Palladium-catalyzed Suzuki-Miyaura cross-coupling reactions of potassium aryl- and heteroaryltrifluoroborates

Article abstract of DOI:10.1021/jo0342368

An extended study of the reactivity of potassium aryl- and heteroaryltrifluoroborates in Suzuki-Miyaura cross-coupling reactions is presented. The coupling of aryl- and electron-rich heteroaryltrifluoroborates with aryl and activated heteroaryl bromides proceeds readily under ligandless conditions. When deactivated aryl- and heteroaryltrifluoroborates are coupled with aryl and heteroaryl bromides and chlorides, a low loading (0.5-2%) of PdCl₂(dppf)·CH₂Cl₂ efficiently catalyzes the reactions. Under either condition, reactions can generally be carried out in an open atmosphere.

Full text of DOI:10.1021/jo0342368

P a lla d iu m -Ca ta lyzed Su zu k i-Miya u r a Cr oss-Cou p lin g Rea ction s
of P ota ssiu m Ar yl- a n d Heter oa r yltr iflu or obor a tes
Gary A. Molander* and Betina Biolatto
Roy and Diana Vagelos Laboratories, Department of Chemistry, University of Pennsylvania,
Philadelphia, Pennsylvania 19104-6323
gmolandr@sas.upenn.edu
Received February 21, 2003
An extended study of the reactivity of potassium aryl- and heteroaryltrifluoroborates in Suzuki-
Miyaura cross-coupling reactions is presented. The coupling of aryl- and electron-rich heteroaryl-
trifluoroborates with aryl and activated heteroaryl bromides proceeds readily under ligandless
conditions. When deactivated aryl- and heteroaryltrifluoroborates are coupled with aryl and
heteroaryl bromides and chlorides, a low loading (0.5-2%) of PdCl2(dppf)‚CH2Cl2 efficiently catalyzes
the reactions. Under either condition, reactions can generally be carried out in an open atmosphere.
In tr od u ction
metal/ligand systems that facilitate the cross-coupling
with different electrophiles have been the most exten-
sively studied. Combinations that range from Pd(0) (Pd/
C, Pd/Al2O3, Pd2dba3) or Pd(II) [PdCl2, Pd(OAc)2] in the
absence of ligands5 to very elaborate catalyst/ligand
systems6 have been applied. Very recently the successful
use of air-stable catalysts has been reported to improve
the catalyst turnover, to achieve the coupling with less
reactive electrophiles, or to permit coupling with hindered
substrates.7
One area that has been underdeveloped involves the
use of alternative organoboron coupling partners. Al-
though the palladium-catalyzed cross-coupling reaction
of arylboronic acids or -esters bearing sp2-hybridized
substituents with electrophiles such as organic halides
and triflates generally proceeds readily,1,3,8 the boronic
The synthesis of biaryl and biheteroaryl systems
through transition metal catalyzed cross-coupling reac-
tions has been well-studied and developed.1 Among all
the possible organometallics used as nucleophilic part-
ners, tin (Stille coupling)2 and boron derivatives (Suzuki-
Miyaura coupling)3 and, to a lesser extent, organozinc
(Negishi coupling)4 and Grignard reagents (Kumada
coupling),1 have been the most frequently used for cross-
coupling reactions of aryl and heteroaryl compounds. The
Suzuki-Miyaura cross-coupling reaction, which involves
the coupling of an organoboron compound with an elec-
trophile, possesses notable advantages over the other
related techniques. Particularly striking is the array of
functionality tolerated in the coupling process. The use
of organoboron compounds is also valued because the
inorganic byproducts of the reaction are nontoxic and can
be readily removed by simple workup procedures, while
many tin compounds are toxic, and complete removal of
tin-containing byproducts is a well-recognized problem.
From the many parameters that can be changed to
ensure success in the Suzuki-Miyaura reaction, the
(5) (a) Kabalka, G. W.; Pagni, R. M.; Hair, C. M. Org. Lett. 1999, 1,
1423-1425. (b) Kabalka, G. W.; Namboodiri, V.; Wang, L. Chem.
Commun. 2001, 775. (c) LeBlond, C. R.; Andrews, A. T.; Sun, Y.; Sowa,
J . R., J r. Org. Lett. 2001, 3, 1555-1557. (d) Sakurai, H.; Tsukuda, T.;
Hirao, T. J . Org. Chem. 2002, 67, 2721-2722. (e) Campi, E. M.;
J ackson, W. R.; Marcuccio, S. M.; Naeslund, C. G. M. Chem. Commun.
1994, 2395. (f) Darses, S.; J effery, T.; Geneˆt, J .-P.; Brayer, J .-L.;
Demoute, J .-P. Tetrahedron Lett. 1996, 37, 3857-3860. (g) Bumagin,
N. A.; Bykov, V. V. Tetrahedron 1997, 53, 14437-14450. (h) Badone,
D.; Baroni, M.; Cardamone, R.; Ielmini, A.; Guzzi, U. J . Org. Chem.
1997, 62, 7170-7173. (i) Goodson, E. F.; Wallow, T. I.; Novak, B. M.
Org. Synth. 1998, 75, 61-68. (j) Blettner, C. G.; Ko¨nig, W. A.; Stenzel,
W.; Schotten, T. J . Org. Chem. 1999, 64, 3885-3890. (k) Bussolari, J .
C.; Rehborn, D. C. Org. Lett. 1999, 1, 965-967. (l) Chi, S. M.; Choi,
J .-K.; Yum, E. K.; Chi, D. Y. Tetrahedron Lett. 2000, 41, 919-922. (m)
Zin, D.; Monteiro, A. L.; Dupont, J . Tetrahedron Lett. 2000, 41, 8199-
8202. (n) Ma, D.; Wu, Q. Tetrahedron Lett. 2001, 42, 5279-5281.
(6) (a) Grasa, G. A.; Viciu, M. S.; Huang, J .; Zhang, C.; Trudell, M.
L.; Nolan, S. P. Organometallics 2002, 21, 2866-2873 and references
therein. (b) Bei, X.; Turner, H. W.; Weinberg, W. H.; Guram, A. S. J .
Org. Chem. 1999, 64, 6797-6803. (c) Yin, J .; Rainka, M. P.; Zhang,
X.-X.; Buchwald, S. L. J . Am. Chem. Soc. 2002, 124, 1162-1163. (d)
Parrish, C. A.; Buchwald, S. L. J . Org. Chem. 2001, 66, 3820-3827.
(e) Feuerstein, M.; Berthiol, F.; Doucet, H.; Santelli, M. Synlett 2002,
1807-1810 and references therein.
(1) (a) Tsuji, J . Palladium Reagents and Catalysis; Wiley and Sons:
Chichester, UK, 1995. (b) Diederich, F.; Stang, P. J . Metal-Catalyzed
Cross-Coupling Reactions; VCH: Weinheim, Germany, 1998. (c) Cross-
Coupling Reactions. A Practical Guide. In Topics in Current Chemistry;
Miyaura, N., Vol. Ed.; Springer-Verlag: Berlin, Germany, 2002; Vol.
219.
(2) (a) Stille, J . K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508-
524. (b) Farina, V.; Krishnamurthy, V.; Scott, W. J . Org. React. 1997,
50, 1-652.
(3) For reviews see: (a) Miyaura, N.; Suzuki, A. Chem. Rev. 1995,
95, 2457-2483. (b) Suzuki, A. J . Organomet. Chem. 1999, 576, 147-
168. (c) Suzuki, A. In Metal-Catalyzed Cross-Coupling Reactions;
Diederich, F., Stang, P. J ., Eds.; VCH: Weinheim, Germany, 1998; pp
49-97. (d) Hassan, J .; Se´vignon, M.; Gozzi, C.; Schulz, E.; Lemaire,
M. Chem. Rev. 2002, 102, 1359-1469.
(4) (a) Giovannini, R.; Knochel, P. J . Am. Chem. Soc. 1998, 120,
11186-11187. (b) Lipshutz, B. H.; Blomgren, P. A. J . Am. Chem. Soc.
1999, 121, 5819-5820. (c) Piber, M.; J ensen, A. E.; Rottla¨nder, M.;
Knochel, P. Org. Lett. 1999, 1, 1323-1326. (d) Lipshutz, B. H.;
Blomgren, P. A.; Kim, S.-K. Tetrahedron Lett. 1999, 40, 197-200. (e)
Dai, C.; Fu, G. C. J . Am. Chem. Soc. 2001, 123, 2719-2724.
(7) (a) Li, G. Y.; Zheng, G.; Noonan, A. F. J . Org. Chem. 2001, 66,
8677-8681. (b) Kataoka, N.; Shelby, Q.; Stambuli, J . P.; Hartwig, J .
F. J . Org. Chem. 2002, 67, 5553-5566. (c) Liu, S.-Y.; Choi, M. J .; Fu,
G. C. Chem. Commun. 2001, 2408-2409. (d) Uozumi, Y.; Nakai, Y.
Org. Lett. 2002, 4, 2997-3000 and references therein.
10.1021/jo0342368 CCC: $25.00 © 2003 American Chemical Society
Published on Web 05/03/2003
4302
J . Org. Chem. 2003, 68, 4302-4314

