
ꢀ
ꢀ
3
-bPhB(OH) /4-bPhB(OH)3
,
4-bPhB(OH) /2-bPhB(OH)3
,
and
2
2
ꢀ
4
-bPhB(OH) /3-bPhB(OH)3 were recorded at 258C (the potassi-
The negative values of DE (0 K) indicate that the given aryl-
2
el
um salts were used as the biphenylborate derivatives). In both
boronic acid is a stronger acid than phenylboronic acid; the
ꢀ
solvent mixtures, the spectra of 2-bPhB(OH) /4-bPhB(OH) and
more negative the DE (0 K) value, the stronger the acid. Ac-
2
3
el
ꢀ
4
-bPhB(OH) /2-bPhB(OH)3 are virtually superimposable, which
cording to both computational and experimental results, the
2
indicates that the boronic acids–borates equilibria are estab-
lished quite fast and according to what should be expected
from a regular acid–base equilibrium. In these cases, the spec-
boronic acids bearing electron-withdrawing groups (EWGs; R=
4-CHO and 4-NO ) are by far the strongest acids and the ones
2
substituted with electron-donating groups (EDGs; R=OCH3
tra are consistent with those of a nearly equimolar mixture of
and CH ) the weakest. For the biphenyl isomers, both compu-
3
ꢀ
2
-bPhB(OH) and 4-bPhB(OH) and no significant characteristic
tational and experimental NMR results indicate that 3-
2
3
ꢀ
signals of 2-bPhB(OH)3 and 4-bPhB(OH) are observed, which
bPhB(OH) and 4-bPhB(OH) have similar pK values. The only
2
2
2
a
thus indicates that 4-bPhB(OH)2 is a stronger acid than
disagreement is the case of 2-bPhB(OH) , which, according to
2
2
-bPhB(OH)2.
The spectra of 3-bPhB(OH) /4-bPhB(OH) and 4-bPhB(OH)2/
the computational results but contrary to the experimental
ones, should be a stronger acid relative to its two isomers. As
mentioned earlier, this discrepancy is due to the less efficient
ꢀ
3
2
ꢀ
3
-bPhB(OH)3 are also identical. In this case, the spectra show
that all four species are present in significant concentrations,
solvation of the borate head in 2-bPhB(OH) . The extrapolation
2
which indicates that the pKa values of 3-bPhB(OH)2 and 4-
of the acid strength from the gas phase and pure aqueous so-
lution to the aqueous organic solvents used in Suzuki reactions
must be made with caution. However, the order of the acid
bPhB(OH) are similar. For detailed NMR analysis, see Section S5
2
of the Supporting Information. These findings are logical con-
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1293

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
strength of boronic acids in aqueous organic mixtures used
herein shall follow the same trend as that of pure water be-
cause of the following: 1) DMF is a solvent with bulk properties
similar to those of water (high dielectric constant and high
[
37]
ability to stabilize ionic species);
2) for biphenylboronic
acids, the NMR results indicate that the acid strength follows
the same trend in D O/[D ]acetone (1:1) and D O/[D ]DMSO
2
6
2
6
(
1:2); and 3) except 2-bPhB(OH) , the computational gas-phase
2
results are in good qualitative agreement with the experimen-
tal evaluation of relative acid strength, which indicates that the
pK trend is comparable even in vacuum. Thus, the combina-
a
tion of the experimental and computational results gives
a semiquantitative scaling of the relative acidities of boronic
acids in aqueous organic mixtures.
Boronic acids behave as Lewis acids, and their equilibrium
[
5]
pK values are calculated based on Equation(1):
a
RꢀBðOHÞ þ H O Ð RꢀBðOHÞ þ Hþ
ꢀ
ð1Þ
2
2
3
Hence, it follows from simple acid–base relationships that
a lower pK value directly implies a greater tendency to cap-
a
ꢀ
ture HO . Because K CO is a weak polyprotic base, the analy-
2
3
sis of the acid–base equilibrium is not as straightforward as in
the case of KOH. To better visualize the effect of base stoichi-
ometry, the equilibrium concentrations of boronic acids and
borates in water were calculated through a simultaneous equi-
librium approach by using the least-squares method. These
concentrations lead to the initial distribution of species in the
reaction medium before transmetalation occurs. This step was
performed for two representative cases: one in which the two
boronic acids have a high pK difference, DpK =1.7 (mimicking
a
a
Figure 2. Calculated fractional initial equilibrium concentration of the stron-
the competition between R /R =4-OCH /4-CHO and 4-CH /4-
1
2
3
3
ger acid’s borate in water as a function of base equivalents for a) DpK
a
=1.7
NO ), and another in which that difference is smaller, DpK =
2
a
and b) DpK =0.6.
a
0
4
.6 (as in competitions between R /R =2-Ph/4-Ph and 4-CHO/
1 2
-NO ). The calculated fractional initial equilibrium concentra-
2
tion (f) of the stronger acid’s borate (borate 1) in water [see
Eq. (2)] as a function of base stoichiometry (considering K =
peting boronic acids, and the catalyst were dissolved in
a given volume of DMF at 308C. After ensuring complete dis-
solution of all reactants, the same volume of a K CO solution
w
ꢀ
14
2ꢀ
ꢀ4
ꢀ
ꢀ8
10
,
Kb(CO3 )=2.1ꢂ10 , and K (HCO )=2.3ꢂ10 ) is
b 3
[
38]
shown in Figure 2.
2
3
in water was added and the mixture was heated at 1008C for
h. To ensure that the addition of the aqueous base does not
½
borate 1ꢁ
3
f ¼
ð2Þ
½
borate 1ꢁ þ ½borate 2ꢁ
cause precipitation of any reactant, some solubility tests were
made, which confirmed that the aqueous organic solvents
used herein can completely solubilize the mixture of all reac-
tants relatively fast (the Pd catalyst was not included in these
tests; otherwise, its presence would initiate the reaction).
Moreover, the first synthesis reaction presented in Table 2 was
repeated in a slightly different way: instead of adding aqueous
K CO to the mixture in DMF, a small amount of solid Pd(OAc)
The figure shows that f first decreases steeply (and more
pronouncedly for higher difference in pK values) and reaches
a
a plateau after the addition of excess base. For KOH, f stabilizes
at 2 base equivalents [base/boronic acids (1:1) stoichiometry],
whereas for K CO , it reaches a plateau at approximately
2
3
3
base equivalents.
2
3
2
was added to the completely solubilized mixture of all reac-
tants in H O/DMF (1:1) at 308C. The reaction was then heated
at 1008C for 3 h as mentioned earlier. The selectivity results
2
Competition reactions between two boronic acids
The results of the competition reactions between two boronic
acids with 1 equiv. (molar equivalent) of bromobenzene,
were similar: the percent selectivity was 24/76 for R /R in the
1
2
Suzuki reaction and 67/23/10 in the S/HC/DB competition,
which thus confirms that the selectivity results are not affected
by solubility issues.
1
equiv. of each boronic acid, and varying amounts of K CO3
2
are summarized in Table 2. Unless otherwise noted, all reac-
tions were performed as follows: bromobenzene, the two com-
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1294

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
acid is preferentially converted to its respective borate and this
Table 2. Results of the Suzuki–Miyaura competition reactions between
two boronic acids.
[
a]
preference is stronger as the difference in pK values between
a
the two boronic acids increases (Figure 2). In this situation, the
amount of the base is not enough to convert both boronic
acids to their respective borates and the initial equilibrium dis-
tribution consists of larger amounts of the weaker acid in the
ꢀ
RꢀB(OH) form and of the stronger acid in the RꢀB(OH) form.
