Cross-coupling reactions of aromatic and heteroaromatic silanolates with aromatic and heteroaromatic halides

Article abstract of DOI:10.1021/ja8091449

The alkali-metal salts (potassium and sodium) of a large number of aryl- and heteroarylsilanols undergo efficient cross-coupling with a wide range of aromatic bromides and chlorides under mild conditions to form polysubstituted biaryls. The critical feature for the success of these coupling reactions and their considerable scope is the use of bis(tri-tert-butylphosphine)palladium. Under the optimized conditions, electron-rich, electron-poor, and sterically hindered arylsilanolates afford cross-coupling products in good yields. Many functional groups are compatible with the coupling conditions such as esters, ketones, acetals, ethers, silyl ethers, and dimethylamino groups. Two particularly challenging substrates, (2-benzofuranyl)dimethylsilanolate and (2,6-dichlorophenyl)dimethylsilanolate prepared as their sodium salts showed excellent activity in the coupling reactions, in the former case also with aromatic chlorides. General methods for the efficient synthesis of a wide range of aromatic silanols are also described.

Full text of DOI:10.1021/ja8091449

Published on Web 02/06/2009
Cross-Coupling Reactions of Aromatic and Heteroaromatic
Silanolates with Aromatic and Heteroaromatic Halides
Scott E. Denmark,* Russell C. Smith, Wen-Tau T. Chang, and Joseck M. Muhuhi
Roger Adams Laboratory, Department of Chemistry, UniVersity of Illinois at
Urbana-Champaign, Urbana, Illinois 61801
Received November 22, 2008; E-mail: sdenmark@illinois.edu
Abstract: The alkali-metal salts (potassium and sodium) of a large number of aryl- and heteroarylsilanols
undergo efficient cross-coupling with a wide range of aromatic bromides and chlorides under mild conditions
to form polysubstituted biaryls. The critical feature for the success of these coupling reactions and their
considerable scope is the use of bis(tri-tert-butylphosphine)palladium. Under the optimized conditions,
electron-rich, electron-poor, and sterically hindered arylsilanolates afford cross-coupling products in good
yields. Many functional groups are compatible with the coupling conditions such as esters, ketones, acetals,
ethers, silyl ethers, and dimethylamino groups. Two particularly challenging substrates, (2-benzofuran-
yl)dimethylsilanolate and (2,6-dichlorophenyl)dimethylsilanolate prepared as their sodium salts showed
excellent activity in the coupling reactions, in the former case also with aromatic chlorides. General methods
for the efficient synthesis of a wide range of aromatic silanols are also described.
Chart 1
Introduction
Biaryls represent a common structural motif in both natural
products and biologically active compounds (Chart 1). For
example, vancomycin is a glycopeptide antibiotic that contains
a configurationally defined biaryl subunit.1 Additionally, a
number of pharmaceutical compounds containing a biaryl moiety
have been developed including Losartan and Irbesartan.2 The
prevalence of the biaryl subunit is further evident in the
development of highly conjugated aromatic polymers that
possess interesting optical and electronic properties.3 Lastly, the
biaryl scaffold is an integral part of the ligand design for many
catalytic processes. Chiral ligands such as BINAP, BINOL,
MOP, SEGPHOS, SYNPHOS, and their derivatives have been
utilized in many asymmetric transformations.4
Given the importance of the biaryl motif, a myriad of methods
have been developed for the synthesis of biaryl compounds.5
In 1901, Ullmann discovered the homocoupling reaction of aryl
halides in the presence of copper powder at elevated temper-
atures (>200 °C).6 Since this seminal report, extensive reaction
development has produced milder conditions for copper-,
nickel-, or palladium-catalyzed couplings.7 Although this method
is useful for the synthesis of symmetrical biaryls, high-yielding
(1) Rama Rao, A. V.; Gurjar, M. K.; Reddy, K. L.; Srinivasa Rao, A.
Chem. ReV. 1995, 95, 2135–2167.
(2) Corbet, J.-P.; Mignani, G. Chem. ReV. 2006, 106, 2651–2710.
(3) (a) Roncali, J. Chem. ReV. 1992, 92, 711–738. (b) Electronic Materials:
The Oligomer Approach; Mu¨llen, K., Wegner, G., Eds.; Wiley-VCH:
Weinheim, Germany, 1998.
preparations of unsymmetrical biaryls often require pairing an
electron-rich aryl halide with an electron-deficient aryl halide.
Among the modern methods of carbon-carbon bond forma-
tion, the transition-metal catalyzed cross-coupling reaction has
ascended to a strategy level reaction in both industrial and
academic sectors.8 For example, a survey of the reactions scaled
in the GMP facilities at Pfizer-Groton shows that the use of
cross-coupling reactions has increased by 10.2% between 1985
and 2002.9 Of particular importance is the palladium-catalyzed,
(4) (a) Berthod, M.; Mignani, G.; Woodward, G.; Lemaire, M. Chem.
ReV. 2005, 105, 1801–1836. (b) Seyden-Penne, J. Chiral Auxiliaries
and Ligands in Asymmetric Synthesis; Wiley: New York, 1995.
(5) (a) Sainsbury, M. Tetrahedron 1980, 36, 3327–3359. (b) Stanforth,
S. P. Tetrahedron 1998, 54, 263–303. (c) Hassan, J.; Se´vignon, M.;
Gozzi, C.; Schulz, E.; Lemaire, M. Chem. ReV. 2002, 102, 1359–
1469.
(6) Ullmann, F.; Bielecki, J. Chem. Ber. 1901, 34, 2174–2185.
(7) Nelson, T. D. Org. React. 2004, 63, 265–555.
9
3104 J. AM. CHEM. SOC. 2009, 131, 3104–3118
10.1021/ja8091449 CCC: $40.75
2009 American Chemical Society

Cross-Coupling Reactions of Silanolates with Halides
A R T I C L E S
cross-coupling reaction for the synthesis of biaryls. Various
organometallic donors have been studied for this purpose,
including organomagnesium (Kumada), organozinc (Negishi),
organotin (Stille), organoboron (Suzuki), and organosilicon
(Hiyama) reagents.8 Although Grignard reagents can couple
effectively, their high nucleophilicity and basicity are not
compatible with many functional groups. Organozinc reagents
are more tolerant of sensitive functionality; however, the
instability of arylzinc reagents has hampered large-scale devel-
opment.10 In contrast, organostannanes are mild, stable, and
readily prepared reagents, but a major drawback is the inherent
toxicity of the tin reagents and byproducts.11 Consequently,
recent investigations have focused on using organoboron-based
reagents and many synthetically challenging biaryls have been
prepared by judicious choice of catalysts and ligands.12 Despite
the broad application of this cross-coupling strategy, these
methods are not without drawbacks. For example, some boronic
acids are difficult to handle, difficult to purify, unstable to
extended storage and suffer protiodeborylation under cross-
coupling conditions. Variants of boron reagents such as borate
esters and trifluoroborates13 have been developed to address
some of these problems.14 In recent years arylsilicon nucleo-
philes have emerged as useful nucleophiles for cross-coupling
reactions. The specific advantages of organosilicon-based donors
have been described in detail in a number of review articles
and will be summarized only briefly below.15
moderate yields (47-71%).18 To address this limitation, stable
triallylarylsilanes were introduced for the cross-coupling of a
range of aryl chlorides.19 A major disadvantage of all the
aforementioned arylsilicon donors is the need for fluoride
activators in superstoichiometric amounts to facilitate trans-
metalation. The widespread use of silicon-based protecting
groups in complex molecule synthesis along with the cost and
aggressive nature of soluble fluoride sources precludes their use
as activators. Accordingly, alternative methods of activation
have been developed for arylsilicon reagents. In 2005, Hiyama
and co-workers reported the use of aryl[2-(hydroxymethyl)-
phenyl]dimethylsilanes which bear a pendant nucleophilic
tether.20 In the presence of potassium carbonate and copper
iodide, these reagents deliver aryl groups to aromatic halides.