Reactivity of Potassium Aryl- and Heteroaryltrifluoroborates
acids are subject to cyclic trimerization, and the resulting
uncertainties in stoichiometry require an excess of these
compounds in the coupling reactions. The use of boronic
esters can partially solve that problem, but this leads to
a lack of atom economy and additional (problematic)
purification steps, especially when catechol or pinacol are
used as the alcohol moiety. Additionally, there are some
relatively important compounds (including many het-
eroaryl derivatives) that do not have an adequate shelf
life or cannot be adequately purified. Finally, many
arylboronic acids are prone to protodeboronation during
the coupling process. This is particularly true for het-
eroarylboronic acids bearing the boron atom ortho to the
heteroatom.9 It is also known that electron-donating
groups in the aryl unit increase the rate of protodebor-
onation, while electron-withdrawing groups decrease the
rate.9 Although the higher reactivity of the former
organoboronic acids aids in preventing the protodebor-
onation from competing with the cross-coupling process,
a high percentage of protodeboronated compound can still
be produced if the cross-coupling is too slow, and in those
cases homocoupling may even become a competing pro-
cess.10 On the other hand, electron-poor arylboronic acids
have a higher tendency to homocouple.11 This, combined
with lower reactivity in the transmetalation process,
leads to lower yields in the cross-coupling, impeding the
use of many interesting compounds in the Suzuki reac-
tion. Therefore, the use of a more reactive/more stable
organoboron system appears essential.
with assurance on small scale. Additionally, the organ-
otrifluoroborates are environmentally friendly. The byprod-
ucts of these reactions convert to harmless inorganic salts
that are readily separable from the desired products. The
trifluoroborates possess a relatively low molecular weight,
so that the mass of material required is kept to a
minimum. Finally, water may be used as a cosolvent or
solvent, minimizing, or in some cases completely elimi-
nating, the need for organic solvent. No protecting groups
are needed for organotrifluoroborate substrates bearing
ketones, esters, amides, amines, nitriles, nitro groups,
halides, alcohols, aldehydes, or carboxylic acids. Finally,
the trifluoroborate system appears less subject to pro-
todeboronation than other boron derivatives.
In 1996, Geneˆt reported the first coupling reactions
involving potassium aryl- and alkenyltrifluoroborates
with arenediazonium tetrafluoroborates.13 In 1999, Xia
applied this method to carry out the coupling with
diaryliodonium salts14 as the electrophilic partners. These
reactions could be performed in the presence of halogen
functionalities on the substrates because no base was
required, and under these conditions aryl halides do not
couple.8b Our discovery that added base was required to
allow coupling with aryl halides and triflates led us to
conduct an exhaustive investigation in the synthesis and
application of potassium aryl and heteroaryl trifluorobo-
rates in Suzuki-Miyaura cross-coupling reactions with
an array of aryl and heteroaryl halides.
In a previous communication,15 we reported the ef-
ficient ligandless palladium-catalyzed cross-coupling re-
actions of some potassium aryltrifluoroborates with aryl
and heteroaryl bromides. The cross-couplings of several
potassium aryltrifluoroborates with simple aryl halides
were found to take place efficiently under extremely
economical conditions, using a low loading of a ligandless
Pd(OAc)2, in alcoholic or aqueous solvents in the air.
Thus, these reactions turned out to be highly atom
economic, simple to carry out, and insensitive to oxygen.
During the course of our investigation, Batey and Quach16
reported the cross-coupling reactions of tetrabutylam-
monium organotrifluoroborates, using 5 mol % of Pd-
(OAc)2 and 5 mol % of dppb as a catalyst in the presence
of 2 equiv of Cs2CO3 in DME/water (1:1). The reactions
usually took between 12 and 24 h. Herein we describe
the advances and scope of the coupling process involving
a variety of aryl- and heteroaryltrifluoroborates with aryl
and heteroaryl halides under conditions with and without
ligands.
Among the most promising alternative boron reagents
are potassium trifluoroborates. These are easily prepared
from organoboronic acids or esters by treatment with an
aqueous solution of KHF2 (eq 1).12 The potassium orga-
notrifluoroborates are monomeric solids, easily prepared
in quantities in excess of 200 g. They are in general
indefinitely stable in the air. They can thus be stored on
the shelf and/or sold commercially. This not only broad-
ens their availability, but also makes them more useful
for employment in combinatorial chemistry and in mul-
tistep syntheses in which a Suzuki-Miyaura coupling to
a valuable partner comprises a key transformation, or
when the organoboron intermediate itself is a very
valuable component. In either case, the critical organo-
trifluoroborate can be isolated and purified, and optimi-
zation of the strategic step can be carried out easily and
Resu lts a n d Discu ssion
We initiated our study of the Suzuki-Miyaura cross-
coupling reaction by optimizing the conditions in terms
of catalysts, ligands, bases, and solvents for the reaction
between potassium phenyltrifluoroborate (1a ) and 1-bro-
monaphthalene (2a ) (eq 2).
(8) (a) Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 1999, 37,
3387-3388. (b) Littke, A. F.; Chaoyang, D.; Fu, G. C. J . Am. Chem.
Soc. 2000, 122, 4020-4028.
(9) (a) Roques, B. P.; Florentin, D.; Callanquin, M. J . Heterocycl.
Chem. 1975, 12, 195-196. (b) Florentin, D.; Fournie´-Zaluski, M. C.;
Callanquin, M.; Roques, B. P. J . Heterocycl. Chem. 1976, 13, 1265-
1272. (c) Kuivila, H. G.; Nahabedian, K. V. J . Am. Chem. Soc. 1961,
83, 2159-2163.
The study was carried out in a reactor block, using 0.25
mmol of each reagent in the presence of 5 mol % of
(13) (a) Darses, S.; J effery, T.; Brayer, J .-L.; Demoute, J .-P.; Geneˆt,
J .-P. Bull. Soc. Chim Fr. 1996, 133, 1095-1102. (b) Darses, S.; Brayer,
J .-L.; Demoute, J .-P.; Geneˆt, J .-P. Tetrahedron Lett. 1997, 38, 4393-
4396. (c) Darses, S.; Michaud, G.; Geneˆt, J .-P. Eur. J . Org. Chem. 1999,
1875-1883.
(14) Chen, Z.-C.; Xia, M. Synth. Commun. 1999, 29, 2457-2465.
(15) Molander, G. A.; Biolatto, B. Org. Lett. 2002, 4, 1867-1870.
(16) Batey, R. A.; Quach, T. D. Tetrahedron Lett. 2001, 42, 9099-
9103.
(10) Moreno-Man˜as, M.; Pe´rez, M.; Pleixats, R. J . Org. Chem. 1996,
61, 2346-2351 and references therein.
(11) Wong, M. S.; Zhang, X. L. Tetrahedron Lett. 2001, 42, 4087-
4089.
(12) (a) Vedejs, E.; Chapman, R. W.; Fields, S. C.; Lin, S.; Schrimpf,
M. R. J . Org. Chem. 1995, 60, 3020-3027. (b) Vedejs, E.; Fields, S. C.;
Hayashi, R.; Hitchcock, S. R.; Powell, D. R.; Schrimpf, M. R. J . Am.
Chem. Soc. 1999, 121, 2460-2470.
J . Org. Chem, Vol. 68, No. 11, 2003 4303