2
3
R
1
/R
2
K
2
CO
3
Selectivity [%]
It is therefore expected that the Suzuki product distribution in
the final mixture will reflect the ratio of the more reactive form
toward transmetalation. Because fewer base equivalents favor
the reactivity of the stronger boronic acid, the most important
competition for transmetalation occurs between the two
Suzuki reaction
S/HC/DB
competition
[
b]
[c]
[d]
[
equiv.]
(R
1
/R
2
)
2
3
2
4
4
4
-Ph/4-Ph
-Ph/4-Ph
-Ph/4-Ph
1.0
24/76
32/68
39/61
48/52
48/52
48/52
36/64
43/57
45/55
35/65
68/32
81/19
34/66
78/22
83/17
46/54
57/43
58/42
58/42
61/23/16
70/17/13
66/21/13
68/31/1
71/27/2
68/30/2
68/29/3
61/30/9
64/29/7
67/24/9
67/23/10
72/15/13
67/33/0
68/32/0
77/23/0
69/31/0
70/30/0
70/30/0
60/40/0
2
4
.5
.7
borate forms. The reactions for R /R =2-Ph/4-Ph were repeat-
1.0
1
2
2
4
.3
.6
ed in H O/acetone (1:1) to reproduce the solvent used in the
2
1
H NMR experiments for the evaluation of relative acid
[
e]
0.9
strength; the same trend in reaction selectivity was observed.
2
5
.4
.0
Ortho-substituted phenylboronic acids were found to be gen-
[3,27]
-OCH
3
/4-CHO
1.0
erally less reactive than the other isomers.
The results pre-
2
4
.5
.6
sented herein for the R /R =2-Ph/4-Ph competition reactions
1
2
also confirm this tendency, and the para isomer is always more
reactive. Steric hindrance differentiates between the reactivity
of the biphenyl isomers; however, this differentiation is expect-
ed to be independent of base stoichiometry, and therefore the
ortho effect need not be considered explicitly herein.
-CH
3
/4-NO
2
1.0
2
4
.6
.9
-CHO/4-NO
2
1.0
2
3
4
.5
.5
.9
The substituent effects play a dual role in the Suzuki reac-
tion. EWGs in the para position of phenylboronic acid decrease
[
a] All reactions were performed in H
boronic acid, 1 equiv. of bromobenzene, and approximately 5 mol% of
Pd(OAc) at 1008C (addition of an aqueous base at 308C) for 3 h; [b] Rela-
2
O/DMF (1:1) with 1 equiv. of each
the pK value because they can delocalize the negative charge
a
of the conjugate borate base by resonance. However, accord-
ing to some previous studies, EWG substitution is associated
with lower transmetalation rates; one possible explanation for
this is the decreased nucleophilicity of boron species, which
can decrease its reactivity toward the electrophilic Pd
2
tive to bromobenzene; [c] Selectivity toward the two possible products
of the Suzuki–Miyaura reaction; [d] Selectivity for the three main reaction
pathways: S, HC, and DB; [e] Reactions performed in H
58C (addition of an aqueous base at 258C) for 2 h.
2
O/acetone (1:1) at
6
[3,39]
center.
EDGs demonstrate the opposite effects. According
to this, and as was observed herein, the Suzuki selectivity to
EWG-substituted phenylboronic acids is higher for lower base
equivalents, at which the concentration of the respective
borate is significantly higher than that of the borate of the
phenylboronic acid bearing the EDG. With the increase in the
amount of the base, the concentrations of the two borates
gradually become similar, and the selectivity in the Suzuki reac-
tion is expected to reflect more the intrinsic reactivity of each
organoboron species toward transmetalation and less the bor-
onic acids–borates equilibria. The higher transmetalation rate
of organoboron species bearing the EDG then leads to its in-
creased Suzuki selectivity. Eventually, a point is reached at
which the amount of the base ensures that all boronic acids
are virtually converted to the borates. From this point onward,
the reaction selectivity no longer depends on the acid–base
equilibrium and reflects only the difference in transmetalation
rates. The assumption that the borate form is the more reac-
tive species in the Suzuki coupling is perfectly consistent with
this observation. Another validating result is the Suzuki selec-
tivity observed for R /R =4-OCH /4-CHO if different amounts
The results summarized in Table 2 indicate that the relative
reactivity of the boronic acid with higher pK increases with
a
the amount of K CO used. It can also be verified that the de-
2
3
pendence of the reaction selectivity on base stoichiometry in-
creases with the increase in the difference in pK values of the
a
competing boronic acids. In the cases of R /R =4-OCH /4-CHO
1
2
3
and 4-CH /4-NO , in which the difference in pK values is much
3
2
a
larger, this effect is so dramatic that it actually inverts the
order of selectivity. However, if the pK values of the two com-
a
peting boronic acids are similar, as in the case of R /R =3-Ph/
1
2
4
-Ph, the Suzuki selectivity is virtually independent of the
amount of the base used. These observations confirm that
acid–base chemistry is a decisive factor affecting the outcome
of the Suzuki–Miyaura reaction. These results also indicate that
under this set of reaction conditions, the borates are the more
reactive species toward transmetalation in Suzuki–Miyaura
cross-coupling, which thus supports the preference for path B.
The amount of the base used dictates the boronic acids–bo-
rates equilibria on the basis of the pK values of the two bor-
a
1
2
3
onic acids involved. If less K CO is used, then the stronger bor-
of two boronic acids were used. With use of 2 equiv. of 4-me-
thoxyphenylboronic acid versus 1 equiv. of 4-formylphenylbor-
onic acid (for 1 equiv. of K CO ), the major Suzuki product was
2
3
onic acid (lower pK ) reacts more extensively in all cases and
a
its selectivity decreases with the increase in base stoichiometry.
For a lower amount of the base used, the stronger boronic
2
3
still 4-phenylbenzaldehyde in a percent selectivity of 48/52
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1295

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
compared with the 35/65 result in Table 2. Although the great-
er boronic acid concentration and the higher transmetalation
Formation of boric acid and base saturation limit
rate were associated with the presence of the EDG 4-OCH , the
The selectivity observed for the formation of para-terphenyl
3
Suzuki reaction for low base stoichiometry still favored the
(R =4-Ph), compared with that observed for ortho-terphenyl
2
boronic acid with lower pK . With use of twice as much 4-for-
(R =2-Ph), as a function of base stoichiometry is shown in
1
a
mylphenylboronic acid as 4-methoxyphenylboronic, the per-
cent selectivity increased to 17/83, which showed that dou-
Figure 3.
bling the amount of the boronic acid with lower pK produces
a
a greater variation in the Suzuki selectivity than does doubling
the amount of the weaker boronic acid.
The selectivity results for the Suzuki reaction with a stronger
base, KOH, and a different catalyst, Pd(PPh ) , are summarized
3
4
in Table 3. The same selectivity behavior relative to base stoi-
Table 3. Effect of the catalyst and the base on the selectivity in the
[
a]
Suzuki reaction between two boronic acids.
/R catalyst Base
equiv.]