The iterative cross-coupling of these organosilicon reagents was
elegantly demonstrated for the synthesis of oligoarenes.21
As part of our longstanding interest in the use of the
organosilanes in cross-coupling reactions,15c-f,h,i we have
investigated use of these compounds in the synthesis of biaryls.
One of the initial illustrations of this class of reagents stemmed
from using arylsiletanes as aryl transfer substrates.22 The
palladium-catalyzed cross-coupling of aryl(halo)silacyclobutanes
could be achieved via a fluoride-promoted reaction with aryl
iodides to provide biaryls in good yields. Mechanistic studies
(with alkenylsilacyclobutanes) demonstrated that treatment of
silacyclobutanes with tetrabutylammonium fluoride (TBAF) led
to cleavage of the siletanes and formation of organosilanols.23
Background
Aryl(dihalo)silanes represent the first organosilicon reagents
used for biaryl synthesis.16 Because of the sensitivity of
halosilanes to moisture, acid, or base, the more robust trialkoxy-
silanes have been advocated as more practical reagents for cross-
coupling with aryl iodides and bromides.17 However, the
reaction with aryl chlorides provides the biaryl products in
The introduction of (organo)dimethylsilanols as coupling
partners has allowed the development of nonfluoride promoted
cross-coupling reactions. An early report from Hiyama and co-
workers described the use of silver(I) oxide as a stoichiometric
promoter in conjunction with (Ph3P)4Pd to prepare biaryls from
silanol precursors.24 Although good yields were obtained, large
amounts of a silver activator and long reactions times are needed.
(8) (a) Metal-Catalyzed Cross-Coupling Reactions, 2nd ed.; de Meijere,
A., Diederich, F., Eds.; Wiley-VCH: Weinheim, Germany, 2004; Vols.
1 and 2. (b) Handbook of Organopalladium Chemistry for Organic
Synthesis; Negishi, E.-i., Ed.; Wiley-Interscience: New York, 2002.
(9) Dugger, R. W.; Ragan, J. A.; Brown Ripin, D. H. Org. Process Res.
DeV. 2005, 9, 253–258.
Subsequent investigations from these laboratories identified
Cs2CO3 as an effective promoter for the cross-coupling of (4-
methoxyphenyl)dimethylsilanol with a range of aryl iodides and
bromides.25 These studies also addressed the problem of
homocoupling from the aryl halide through the judicious choice
of ligands. Unfortunately, attempts to expand the scope of the
arylsilanol afforded lower yields. More forcing conditions (110
°C) and other bases (CsOH) were required because of the lower
rate of transmetalation of other arylsilanols and their tendency
to exist as their thermodynamically more stable disiloxanes.
(10) Tucker, C. E.; Majid, T. N.; Knochel, P. J. Am. Chem. Soc. 1992,
114, 3983–3985.
(11) (a) Krigman, M. R.; Silverman, A. P. Neurotoxicology 1984, 5, 129–
139. (b) Cragg, S. T. In Patty’s Toxicology; Bingham, E., Cohrssen,
B., Powell, C. H., Eds.; Wiley: Hoboken, NJ, 2001; DOI: 10.1002/
0471435139.tox.093.
(12) (a) Bedford, R. B.; Cazin, C. S. J.; Holder, D. Coord. Chem. ReV.
2004, 248, 2283–2321. (b) Nolan, S. P.; Navarro, O. ComprehensiVe
Organometallic Chemistry III, 2007, Chapter 11.01, pp 1-37.
(13) Miyaura, N. In Metal-Catalyzed Cross-Coupling Reactions, 2nd ed.;
de Meijere, A., Diederich, F., Eds.; Wiley-VCH: Weinheim, 2004;
Chapter 2.
In the course of mechanistic studies on the cross-coupling
reactions of arylsilanols, it was discovered that arylsilanolate
salts are stable, isolable compounds that are resistant to
(14) Molander, G. A.; Ellis, N. Acc. Chem. Res. 2007, 40, 275–286.
(15) (a) Hiyama, T. In Metal-Catalyzed Cross-Coupling Reactions; Dieder-
ich, F., Stang, P. J., Eds.; Wiley-VCH: Weinheim, 1998; Chapter 10.
(b) Hiyama, T. J. Organomet. Chem. 2002, 653, 58–61. (c) Denmark,
S. E.; Sweis, R. F. Chem. Pharm. Bull. 2002, 50, 1531–1541. (d)
Denmark, S. E.; Sweis, R. F. Acc. Chem. Res. 2002, 35, 835–846. (e)
Denmark, S. E.; Ober, M. H. Aldrichim. Acta 2003, 36, 75–85.
(f) Denmark, S. E.; Sweis, R. F. In Metal Catalyzed Cross-Coupling
Reactions, 2nd ed.; de Meijere, A., Diederich, F., Eds.; Wiley-VCH:
Weinheim, 2004; Chapter 4. (g) Handy, C. J.; Manoso, A. S.; McElroy,
W. T.; Seganish, W. M.; DeShong, P. Tetrahedron 2005, 61, 12201–
12225. (h) Denmark, S. E.; Regens, C. S. Acc. Chem. Res. 2008, 41,
1486–1499. (i) Denmark, S. E.; Butler, C. R. Chem. Commun. 2009,
20–33.
(18) Mowery, M. E.; DeShong, P. Org. Lett. 1999, 1, 2137–2140.
(19) (a) Nakao, Y.; Oda, T.; Sahoo, A. K.; Hiyama, T. J. Organomet. Chem.
2003, 687, 570–573. (b) Sahoo, A. K.; Nakao, Y.; Hiyama, T. Chem.
Lett. 2004, 33, 632–633. (c) Sahoo, A. K.; Oda, T.; Nakao, Y.; Hiyama,
T. AdV. Synth. Catal. 2004, 346, 1715–1727.
(20) Nakao, Y.; Imanaka, H.; Sahoo, A. K.; Yada, A.; Hiyama, T. J. Am.
Chem. Soc. 2005, 127, 6952–6953.
(21) Nakao, Y.; Chen, J.; Tanaka, M.; Hiyama, T. J. Am. Chem. Soc. 2007,
129, 11694–11695.
(22) Denmark, S. E.; Wu, Z. Org. Lett. 1999, 1, 1495–1498.
(23) Denmark, S. E.; Wehrli, D.; Choi, J. Y. Org. Lett. 2000, 2, 2491–
2494.
(16) (a) Hatanaka, Y.; Fukushima, S.; Hiyama, T. Chem. Lett. 1989, 18,
1711–1714. (b) Gouda, K.-I.; Hagiwara, E.; Hatanaka, Y.; Hiyama,
T. J. Org. Chem. 1996, 61, 7232–7233.
(24) Hirabayashi, K.; Mori, A.; Kawashima, J.; Suguro, M.; Nishihara, Y.;
Hiyama, T. J. Org. Chem. 2000, 65, 5342–5349.
(25) (a) Denmark, S. E.; Ober, M. H. Org. Lett. 2003, 5, 1357–1360. (b)
Denmark, S. E.; Ober, M. H. AdV. Synth. Cat. 2004, 346, 1703–1714.
(17) Mowery, M. E.; DeShong, P. J. Org. Chem. 1999, 64, 1684–1688.
9
J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009 3105

A R T I C L E S
Denmark et al.
dimerization to the corresponding disiloxane.26 The ease of
preparation of these salts made them attractive targets for
investigating their utility as organosilicon nucleophiles.27 As
part of this study, the use of phosphine oxides as stabilizing
ligands was also discovered. A significant rate enhancement was
observed using bisphosphine oxides derived from common
bidentate phosphines, as well as triphenylphosphine oxide.27
These results were transformed into a useful and effective set
of reaction conditions that could be employed for the cross-
coupling of potassium (4-methoxyphenyl)dimethylsilanolate
with a range of aryl bromides.27 However, these conditions were
not uniformly applicable. Because no strongly coordinating
ligands are present, significant amounts of homocoupling
products from the aryl bromides are formed and the overall
yields of the desired biaryls correspondingly decreases. More-
over, the extension of this method to other arylsilanolates was
unsuccessful as unproductive dimerization to the disiloxane was
observed.