Molander and Biolatto
A reduction in the amount of base (either amine or
K2CO3) to 2 equiv lowered the yield substantially, or
required longer reaction times to achieve similar yields
(compare entries 11 and 14). Different solvent mixtures
were also analyzed. The use of 1:1 ethanol/water mixtures
led to slightly lower yields under both ligand-added and
ligandless conditions (compare entries 3 and 8). Less
polar mixtures, such as 20:1 methanol/THF or ethanol/
THF (entries 16 and 17), led to even lower yield or cross-
coupled product and a higher recovery of unreacted aryl
bromide. This reveals a reduced tendency for protodeha-
logenation under less polar conditions, but a concomitant
retardation of the cross-coupling process as well.
Upon choosing the ligandless conditions [0.5 mol % of
Pd(OAc)2, 3 equiv of K2CO3 in methanol] to perform the
cross-coupling reaction, electrophiles incorporating a full
palette of functionalities were analyzed in an effort to
display compatibility with the process developed. At the
outset, the reaction was tested by using aryl bromides
possessing electron-withdrawing groups. The results are
shown in Table 2.
All of the reactions led to very high yields of cross-
coupling product, with reaction times between 45 min
and 2 h. The dihaloarene in entry 7 showed reduced yield.
This may be ascribed in part to protodehalogenation on
both bromine and chlorine positions, because only traces
of triphenyl due to double coupling were detected. A
variety of functional groups were tolerated, even in the
ortho position. Interestingly, the reaction involving 4-bro-
mobenzoic acid, 2i (entry 8), can be performed in water
at room temperature and is more efficient in that solvent
than methanol. (The reaction in water could be performed
at ambient temperature, while that in methanol required
heating at reflux. This is perhaps an issue of solubility.5g)
When coupling potassium 1-phenytrifluoroborate (1a )
with electron-rich bromides (Table 3), unexpected results
were obtained. 4-Bromoanisole (2j) reacted very well
under ligandless conditions, providing 95% of product
after 2 h of reaction (Table 3, entry 1). However, when a
second methoxy group was present in the ortho position
of the aryl halide 2n (entry 5), both electronic and steric
effects conspired to require the presence of a ligand to
achieve high yields of the cross-coupling product and
minimize the homocoupling of the aryl bromide. The use
of a simple ligand such as PPh3 in a 1:1 ratio with respect
to the catalyst proved to be sufficient, providing higher
yields than previously reported procedures using more
complex ligands.19,20
In terms of the preferred solvent for coupling, 4-bro-
mophenol (2k ) behaved the same as benzoic acid deriva-
tive 2i: in both cases the reaction is more efficient in
water than in methanol. However, the difference in yields
between methanolic and aqueous conditions was more
pronounced. Only under aqueous conditions did the
reaction proceed with good yields (Table 3, entry 2).5g For
halides 2l and 2m (entries 3 and 4), the amide or amine
themselves may be acting as ligands, presumably lower-
ing the activity of the catalyst in the catalytic cycle.21 As
expected, the N-acetyl derivative 2l was more reactive
catalyst/ligand, 3 equiv of base, and 2 mL of solvent
heating at reflux overnight under a blanket of nitrogen
(Table 1). The reactions were analyzed by quantitative
gas chromatography, determining the presence of 1-phe-
nylnaphthalene (3a ), unreacted 2a , as well as the amount
of binaphthyl and biphenyl formed by homocoupling,
using the internal standard method (hexadecane used as
standard). Naphthalene formed by reduction of 2a (pro-
todehalogenation) was determined by difference.
Considering commercially available catalysts or those
reported in the literature previously determined to be
successful in the Suzuki coupling of arylboronic acids,
Pd(OAc)2, Pd2(dba)3, and PdCl2(dppf)‚CH2Cl2 were chosen
and combined with ligands such as 1,1′-bis(diphenylphos-
phino)ferrocene (dppf) or t-Bu3P. As bases, Et3N, i-Pr2-
NEt (Hu¨nig’s base), t-BuNH2, K2CO3, Cs2CO3, KF, and
K3PO4 were used. Environmentally friendly solvent
systems were preferred (water, MeOH, EtOH, n-PrOH,
i-PrOH, dioxane, THF).
In general, optimal conditions were found to involve
the use of alcohols as solvents. This can be attributed to
the higher solubility of the aryltrifluoroborate salts in
methanol and ethanol at refluxing temperatures, which
allows a homogeneous reaction. The best system, as for
the coupling of alkenyltrifluoroborates,17 involved the use
of PdCl2(dppf)‚CH2Cl2 and Hu¨nig’s base in ethanol as
solvent (Table 1, entry 1). Auspiciously, ligandless condi-
tions were found to be quite efficient in the coupling
process (Table 1, entry 10).
Surprisingly, the reaction carried out in the presence
of PdCl2(dppf)‚CH2Cl2 and Cs2CO3 in THF/water (entry
9) (used in the coupling of other organotrifluoroborates)18
led to only 12% yield. A system similar to the one used
by Fu8 to perform the coupling with aryl chlorides [Pd2-
(dba)3/(t-Bu3P)] gave only moderate yields (entry 19).
In optimizing the two most efficient conditions [using
PdCl2(dppf)‚CH2Cl2 and Pd(OAc)2] in terms of catalyst
loading and reaction time, the advantages of the ligan-
dless condition became evident. The reaction reached
completion in shorter reaction times with only a slight
decrease in the yield. A reaction scaled up to 15 mmol
(entry 13) led to an 80% isolated yield of 3a and 14% of
unreacted 1-bromonaphthalene (2a ), using 0.2 mol % of
catalyst loading after heating at reflux for 9 h. To reduce
the cost of the reaction, other amines (Et3N and t-BuNH2)
and even K2CO3 could be used instead of Hu¨nig’s base
under ligand-added conditions (entries 4-6) with similar
results. Under both ligand and ligandless conditions, the
reactions could be performed in the air without a reduc-
tion in the yield of the cross-coupled biaryls.
(17) (a) Molander, G. A.; Rodriguez Rivero, M. Org. Lett. 2002, 4,
107-109. (b) Molander, G. A.; Bernardi, C. R. J . Org. Chem. 2002, 67,
8424-8429.
(18) (a) Molander, G. A.; Ito, T. Org. Lett. 2001, 3, 393-396. (b)
Molander G. A.; Katona, B. W.; Machrouhi, F. J . Org. Chem. 2002,
67, 8416-8423.
(19) Feuerstein, M.; Laurenti, D.; Bougenat, C.; Doucet, H.; Santelli,
M. Chem. Commun. 2001, 325-326.
(20) The reaction between the corresponding diazonium salt and
phenyltrifluoroborate was performed under ligandless conditions with
similar results. See ref 13a.
4304 J . Org. Chem., Vol. 68, No. 11, 2003

Reactivity of Potassium Aryl- and Heteroaryltrifluoroborates
TABLE 1. Op tim a l Com bin a tion s of Ca ta lyst, Ba se, a n d Solven t for th e Rea ction betw een P ota ssiu m
P h en yltr iflu or obor a te (1a ) a n d 1-Br om on a p h th a len e (2a )
reaction conditionsa
% yield 3a ;
entry
catalyst/ligand
solvent
time
base
% unreacted 2a b
1
2
3
4
5
5 mol % of PdCl2(dppf)‚CH2Cl2
(2 mol %)
(1 mol %)
(5 mol %)
(1 mol %)
EtOH
12 h
6 h
7 h
12 h
12 h
6 h
i-Pr2NEt
97; 0.8
95; 0.6
99; -
90; 8
82; 11
78; 8
Et3N
t-BuNH2
K2CO3
6
(1 mol %)
7
8
9
(5 mol %)
(1 mol %)
(5 mol %)
5 mol % of Pd(OAc)2
(2 mol %)
(0.5 mol %)
(0.2 mol %)
(2 mol %)
(2 mol %)
THF/H2O 10:1
EtOH/H2O 1:1
THF/H2O 10:1
MeOH
12 h
6 h
12 h
12 h
2 h
2 h
9 h
7 h
6 h
6 h
3 h
12 h
12 h
i-Pr2NEt
i-Pr2NEt
Cs2CO3
K2CO3
79; 6
90; 3
12; 81
84; 2
10
11
12
13
14
15
16
17
18
19
90 (88)c
(75)c
80; 14c,d
80; 15e
85; 13
78; 13
70; 29
79; 12
51; 3
EtOH/H2O 1:1
(0.5 mol %)
(0.5 mol %)
MeOH/THF 20:1
EtOH/THF 20:1
H2O + 1 equiv of Bu4N+Br-
dioxane
(5 mol %)
K2CO3
i-Pr2NEt
5 mol % of Pd2(dba)3/(t-Bu3P)f
a
Conditions: 0.25 mmol of aryltrifluoroborate 1a and aryl bromide 2a are reacted in the presence of catalyst/ligand, 3 equiv of base
b
and 2 mL of solvent, heating at reflux overnight under nitrogen. The reactions were analyzed by GC with hexadecane as an internal
standard. c After purification by chromatography on silica gel. Reaction performed on a 15-mmol scale. e Two equivalents of base were
d
used. f 1.2:1 P:Pd ratio.
than dimethylamine 2m , leading to higher yields, al-
though incomplete conversion,22 under ligandless condi-
tions (entry 3). In both cases, the use of water as a solvent
proved to be detrimental. For the amine-substituted aryl
bromide 2m , as in the case of 2n , the addition of PPh3
was necessary to achieve good yields of product (Table
3, entry 4).19
ligandless conditions (Table 4, entry 3). Previously,
couplings involving the corresponding boronic acids or
esters under ligandless conditions had been performed
in the presence of Bu4N+Br- in water,5k by applying
microwave irradiation,5j,23 or with aryl iodides to obtain
similar yields.5b,e,g,23
In applying these conditions to the coupling of other,
more electron-deficient trifluoroborates (Table 4, entries
4-8), a high percentage of trifluoroborate homocoupling
was observed.24 The use of the in situ formed catalyst
from Pd(OAc)2/(PPh3) allowed improvement of the yields
for the coupling of 4-acetylphenyltrifluoroborate (1e)
(Table 4, entry 4). However, for the aryltrifluoroborates
bearing the 3-nitro, 3,5-bis(trifluoromethyl), and 2,6-
difluoro substituents (1f, 1g, and 1h , respectively), only
the use of PdCl2(dppf)‚CH2Cl2 with triethylamine or K2-
CO3 in ethanol allowed acquisition of the coupling
products in high yields (entries 5-7). These results are
comparable to those achieved previously with the corre-
sponding arylboronic acids or esters and highly air-
sensitive Pd(PPh3)4, and better than those for 2,6-
difluorophenylboronic acid, which has been reported to
fail to couple under different conditions.25
After performing a number of reactions, it became
evident that the high yields and the nature of the
byproducts allowed very simple workup techniques to be
applied. In many cases, after the reactions were cooled
and water was added, the products precipitated directly
from the aqueous solution (after acidification in the case
of acidic substrates). In other cases, the reaction was
subjected to a standard aqueous workup. After concen-
tration, the product could be purified either by recrys-
tallization or by being passed through a short pad of silica
gel eluting with pentane or pentane-dichloromethane
mixtures to obtain the desired product (boronated and
catalyst impurities were retained in the silica gel). These
two convenient procedures allowed the products to be
obtained in most cases with 95% or higher purity.
The reactions of electron-rich aryltrifluoroborates 1b
and 1c with 4-bromobenzonitrile (2e) (entries 1 and 2,
Table 4) proceeded with very good yields. The combina-
tion of electron-poor aryl bromide and electron-rich
trifluoroborate was the most favorable, and therefore
there was no need for special ligands for the coupling.
Even the slightly electron-poor 4-fluorophenyltrifluorobo-
rate (1d ) gave nearly quantitative yields of product under
During the preparation of this paper, Frohn et al.26
reported the high-yielding coupling of pentafluorophe-
nyltrifluoroborate (1i). The conditions developed required
the use of 10 mol % of Pd(OAc)2 and 20 mol % of PPh3 in
toluene at 100 °C for 3 h, with the addition of 2 equiv of
K2CO3 and 1.2 equiv of Ag2O, to afford 76-93% for the
reactions with electron-poor aryl iodides. Frohn tested
(21) Ligands containing amine moieties have proved to provide
active catalysts for Suzuki couplings when a phosphine group is also
present in the molecule: Yin, J .; Buchwald, S. L. J . Am. Chem. Soc.
2000, 122, 12051-12052. Other N-containing ligands, for example, bis-
(pyrimidyl)-based palladium complexes, do not show a high efficiency
in Suzuki couplings: Buchmeiser, M. R.; Schareina, T.; Kempe, R.;
Wurst, K. J . Organomet. Chem. 2001, 634, 39-46.
(23) Villemin, D.; Caillot, F. Tetrahedron Lett. 2001, 42, 639-642.
(24) Conditions involving nonpolar solvents, open atmosphere, phos-
phine or phosphite added, or ligandless Pd(OAc)2 have been used to
perform the homocoupling of both electron-poor and electron-rich
arylboronic acids. See ref 11.
(25) Thiemann, T.; Umeno, K.; Ohira, D.; Inohae, E.; Sawada, T.;
Mataka, S. New J . Chem. 1999, 23, 1067-1070.
(26) Frohn, H.-J .; Adonin, N. Y.; Bardin, V. V.; Starichenko, V. F.
Tetrahedron Lett. 2002, 43, 8111-8114.
(22) The ligandless coupling reaction in water requires the addition
of 1 equiv of Bu4NBr to produce quantitative yields. See ref 5f.
J . Org. Chem, Vol. 68, No. 11, 2003 4305