R
1
2
Selectivity [%]
Suzuki reaction
(R /R
S/HC/DB
competition
[
b]
[
1
2
)
2
-Ph/4-Ph–Pd(OAc)
2
K
K
K
2
2
2
CO
CO
CO
3
3
3
(1.0)
(2.5)
(4.7)
24/76
32/68
39/61
28/72
41/59
25/75
33/67
39/61
45/55
35/65
68/32
81/19
39/61
54/46
74/26
40/60
74/26
77/23
61/23/16
70/17/13
66/21/13
76/22/2
69/25/6
64/17/19
75/19/6
71/19/10
70/18/12
67/24/9
67/23/10
72/15/13
76/22/2
61/31/8
73/17/10
74/18/8
76/15/9
72/17/11
Figure 3. Selectivity toward the para-terphenyl as a function of K
alents in the Suzuki reaction between 2- and 4-biphenylboronic acids in
O/DMF (1:1) with 1 equiv. of each boronic acid and 5 mol% of Pd(OAc)
2
CO
3
equiv-
KOH (1.1)
KOH (2.7)
KOH (5.0)
H
2
2
at
1
2
358C for 40 min; R =2-Ph and R =4-Ph.
2-Ph/4-Ph–Pd(PPh
3
)
4
K
2
K
2
K
2
K
2
K
2
K
2
CO
3
CO
3
CO
3
CO
3
CO
3
CO
3
(1.0)
(2.3)
(5.2)
(1.0)
(2.5)
(4.6)
These reactions were performed in an analogous way but at
4-OCH
3 2
/4-CHO–Pd(OAc)
358C for only 40 min of the reaction time. These conditions
were shown to be sufficient for the reactions to occur com-
pletely; no significant amounts of the starting reactants were
observed in the final reaction mixtures and the yields of the
two terphenyl products were similar to the ones observed ear-
lier. However, because these reactions were performed at
a lower temperature, DB was less extensive (ꢂ2–3 mol% for
the cases with less amount of the base) and a little more
ortho-terphenyl was formed; for 1 equiv. of K CO under these
KOH (0.9)
KOH (2.3)
KOH (4.6)
4-OCH
3
/4-CHO–Pd(PPh
3
)
4
K
2
K
2
K
2
CO
3
CO
3
CO
3
(1.0)
(2.5)
(5.2)
[
2
a] All reactions were performed in H O/DMF (1:1) with 1 equiv. of each
boronic acid, 1 equiv. of bromobenzene, and approximately 5 mol% of
the Pd catalyst at 1008C (addition of an aqueous base at 308C) for 3 h;
2
3
conditions, the Suzuki selectivity was 29/71 compared with 24/
6 observed at 1008C (Table 2). How the selectivity to the
[b] Relative to bromobenzene.
7
strongest acid, 4-bPhB(OH) , decreases with the use of more
2
chiometry was observed as in the reactions with K CO base.
base and tends to stabilize at approximately 4–5 equiv. of
K CO is shown in Figure 3. This behavior is consistent with the
2
3
The use of Pd(PPh ) does not change the selectivity trend, but
3
4
2
3
produces a slight change in the individual results, which prob-
ably reflects the different intrinsic reactivity of each organobor-
on species toward transmetalation if a different Pd catalyst is
used. Nevertheless, for the R /R =2-Ph/4-Ph competition, the
rationale presented earlier; if enough base is used, both boron-
ic acids are completely converted to the borates and no more
differentiation in reaction selectivity can be caused by the con-
centration ratio of the reactive species. Another factor that af-
fects the Suzuki selectivity is the formation of boric acid
1
2
use of excess KOH (5.0 equiv.) results in a higher than expected
selectivity toward para-terphenyl. This result can be due to
more extensive DB, which is expected to be favored in highly
[5]
[B(OH) ; pKa 9.0] as a byproduct of all reaction pathways
3
[3–5,17]
shown in Scheme 2.
The concentration of this compound
[
4,5,17]
basic media and for ortho-substituted boronic acids.
In ad-
increases as the reaction proceeds, and because its pK value is
a
dition, DB is considerably more important if R /R =2-Ph/4-Ph
similar to those of some of the boronic acids used herein, its
effect on the boronic acids–borates equilibria must be consid-
1
2
than if R /R =3-Ph/4-Ph (Table 2), which indicates that
1
2
2
-bPhB(OH)2 deboronates more than its isomers. This result
ered. In this way, newly formed B(OH) partially reacts with
3
ꢀ
can also contribute to the lower selectivity toward ortho-ter-
both borates and forms B(OH)4 and the respective boronic
phenyl observed in H O/DMF (1:1) than in H O/acetone (1:1). If
acids; the extent of these reactions depends on the pK values
2
2
a
2
-bPhB(OH) deboronates to a higher extent, less of it reacts
of all species. The involvement of B(OH) can explain the fact
2
3
through Suzuki coupling.
that the Suzuki selectivity to the stronger acid continues to de-
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1296

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
crease after the use of 2–
3
equiv. of the base, a stoichiom-
etry that would be enough to
ensure quantitative conversion
of both boronic acids to their re-
spective
borates
(Figure 2).
Hence, no significant change in
selectivity would be expected
from this point onward. For
most competition reactions pre-
sented herein, the selectivity
continues to decrease after the
use of 2.5 equiv. of the base,
and for the case presented in
Figure 3, the selectivity reaches
a plateau only at approximately
4
–5 equiv. of K CO . This finding
2 3
is logical considering that
equiv. of both boronic acids
2
Scheme 3. General Suzuki competition scenario. EDG=Electron-donating group; EWG=electron-withdrawing
used in these reactions yields
approximately 2 equiv. of B(OH)3
because B(OH)3 is a byproduct
group.
of all the three main reaction pathways shown in
Competition reactions with biphenylborates
[
3–5,17]
Scheme 2.
Therefore, a total of 4 equiv. of acids is re-
quired to react [2ꢂboronic acids+2ꢂB(OH) ], and if more
The results of the Suzuki competition reactions with potassium
biphenylborates as starting reactants are summarized in
Table 4. These results indicate that the Suzuki selectivity is not
significantly affected if the base is replaced by a similar
3
than 4 equiv. of the base is used, only then the complete con-
version of both boronic acids to the borates occurs. Below
4
equiv., the base is not enough to convert all B(OH) formed
3
ꢀ
to B(OH)4 and thus B(OH) partially reacts with the organobo-
amount of a borate conjugate base. The only exception is if 2-
3
ꢀ
rates to form the less reactive boronic acids, an effect that is
more pronounced for the weaker boronic acid. This effect
leads to a renewed concentration gradient between the two
borates, which again favors the Suzuki coupling of the boronic
bPhB(OH)3 is used and reacted against 4-bPhB(OH) ; however,
2
the slightly higher selectivity for the formation of ortho-ter-
phenyl can again be related with the lower extent of DB. The
ꢀ
ꢀ
reaction between 2-bPhB(OH)3 and 4-bPhB(OH)3 was repeat-
acid with lower pK . Before this saturation limit, the reaction
a
selectivity is controlled by base stoichiometry. After the satura-
tion limit, the base is enough to react with all the acids in the
reaction medium and no more change in selectivity with base
stoichiometry is observed. As shown in Figure 2, the newly
Table 4. Results of the Suzuki–Miyaura competition reactions with biphe-
nylborates.
[
a]
Entry
Selectivity [%]
Suzuki reaction S/HC/DB
(R /R
formed B(OH) partially consumes the basic species in the solu-
3
ꢀ
2ꢀ
ꢀ
1
2
)
competition
61/23/16
tion (the two borates and HO /CO /HCO ), shifting the pla-
3
3
teau to higher base stoichiometry, which is in accordance with
the experimental findings.