The variable levels of success achieved with palladium-
catalyzed cross-coupling of aryldimethylsilanols and its deriva-
tives meant that a more general set of reaction conditions for
the preparation of biaryls is still required. In light of these
limitations, the following objectives were formulated: (1) prepare
a wide range of arylsilanols and their silanolate salts, (2)
determine the stability of these reagents, (3) evaluate the
competence of these alkali-metal silanolates for the palladium-
catalyzed cross-coupling reactions, and (4) develop conditions
that allow the facile cross-coupling of a range of aryl halides.
We report herein the successful realization of these goals in
the development of a practical fluoride-free, cross-coupling of
a wide range of arylsilanolates with a wide range of aryl halides
that provides unsymmetrical biaryls. Moreover, this reaction has
been implemented in the cross-coupling of heterocyclic silano-
lates that have not been disclosed previously.
accomplished by treatment of an aryl iodide (e.g., 4-iodoanisole
or 1-iodonaphthalene) with n-butyllithium at low temperature
followed by trapping the organolithium intermediate with
hexamethylcyclotrisiloxane (D3) to afford (4-methoxyphen-
yl)dimethylsilanol 1 and (1-naphthyl)dimethylsilanol 2 in 88%
and 84% yields, respectively (Table 1, entries 1-2). Although
aryl iodides readily participate in halogen-metal exchange, the
respective aryl bromides are more attractive substrates because
of their greater availability and lower cost.
Thus, the scope of the direct process was further illustrated
with a diverse array of aryl bromides. In these cases tert-
butyllithium was used to generate the organolithium intermedi-
ate. Electron-rich aryl bromides such as tert-butyldimethylsilyl
(TBS)-protected 4-bromophenol, 3- and 2-bromoanisole, and
4-bromostilbene all furnished the aryldimethylsilanols in good
to excellent yields of analytically pure products (72-95%)
(Table 1, entries 3-6). Similarly, electron-deficient aryl bro-
mides containing trifluoromethyl, chloro, and fluoro substituents
showed comparable reactivity and overall yields (79-87%)
(Table 1, entries 9-11).
Table 1. Preparation of Arylsilanols and the Corresponding
Alkali-Metal Arylsilanolates
Results
1. Preparation of Silanols and Silanolates.28 To set the stage
for our studies, a large number of arylsilanols and their derived
silanolates were needed to determine their chemical stability,
ease of handling, and suitability for cross-coupling reactions
with aryl electrophiles. In general, the arylsilanols used herein
were prepared by three different protocols: (1) direct installation
of the dimethylsilanol unit through halogen-lithium exchange
and trapping with a silicon electrophile,29 (2) catalytic, oxidative
hydrolysis of dimethylsilanes,30 and (3) palladium-catalyzed
silylation of aryl halides.31 The selection of the most appropriate
method was based on a number of considerations including the
availability of a suitable starting material and compatibility of
any functional groups in the molecule. Each approach is
described in detail below.
1.1. Preparation of Arylsilanols via Halogen-Lithium Ex-
change. Direct installation of the dimethylsilanol moiety was
(26) (a) Tymonko, S. A. Ph.D. Thesis, University of Illinois at Urbana-
Champaign, 2007. (b) Denmark, S. E.; Tymonko, S. E.; Smith, R. C.;
Ober, M. H. Unpublished results from these laboratories.
(27) Denmark, S. E.; Smith, R. C.; Tymonko, S. A. Tetrahedron 2007, 63,
5730–5738.
a Methods for silanol preparations: A. 1. n-BuLi (1.0 equiv) or t-BuLi
(2.0 equiv), Et2O, -78 °C 2. D3 (0.33 equiv), Et2O, -78 °C to rt. B. 1.
n-BuLi (1.0 equiv) or t-BuLi (2.0 equiv), Et2O, -78 °C 2. Me2SiCl(H),
-78 °C to rt 3.[RuCl2(p-cymene)]2 or [Ir(1,5-cyclooctadiene)Cl]2 (1-2
mol %), H2O (2-5 equiv), CH3CN, rt or aq TBAOH (1.5 equiv, 1 M),
CH3CN, 0 °C C. 1. (EtO(CH3)2Si)2 (1.5 equiv), [allylPdCl]2 (5 mol %),
BPTBP (10 mol %), NMP, 50 °C 2. pH 5 acetate buffer, HS(CH2)2-
NH2 ·HCl, CH3CN. b Yields correspond to isolated, purified products
c Methods for silanolate preparation: D. KH (1.2 equiv), THF or C6H6,
rt. E. NaHMDS, THF, rt. F. NaH (1.2 equiv), C6H6, rt.
(28) Lickiss, P. D. AdV. Inorg. Chem. 1995, 42, 147–262.
(29) Frye, C. L. S. R. M.; Fearon, F. W. G.; Klosowski, J. M.; DeYoung,
T. J. Org. Chem. 1970, 35, 1308–1314.
(30) (a) Lee, M.; Ko, S.; Chang, S. J. Am. Chem. Soc. 2000, 12011–12012.
(b) Lee, Y.; Seomoon, D.; Kim, S.; Han, H.; Chang, S.; Lee, P. H. J.
Org. Chem. 2004, 69, 1741–1743.
(31) Denmark, S. E.; Kallemeyn, J. M. Org. Lett. 2003, 5, 3483–3486.
9
3106 J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009

Cross-Coupling Reactions of Silanolates with Halides
A R T I C L E S
The direct metalation of dibenzothiophene using n-BuLi in
refluxing ether gave the 4-dibenzothienyllithium intermediate
that underwent further reaction with D3 to afford analytically
pure (4-dibenzothienyl)dimethylsilanol 12 in 54% yield (Table
1, entry 12). These procedures were easily scalable, e.g., the
preparation of 12 was routinely performed on a 20 mmol scale
(5 g of product).
1.2. Catalytic Oxidative Hydrolysis of Arylsilanes. For less
reactive or less stable aryllithium intermediates (such as those
from 4-bromo-N,N-dimethylaniline, benzofuran, and 2,6-dichlo-
robromobenzene) the more reactive electrophile dimethylchlo-
rosilane was employed to provide the arylsilanes in high yields.
Selective oxidation of these silyl hydrides was accomplished
with [RuCl2(p-cymene)]2 (1-2 mol %) in MeCN/H2O to yield
the desired silanols in 79-95% yield (Table 1, entries 7-8,
13-14). Although, successful oxidative addition could be
effected under these conditions with (2,6-dichlorophenyl)di-
methylsilane, the reaction was quite sluggish (>8 h). The more
active catalyst, [Ir(COD)Cl]2 led to the formation of the desired
silanol in 45 min (in the presence of 5 equiv of H2O). These
processes are reproducible, e.g., (2-benzofuranyl)dimethylsilane
can be prepared consistently on an 85 mmol scale, while the
oxidative hydrolysis of this silane is routinely performed on
>20 mmol scale.
Another silanol that initially proved challenging to prepare
was (4-(1,3-dioxolan-2-yl)phenyl)dimethylsilanol 7. The lower
reactivity of the aryllithium intermediate (from the correspond-
ing bromide) with D3 led to significant amounts of disiloxane
formation. Preparation of the (4-(1,3-dioxolan-2-yl)phenyl)di-
methylsilane as described above was straightforward, but
attempts to effect the oxidative hydrolysis using either Ru or Ir
catalysis was problematic. Significant amounts of 4-(hydroxy-
dimethylsilyl)benzaldehyde (4-10%) were observed and sepa-
ration from the desired silanol was not easily achieved.