Molander and Biolatto
TABLE 2. Cr oss-Cou p lin g Rea ction s of P ota ssiu m
P h en yltr iflu or obor a te (1a ) w ith Electr on -P oor Ar yl
Ha lid es
TABLE 3. Cr oss-Cou p lin g Rea ction s of P ota ssiu m
P h en yltr iflu or obor a te (1a ) w ith Electr on -Rich Ar yl
Ha lid es
a
Condition A: Pd(OAc)2 (0.5 mol %), K2CO3 (3 equiv), MeOH,
reflux. Condition B: Pd(OAc)2 (0.5 mol %), K2CO3 (3 equiv), water,
65 °C. Condition C: Pd(OAc)2 (0.5 mol %), Ph3P (0.5 mol %), K2CO3
b
(3 equiv), MeOH, reflux. Determined by NMR analysis.
out with dry THF, leading to a 38% yield after 36 h (entry
9). To establish the difference in reactivity between the
trifluoroborate salt and the corresponding boronic acid,
the reaction involving pentafluorophenylboronic acid was
carried out under the same conditions. Analysis by GC-
MS after 36 h showed the presence of only traces of
product, with a high proportion of unreacted 4-bro-
mobenzonitrile (2e). This is the clearest indication yet
of the enhanced suitability of the organotrifluoroborate
system compared to the corresponding boronic acids in
many coupling reactions.
We next examined ortho-substituted coupling partners
to assess steric effects in the process (Table 5). Under
ligandless conditions, 1-naphthyltrifluoroborate (1j) pro-
duced a 60% yield of product in the reaction with 2e after
9 h (90% conversion by GC analysis), revealing the
increased steric hindrance of the naphthalene ring (Table
5, entry 1). The presence of only one methyl group in the
ortho position of the trifluoroborate 1k also permitted the
use of ligandless conditions for the couplings (Table 5,
entries 2-4), although the reaction with electron-rich
4-bromoanisole (2j) or hindered o-bromobenzaldehyde (2f)
required 12-21 h to reach completion. When two methyls
were situated ortho to the trifluoroborate group [2,6-
dimethylphenyltrifluoroborate (1l), entries 5 and 6], the
reaction required more time and did not reach completion
under ligandless conditions. The use of Pd(OAc)2/(PPh3)
a
Condition A: Pd(OAc)2 (0.5 mol %), K2CO3 (3 equiv), MeOH,
reflux, under N2. Condition B: Pd(OAc)2 (0.5 mol %), K2CO3 (3
b
equiv), water, 65 °C. Reaction carried out in open atmosphere.
c Reaction carried out at room temperature. Determined by NMR
d
analysis.
some aryl bromides as well. For example, 4-bromoni-
trobenzene (2b) produced a 62% yield after 24 h. When
we performed the coupling of 1i with 4-bromobenzonitrile
(2e) under condition D using 3 equiv of K2CO3 or Et3N
as base, the reaction led to no product, with a high
percentage of homocoupling of the 4-bromobenzonitrile
(2e) after 36 h. Neither the use of Pd(PPh3)4 and K2CO3
in toluene/ethanol/water 10:10:2 nor aprotic conditions
[PdCl2(dppf)‚CH2Cl2, K2CO3, DMF, 90 °C, 24 h] produced
the expected product, but rather homocoupling of 2e and
an increased amount of hydrolyzed 4-bromobenzonitrile
(under the protic conditions) were observed. When a
mixture of methanol/THF 5:1 was used as a solvent in
the presence of 5 mol % of PdCl2(dppf)‚CH2Cl2 and (i-
Pr)2NEt as a base, after 36 h 10% of product was detected
by GC-MS analysis (entry 8). The crucial role played by
solvent was then evident, and the reaction was carried
4306 J . Org. Chem., Vol. 68, No. 11, 2003

Reactivity of Potassium Aryl- and Heteroaryltrifluoroborates
TABLE 4. Cr oss-Cou p lin g Rea ction s of Electr on -Rich
a n d Electr on -P oor P ota ssiu m Ar yltr iflu or obor a tes w ith
4-Br om oben zon itr ile (2e)
Cl2 improved the yields only when used in a mixture of
MeOH/THF 5:1, which again suggested the important
role of the solvent in the reaction. Because the oxidative
addition step was impeded, the use of the more reactive
4-methoxyphenyltrifluoroborate (1b) under ligandless
conditions did not lead to high yields of product but
rather a high proportion of homocoupling of the trifluo-
roborate after long reaction times. In this case, the use
of Pd(OAc)2/(PPh3) allowed moderate yields to be achieved
(entry 8). Surprisingly, the use of PdCl2(dppf)‚CH2Cl2 did
not improve the yields. Batey and Quach had previously
used the tetrabutylammonium salt of the phenyltrifluo-
roborate in the presence of 5 mol % of Pd(OAc)2(dppb),
Cs2CO3, and DME/water 1:1 to obtain an 85% yield of
3a c after 24 h at 50 °C.16 In that case the Bu4N+ cation
may not only stabilize the anionic palladium species,3d,5m,27
but may also act as a phase-transfer catalyst for the
solubilization of both reagents in the aqueous-organic
medium. We also observed that after similar reaction
times, the reactions of 2o in methanol/THF led to less
protodehalogenation of the aryl halide, and consequently
higher yields were realized. To achieve good yields in
more hindered situations (e.g., steric hindrance in both
the trifluoroborate and the bromide) it might be neces-
sary to optimize the reaction further by using other
ligands and/or bases under aprotic conditions that have
been demonstrated to be efficient in similar situations
with arylboronic acids.6c,8b,28
Considering the preliminary results obtained in the
ligandless couplings of phenyltrifluoroborate (1a ) with
2-acetyl-5-bromothiophene (2s) (Table 6, entry 4) and the
coupling between 3-thiophenetrifluoroborate (1m ) and
4-bromobenzonitrile (2e) (eq 3),15 we intended to deter-
mine the scope of the ligandless protocol in couplings
involving heteroaryl halides and heteroaryltrifluorobo-
rates, as well as the reactivity of the heteroaryltrifluo-
roborates in the Suzuki couplings.
a
Condition A: Pd(OAc)2 (0.5 mol %), K2CO3 (3 equiv), MeOH,
reflux. Condition C: Pd(OAc)2 (0.5 mol %), Ph3P (0.5 mol %), K2CO3
(3 equiv), MeOH, reflux. Condition D: PdCl2(dppf)‚CH2Cl2 (0.5 mol
Initially, electron-deficient heteroaryl bromides were
chosen. When 1-phenyltrifluoroborate (1a ) was reacted
with 2-bromopyridine (2p ) in the presence of 1 mol % of
Pd(OAc)2, the reaction did not reach completion even
after 24 h. It was then observed (running a 0.1-mmol
reaction in the reactor block) that after 5 h the reaction
halted after 90% conversion. Only 70% of product was
recovered after workup and purification by chromatog-
raphy29 [the reaction was run on a 0.5-mmol scale, with
10% of unreacted bromopyridine (by GC analysis) also
present] (Table 6, entry 1). The use of longer reaction
times produced the homocoupling product of the remain-
b
%), Et3N (3 equiv), EtOH, reflux. Detected by GC-MS analysis.
c i-Pr2NEt used as base. Reaction carried out under inert atmo-
d
sphere.
in 1:1 or 1:2 ratio, which normally allowed an improve-
ment in the oxidative addition of electron-rich aryl
bromides, was deleterious in these couplings because it
increased the amount of aryl bromide homocoupling.
Finally, PdCl2(dppf)‚CH2Cl2 produced good yields in
couplings with 4-bromobenzonitrile (2e) after 16 h (entry
5), but only partial conversion for the coupling with
4-bromoanisole (2j) after 21 h of reaction (entry 6).
To evaluate the steric hindrance on the halide partner,
2-bromomesitylene (2o) was reacted with phenyltrifluo-
roborate (1a ) under ligandless conditions. The reaction
afforded moderate yields after very long reaction times
and high catalyst loading. The use of PdCl2(dppf)‚CH2-
(27) Amatore, C.; J utand, A. Acc. Chem. Res. 2000, 33, 314-321.
(28) (a) Watanabe, T.; Miyaura, N.; Suzuki, A. Synlett 1992, 207-
210. (b) Castanet, A.-S.; Colobert, F.; Broutin, P.-E.; Obringer, M.
Tetrahedron: Asymmetry 2002, 13, 659-665 and references therein.
(29) Colacot, T. J .; Gore, E. S.; Kuber, A. Organometallics 2002, 21,
3301-3304.
J . Org. Chem, Vol. 68, No. 11, 2003 4307