2-bPhB(OH)
2
2
2
vs.
24/76
30/70
23/77
32/68
34/66
48/52
49/51
49/51
4
2
4
-bPhB(OH)
-bPhB(OH)
-bPhB(OH) (no base)
(1.0 equiv. of K
vs.
3ꢀ
2
CO
3
)
81/17/2
68/26/6
70/17/13
71/17/12
68/31/1
73/26/1
68/31/1
However, if the pK values of both boronic acids are signifi-
a
ꢀ
cantly lower than that of B(OH)3 (R /R =4-CHO/4-NO in
2-bPhB(OH)3 vs.
4-bPhB(OH)
1
2
2
Table 2), the dependence of the Suzuki selectivity on base stoi-
chiometry after 2.5 equiv. of K CO is negligible. In this case,
2
2
2
(no base)
vs.
(2.5 equiv. of K
3ꢀ
ꢀ
2
4
2
-bPhB(OH)
-bPhB(OH)
-bPhB(OH) vs.
2
3
2
CO
CO
3
)
)
B(OH) is a much weaker acid than the two boronic acids and
3
thus its formation in the mixture could not significantly affect
the boronic acids–borates equilibria, which results in negligible
changes in the relative quantities of the two competing bo-
rates. Hence, in this case, the saturation limit is attained as ex-
pected (Figure 2). The general competition reaction between
two boronic acids in Suzuki–Miyaura cross-coupling and the
processes that are proposed as the most relevant for reaction
selectivity considering the reaction conditions used herein are
illustrated in Scheme 3.
4-bPhB(OH)3 (no base)
3-bPhB(OH)
2
2
2
vs.
4
3
4
-bPhB(OH)
-bPhB(OH)
-bPhB(OH) (no base)
(1.0 equiv. of K
vs.
3ꢀ
2
3
ꢀ
3-bPhB(OH)3 vs.
4-bPhB(OH) (no base)
3ꢀ
[
a] All reactions were performed in H
2
O/DMF (1:1) with 1 equiv. of each
organoboron species, 1 equiv. of bromobenzene, and approximately
5
mol% of Pd(OAc)
2
at 1008C for 3 h. Base stoichiometry is relative to
ChemCatChem 2014, 6, 1291 – 1302 1297
bromobenzene.
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[
1–3,17]
ed at 358C for 45 min, and the selectivity results were similar
base in this reaction is in the transmetalation step,
and
to the ones summarized in Table 4: the percent selectivity of
thus varying the amount of the base mainly affects the reactiv-
ity of the species involved in this step of the mechanism. In
this mechanistic study, to maintain the routine reaction condi-
tions, the precursor of the active Pd species Pd(OAc)2 or
Pd(PPh ) was used instead of adding the active Pd species di-
34/66 in the Suzuki reaction and 75/16/9 in the S/HC/DB com-
petition. The reaction yields of terphenyl isomers were also
comparable: 81% at 358C and 85% at 1008C. This result sug-
gests that the effect of temperature on selectivity is small and
that the reactions proceed relatively fast.
3
4
rectly to the reaction medium, a method that would increase
the confidence that no other unknown intermediates are
formed in the solution. However, the selectivity results are sim-
The fact that the Suzuki selectivity results are similar for the
reactions that are compared in Table 4 proves that the acid–
base equilibrium is actually established in the reaction
medium. In the cases in which one borate reactant is used, re-
placing the reactant that is used as the borate produces similar
results; these cases compare to the reactions in which 1 equiv.
of the base is used. In these reactions, the Suzuki coupling
II
0
ilar for the two catalysts tested, Pd (OAc) and Pd (PPh ) , and
2
3 4
II
the main R’ꢀPd ꢀX intermediate has already been shown to be
formed in the reaction medium under these reaction condi-
[3]
tions, making it safe to assume that the main reaction path-
way yielding the desired Suzuki products is the reaction be-
II
product corresponding to the boronic acid with lower pK is
tween R’ꢀPd ꢀX and organoboron species. Moreover, the dif-
a
formed preferentially. If the borates of both boronic acids are
used, the results are comparable to those of the case in which
ferential selectivity analysis used herein ensures that the reac-
tions of the two competing organoboron species proceed and
are evaluated under the same experimental conditions. These
facts reduce the possibility that some unknown intermediates
significantly affect the presented analysis.
2.5 equiv. of the base is used. Here, the preference for the
strongest boronic acid decreases, which reflects the higher
ꢀ
equilibrium proportion of 2-bPhB(OH)3 . The Keq values for the
general reaction for the boronic acids used herein [Eq. (3)]
This study shows that increasing the amount of the base
ꢀ5
ꢀ7
range from approximately 10 to 10 (considering the pKa
values and Kw in pure water at 258C), which indicates that in
favors the reactivity of the boronic acid with higher pK ; this
a
observation indicates that the competition for transmetalation
is dictated by the relative rates of attack of the two borate spe-
all cases the concentration of RꢀB(OH) is low. These results
2
ꢀ
II
further support the preference for path B.
cies, ꢀB(OH)3 , toward the R’ꢀPd ꢀX complex (path B). If path
A was preferred (the reaction of boronic acids with the R’ꢀ
RꢀBðOHÞ3ꢀ
Ð RꢀBðOHÞ þ HOꢀ
2
II
ð3Þ
Pd ꢀOH intermediate), the use of many base equivalents
would result in lower Suzuki reaction conversions because the
concentrations of boronic acids in the solution would be too
low to ensure a fast and complete reaction with high selectivi-
ty toward the desired Suzuki products. The increase in the
amount of the base does not significantly decrease the reac-
tion selectivity toward the Suzuki products (Tables 2 and 3).
Moreover, as mentioned earlier, unreacted bromobenzene and
boronic acids were never observed in significant amounts and
in all cases high reaction yields were obtained. Simple acid–
Dependence of transmetalation on reaction conditions
One of the major concerns of this study was to analyze the
role of the base under the experimental conditions usually
used in Suzuki reactions by avoiding any perturbation or modi-
fication of the reactions as they proceed. Because it is known
that, contrary to oxidative addition and reductive elimination,
the transmetalation step in cross-coupling reactions is general-
ꢀ
base equilibrium calculations indicate that for 1 equiv. of HO ,
[
15–17]
ly sensitive to experimental conditions,
it is justifiable to
the fraction of borate [calculated by using Eq. (4)] in water for
[33]
reproduce the actual reaction conditions as closely as possible.
The side reactions are unavoidable if our competition reaction
strategy is used; significant DB and HC products are typical of
many Suzuki reactions performed in the laboratory and indus-
phenylboronic acid (pK 8.8), considering the typical initial
a
ꢀ
3
concentrations used in Suzuki reactions (ꢂ0.1–0.01 moldm ),
is greater than 0.97.
[
3,5,17]
½
borateꢁ
try for synthetic requirements.