Gratifyingly, simply treating the (4-(1,3-dioxolan-2-yl)phen-
yl)dimethylsilane with 1.5 equiv of tetrabutylammonium hy-
droxide32 led to quantitative formation of the desired silanol in
79% yield (Table 1, entry 7).
of the arylsilanolate has a number of advantages that deserve
mention, the most important of which is operational simplicity.
The salt can be charged directly into the reaction mixtures, and
no additional base is needed. Moreover, isolation of the
arylsilanolate avoids the formation of disiloxane from the parent
silanol under the reaction conditions. Accordingly, metal
hydrides and disilazanes were chosen as the bases to promote
rapid and quantitative deprotonation that allowed the isolation
of the arylsilanolate salts. For the most part, potassium silanolate
salts were selected because they are more nucleophilic than
sodium silanolates, thus avoiding the complicating factor of a
rate-limiting displacement of the palladium halide in the catalytic
cycle.36
The potassium arylsilanolates were prepared by the dropwise
addition of a solution of the arylsilanol (in THF or benzene) to
either KH or KHMDS in the same solvent. After 15 min, the
reaction mixture was Schlenk filtered to remove any particulates.
The solution was concentrated in vacuo to afford solids that
were triturated with hexanes and dried to afford the arylsilanolate
salts as stable powders in most cases (Table 1, entries 1-12,
15).37 Overall, excellent yields of the potassium salts were
obtained but the less than quantitative recovery was occasionally
seen because of partial solubility of these compounds in hexanes.
Unfortunately, the preparation of both potassium (2-benzo-
furanyl)dimethylsilanolate and (2,6-dichlorophenyl)dimethyl-
silanolate by the addition of the corresponding silanol to KH
or KHMDS in THF led to rapid desilylation to benzofuran and
1,3-dichlorobenzene, respectively. This problem has been
encountered previously in the preparation of potassium (2-
indolyl)dimethylsilanolates.38 The instability of the potassium
salts of heterocyclic compounds was circumvented by preparing
the sodium silanolate Na+13- and Na+14- using either NaH or
NaHMDS. Generally, NaHMDS is the preferred base because
it provides higher yields and no filtration is required prior to
concentration (Table 1, entry 13). These sodium salts were stable
solids as well.
2. Optimization of Cross-Coupling Reaction Conditions.
2.1. Substrate Choice. The selection of an appropriate substrate
pair for the initial optimization was crucial in view of the limited
substrate scope seen in previous studies.25 Whereas highly
reactive silanolates from electron-rich silanols did react smoothly,
less activated silanolates required more vigorous reaction
conditions and gave poorer yields. Thus, to ensure a successful
demonstration of broad scope, we selected very challenging
substrates for optimization studies with the expectation that
conditions that work well for the poorest substrates, should be
applicable to better ones in general. Accordingly, potassium (4-
trifluoromethylphenyl)dimethylsilanolate K+9- and 4-bromo-
anisole 16a were selected because of (1) the poor nucleophilicity
of K+9-, (2) the slower oxidative addition by Pd(0) catalysts
to 16a, and (3) the increased propensity for 16a to undergo
homocoupling. Orienting experiments with K+9- (1.5 equiv)
and 16a under conditions developed for couplings of aryl-
1.3. Palladium-Catalyzed Silylation of Aryl Halides. The
installation of the dimethylsilanol unit into substrates possessing
functional groups that are incompatible with organolithium
reagents (esters, ketones, nitriles, etc.) employed previously
developed catalytic reaction with 1,2-diethoxy-1,1,2,2-tetra-
methyldisilane.31 For example, tert-butyl-4-(hydroxydimethyl-
silyl)benzoate was prepared from tert-butyl-4-bromobenzoate,
allylpalladium chloride dimer ([allylPdCl]2), 2-(di-tert-butylphos-
phino)biphenyl (BPTBP),33 and diisopropylethylamine and the
disilane in NMP. The silyl ether was hydrolyzed to the silanol
15 with 1.0 M acetic acid in 82% yield, over two steps (Table
1, entry 15). The preparation of 15 was routinely carried out
on a 10 mmol scale.
1.4. Preparation of Arylsilanolates from the Parent Sila-
nols. Alkali metal arylsilanolate salts can be prepared from
arylsilanols (pKa ≈ 9-11)28 by any of three different protocols:
(1) irreversible deprotonation of the arylsilanol in situ with a
strong Brønsted base, (2) irreversible deprotonation of the
arylsilanol with a strong base and isolation of the silanolate
salt,34 and (3) silanolate exchange with disiloxanes.35 Isolation
(35) Denmark, S. E.; Butler, C. R. J. Am. Chem. Soc. 2008, 130, 3690–
3704.
(36) The cesium and rubidium salts could be prepared also and are
competent coupling partners. However, the difficulty in handling these
metals renders these reagents less attractive. For a discussion of the
working mechanistic hypothesis see ref 53.
(32) Denmark, S. E.; Neuville, L. Org. Lett. 2000, 2, 3221–3224.
(33) Wolfe, J. P.; Singer, R. A.; Yang, B. H.; Buchwald, S. L. J. Am. Chem.
Soc. 1999, 121, 9550–9561.
(37) All aryldimethylsilanolates were isolated as stable powders except for
potassium 3-(methoxyphenyl)dimethylsilanolate which was a viscous
oil.
(38) Denmark, S. E.; Baird, J. D.; Regens, C. S. J. Org. Chem. 2008, 73,
1440–1455.
(34) (a) Denmark, S. E.; Baird, J. D. Chem. Eur. J. 2006, 12, 4954–4963.
(b) Denmark, S. E.; Baird, J. D. Org. Lett. 2006, 8, 793–795.
9
J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009 3107

A R T I C L E S
Denmark et al.
Table 2. Optimization of the Ligand/Palladium Combination for the Cross-Coupling of K+9- with 16aa
entrya
Pd source, 5 mol % Pd
ligand, 1:1 Pd/L
time, h
conversion, %b
product yield, %c
1
2
3
4
5
6
7
8
Pd(dba)2
Ph3P(O)
dppp
7
20
20
20
7
7
7
3
7
24
100
43
68
37
95
54
69
7
30
10
37
9
45
30
51
12
(C3H5)CpPd
PdCl(C3H5)(Ii-Pr)
(Ph3P)4Pd
-
-
[allylPdCl]2
[allylPdCl]2
[allylPdCl]2
[allylPdCl]2
[allylPdCl]2
[allylPdCl]2
[allylPdCl]2
[allylPdCl]2
18
-
dppp(O)2
dppp(O)
dppp
9
Ph3As
SPhos
t-Bu3P
t-Bu3P·HBF4
-
22
71
10
11
12
13
14
7
46
5.5
20
7
99 (100)d
88
79 (89)d
68
90
73
(t-Bu3P)2Pd (19)
-
5
100 (100)e
88 (92)e
a Reactions conditions: 1.5 equiv of arylsilanolate K+9- and 1.0 equiv of 16a. b Conversion was based on consumption of aryl bromide as determined
by GC analysis using an internal standard. c Yield determined by GC analysis using an internal standard. d Yield in parentheses based on 2:1 ratio of
ligand/Pd. e Conversion and yield in parentheses refers to the use of (t-Bu3P)2Pd purchased from Aldrich Chemical Co.
silanolates previously (Pd(dba)2 and Ph3P(O) in toluene at 90
°C)27 provided low conversion and product yield (Table 2, entry
1). Other Pd(II) and Pd(0) sources successfully catalyzed the
cross-coupling reaction albeit in low yields (Table 2, entries 2
and 4). Moreover, the use of an N-heterocyclic carbene-based
palladium catalyst gave a poor yield of 17a (entry 3).
also be effective. However, the more air and moisture stable
t-Bu3P·HBF4 salt led to a slightly diminished overall yield (entry
12). The use of the palladacycle catalyst 18 (previously
described for the cross-coupling of indolyl-2-silanols)34b pro-
vided a comparable conversion and product yield at elevated
temperatures (Table 1, entry 13). Interestingly, increasing the
ligand/Pd loading from 1:1 to 2:1 using t-Bu3P and [allylPdCl]2
provided an additional increase in yield of 17a to 89% (Table
2, entry 11). Finally, the preformed “Fu’s catalyst” ((t-Bu3P)2Pd
(19)) with the desired ligand/Pd stoichiometry,42 led to complete
consumption of both aryl halide and arylsilanolate to afford 17a
in 88% yield (entry 14). At this point, only minor, substrate-
specific, modifications (solvent, temperature, and stoichiometry)
were needed to create a general set of reaction conditions.