Molander and Biolatto
TABLE 5. Cr oss-Cou p lin g Rea ction s of Ster ica lly Hin d er ed P ota ssiu m Ar yltr iflu or obor a tes a n d Ar yl Ha lid es
a
Condition A: Pd(OAc)2 (0.5 mol %), K2CO3 (3 equiv), MeOH, reflux. Condition C: Pd(OAc)2 (0.5 mol %), Ph3P (0.5 mol %), K2CO3 (3
equiv), MeOH, reflux. Condition D: PdCl2(dppf)‚CH2Cl2 (0.5 mol %), Et3N (3 equiv), EtOH, reflux.
ing 2-bromopyridine. Presumably, the capacity for 2-bro-
mopyridine to complex to the catalyst prevented comple-
tion of the coupling.21 When 3-bromopyridine (2q) was
used under similar conditions, complete conversion was
observed after 2 h, and a 90% yield of product was
obtained after workup and purification (Table 6, entry
2). The reaction with 5-bromopyrimidine (2r ), because it
is also an electron-poor heteroaryl bromide, proceeded
with very good yields after 5 h under ligandless condi-
tions (Table 6, entry 3).
reaction mixtures, thus providing capricious isolated
yields. The progress of the reactions was then monitored
1
by GC and H NMR to determine the relative reactivity
of the 2- and 3-halosubstituted substrates. It was ob-
served that 2-bromothiophene led to high conversion
(between 70 and 80%, with 5% unreacted 2v after 6 h),
while 3-bromothiophene (2w ) and 3-bromofuran (2x)
showed only traces of products under ligandless condi-
tions. In the case of 3-bromothiophene (2w ), even the use
of ligands such as PPh3, PCy3, or dppf did not afford more
than 50% conversion.27,30 When the more activated
thiophene 2s and furan 2t were used, the reactions took
place under ligandless conditions in relatively short
reaction times (Table 6, entries 4 and 5). Cross-couplings
The coupling products of 2-bromothiophene (2v), 3-bro-
mothiophene (2w ), and 3-bromofuran (2x) with 1-phe-
nyltrifluoroborate (1a ) sublime at room temperature,
which led initially to low recoveries of products from the
4308 J . Org. Chem., Vol. 68, No. 11, 2003

Reactivity of Potassium Aryl- and Heteroaryltrifluoroborates
TABLE 6. Cr oss-Cou p lin g Rea ction s of P ota ssiu m
P h en yltr iflu or obor a te (1a ) a n d Heter oa r yl Ha lid es
TABLE 7. Cr oss-Cou p lin g Rea ction s of P ota ssiu m
1-Na p h th yltr iflu or obor a te (1j) a n d Heter oa r yl Ha lid es
a
Condition D: PdCl2(dppf)‚CH2Cl2 (0.5 mol %), Et3N (3 equiv),
b
EtOH, reflux. K2CO3 used as a base.
a
Condition A: Pd(OAc)2 (0.5 mol %), K2CO3 (3 equiv), MeOH,
Cl2 allowed good-to-high conversion to products (Table
7, entries 1, 2, and 4, respectively). The same trend in
the reactivity between substitution in the 2- and 3-posi-
tions was observed. The reaction with 3-bromofuran (2x)
had to be carried out in the dark (using an excess of 2x),
due to the tendency of both 2x and the coupling product
3a n to photodecompose. Under these conditions, a 68%
yield of a clear oil was obtained (Table 7, entry 3), which
quickly turned yellow even while kept in the dark. In all
these reactions, the formation of the homocoupled prod-
uct from the trifluoroborate, and even from the heteroaryl
bromide, was observed, which greatly complicated the
purification of the cross-coupling products and reduced
the isolated yield to some extent.
reflux. Condition B: Pd(OAc)2 (0.5 mol %), K2CO3 (3 equiv), water,
65 °C.
with thiophene and furan derivatives have been studied
by Bussolari and Rehborn5k with use of ligandless condi-
tions in water as a solvent. By adding 1 equiv of
tetrabutylammonium bromide those reactions could be
carried out at room temperature in shorter times. In our
studies, only the reaction of 1a with 5-bromo-2-furoic acid
(2t) can be carried out in water under ligandless condi-
tions, affording 67% of product 3a j (Table 6, entry 5),
without addition of tetrabutylammonium bromide. As
expected, the reaction in methanol furnished only 49%
yield after 8 h.
The value of the trifluoroborates to improve the cross-
couplings is again exemplified in the reaction with
electron-deficient heteroaryl chlorides. When 2-chloro-
pyrazine (2z) was reacted with 1-naphthyltrifluoroborate
(1j), an 85% yield of the desired product 3a p was
obtained after 9 h (Table 7, entry 5). For comparison, the
reaction involving a similar arylboronic acid (2-methoxy-
1-naphthylboronic acid) and 3,6-dimethyl-2-chloropyra-
zine requires the use of 3 mol % of Pd(PPh3)4 in DME
(Na2CO3 as base) to obtain a 61% yield of product after
48 h (refluxing temperature),31 and the reaction with the
more activated 2-propoxyphenylboronic acid in the pres-
ence of 5 mol % of PdCl2(dppb) in toluene/ethanol/water
4:1:2 requires 24 h at reflux to yield 78% of product.32
The reaction with the unprotected 7-bromoindole (2u )
under ligandless conditions produced partial conversion
of the heteroaryl bromide. Only 44% of product 3a k was
isolated after purification by column chromatography
(Table 6, entry 6). Under the basic conditions the indole
may form the nitrogen anion. This not only increases the
electron density of the indole ring, but also may increase
complexation to the catalyst, thus reducing its activity
in the catalytic cycle.21
When more sterically hindered 1-naphthyltrifluorobo-
rate (1j) was coupled with thiophenes 2v and 2w , and
2-bromothiazole (2y), only the use of PdCl2(dppf)‚CH2-
(30) It has been shown that 3-bromothiophene requires special
ligands to afford good yields of coupling products. Feuerstein, M.;
Doucet, H.; Santelli, M. Tetrahedron Lett. 2001, 42, 5659-5662.
(31) McCarthy, M.; Gury, P. J . Tetrahedron 1999, 55, 3061-3070.
J . Org. Chem, Vol. 68, No. 11, 2003 4309