However, envisaging a dif-
f ¼
ð4Þ
½
borateꢁ þ ½boronic acidꢁ
ferent approach to reduce side reactions and other possible
side effects inevitably leads to some oversimplification of the
real process and/or some modification of the reaction
medium, which invalidates the requirement of using routine
reaction conditions and adds more uncertainty if the derived
conclusions are extrapolated to the real situation. The fact that
a given product is formed in higher yield than the other can
be directly related to the greater relative reactivity of the re-
spective organoboron species toward transmetalation because
of the following reasons: 1) the extent of side reactions (HC
and DB) and the overall reaction yield do not change signifi-
cantly and show no correlation with base stoichiometry; the
major effect of changing the amount of the base is on the se-
lectivity toward two Suzuki products; 2) the main role of the
In aqueous organic solvents, the pK value is slightly differ-
a
ent from that in pure water. The fraction of borate is expected
to be even higher as the water content of the solvent decreas-
ꢀ
es, reducing the concentration of the highly hydrophilic HO
anion by allowing it to form the more hydrophobic organobor-
on anion. This simple calculation confirms the low concentra-
tions of the two boronic acids in Suzuki–Miyaura reactions per-
formed under routine laboratory conditions. Suzuki reactions
are relatively fast and usually performed under mild conditions,
thus giving high yields (1–2 h are usually enough to obtain
[3,17,27,39–44]
yields above 80%),
which again indicates that the
main role of the base in these reactions is to convert the
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1298

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
ꢀ
[8,14]
poorly reactive boronic acid to the more reactive borate for
subsequent transmetalation (path B). This rationale can also
help explain why sterically hindered boronic acids generally
lead to lower Suzuki reaction yields and why those yields as
well as coupling rates can be improved by using stronger
HO .
This can be explained as the acidic species of one of
the two possible reactive pairs, RꢀB(OH) /R’ꢀPdꢀOH (for
2
ꢀ
path A) or RꢀB(OH) /R’ꢀPdꢀX (for path B), is significantly con-
3
sumed by the excess base. Hence, if K [Eq. (5)]>K [Eq. (6)],
eq
eq
path B probably becomes the main route for transmetalation;
[
3,17,44,45]
bases such as NaOH or Ba(OH)2.
of these boronic acids tend to be higher in aqueous solutions
ortho groups prevent the efficient stabilization of the borate
Because the pK values
however, if K [Eq. (5)]<K [Eq. (6)], path A becomes the main
a
eq
eq
route. The reported pK values of boronic acids in pure water
a
5
[5,33]
(
indicate that K (5) is in the order of 10 .
The NMR spectra
eq
head through solvation), a more basic medium is needed to
ensure a significant concentration of the borate form. More-
of the three isolated biphenylborates demonstrate that in D O/
2
[D ]acetone (1:1) and D O/[D ]DMSO(1:2), only the peaks for
6
2
6
over, if the pK value of the boronic acid is higher than that of
the borate form can been observed, which means that in
a
B(OH) , the reaction tends to become self-quenching; the frac-
these solvents, K [Eq. (5)]@1 (considering the 1:1 stoichiome-
3
eq
ꢀ
tion of the reactive borate starts to decrease substantially with
try of boronic acid and HO presented in the borate sample).
the formation of B(OH) . A more basic medium also ensures
The results published by Carrow and Hartwig indicate that for
3
that more of B(OH) is converted to the non-borate-scavenger
the reaction depicted in Equation (7) in THF/H O (50:1) and
3
2
ꢀ
B(OH)4 form. Moreover, the results presented herein support
THF/H O (25:1), the Keq(208C) value is 0.77 and 0.11,
2
[7]
the preference for path B because of the following: 1) the se-
respectively.
lectivity to the stronger boronic acid is higher for low base sto-
ꢀ
PhðPCy Þ ꢀPdꢀBr þ NBu OH Ð PhðPCy Þ ꢀPdꢀOH þ NBu Br
ichiometry, at which the concentration of its borate RꢀB(OH)
3
2
4
3
2
4
3
relative to the weaker acid is higher and the absolute concen-
ð7Þ
tration of RꢀB(OH) of the weaker acid is the highest; 2) the se-
2
lectivity to the weaker boronic acid is higher for higher base
stoichiometry, at which the relative concentration of its borate
Hence, in 1:1 aqueous organic mixtures, the K [Eq. (6)]
eq
value is expected to be very small. Notably, the negatively
charged borate species and the Pd–halide complex, in contrast
to the PdꢀOH intermediate, were detected in reaction mixtures
ꢀ
RꢀB(OH)3 is more leveled, and the absolute concentration of
its RꢀB(OH) is lower; 3) the reaction is not quenched at high
2
[30–32]
base stoichiometry, at which the concentrations of both boron-
ic acids are negligible; instead, it proceeds as usual, which
gives high isolated reaction yields in relatively fast reactions;
by using ESIMS.
These findings suggest that if the reaction
conditions used herein are considered, then K [Eq. (5)]@
eq
Keq [Eq. (6)], which supports the preference for path B. As re-
4
) larger difference in the pK values of the two competing
gards the byproduct B(OH) , it shall abstract preferentially the
a
3
ꢀ
boronic acids leads to greater variations in the Suzuki selectivi-
ty with K CO equivalents, which follows the greater variations
HO anion from the strongest conjugate base—in this case
PdꢀOH—which further reduces its equilibrium concentration.
All these observations confirm the low equilibrium concentra-
tions of both the PdꢀOH complex and the boronic acid under
2
3
ꢀ
in the relative concentrations of the two borates RꢀB(OH)
3
with base stoichiometry (Figure 2); however, the dependence
of the relative concentrations of the two boronic acids, Rꢀ these conditions. Although the use of higher amounts of the
B(OH) , on K CO equivalents is less in all cases (see details in
base increases the concentration of PdꢀOH, it decreases the
2
2
3
Section S7 of the Supporting Information).
concentration of RꢀB(OH) to marginal values. Because the
2
However, the main transmetalation pathway followed in
a Suzuki reaction may greatly depend on the reaction condi-
tions. Our results do not eliminate the possibility that path A is
followed if a set of different conditions is used, either when
performing other types of mechanistic studies or just when
carrying out the synthesis in a different way. It is conceivable
borate is the more reactive species does not imply that it is
the B atom, or the C atom directly bonded to it, that attacks
the Pd center. The attacking group can be one of the three nu-
cleophilic ꢀOH species of the borate head. If a precomplexation
of the B atom of RꢀB(OH) with ꢀOH species of PdꢀOH occurs
2
easily, then a precomplexation of one of the three ꢀOH species
ꢀ
ꢀ
that RꢀB(OH) and RꢀB(OH) are virtually unreactive toward
of RꢀB(OH) with PdꢀX is also feasible (see Ref. [10]). To date,
2
3
3
II
II
[8,10,14]
R’ꢀPd ꢀX and R’ꢀPd ꢀOH, respectively.
Therefore, which
and considering all the divergences presented in the literature,
the most reasonable rationalization for the role of the base in
the Suzuki coupling is accepting that because transmetalation
is sensitive to experimental conditions, either path is possible,
depending on those conditions. The results presented herein
indicate that if the reaction conditions similar to those routine-
ly used in Suzuki synthesis are used, path B is the preferred
route for transmetalation.
path is preferred depends on the relative magnitude of K for
eq
the reactions depicted in Equations (5) and (6) (and how those
Keq values vary with solvent, temperature, Pd catalyst, R group,
etc.):
ꢀ
ꢀ
3
RꢀBðOHÞ þ HO Ð RꢀBðOHÞ
R’ꢀPdꢀX þ HO Ð R’ꢀPdꢀOH þ Xꢀ
ð5Þ
ð6Þ
2
ꢀ
Under the routine synthesis conditions, either the amount of
Conclusions
the base is not enough for both reactions to reach completion
ꢀ
or the reaction is quenched because RꢀB(OH) does not react
The role of the base in the Suzuki–Miyaura cross-coupling reac-
tion was evaluated experimentally by using a competition re-
action methodology. This approach involved the coupling of
3
II
with R’ꢀPd ꢀOH. As observed by Amatore et al., the transmeta-
lation rate starts to decrease at very high concentrations of
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1299

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
relatively simple arylboronic acids with bromobenzene with
use of common Pd catalysts and solvents and varying amounts
of the base. The reaction selectivity was evaluated after the
completion of the reaction, avoiding any perturbation or modi-
fication of the reaction medium, to study the role of the base
under the actual and routinely used reaction conditions. The
results presented herein can only be rationalized completely if
was washed first with water and then with NaOH(aq) and dried over
anhydrous Na SO . The solvent was evaporated, and the product
2
4
was purified by using column chromatography with the CH Cl /
2
2
light petroleum (1:1) mixture as an eluent. After the solvent was
evaporated to dryness, the 1,1’-biphenyl derivatives were obtained
as white to yellow solids in good yields.