2.3. Solvent, Temperature, and Stoichiometry. Solvents
representing a wide range of polarities were tested (Table 3).
Toluene (ε ) 2.4) was an effective solvent for the cross-coupling
of K+9- (1.5 equiv), providing the biaryl in 92% yield within
5 h (entry 1). Benzotrifluoride, a slightly more polar aromatic
solvent (ε ) 9.2), afforded 17a in 90% yield in 3 h (entry 2).
Ethereal solvents, such as dioxane (entry 5), led to productive
cross-coupling albeit in slightly lower yields. The use of highly
polar solvents, such as NMP or DMF, led to a rapid, unproduc-
2.2. Palladium Source and Ligands. Clearly, a thorough
optimization of reaction conditions was needed which began
with studies on the influence of ligands. Because [allylPdCl]2
is cleanly and quantitatively reduced to ligandless Pd(0) in the
presence of arylsilanolates, this complex was chosen for the
ligand studies.39 When [allylPdCl]2 was employed in the absence
of ligands for the cross-coupling of K+9- (1.5 equiv) and 16a
only a 9% yield of 17a was obtained (Table 2, entry 5).
Employing [allylPdCl]2 and dppp(O)2 afforded a substantial
increase in conversion and yield (entry 6) compared to the
ligandless conditions. Under the reasonable assumption that a
more coordinating ligand was required to prolong the catalyst
lifetime, both dppp(O) and dppp were tested but proved to be
ineffective (Table 2, entries 7-8). Furthermore, the use of
ligands that have been successful for the cross-coupling of
aryl-25 and alkenylsilanols,40 including Ph3As and SPhos,41
provided poor yields of 17a (entries 9 and 10, respectively).
Gratifyingly, the use of t-Bu3P in combination with [allylPdCl]2
provided complete conversion of 16a and a high yield of biaryl
17a (entry 11). Although an acceptable yield of 17a was
achieved, additional studies were needed to clarify the optimal
catalyst loading and ligand stoichiometry.
Table 3. Optimization of Solvent for the Cross-Coupling of K+9-
with 4-Bromoanisolea
If indeed the active catalytic species is t-Bu3P-Pd(0), other
catalyst combinations that could produce this complex should
(39) Denmark, S. E.; Smith, R. C. Synlett 2006, 18, 2921–2928.
(40) Denmark, S. E.; Kallemeyn, J. M. J. Am. Chem. Soc. 2006, 128,
15958–15959.
entrya
solvent
dielectric constant
time, h
conversion, %b
yield, %b
1
2
3
4
5
toluene
C6H5CF3
NMP
DMF
dioxane
2.4
9.2
32.2
36.7
2.2
5
3
0.25
0.25
5
100
98
100
100
88
92
90
0
0
74
(41) 2-Dicyclohexylphosphino-2′,6′-dimethoxybiphenyl: Barder, T. E.;
Walker, S. D.; Martinelli, J. R.; Buchwald, S. L. J. Am. Chem. Soc.
2005, 127, 4685–4696.
(42) (a) Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2001, 123, 2719–2724. (b)
Littke, A. F.; Fu, G. C. Org. Synth. 2005, 81, 63–70. (c) (t-Bu3P)2Pd
can also be purchased from either Aldrich Chemical Co. or Strem
Chemicals.
a Reactions employed 1.5 equiv of K+9-. b Conversion and yield
were determined by GC analysis using an internal standard.
9
3108 J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009

Cross-Coupling Reactions of Silanolates with Halides
A R T I C L E S
Table 5. Optimization of the Catalyst Loading for the Preparation
of 17aa
tive consumption of aryl bromide in only 15 min (entries 3-4).
No biaryl product was obtained, and multiple products were
observed including reduction of the aryl bromide to anisole.
Ultimately, toluene was chosen as the solvent because of its
widespread use and lower cost compared to benzotrifluoride.
Next, the reaction temperature was optimized using toluene
as the solvent (Table 4). Initial studies began at 90 °C following
earlier experiences with aryldimethylsilanols. At this tempera-
ture, the cross-coupling of K+9- (1.5 equiv) afforded 17a in
excellent yield and short reaction time (Table 4, entry 3). A 10
or 20 °C decrease in the temperature significantly decreased
both conversion (72% and 66%) and yield to only 66% or 61%
in 20 h, respectively (entries 1-2). Although increased tem-
peratures led to more rapid consumption of 16a, no increase in
yield was obtained (entries 4 and 5).
entry cat. loading, equivb concentration, Mb time, h conversion, %c yield, %c
1
2
0.01
0.25
0.25
0.50
0.50
0.50
0.25
0.50
20
11
3
5
5
55
89
98
79
91
46
77
92
74
82
92
91
0.025
0.025
0.025
0.025
0.05
3
4d
5e
6
5
3
100
100
7
0.05
a Reactions employed 1.5 equiv of K+9-. b Calculated based upon 1.0
equiv of aryl bromide. c Conversion and yield were determined by GC
analysis using an internal standard. d Reaction employed 1.1 equiv of
a
Table 4. Optimization of Temperature for Cross-Coupling of K+9-
e
K+9-. Reaction employed 1.3 equiv of K+9-.
reaction conditions developed above. Indeed, both electron-rich
(entries 1-2) and electron-deficient (entry 3) aryl bromides
reacted smoothly with K+1- to afford biaryls in good yields
and reasonable reaction times (3-4 h). 2-Bromotoluene was
also reactive (entry 2) and provided the biaryl product efficiently
within that reaction time. Heterocyclic bromides were of
particular interest because these substrates previously provided
only modest success in the coupling with arylsilanolates.25,44
Gratifyingly, 3-bromoquinoline 16e afforded the coupling
product in high yield (entry 4).
entry
temperature, °C
time, h
conversion, %b
yield, %b
1
2
3
4
5
70
80
90
100
110
20
20
5
3
1
66
72
100
98
61
66
92
86
86
100
a Reaction employed 1.5 equiv of K+9-. b Conversion and yield were
determined by GC analysis using an internal standard.
As has now become the state of the art, aryl chlorides are
recognized as more useful substrates in view of their greater
availability and lower cost, making them attractive candidates
for these studies.45 Although aryl chlorides are less reactive
toward oxidative addition by palladium because of the stronger
C-Cl bond,46 the use of strong electron-donating phosphine
ligands such as t-Bu3P has overcome this challenge.42,45 Thus,
the examination of aryl chlorides under identical conditions was
undertaken. When arylsilanolate K+1- was combined with a
representative set of aryl chlorides, the biaryls were obtained
in yields equivalent to those obtained from aryl bromides (Table
6, entries 5-7). Whereas, chlorobenzene reacted in 5 h to
provide an 88% yield of 20f (entry 5), aryl chlorides bearing a
methyl group at either the 2- or the 4-position required slightly
longer reaction times (7 h, entries 6-7).