Molander and Biolatto
TABLE 8. Cr oss-Cou p lin g Rea ction s of P ota ssiu m
3-Th iop h en etr iflu or obor a te (1m ) a n d Heter oa r yl Ha lid es
F IGURE 1. Unprotected bromo azoles.
In further studies, potassium 3-thiophenetrifluorobo-
rate (1m ) was challenged with different heteroaryl
halides. Being an electron-rich heteroaromatic compound,
1m reacted under ligandless conditions with activated
(electron deficient) heteroaryl bromides such as pyridines
2p and 2q (Table 8, entries 1 and 2). The higher
reactivity of the 3-substituted pyridine 2q was evidenced
by the shorter reaction times needed to reach completion.
Surprisingly, the use of PdCl2(dppf)‚CH2Cl2 did not seem
to improve the results. Also 5-bromopyrimidine (2r )33
(Table 8, entry 3), and even thiophene 2s (Table 8, entry
5), can be coupled under ligandless conditions with
moderate to high yields. For heteroaryl halide 2r , partial
conversion was observed after 10 h, and the use of
PdCl2(dppf)‚CH2Cl2 provided a slightly higher yield of
product, without recovery of starting bromide 2r , after
longer reaction times (Table 8, entry 3). Unfortunately,
coupling involving water-soluble heteroaryl halides (such
as furan 2t or some pyridinecarboxylic acids) cannot be
performed in water, possibly because of decomposition
of the 3-thiophenetrifluoroborate (1m ) under these condi-
tions. Thus, the reaction with furancarboxylic acid 2t
requires the use of PdCl2(dppf)‚CH2Cl2 in ethanol to
afford 70% yield of product (Table 8, entry 7). Again, the
coupling with unactivated thiophene 2v and thiazole 2y
under ligandless conditions led to low-to-moderate yields,
and the use of PdCl2(dppf)‚CH2Cl2 is required (entries 4
and 8, respectively). For 3-bromofuran (2x), because of
its lower reactivity, the proportion of trifluoroborate
homocoupling is especially high. The tendency of the
product to photodecompose and sublime and the impos-
sibility of separation by chromatography forced us to
analyze the mixture only by GC-MS and 1H NMR,
revealing a poor yield of product 3a v (24%), with 16% of
3,3′-dithiophene (3a w ).34 Interestingly, the reaction with
the unprotected 7-bromoindole (2u ) under ligand-added
conditions produced a high yield of product (Table 8,
entry 9), showing that under basic conditions the unpro-
tected nitrogen cannot compete with the dppf ligand for
complexation of the catalyst (compare with Table 6,
entry 6). When 2-chloropyrazine (2z) was reacted with
3-thiophenetrifluoroborate (1m ), an 83% yield of the
desired product 3ba was obtained after 12 h (Table 8,
entry 10).
a
Condition A: Pd(OAc)2 (0.5 mol %), K2CO3 (3 equiv), MeOH,
reflux. Condition D: PdCl2(dppf)‚CH2Cl2 (0.5 mol %), Et3N (3
equiv), EtOH, reflux. b Determined by NMR analysis. c K2CO3 used
as a base.
In the case of these azoles not only complexation of the
nitrogen anion under the basic conditions35 but also the
more electron-rich nucleus may conspire to decrease the
rate of the oxidative addition. Protection of the nitrogen
with electron-withdrawing groups may allow the desired
coupling products to be accessed.
The reactions of 1m with 4-bromo-1H-imidazole (2a a )
and 3,5-dimethyl-4-bromo-1H-pyrazole (2a b) (Figure 1)
led to traces of products detected only by GC-MS analysis.
The selectivity in the cross-coupling between 3-thiophen-
etrifluoroborate (1m ) and 2,4,6-trichloropyrimidine (2a c)
(Figure 2) was analyzed under different conditions (Table
9). Assuming that the higher reactivity of the heteroaryl
(32) Ali, N. M.; McKillop, A.; Mitchell, M. B.; Rebelo, R. A.;
Wallbank, P. J . Tetrahedron 1992, 48, 8117-8126.
(33) The coupling with 3-thiopheneboronic acid has been performed
in the presence of 3 mol % of Pd(PPh3)4, Na2CO3, in DME-water,
affording 72% yield of product after heating at reflux overnight.
Gronowitz, S.; Hoernfeldt, A. B.; Kristjansson, V.; Musil, T. Chem. Scr.
1986, 26, 305-309.
(35) Different pyrazoles have been employed as nonchelating and
chelating ligands. Grotjahn, D. B.; Van, S.; Combs, D.; Lev, D. A.;
Schneider, C.; Rideout, M.; Meyer, C.; Hernandez, G.; Mejorado, L. J .
Org. Chem. 2002, 67, 9200-9209.
(34) The cross-coupling and homocoupling products are highly
sensitive to light and sublime at room temperature, accounting for the
low recovery of product.
4310 J . Org. Chem., Vol. 68, No. 11, 2003

Reactivity of Potassium Aryl- and Heteroaryltrifluoroborates
TABLE 10. Cr oss-Cou p lin g Rea ction s of P ota ssiu m
Heter oa r yltr iflu or obor a tes a n d 5-Br om op yr im id in e (2r )
FIGURE 2. Products of coupling between potassium 3-thiophe-
nyltrifluoroborate 1m and 2,4,6-trichloropyrimidine 2a c in
Table 9.
a
Condition D: PdCl2(dppf)‚CH2Cl2 (0.5 mol %), Et3N (3 equiv),
EtOH, reflux.
TABLE 9. Cr oss-Cou p lin g Rea ction s of P ota ssiu m
3-Th iop h en etr iflu or obor a te (1m ) a n d
2,4,6-Tr ich lor op yr im id in e (2a c)
uct 3bc was obtained, along with 10% of bicoupled
pyrimidine 3bd (Table 9, entry 2). The reaction with 2
equiv of trifluoroborate 1m and 6 equiv of base allowed
the isolation of 50% of the bicoupled product 3bd and
25% of tricoupled pyrimidine 3be (Table 9, entry 3), and
with 3 equiv of 1m and 9 equiv of base, the tricoupled
product 3be was accessed in 65%, with only 5% of
bicoupled product 3bd (Table 9, entry 4).37 It is thus
possible to couple a heteroaryl or aryl moiety selectively
to the trichloropyrimidine 2a c in the 4-position with use
of just 1 equiv of trifluoroborate. The selectivity might
be enhanced by the use of a less reactive trifluoroborate,
as well as a lower catalyst loading. These results are in
agreement with those reported by Schomaker and Delia38
in the coupling of phenylboronic acid and 2,4,6-trichlo-
ropyrimidine in the presence of 5 mol % of Pd(OAc)2-
(PPh3)2, using Na2CO3 as a base after 18 h in boiling
DME/water. It is also evident from comparison that the
heteroaryltrifluoroborate system has a higher reactivity
than phenylboronic acid, because it undergoes the cou-
pling process with a lower loading of catalyst in shorter
reaction times.
For the reactions of other heteroaryltrifluoroborates,
5-bromopyrimidine (2r ) was chosen as the coupling
partner (Table 10). Unlike cross-couplings with the
thiophenetrifluoroborate 1m , reactions involving the
furantrifluoroborate 1n and pyridinetrifluoroborate 1o
did not afford good yields of products under ligandless
conditions. In the case of 1n , the reaction in the presence
of Pd(OAc)2 led to incomplete conversion (23%) while for
1o, the reaction did not proceed at all. On the other hand,
the reaction with PdCl2(dppf)‚CH2Cl2 led to 72% yield of
product (Table 10, entry 1) for electron-rich trifluorobo-
a
Condition A: Pd(OAc)2 (2 mol %), K2CO3 (3 equiv), MeOH,
reflux. Condition D: PdCl2(dppf)‚CH2Cl2 (1 mol %), Et3N (3 equiv),
EtOH, reflux. Determined by GC-MS analysis.
b
halide would allow us to perform the reaction under
ligandless conditions, the coupling was carried out with
1 equiv of trifluoroborate and 3 equiv of K2CO3 in the
presence of 2 mol % of Pd(OAc)2. Surprisingly, the crude
reaction mixture showed a mixture of products arising
from mono-, di-, and trisubstitution of the chlorine with
methoxy groups,36 along with a high percentage of
disubstituted and monocoupled product 3bb (Table 9,
entry 1).
(36) Trichloropyrimidine is known to undergo nucleophilic substitu-
tion by phenolate ions. Delia, T. J .; Nagarajan, A. J . Heterocycl. Chem.
1998, 35, 269-273. Whether the palladium catalyst plays a role in
favoring this reaction under our conditions is not known.
(37) Several byproducts arising from substitution of the chlorine
atoms by ethoxy groups and/or coupling with the aryltrifluoroborates
were detected by GC-MS analysis, in trace levels, which accounts for
the conversion of the remaining 20-30% of halide 2a c, for entries 2-4,
Table 9.
When the reaction was performed in the presence of 1
mol % of PdCl2(dppf)‚CH2Cl2, 70% of monocoupled prod-
(38) Schomaker, J . M.; Delia, T. J . J . Org. Chem. 2001, 66, 7125-
7128.
J . Org. Chem, Vol. 68, No. 11, 2003 4311