ꢀ
the organoborate form, RꢀB(OH)3 , is considered as the most
Synthesis of ortho-quaterphenyl
reactive species toward transmetalation in the Suzuki coupling.
These results do not completely eliminate the possibility that
both the boronic acid and the borate are reactive toward
transmetalation; however, they strongly indicate that the
borate is more reactive if the synthesis is performed in aque-
ous organic mixtures and using standard conditions and meth-
ods. It can therefore be concluded that under such conditions,
the main role of the base in the Suzuki–Miyaura reaction is to
convert the boronic acid to the more reactive organoborate,
which facilitates transmetalation with the Pd–halide intermedi-
ate (path B). The most important factors that affect the Suzuki
reaction selectivity are as follows: 1) the boronic acids–borates
A solution of 2-bromobiphenyl (1.0 equiv.), 2-biphenylboronic acid
(
1.5 equiv.), and Pd(PPh ) (2 mol%) in toluene (30 mL) was heated
3 4
at 708C and stirred under N . Then, a solution of K CO (1.5 equiv.)
2
2
3
in water (15 mL) was added to the previous solution, and the resul-
tant mixture was maintained at 708C for 10 h under stirring. The
crude product was extracted with toluene. The organic layer was
washed with water, NaOH(aq), and HCl(aq). The organic layer was
dried over anhydrous Na SO4 and evaporated, which produced
2
a white solid. The compound was recrystallized in methanol and
purified by sublimation under reduced pressure; the final product
1
was obtained in 45% yield. H NMR (300.13 MHz, CDCl ): d=6.63
3
(d, J=6.9 Hz, 4H; H-3), 7.01 (t, J=7.4 Hz, 4H; H-2), 7.10 (t, J=
7
.4 Hz, 2H; H-1), 7.16–7.22 (m, 2H), 7.31–7.46 ppm (m, 6H);
C NMR (75.47 MHz, CDCl ): d=125.9, 127.0, 127.4, 127.4, 129.2,
3
1
3
equilibria, which depend on the pK values of the two compet-
a
ing boronic acids and on the amount of the base used; 2) the
129.9, 131.6, 140.0, 140.9, 141.0 ppm (see Figure S14 for atom
formation of B(OH) as the reaction proceeds, which disturbs
numbering in ortho-quaterphenyl).
3
the acid–base equilibrium; and 3) the intrinsic reactivity of
each organoboron species toward transmetalation. From the
practical point of view, the rationale derived from this study
enables one to control the selectivity and outcome of the
Suzuki–Miyaura reaction by varying the amount and/or type of
Synthesis of meta-quaterphenyl
A solution of 3-bromobiphenyl (1.0 equiv.), 3-biphenylboronic acid
(1.5 equiv.), and PdCl
heated at 708C. Then, a solution of K CO3 (1.8 equiv.) in water
(dppe) (2 mol%) in toluene (30 mL) was
2
the base used and considering the pK values of the two com-
2
a
(15 mL) was added to the previous solution, and the resultant mix-
peting boronic acids.
ture was maintained at 708C for 10 h under stirring. The crude
product was extracted with toluene. The organic layer was washed
with water, NaOH(aq), and HCl(aq). The organic layer was dried over
anhydrous Na SO and evaporated, which produced a white solid.
Experimental Section
2
4
The compound was recrystallized in methanol and purified by sub-
Materials and supporting substances
limation under reduced pressure; the final product was obtained
1
Most reagents and solvents were purchased from commercial sour-
ces and used as supplied. The compounds bromobenzene, biphen-
yl, ortho-terphenyl, meta-terphenyl, para-terphenyl, para-quater-
phenyl, anisole, benzaldehyde, toluene, nitrobenzene, 4,4’-dime-
thylbiphenyl, 4,4’-dimethoxybiphenyl, 4,4’-diformylbiphenyl, and
in 76% yield. H NMR (300.13 MHz, CDCl ): d=7.42 (t, J=7.4 Hz,
3
2
7
H; H-1), 7.51 (t, J=7.4 Hz, 4H; H-2), 7.59 (d, J=7.8 Hz, 2H; H-7),
.62–7.74 (m, 8H), 7.91 ppm (t, J=1.6 Hz, 2H; H-10); C NMR
1
3
(75.47 MHz, CDCl ): d=126.2, 126.2, 126.2, 127.3, 127.4, 128.8,
3
1
29.2, 141.1, 141.7, 141.8 ppm (see Figure S16 for atom numbering
4
1
,4’-dinitrobiphenyl were obtained from commercial sources. The
,1’-biphenyl derivatives 4-methoxybiphenyl, 4-phenylbenzalde-
in meta-quaterphenyl).
hyde, 4-methylbiphenyl, and 4-nitrobiphenyl were synthesized
through the Suzuki coupling by reacting the corresponding aryl-
Synthesis of 1,1:3’,1”:4”,1”’-quaterphenyl
1
13
boronic acids with bromobenzene. The H and C NMR spectra
were recorded on a Bruker Avance 300 spectrometer, except for 4-
methoxybiphenyl, the spectra of which were recorded on a Bruker
Avance III 400 spectrometer, with TMS as an internal reference and
A solution of 4-bromobiphenyl (1.0 equiv.), 3-biphenylboronic acid
(
1.5 equiv.), and Pd(PPh3)4 (2 mol%) in DMF (20 mL) was stirred
under N at 08C. After complete dissolution, a solution of K CO
3
2
2
(1.8 equiv.) in water (10 mL) was added to the previous solution
CDCl as a solvent unless otherwise stated. The detailed description
3
and the resultant mixture was maintained at 08C for 5 h under stir-
of the syntheses and the detailed compound characterization are
presented in the Supporting Information.
ring. The crude product was extracted with CHCl . The organic
3
layer was washed with water, NaOH(aq), and HCl(aq). The organic
layer was dried over anhydrous Na SO and evaporated, which pro-
2
4
duced a white solid. The compound was purified by sublimation
Synthesis of 1,1’-biphenyl derivatives
under reduced pressure; the final product was obtained in 80%
1
K CO (1.4 equiv.) dissolved in water (3.5 mL) was added to a solu-
yield. H NMR (300.13 MHz, CDCl ): d=7.33–7.41 (m, 2H), 7.47 (t,
2
3
3
tion of bromobenzene (1.0 equiv.), boronic acid (1.2 equiv.), and
J=7.4 Hz, 4H; H-2 and H-17), 7.54 (d, J=7.5 Hz, 1H; H-13), 7.63–
7.77 (m, 10H), 7.86 ppm (t, J=1.6 Hz, 1H; H-10); C NMR
1
3
Pd(OAc) (2 mol%) in DMF (3.5 mL). The mixture was heated and
2
stirred at 1008C for 4 h. Then, CH Cl was added to the reaction
(75.47 MHz, CDCl ): d=126.0, 126.0, 126.2, 127.1, 127.3, 127.4,
2
2
3
mixture at RT, which was filtered through Celite. The organic layer
127.4, 127.5, 127.6, 128.8, 129.2, 140.0, 140.3, 140.6, 141.1, 141.2,
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1300

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
1
41.8 ppm (see Figure S18 for atom numbering in 1,1’:3’,1”:4”,1”’-
and washed with water and KOH(aq). The resulting organic phase
was dried over anhydrous Na SO and analyzed by using GC.
quaterphenyl).