The final dimension of optimization concerned the catalyst
loading, reaction concentration, and silanolate stoichiometry
(Table 5). Decreasing the catalyst loading to 0.01 equiv relative
to the aryl halide dramatically decreased the reaction rate (20
h) and product yield 46% (entry 1). Using 0.025 equiv of the
catalyst gave a comparable yield of 17a, but the reaction stalled
at 89% conversion (entry 2). Performing the reaction at a higher
concentration (0.5 M with respect to 16a) resulted in a
heterogeneous mixture at room temperature (entry 3). However,
upon heating to 90 °C, the mixture became homogeneous and
within 3 h, a 92% yield of 17a was recorded (entry 3, GC
analysis). An increase in the catalyst loading at the higher
concentration (0.5 M) did not provide an increase in product
yield. Attempts at decreasing the amount of K+9- (from 1.5 to
1.1 or 1.3 equiv) for the cross-coupling afforded lower yields
(entries 4-5). The optimized cross-coupling conditions (1.5
equiv of K+9-, per aryl bromide, (t-Bu3P)2Pd (2.5 mol %) in
toluene at 90 °C) provide a simple and practical protocol that
will be used in the substrate survey that follows.43
Two other substituted arylsilanolates, K+4- and K+5-, which
bear a methoxy group at either the 3- or the 2-position, led to
good yields with all of the aryl bromides tested (Table 6, entries
8-13). Interestingly, K+4- reacted with 4-bromochlorobenzene
to provide a 4-chlorobiaryl (entry 10). Although the reaction
times for (2-methoxyphenyl)dimethylsilanolate K+5- were
longer (8 h), the biaryl products were formed efficiently
(including the 2,2′-diortho-substituted biaryl from coupling with
2-bromotoluene (entry 13)).47
3. Preparative Cross-Coupling Reactions of Aryl
Silanolates. 3.1. Cross-Coupling of Electron-Rich Alkali-Metal
Silanolates with Aryl Halides. For the initial survey of cross-
coupling reactions with aryl bromides under the newly devel-
oped conditions, K+1- was chosen as the silanolate coupling
partner (Table 6). This arylsilanolate was used successfully in
previous studies and should react in a similar fashion under the
Next, other electron-rich arylsilanolates bearing different
substituents, as well as electron-neutral silanolates, were
(44) Denmark, S. E.; Wu, Z. Org. Lett. 1999, 1, 1495–1498.
(45) For a recent review on the development of ligands for the cross-
coupling of aryl chlorides see: Bedford, R. B.; Cazin, C. S. J.; Holder,
D. Coord. Chem. ReV. 2004, 248, 2283–2321.
(43) Small amounts of disiloxane were detected during the optimization
that could account for the need for 1.5 equiv of silanolate. However,
the general use of 1.5 equiv of silanolate salts assured that all
subsequent reactions proceeded to completion. Most of the substrates
in this study were not further optimized, so it is unclear whether 1.5
equiv is needed for all combinations of substrates.
(46) Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 2002, 41, 4176–4211.
(47) In none of the coupling reactions was the product of homocoupling
detected by GC-MS or NMR analysis.
9
J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009 3109

A R T I C L E S
Denmark et al.
Table 7. Cross-Coupling of Electron-Rich and Electron-Neutral
examined (Table 7). Thus, potassium (N,N-dimethylaminophen-
yl)dimethylsilanolate K+8- reacted with a number of aryl
electrophiles in short reaction times. No immediate difference
was encountered as electron-rich and electron-deficient aryl
bromides smoothly reacted to give the desired biaryl products
in good yields (entries 1-3).
Potassium Arylsilanolates with Aryl Bromides Using (t-Bu3P)2Pda
The cross-coupling of the electron-rich arylsilanolate
derived from TBS-protected phenol (K+3-) was investigated
with varying results. Although a direct reaction involving
phenol-based nucleophiles was unsuccessful because of the
similarity of the pKa’s of the silanol and phenol, simply
protecting the phenol as a silyl ether was effective. The cross-
coupling of K+3- and 16a afforded a modest yield of 24a.
However, reactions involving sterically hindered and electron-
deficient aryl bromides such as tert-butyl 4-bromobenzoate
(16l) were plagued with significant amount of deprotection
to phenol-based products.
The cross-coupling of electronically neutral silanolates K+2-
proceeded uneventfully. Silanolate K+2- was tested with a
number of aryl bromides including 4-(1,3-dioxolane)bromobenz-
ene (entry 8) providing further evidence for the mildness of
these conditions. Interestingly, the cross-coupling of 2-bro-
moanisole with an alkenyl-substituted arylsilanolate K+6-
required 10 h to reach completion affording 51% of 26j. Other
aryl bromides smoothly coupled with K+6-, but difficulty in
Table 6. Cross-Coupling of Methoxy-Substituted Potassium
Arylsilanolates with Aryl Halides using (t-Bu3P)2Pda
a Reactions employed 1.5 equiv of arylsilanolate. b Yields are for
isolated, chromatographically homogeneous products. c Yields are for
analytically pure products. d Potassium (1-naphthyl)dimethylsilanolate.
purification prevented including these substrates in this report.
Overall, a variety of substituents on the electrophile were
compatible and only minor variations were observed between
substrates. With these promising results in hand, we were
encouraged to investigate the more challenging electron-deficient
aryl silanolates.
3.2. Cross-Coupling of Electron-Deficient Alkali-Metal
Silanolates with Aryl Halides. Previous reports from these
laboratories revealed that electron-deficient silanolates per-
formed poorly in cross-coupling reactions, most likely
because of a slow transmetalation of the silanolate.25 This
failure was a major motivation for the current study and the
reason that K+9- was selected for the optimization. The
demonstration that K+9- performed well under the modified
reaction conditions stimulated the investigation of other
electron-deficient arylsilanolates to determine their compe-
tency as nucleophilic partners for the preparation of biaryls.
Under the optimized conditions, K+9- combined with a
number of aryl bromides (Table 8) encompassing electron-
rich (entries 1-3), electron-poor (entry 4), sterically hindered
(entry 3), and heterocyclic substrates (entry 5) in moderate
to excellent yields (68-92%). Moreover, aryl chlorides
bearing reactive functionalities such as phenols with silicon-
based protecting groups, esters, and ketones underwent
smooth cross-coupling with K+9- (entries 6-8).
Additionally, (4-fluorophenyl)dimethylsilanolate, K+11-, also
afforded good yields of the desired products for a range of
electrophiles (Table 8, entries 9-14). Both electron-rich and
electron-deficient aryl bromides afforded biaryls in good yields
(entries 9-12). Heteroaryl bromides were also competent
electrophiles in coupling with K+11- and provided similar yields
compared to those obtained from reactions with electron-rich
aryl silanolates (entries 13 and 14).
a Reactions employed 1.5 equiv of arylsilanolate. b Yields are for
isolated, chromatographically homogeneous products. c Yields are for
analytically pure products.
9
3110 J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009

Cross-Coupling Reactions of Silanolates with Halides
A R T I C L E S
Table 8. Cross-Coupling of Fluorine-Containing Potassium
Arylsilanolates with Aryl Halides Using (t-Bu3P)2Pda
Table 9. Cross-Coupling of Electron-Deficient Potassium
Arylsilanolates with Aryl Bromides Using (t-Bu3P)2Pda
a Reactions employed 1.5 equiv of arylsilanolate. b Reaction
employed 2.0 equiv of arylsilanolate in THF. c Yields are for isolated,
chromatographically homogeneous products. d Yields are for analytically
pure products. e Products were determined to be >99% pure by 1H
NMR spectroscopy.
a Reactions employed 1.5 equiv of arylsilanolate. b Yields are for
isolated, chromatographically homogeneous products. c Yields are for
analytically pure products.
making recourse to the more bulky tert-heptyl ester 16s which
provided the biaryl in 56% yield (entry 10).