Molander and Biolatto
of only 2 equiv of amine or K2CO3 for the ligand-added
and ligandless conditions, respectively, forced us to apply
longer reaction times to achieve similar yields. To
understand the role of the amount of base required, and
considering that a tetracoordinate species was involved
in the transmetalation step,13,43 several experiments were
performed. We conducted experiments heating PhBF3K
in methanol at reflux with the addition of 0, 1, 2, and 3
equiv of base. After 2 h, all of the reaction mixtures were
filtered, equal amounts of deuterated acetone was added
to each one, and the resulting solutions were then
analyzed by using 11B and 19F NMR. The results are
summarized in Table 11.
F IGURE 3. Potassium R-heteroaryltrifluoroborates.
rate 1n . The yield in this case again may be reduced
because of sublimation during the workup process. For
the electron-deficient 3-pyridyl derivative 1o, the reaction
afforded 64% of the desired product (entry 2). Previous
coupling reactions of the corresponding heteroarylboronic
acids or esters have been performed in general under
more severe conditions, such as using Pd(PPh3)4 with
moderate yields39 or with Pd(OAc)2(dppf) or Pd(OAc)2[P(o-
tolyl3)]2 in DMF at 100 °C.40
The addition of 3 equiv of K2CO3 led to the gradual
displacement of fluoride from the trifluoroborate salt,
eventually leading to a different “ate” complex containing
no fluorine substituents (Table 11, entry 4, 5.47 ppm
compared to entry 5, 28.48 ppm of the boronic acid, by
11B NMR; the signal at -149.29 ppm in the 19F NMR is
a sharp singlet, which corresponds to a fluorinated
species with no boron attached). On the basis of the shift
in the 11B NMR (4.35 ppm in entry 1 to 5.47 ppm in entry
4), the formation of a tetracoordinate species could be
postulated wherein all of the fluorides have been dis-
placed for less electron withdrawing groups. It has been
claimed that water as a cosolvent is necessary for the
cross-coupling reaction to proceed, and that one or more
hydroxyl groups in this case would be attached to the
boron.42 Because we found that water, methanol, or
ethanol in the presence of a base could be used in the
coupling process, the tetracoordinate species may arise
from the substitution of fluoride by either hydroxy,
Unfortunately, when the trifluoroborate group is R to
the heteroatom in the furans and pyridines (Figure 3),
the tendency for protodeboronation is higher than that
for the electron-poor aryltrifluoroborates previously ana-
lyzed and even for the corresponding boronic acids.
Furthermore, the synthesis of the corresponding trifluo-
roborates (1p and 1q) is complicated by solubility issues
and the fact that the products do not show an adequate
(extended) shelf life. Therefore, under protic conditions
the reactions of 1p and 1q led to no coupling products.
Furthermore, an extensive analysis of diverse combina-
tions of bases (Et3N, K2CO3, K3PO4, KOTf, CsF, NaOMe),
protic (methanol, ethanol/water, toluene/ethanol, DME/
water) and aprotic solvents (ethyl acetate, THF, dioxane),
and even different catalysts [Pd(OAc)2, PdCl2(dppf)‚CH2-
Cl2, Pd(PPh3)4] for the reaction between 2-pyridyltrifluo-
roborate (1q) and 4-bromobenzonitrile (2e) or 5-bromopy-
rimidine (2r ) led to no success. No high-yielding
cross-coupling reactions have been reported with 2-py-
ridylboronic acid. The only successful reported reaction
of this type involved the coupling between the corre-
sponding dimethyl boronic ester (5 equiv) and 1,4-
dibromobenzene in the presence of 7 mol % of Pd(PPh3)4
in dry DME, and this combination produced 64% of
bicoupled product after 20 h.41 Careful examination of
the experimental procedure for the preparation of this
boron compound42 shows that the actual species may be
the “ate” complex, because no acidic workup was applied
after the lithium-boron transmetalation step. Under
these conditions (having a preformed “ate” complex), no
base would be required for the coupling, and the trans-
metalation step could be accelerated, with a decreased
tendency to protodeboronate owing to the use of dry DME
as solvent.
-
alkoxy, or HCO3 groups. Two additional experiments
were then carried out: PhBF3K and PhB(OH)2 were
heated separately at reflux in wet deuterated methanol
in the presence of 3 equiv of base (entries 6 and 7). After
45 min, both reactions were analyzed by 1H, 13C, 19F, and
1
11B NMR. The experiment involving 1a by H and 13C
NMR showed the presence of a single species, which by
19F NMR showed the absence of fluorine attached to the
boron derivative (Table 11, entry 6), and by 11B NMR
corresponded to an “ate” complex. In no case was the
signal of the phenylboronic acid detected. It is then
possible to conclude that under our ligandless conditions,
the use of 3 equiv of base provides an “ate” complex from
the corresponding potassium organotrifluoroborate with
complete substitution of the fluoride by hydroxyl groups.
This species is the most predominant in the reaction,
although it does not exclude the possibility that mono-,
di-, or trifluorinated species could also be involved in the
catalytic cycle. Thus, the fluoride, which has been shown
to have a beneficial effect on some coupling processes,8b
plays a complicated role in these reactions.25 The fluoride
may be part of the counterion sphere, stabilizing various
catalytic intermediates, in addition to serving as a ligand
on the “ate” complex. Surprisingly, the experiment
involving phenylboronic acid showed the presence of the
Mech a n istic Stu d ies
In analyzing the optimal amount of base necessary to
carry out the coupling process, we observed that the use
(39) Gronowitz, S.; Peters, D. Heterocycles 1990, 30, 645-58.
(40) Under these conditions, the reactions with the more electron
deficient methyl 5-bromonicotinate afford high yields of products after
2-6 h. (a) Thompson, W. J .; Gaudino, J . J . Org. Chem. 1984, 49, 5237-
5243. (b) Thompson, W. J .; J ones, J . H.; Lyle, P. A.; Thies, J . E. J .
Org. Chem. 1988, 53, 2052-2055.
(41) Sindkhedkar, M. D.; Mulla, H. R.; Wurth, M. A.; Cammers-
Goodwin, A. Tetrahedron 2001, 57, 2991-2996.
(42) Bromopyridine in dry diethyl ether at -78 °C was treated with
1 equiv of n-BuLi, then with 2 equiv of B(OMe)3, and then allowed to
warm to room temperature overnight. The solvent was evaporated,
and volatile impurities were removed as an azeotrope with dry MeOH.
1
same single species by H, 13C, and 11B NMR (Table 11,
entry 7). It therefore appears that both boronated re-
agents undergo the formation of the same “ate” complex.
By contrast, the feasibility of performing the cross-
coupling reaction of potassium pentafluorophenyltrifluo-
(43) Matos, K.; Soderquist, J . A. J . Org. Chem. 1998, 63, 461-470.
4312 J . Org. Chem., Vol. 68, No. 11, 2003

Reactivity of Potassium Aryl- and Heteroaryltrifluoroborates
TABLE 11. Effect of th e Ad d ition of Ba se to P ota ssiu m P h en yltr iflu or obor a te (1a )
entry
1
experiment
19F NMR
11B NMR
4.35 (q)
a
1a + 0 equiv of K2CO3 in methanol/acetone-d6
-143.58 (q)
-151.51 (q)
-148.72 (br s)
-150.82 (q)
-147.9 (br s)
-150.38 (q)
-149.29 (s)
2
3
1a + 1 equiv of K2CO3 in methanol/acetone-d6
1a + 2 equiv of K2CO3 in methanol/acetone-d6
5.74 (br s)
5.59 (br s)
4
5
6
7
1a + 3 equiv of K2CO3 in methanol/acetone-d6
phenylboronic acid in methanol/acetone-d6
1a + 3 equiv of K2CO3 in methyl-d3 alcohol-d
5.47 (br s)
28.48 (br s)
4.98 (br s)
4.83 (br s)
-150.21(s)
phenylboronic acid + 3 equiv of K2CO3 in methyl-d3 alcohol-d
a
After the reaction was cooled, a precipitate was formed that dissolved in acetone-d6.
roborate (1i) under aprotic conditions (dry THF as
solvent, dry i-Pr2NEt as base) suggests that fluorinated
species can be the key intermediates in the catalytic
process. To assess this possibility and the different
reactivity of 1i and its boronic acid analogue, two
additional experiments were carried out with these
reagents in dry THF, adding 3 equiv of Hu¨nig’s base, and
heating to reflux during 12 h. Both crude reactions were
then analyzed by 19F and 11B NMR after deuterated
acetone was added. Surprisingly, the pentafluorophenyl-
boronic acid showed no signal of the boronic acid in 11B
NMR, and the presence of at least 5 different signals in
the 19F NMR spectrum. Trifluoroborate 1i remained as
a single “ate” complex (quartet at 3.72 ppm in 11B NMR,
which suggests the presence of BF3-), with four signals
still present in 19F NMR. These preliminary results
suggest that if, as observed for phenyltrifluoroborate (1a ),
partial or completely fluorinated intermediates are not
predominant in the catalytic cycle, the trifluoroborates
could still be involved in the cross-coupling process. A
more exhaustive analysis must be done to determine the
actual species involved in the catalytic cycle, as well as
the degree of protodeboronation and homocoupling of the
aryltrifluoroborate that compete with the cross-coupling
reaction.
It is important to point out the use of PdCl2(dppf)‚CH2-
Cl2 instead of Pd(PPh3)4 in the reactions involving
electron-poor aryltrifluoroborates as well as heteroaryl-
trifluoroborates, which permits reactions to be carried
out in an open atmosphere and with a lower loading of
catalyst. The ability to utilize electron-deficient arylbo-
rons routinely is another hallmark advantage of the
trifluoroborate technology over that of boronic acid-based
methods.
Unfortunately, the same factors that contribute to the
higher reactivity of these trifluoroborates (e.g., the higher
electron-withdrawing strength of the trifluoroborate
group and the preformed “ate” complex) leads to facile
protodeboronation for some of the R-heteroaryltrifluo-
roborates examined. This may be considered one of the
few disadvantages in the use of these boron reagents.
Biaryls are found in many compounds of pharmaceuti-
cal interest, and many recently reported drug candidates
and analogues have been synthesized via Suzuki coupling
reactions. What remains then is a study to demonstrate
the applicability of the trifluoroborate syntheses and
reactions for valuable intermediates in the natural
product synthesis. These studies currently are being
undertaken in our group.
Exp er im en ta l Section
Con clu sion s
Gen er a l P r oced u r e for P ota ssiu m Ar yl- a n d Het-
er oa r yltr iflu or obor a te Syn th esis. P r ep a r a tion of P ota s-
siu m 3-Th iop h en yltr iflu or obor a te (1m ). 3-Thiophenylbo-
ronic acid (26.2 mmol, 4.9757 g) and potassium hydrogen
fluoride (65.8 mmol, 5.1419 g) were placed in a 100-mL
Nalgene container and were vigorously stirred in a mixture
of methanol (7.5 mL) and water (14 mL) for 2 h. The resulting
amber solid was allowed to stand for 2 h at 4 °C and was then
filtered and washed with a minimum amount of cold methanol.
The solid was then taken up in hot acetone and filtered again.
The filtrate was cooled to room temperature and ethyl ether
was added in portions, with stirring, until no cloudiness was
observed in the supernatant. The mixture was allowed to stand
for 1 h at 4 °C to complete crystallization. The solid was filtered
and washed with cold ethyl ether until white crystals were
obtained. The crystalline solid was dried on a high-vacuum
Schlenk line to give 4.6793 g (94%) of the desired material.
The spectral data obtained were in accordance with those
described in the literature.11c mp >260 °C dec. 1H NMR
(acetone-d6, 500 MHz) δ 7.20 (s, 1 H), 7.14 (m, 2 H). 13C NMR
(acetone-d6, 125 MHz) δ 131.8, 125.2, 122.3. 11B NMR (acetone-
d6, 64.2 MHz) δ 4.91 (br s). 19F NMR (acetone-d6, 470.5 MHz)
δ -139.5 (d, J ) 75 Hz). Anal. Calcd for C4H3BF3KS: C, 25.28,
H, 1.59. Found: C, 25.36, H, 1.66.
It has been demonstrated that there are inherent
advantages in utilizing organotrifluoroborates for Suzuki
coupling reactions. In general, they are more robust, more
easily purified, easier to handle, and less prone to
protodeboronation. A wide array of electron-withdrawing
and electron-donating groups in both coupling partners
is tolerated under the ligandless and ligand-added pro-
tocols.
Generally, the trifluoroborate coupling system proved
to be more reactive than the corresponding boronic acids
or esters under standard reaction conditions. Despite
several reports applying ligandless conditions to the
reactions of arylboronic acids or esters,5 there is no single
contribution accounting for as many successful couplings
of reagents bearing such a variety of functional groups
as is reported herein. In general a lower loading of
catalyst, lower temperatures, and shorter reaction times
can be utilized to match the yields obtained in previously
reported couplings involving aryl- and heteroarylboronic
acids, even when PdCl2(dppf)‚CH2Cl2 was required as
catalyst. Most importantly, in all of the examples shown,
the reactions can be carried out in air with no attenuation
in the yields.
Gen er a l P r oced u r e for Su zu k i-Miya u r a Cr oss-Cou -
p lin g Rea ction s u n d er Con d ition A. P r ep a r a tion of
1-P h en yln a p h th a len e (3a ). To a suspension of potassium
J . Org. Chem, Vol. 68, No. 11, 2003 4313