2
4
Synthesis of 1,1’:2’,1”:4”,1”’-quaterphenyl
Representative Suzuki–Miyaura competition reaction of
-methylphenylboronic acid and 4-nitrophenylboronic acid
4
A solution of 2-bromobiphenyl (1.0 equiv.), 4-biphenylboronic acid
(
0
1.5 equiv.), and Pd(PPh ) (2 mol%) in DMF (20 mL) was stirred at
A mixture of bromobenzene (0.50 mmol), 4-methylphenylboronic
acid (0.51 mmol), 4-nitrophenylboronic acid (0.71 mmol), and
Pd(OAc)2 (5 mol%) in DMF (15 mL) was prepared in a round
bottom flask and heated at 308C under stirring. After complete dis-
3
4
8C. After complete dissolution, a solution of K CO (1.8 equiv.) in
2 3
water (10 mL) was added to the previous solution and the resul-
tant mixture was maintained at 08C for 5 h under stirring. The
crude product was extracted with toluene. The organic layer was
washed with water and NaOH(aq). The organic layer was dried over
anhydrous Na SO and evaporated, which produced a white solid.
solution, a solution of K CO3 (0.51 mmol) in water (15 mL) was
2
added and the mixture was heated at 1008C for 3 h. The final mix-
2
4
ture was left to cool to RT and extracted twice with CH Cl . The or-
2
2
The compound was purified by sublimation under reduced pres-
sure; the final product was obtained in 90% yield. H NMR
ganic phase was washed once with water, twice with KOH(aq), and
once with water and then dried over anhydrous Na SO . An aliquot
1
2
4
(
300.13 MHz, CDCl ): d=7.15–7.25 (m, 7H), 7.32 (tt, J =7.2 Hz, J =
of this final organic solution was injected into the gas chromato-
graph for selectivity analysis. The solvent was then removed by
evaporation, and a brown viscous oil was obtained, which con-
tained the two Suzuki–Miyaura products in 94% yield (GC-correct-
ed isolated yield).
3
1
2
1
3
1
1
1
.7 Hz, 1H; H-1), 7.38–7.50 (m, 8H), 7.58 ppm (d, J=7.2 Hz, 2H; H-
1
3
); C NMR (75.47 MHz, CDCl ): d=126.5, 126.5, 126.9, 127.2, 127.5,
3
27.9, 128.7, 129.9, 130.3, 130.6, 130.7, 139.0, 140.0, 140.5, 140.6,
40.6, 141.5 ppm (see Figure S20 for atom numbering in
,1’:2’,1”:4”,1”’-quaterphenyl).
Representative Suzuki–Miyaura competition reaction of
2-biphenylboronic acid and 4-biphenylboronic acid
Preparation of potassium biphenylborates
+
ꢀ
The potassium borate derivatives, K ·ArꢀB(OH) , of 2-, 3-, and 4-
A mixture of bromobenzene (0.70 mmol), 2-biphenylboronic acid
(0.71 mmol), 4-biphenylboronic acid (0.50 mmol), and Pd(OAc)2
(5 mol%) in DMF (20 mL) was prepared in a round bottom flask
and heated at 358C under stirring. After complete dissolution, a so-
3
biphenylboronic acids were prepared by dissolving the corre-
sponding boronic acids in water with a small excess of KOH, ac-
[46]
cording to a previously reported method. The solution was then
filtered to remove undissolved boronic acid, and the water was re-
moved by evaporation. The resulting precipitate was washed first
with small droplets of water and then with acetone and ethanol.
The potassium arylborates were characterized by using attenuated
lution of K CO3 (1.2 mmol) in water (20 mL) was added and the
2
mixture was heated at 358C for 40 min. The final mixture was ex-
tracted twice with CH Cl . The organic phase was washed once
2
2
with water, twice with KOH(aq), and once with water and then dried
over anhydrous Na SO . An aliquot of this final organic solution
1
total reflectance FTIR spectroscopy and H NMR spectroscopy in
2
4
D O/[D ]acetone (1:1) and D O/[D ]DMSO (1:2) at 258C. The biphe-
was injected into the gas chromatograph for selectivity analysis.
The solvent was then removed by evaporation, and a yellow vis-
cous oil was obtained, which contained the two Suzuki–Miyaura
products in 94% yield (GC-corrected isolated yield).
2
6
2
6
nylborates were found to be poorly soluble in water and acetone
but highly soluble in an acetone-rich mixture of both solvents,
which suggested that acetone solvates the larger nonpolar biphen-
yl region and water tended to solvate the smaller borate head.
GC analyses
Representative synthesis of potassium 2-biphenylborate
All GC analyses were performed with an HP 4890 apparatus
equipped with an HP-5 column, cross-linked, diphenyl (5%), and
dimethylpolysiloxane (95%), and a flame ionization detector. Care
was taken to ensure that the final CH Cl organic phase was homo-
Solid 2-biphenylboronic acid (3.1 mmol) was added slowly to a so-
lution of KOH (3.4 mmol) in water (20 mL). The aqueous mixture
was filtered by gravity, and the water was removed by evaporation.
The white solid obtained was washed first with small droplets of
water and then with acetone and ethanol and dried under
vacuum, which produced a white powder of potassium 2-biphenyl-
borate in 97% yield.
2
2
geneous, and no precipitates or suspensions were present. Sample
injections (1 mL) were performed with a microsyringe. The reaction
selectivity was evaluated by analyzing the full chromatograms; all
the relevant reagents, products, and byproducts of the Suzuki reac-
tion (unreacted bromobenzene as well as coupling, HC, and DB
products) could be detected and quantified in this way. Unreacted
boronic acids were never observed. To convert the observed rela-
tive peak areas into relative mole fractions, calibration injections
were performed with samples of known concentrations of each
compound. These samples were prepared from the originally syn-
thesized or commercially purchased compounds to ascribe each
peak to each given compound on the basis of their retention
times.
Suzuki–Miyaura competition reactions
Unless otherwise noted, all reactions were performed using bromo-
benzene (1.0 equiv.), each boronic acid (1.0 equiv.), a base (K CO3
or KOH, in varying amounts), and Pd(OAc)2 or Pd(PPh3)4 catalyst
2
(
ꢂ5 mol%) in H O/DMF (1:1) at 358C or 1008C for 40 min or 3 h,
2
respectively. Most reactions were initiated by adding the aqueous
solution of the base to the mixture of the other reactants in DMF
at 308C after the complete dissolution of all species in the solution.
In the reactions in which at least one boronic acid was replaced by
its respective potassium borate, no base was used. These reactions
were initiated by adding the Pd catalyst to the solvent mixture of
1
H NMR evaluation of acid strength
The relative acidities of 2-, 3-, and 4-biphenylboronic acids were
1
all reactants. The final reaction mixture was extracted with CH Cl2
evaluated by recording the H NMR spectra of equimolar mixtures
2
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1301

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
of one biphenylboronic acid and one potassium biphenylborate of
another biphenyl isomer in D O/[D ]acetone (1:1) and D O/
[14] C. Amatore, G. Le Duc, A. Jutand, Chem. Eur. J. 2013, 19, 10082.