Because both K+9- and K+11- were effective nucleophiles
under the optimized conditions, the use of other electron-
deficient arylsilanolates bearing more sensitive functional groups
was of interest (Table 9). Both tert-butyl ester K+15- (entries
1-3) and acetal-derived K+7- arylsilanolates (entries 4-5)
reacted smoothly and in good yields without any observable
saponification or deprotection. In each case, 4-bromochloroben-
zene 16h was used and preferential cross-coupling reaction
occurred at the more reactive bromine bearing carbon, leaving
the aryl chloride available for further manipulations. Moreover,
the failed cross-coupling of the silyloxy-substituted silanolate
(K+3-) and tert-butyl 4-bromobenzoate (16l) could be remedied
by reversing the roles of the substrates. Thus, combining K+15-
with 16n and 2.5 mol % of (t-Bu3P)2Pd provided 28n in
moderate yield in 6 h (entry 1). Even potassium (4-chlorophen-
yl)dimethylsilanolate K+10- was competent under the cross-
coupling conditions. Biaryls containing an aryl chloride were
formed in good yields and no products arising from multiple
cross-coupling events were detected (entries 6-11). However,
employing 16l as the electrophile led to a low yield of the cross-
coupling product. The yield could be improved slightly by
Another interesting class of compounds that has gained
attention in the medicinal chemistry sector are multisubstituted
chlorinated arenes. These compounds have been investigated
for VEGFr2 and the Src family (Src and YES) kinase inhibition-
based studies and required the use of 2,6-dichlorophenyl
nucleophiles.48 Although the arylboronic acid is the most
commonly used donor, these couplings lead to varying efficien-
cies and overall yields. To improve the effectiveness of this
class of nucleophiles, (2,6-dichlorophenyl)dimethylsilanolate
Na+14- was prepared and investigated. Orienting experiments
with 4-tert-butylbromobenzene 16g under the optimized reaction
conditions led to modest conversions (ca. 80%) as the reactions
stalled (palladium black observed from catalyst death). Conver-
sions were only moderately improved by increasing both the
amount of Na+14- and (t-Bu3P)2Pd, respectively, as unaccept-
able amounts of protiodesilylation were also observed. To
improve the nucleophilicity of the sodium silanolate, 1,4-dioxane
was used as the solvent. Surprisingly, the reaction rate dimin-
ished dramatically but 14 was detected by GC analysis after
(48) Palaniki, M. S. S.; et al. J. Med. Chem. 2008, 51, 1546–1559.
9
J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009 3111

A R T I C L E S
Denmark et al.
2 h and only trace amounts of protiodesilylation were observed.
Unpredictably in THF at reflux (66 °C), no desired product was
observed. However, when the reaction was performed in THF
in a Teflon-sealed vessel at 90 °C using 2.0 equiv of Na+14-
and 7.5 mol % of (t-Bu3P)2Pd, the complete consumption of
16g was observed. This process was successfully employed for
a number of aryl bromides to provide the desired biaryl in
68-76% yields (Table 9, entries 12-14).
bromoindole coupled in high yield thereby highlighting the
mildness of these reaction conditions (entry 5).
Table 11. Cross-Coupling of Sodium (2-Benzofuranyl)-
dimethylsilanolate Na+13- with Aryl Bromides Using (t-Bu3P)2Pda
3.3. Cross-Coupling of Heterocyclic Silanolates. In view of
the enormous interest in heterocycles in the pharmaceutical
industry and to extend the scope of silanolates developed in
these laboratories,38 the cross-coupling of heterocyclic silano-
lates adopting the standard reaction conditions for the synthesis
of biaryls was investigated. Orienting studies were conducted
with the readily prepared potassium (dibenzothienyl)dimethyl-
silanolate K+12-. The scope of the reaction was illustrated using
a representative set of aryl bromides (Table 10). In general, the
cross-coupling products from electron-rich and electron-poor
bromides were obtained in excellent yields (73-96%) in 3 h.
entry
R
no.
solvent
temp., °C time, h product yield, %b
1c 4-OCH3
16a toluene
16n toluene
60
60
60
60
60
60
70
5
5
5
8
5
8
3.5
33a
33n
33r
33c
33t
33u
33v
82
2
3
4
5
6
7
4-OTBS
3-CH2OTBS 16r toluene
2-CH3
99d
71e
95
16c toluene
16t toluene
16u THF
91d
58e,g
59e,g
f
4-NO2
4-CN
16v dioxane
a Reactions employed 1.5 equiv of arylsilanolate. b Yields are for
isolated, chromatographically homogeneous products. c 2.5 mol
%
Table 10. Cross-Coupling of Potassium (Dibenzothienyl)-
dimethylsilanolate K+12- with Aryl Bromides Using (t -Bu3P)2Pda
(t-Bu3P)2Pd was used. d Yields are for analytically pure products. e The
yield is low because of sacrificial purification from polysiloxane side
products. f N-Boc-5-bromoindole. g See ref 49.
Preliminary experiments with electron-poor aryl bromides and
Na+13- did not provide the biaryl products although both
silanolate and aryl bromide were gradually consumed. In
conjunction with the results for the cross-coupling of Na+14-,
ethereal solvents appear to provide a beneficial effect with
sodium silanolates. Therefore, simply changing the solvent to
THF successfully provided 2-(4′-nitrophenyl)benzofuran 33u in
moderate yield (Table 11, entry 6).49 For the cross-coupling
involving 16v, a significant amount of protiodesilylation of the
silanolate was observed in both THF and DME. However,
performing the cross-coupling in dioxane at 70 °C provided 33v
in 58% yield (entry 7).50
entry
R
no.
product
yield, %b
1
2
3
4
4-OCH3
4-CF3
2-OCH3
16a
16i
16j
16q
32a
32i
32j
32q
96 (84)c
93c
89 (81)c
73 (65)c
3-bromothiophene
a Reactions employed 1.5 equiv of arylsilanolate. b Yields are for
isolated, chromatographically homogeneous products. c Yields in
parentheses are for analytically pure products.
With the success of the cross-coupling in the aryl bromide
series, the focus shifted toward aryl chloride substrates. Surpris-
ingly, employing the optimized conditions for the cross-coupling
of Na+13- with 4-chloroanisole 16a′ yielded no desired product.
Although the majority of the aryl chloride remained, a significant
amount of benzofuran from protiodesilylation of Na+13- was
observed. Even executing the reaction at elevated temperatures
(90 °C) did not provide the desired cross-coupling product in
any appreciable amount. At this stage, further optimization was
warranted to identify appropriate conditions for this class of
substrates.
The success of the cross-coupling of alkenyldimethylsilano-
lates with a broad range of aryl chlorides, inspired the
application of these reaction conditions.40 To our delight,
employing SPhos41 and APC in THF smoothly promoted the
cross-coupling of a variety of aryl chlorides with Na+13- (Table
12). Both the silanolate and the aryl chloride were completely
consumed within 5 h to afford good yields (64-85%) of the
biaryl products (entries 1-4). Although the coupling of 3-chlo-
ropyridine did not reach completion in 6.5 h (entry 5), the
desired product was still obtained in 67% yield.
The successes described above for a wide range of
silanolate nucleophiles, encouraged the extension of these
conditions to other substrates that have been elusive to date.
Sodium (2-benzofuranyl)dimethylsilanolate Na+13-, which
failed to provide cross-coupled products using previously
reported conditions, was reinvestigated.38 Using Na+13- and
(t-Bu3P)2Pd as the catalyst, a significant amount of benzofuran
(from protiodesilylation) was observed in the coupling with
aryl bromides at 90 °C. However, simply lowering the
temperature to 60 °C provided good yields of the cross-
coupling products with significantly less protiodesilylation
(Table 11). Both electron-rich (entries 1-3) and sterically
hindered electrophiles (entry 4) afforded the biaryl products
in good to excellent yields (71-99%). Notably, the TBS-
protected 4-bromophenol reacted with Na+13- providing 33n
in nearly quantitative yield (entry 2). Finally, N-Boc-5-
(49) The lower than expected yields were ascribed to problems in
purification. Presumably, 33u and 33v can coordinate to the palladium
which cannot be completely removed by aqueous workup and column
chromatography. Sublimation was required to provide the palladium-
free products judging by the colors of the products. For example, the
yellow solid 33u became an off-white solid after sublimation even
though no change in 1H NMR spectra can be identified. See Supporting
Information for purification procedures.