Molander and Biolatto
phenyltrifluoroborate (1a) (0.0923 g, 0.50 mmol), 1-bromonaph-
thalene (2a ) (0.1035 g, 0.50 mmol), and K2CO3 (0.2045 g, 1.5
mmol) in 0.75 mL of methanol was added 1.25 mL of a 2 ×
10-3 M solution of Pd(OAc)2 in methanol (0.5 mol %), and the
reaction mixture was stirred and heated at reflux for 2 h, then
cooled to room temperature and diluted with water. The
aqueous phase was then extracted with dichloromethane. The
organic solution was washed with 1 N HCl and brine and dried
over MgSO4. The solvent was removed under vacuum. Puri-
fication by chromatography on silica gel (elution with hexane)
yielded 1-phenylnaphthalene (3a ) (0.0761 mg, 0.37 mmol,
75%). The product was identical with authentic material by
1H NMR, 13C NMR, and GC retention time.
Gen er a l P r oced u r e for Su zu k i-Miya u r a Cr oss-Cou -
p lin g Rea ction s u n d er Con d ition B. P r ep a r a tion of
4-Hyd r oxybip h en yl (3k ). To a mixture of potassium phe-
nyltrifluoroborate (1a ) (0.1834 g, 1 mmol), 4-bromophenol (2k )
(0.1764 g, 1 mmol), K2CO3 (0.4153 g, 3 mmol), and Pd(OAc)2
(0.0011 g, 5 × 10-3 mmol, 0.5 mol %) was added water (4 mL).
The reaction was heated at 65 °C with stirring for 2 h. The
reaction mixture was then cooled to room temperature and
treated with 1 N HCl until it was acidic to litmus. The
precipitate was filtered off, washed with water, dissolved with
dichloromethane, and filtered again to eliminate the palladium
black formed during the reaction. The organic solution was
dried over Na2SO4 and filtered. The solvent was removed
under vacuum and the product 3k was analyzed by NMR after
recrystallization from dichloromethane/hexane (82% yield).
The spectral data obtained were in accordance with those
described in the literature.44
Gen er a l P r oced u r e for Su zu k i-Miya u r a Cr oss-Cou -
p lin g Rea ction s u n d er Con d ition C. P r ep a r a tion of 2,4-
Dim eth oxybip h en yl (3n ). A mixture of 0.0088 g of Pd(OAc)2
and 0.0263 g of Ph3P was stirred in 10 mL of methanol for 30
min to obtain a 4 × 10-3 M solution of Pd(OAc)2/(Ph3P). Then
1.25 mL of the solution (5 × 10-3 mmol, 1 mol %) was added
to a mixture of potassium phenyltrifluoroborate (1a ) (0.0925
g, 0.5 mmol), 2,4-dimethoxy-1-bromobenzene (2n ) (0.1080 g,
0.5 mmol), and K2CO3 (0.2127 g, 1.5 mmol) in 0.75 mL of
methanol. The reaction mixture was stirred and heated at
reflux in an open atmosphere for 18 h, then cooled to room
temperature and diluted with water. The aqueous phase was
then extracted with dichloromethane. The organic solution was
washed with brine and dried over MgSO4. The solution was
then filtered through a short pad of silica gel, using a sequence
of pentane (to eliminate any nonpolar byproducts) and pen-
tane/dichloromethane (to elute the product). Further addition
of pure dichloromethane eluted polar byproducts and the
catalyst residue. The solvent was then removed under vacuum
to yield 3n as a colorless oil (0.0809 g, 75%). The spectral data
obtained were in accordance with those described in the
literature.11b
Gen er a l P r oced u r e for Su zu k i-Miya u r a Cr oss-Cou -
p lin g Rea ction s u n d er Con d ition D. P r ep a r a tion of 3′-
Nitr obip h en yl-4-ca r bon itr ile (3s). To a mixture of potas-
sium 3-nitrophenyltrifluoroborate (1f) (0.1150 g, 0.5 mmol),
4-bromobenzonitrile (2e) (0.0912 g, 0.5 mmol), Et3N (0.21 mL,
1.5 mmol), and PdCl2(dppf)‚CH2Cl2 (0.0020 g, 5 × 10-3 mmol,
0.5 mol %) was added ethanol (2 mL). The reaction was heated
at reflux with stirring in an open atmosphere for 12 h, then
cooled to room temperature. Upon addition of water, a
precipitate was formed. The precipitate was filtered off,
thoroughly washed with water, dissolved in dichloromethane,
and dried over MgSO4. The solution was then filtered through
a short pad of silica gel, using a sequence of pentane (to
eliminate nonpolar byproducts) and pentane/dichloromethane
(to elute the product). The solvent was removed under vacuum
and the crude material was purified by recrystallization with
dichloromethane/hexane, to yield 3s as a white solid (0.0999
1
g, 89%). mp 162-164. H NMR (CDCl3, 500 MHz) δ 8.46 (t, J
) 1.9 Hz, 1 H), 8.29 (dd, J ) 8.1, 1.3 Hz, 1 H), 7.93 (dd, J )
7.8, 1.3 Hz, 1 H), 7.81 (dd, J ) 6.7, 1.7 Hz, 2 H), 7.75 (dd, J )
8.4, 1.9 Hz, 2 H), 7.69 (t, J ) 8.0 Hz, 1 H). 13C NMR (CDCl3,
125 MHz) δ 143.0, 140.8, 133.0, 132.9, 130.2, 127.9, 123.3,
122.1, 118.3, 112.4. HRMS (ESI): m/z calcd for C13H8N2O2 (M+)
224.0586, found 224.0579.
Ack n ow led gm en t. We thank J ohnson & J ohnson,
Merck Research Laboratories, Aldrich Chemical Co.,
Array BioPharma, and J ohnson Matthey for their
generous support. B.B. thanks the Fundacio´n Antorchas
(Argentina) for a postdoctoral fellowship.
Su p p or tin g In for m a tion Ava ila ble: Full experimental
details and copies of all NMR spectra (1H, 13C, 19F, and 11B).
This material is available free of charge via the Internet at
http://pubs.acs.org.
(44) Wolfe, J . P.; Singer, R. A.; Yang. B. H.; Buchwald, S. L. J . Am.
Chem. Soc. 1999, 121, 9550-9561.
J O0342368
4314 J . Org. Chem., Vol. 68, No. 11, 2003