[
[
15] K. Tamao, N. Miyaura, Top. Curr. Chem. 2002, 219, 1.
16] P. Espinet, A. M. Echavarren, Angew. Chem. Int. Ed. 2004, 43, 4704;
Angew. Chem. 2004, 116, 4808.
2
6
2
[
D ]DMSO (1:2) at 258C; the initial concentration of each compo-
6
3
ꢀ
nent in the mixture was approximately 0.03 moldm
.
[
[
[
17] A. Meijere, F. Diederich, Metal-Catalyzed Cross-Coupling Reactions, Vol. 1,
2
ed., Wiley-VCH, Weinheim, 2004.
18] H. C. Brown, C. Verbrugge, C. H. Snyder, J. Am. Chem. Soc. 1961, 83,
001.
19] H. C. Brown, C. H. Snyder, J. Am. Chem. Soc. 1961, 83, 1002.
Computational details
1
All theoretical calculations were performed with the Gaussian 03
[47]
[20] R. C. Larock, J. Organomet. Chem. 1974, 72, 35.
software. The full geometry optimizations were performed at the
MP2/cc-pVDZ and B3LYP/6-311+ +G(d,p) levels of theory. The
spin-component-scaled second-order Møller–Plesset approach was
also used for the calculation of the ground-state electronic
[
[
21] N. Miyaura, J. Organomet. Chem. 2002, 653, 54.
22] P. M. Maitlis, The Organic Chemistry of Palladium, Vol. 2, Academic Press,
New York, 1971.
[
23] N. Miyaura, K. Yamada, H. Suginome, A. Suzuki, J. Am. Chem. Soc. 1985,
107, 972.
[36]
energies, E (0 K).
el
[
[
24] F. Sasaya, N. Miyaura, A. Suzuki, Bull. Korean Chem. Soc. 1987, 8, 329.
25] J. H. Brewster, Aspects of Mechanism and Organometallic Chemistry,
Plenum Press, New York, 1978.
Acknowledgements
[
26] A. Pelter, K. Smith, H. C. Brown, Borane Reagents, Academic Press, New
York, 1988.
We thank the Fundażo para a Ciencia e Tecnologia (FCT),
Lisbon, Portugal, and the European Social Fund (ESF) for financial
support to CIQ, the University of Porto (strategic project PEst-C/
QUI/UI0081/2013), and the Organic Chemistry Research Unit
[
[
27] T. I. Wallow, B. M. Novak, J. Org. Chem. 1994, 59, 5034.
28] R. Glaser, N. Knotts, J. Phys. Chem. A 2006, 110, 1295.
[29] A. A. C. Braga, G. Ujaque, F. Maseras, Organometallics 2006, 25, 3647.
[30] C. A. Hutton, A. K. L. Yuen, Tetrahedron Lett. 2005, 46, 7899.
[
[
[
31] M. Nunes, A. L. Monteiro, J. Braz. Chem. Soc. 2007, 18, 1443.
32] A. O. Aliprantis, J. W. Canary, J. Am. Chem. Soc. 1994, 116, 6985.
33] J. Yan, G. Springsteen, S. Deeter, B. Wang, Tetrahedron 2004, 60, 11205.
(
project PEst-C/QUI/UI0062/2013). C.F.R.A.C.L. and A.S.M.C.R. also
thank the FCT and the ESF within the Third Community Support
Framework for the research grants SFRH/BPD/77972/2011 and
SFRH/BD/81261/2011, respectively.
[34] C. F. R. A. C. Lima, M. A. A. Rocha, A. Melo, L. R. Gomes, J. N. Low,
L. M. N. B. F. Santos, J. Phys. Chem. A 2011, 115, 11876.
35] M. A. V. Ribeiro da Silva, L. M. N. B. F. Santos, L. M. S. S. Lima, J. Chem.
Thermodyn. 2008, 40, 375.
[
Keywords: basicity
CꢀC coupling · transmetalation
·
cross-coupling
·
organoborates
·
[36] S. Grimme, J. Chem. Phys. 2003, 118, 9095.
[
[
[
37] E. V. Anslyn, D. A. Dougherty, Modern Physical Organic Chemistry, Univer-
sity Science Books, 2006, chap. 3, p. 147.
38] E. Hogfelt, D. D. Perrin, Stability Constants of Metal Ion Complexes,
[
[
[
[
[
1] N. Miyaura, K. Yamada, A. Suzuki, Tetrahedron Lett. 1979, 20, 3437.
2] N. Miyaura, T. Yanagi, A. Suzuki, Synth. Commun. 1981, 11, 513.
3] N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457.
1
st ed., New Pergamon, Oxford, 1979–1982.
39] C. F. R. A. C. Lima, J. E. Rodriguez-Borges, L. M. N. B. F. Santos, Tetrahe-
dron 2011, 67, 689.
4] N. Miyaura, Top. Curr. Chem. 2002, 219, 11.
[40] L. Liu, Y. Zhang, B. Xin, J. Org. Chem. 2006, 71, 3994.
[41] L. Liu, Y. Zhang, Y. Wang, J. Org. Chem. 2005, 70, 6122.
[42] J. P. Wolfe, R. A. Singer, B. H. Yang, S. L. Buchwald, J. Am. Chem. Soc.
5] D. G. Hall, Boronic Acids: Preparation and Applications in Organic Synthe-
sis Medicine and Materials, 2 ed., Wiley-VCH, Weinheim, 2011, chap. 1.
6] A. Suzuki, Angew. Chem. Int. Ed. 2011, 50, 6722–6737; Angew. Chem.
[
1
999, 121, 9550.
2011, 123, 6854–6869.
[
[
[
[
[
43] D. J. Sinclair, M. S. Sherburn, J. Org. Chem. 2005, 70, 3730.
44] A. Suzuki, Chem. Commun. 2005, 4759.
45] T. Watanabe, N. Miyaura, A. Suzuki, Synlett 1992, 207.
46] S. Bhat, S. Punachithaya, S. Ganesh, WO2005/019229A1, 2005.
47] Gaussian 03, Revision C.02, M. J. Frisch, et al., Gaussian, Inc., Wallingford
CT, 2004.
[7] B. P. Carrow, J. F. Hartwig, J. Am. Chem. Soc. 2011, 133, 2116.
[8] C. Amatore, A. Jutand, G. Le Duc, Chem. Eur. J. 2011, 17, 2492.
[9] C. Adamo, C. Amatore, I. Ciofini, A. Jutand, H. Lakmini, J. Am. Chem. Soc.
2
006, 128, 6829.
10] A. A. C. Braga, N. H. Morgon, G. Ujaque, F. Maseras, J. Am. Chem. Soc.
005, 127, 9298.
[
2
[11] Y. Suzaki, K. Osakada, Organometallics 2006, 25, 3251.
[12] K. Matos, J. A. Soderquist, J. Org. Chem. 1998, 63, 461.
[13] A. J. J. Lennox, G. C. Lloyd-Jones, Angew. Chem. Int. Ed. 2013, 52, 7362;
Angew. Chem. 2013, 125, 7506.
Received: December 16, 2013
Published online on March 12, 2014
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1291 – 1302 1302