4. Discussion. The results above represent a significant
expansion of substrate scope for the aryl-aryl cross-coupling
process. Two critical advances are responsible for these
improvements: (1) the introduction and use of stable arylsil-
anolate salts and (2) the discovery that (t-Bu3P)2Pd efficiently
(50) In toluene, the cross-coupling reaction of Na+13- with 4-bromoben-
zonitrile 16v provided a product whose identity cannot be unambigu-
ously confirmed at this time.
9
3112 J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009

Cross-Coupling Reactions of Silanolates with Halides
A R T I C L E S
tions.52 More importantly, with only one exception, the reagents
are isolable, free-flowing solids that are indefinitely stable and
require no additional activation. The majority of arylsilanolates
were prepared by simple deprotonation of the silanols with KH
in THF, except silanolate K+3- which was prepared in benzene
because of its instability in THF. Although the cause of the
decomposition is unknown, the higher coordinating ability of
THF most likely leads to greater charge separation and the
ability to cleave the TBS group.
Table 12. Cross-Coupling of Sodium (2-Benzofuranyl)-
dimethylsilanolate Na+13- with Aryl Chloridesa
entry
R
no.
time, h
product
yield, %b
Two other silanolates were prepared as their sodium salts
for similar reasons. (2-Benzofuryl)dimethylsilanolate (Na+13-)
was prepared by deprotonation of 13 with NaH or NaHMDS
in THF. Attempts to prepare the corresponding potassium
silanolate led to rapid protiodesilylation to afford benzofuran.
More surprising was the inability to prepare the potassium
silanolate from silanol 14. Treatment of 14 with either KH or
KHMDS in a number of solvents provided a dark, viscous oil
which contained no identifiable products. Here again, making
recourse to the sodium silanolate led to the isolation of Na+14-
as a stable, colorless solid. The stability and ease handling of
the arylsilanolate salts were significant because most arylsilanols
are liquids and are prone to dimerization to form disiloxanes in
the presence of trace amounts of acid or base. However, the
silanolate salts are indefinitely stable at room temperature in a
moisture free environment (similar to sodium methoxide) and
their physical properties facilitate handling, transportation, and
storage.
4.3. Mechanistic Insights Into the Cross-Coupling of Aryl-
silanolates. To facilitate the discussion of the influence of
reaction parameters on the optimization and outcome of the
coupling process, a brief summary of the current mechanistic
hypothesis is warranted. On the basis of previous52 and
ongoing26 mechanistic investigations, the current formulation
of the catalytic cycle is illustrated in Figure 1. This scheme has
been adapted to incorporate the catalyst activation step relevant
to the discussion below. The general picture for coupling of
metal silanolate salts involves the displacement of the X group
on arylPdLnX (i) to form a discrete arylPd(L)-O-SiMe2(aryl)
(ii) species. These compounds have been implicated by kinetic
studies,52 as well as from independent synthesis and isolation.26
Transmetalation from ii to arylPd(L)aryl (iii) is generally
believed to be turnover limiting. The reductive elimination from
iii to form the biaryl product and regenerate the catalyst
completes the catalytic cycle.
Although in-depth mechanistic studies are still incomplete,
the results obtained in this investigation are consistent with this
picture. Among the most important observation is the effective-
ness of the current catalyst/ligand combination compared with
other reports involving arylsilanols. Previous conditions em-
ployed APC in combination with bidentate phosphine ligands
to promote the cross-coupling reaction, as well as to prevent
unwanted homocoupling processes. However, we have recently
discovered the crucial requirement of an open coordination site
for promoting transmetalation from ii.26 Thus, the use of a
bidentate phosphine effectively decreases the opportunity for
transmetalation to occur and slows the overall catalytic process.
Fortunately, the use of a bulky ligand such tri(tert-butyl)phos-
phine prevents a second phosphine from coordinating to the
palladium center thus providing an open coordination site
1
2
3
4
5
4-OCH3
2,6-(CH3)2
4-CF3
16a′
16w′
16i′
16v′
16m′
3.5
5
3
3.5
6.5
33a
33w
33i
33v
33m
77
85
82
64d
67
4-CN
c
a Reactions employed 1.5 equiv of arylsilanolate relative to aryl
chloride (1.0 mmol). b Yields are for isolated, chromatographically
homogeneous products. c 3-Chloropyridine. d See ref 49.
catalyzes the coupling while substantially suppressing the
homocoupling, reduction and desilylation that plagued the earlier
protocols. The ease of preparation of a wide variety of
arylsilanolate salts and the simplicity of the coupling procedure
makes these reagents viable alternatives to the more common
organostannanes and -boranes used for similar transformations.
The broad generality of the method has been demonstrated
through an extensive survey of both nucleophilic and electro-
philic partners bearing a variety of electron demand and
substitution patterns.
4.1. Preparation of Arylsilanols. The methods employed for
the synthesis of the arylsilanol component employed two
different kinds of silicon partners, electrophiles (D3, Me2SiClH)
and a nucleophile ((EtO)Me2Si)2. The preparation of arylsilanols
via halogen-lithium exchange and capture with D3 proceeded
uneventfully to yield the products in good to excellent yields
in most cases. For substrates that gave either unstable or
unreactive organolithium compounds, the use of the more
electrophilic dimethylchlorosilane was needed. In general, the
resulting silanes were readily oxidized to the silanols with water
under catalysis by a Ru or Ir complex. However, one unexpect-
edly challenging case warrants further comment. Initial attempts
to prepare (4-[1,3-dioxolane]phenyl)dimethylsilanol 7 by trap-
ping the organolithium species with D3 led to low yields and
dimerization to the disiloxane. Nevertheless, trapping the
aryllithium with dimethylchlorosilane furnished the arylsilane
in 99% yield. Unfortunately, further difficulties were encoun-
tered in the oxidative hydrolysis using either [RuCl2(cymene)]2
or [IrCl(COD)]2 in MeCN/H2O in which significant amounts
of acetal cleavage (4-10%) were observed. Although the
oxidative hydrolysis was successful in basic medium (0.1 M
NaHCO3 solution), deprotection of the acetal was still observed.
The problem was easily solved by performing the oxidative
hydrolysis with tetrabutylammonium hydroxide in aqueous
acetonitrile.32,51 The ability to maintain a basic, metal-free
medium was crucial to afford the desired silanol in excellent
yield.
4.2. Preparation of Silanolate Salts. The potassium salts of
the arylsilanols were used almost exclusively in this study
because they allowed for rapid combination with the palla-
dium(II) halide intermediates to form the requisite palladium(II)
silanolates without promoting undesirable base-induced reac-
(52) Previous mechanistic studies have demonstrated that palladium(II)
silanolates are the intermediates that undergo transmetalation: Den-
mark, S. E.; Sweis, R. F. J. Am. Chem. Soc. 2004, 126, 4876–4882,
ref 26.
(51) Eaborn, C. Organosilicon Compounds; Butterworth: London, 1960;
p 200.
9
J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009 3113

A R T I C L E S
Denmark et al.
Performing optimization studies with K+9- proved critical
to the development of a general set of conditions for the cross-
coupling of arylsilanolates. The diminished reactivity of this
electron-deficient silanolate should affect two key steps in the
catalytic cycle. The rate of both the displacement of the
palladium halide and the transmetalation of the palladium
silanolate would be affected by the lower electron density of
this benzene ring.
The identification of (t-Bu3P)2Pd as the catalyst was crucial
to the success of this program. “Fu’s catalyst” allowed for rapid
oxidative addition with aryl bromides and rapid transmetalation,
thus avoiding unwanted side processes. The use of a pal-
ladium(0) source avoided the necessity for the reduction of a
palladium(II) precatalyst (e.g., APC) and proved advantageous.
Recent studies from these laboratories have demonstrated that
the reduction of APC requires a stoichiometric amount of the
arylsilanolate (with respect to palladium(II)) to produce pal-
ladium(0).39 Alternatively, phosphines can serve as the reducing
agent for the palladium(II) precatalyst by formation of the
corresponding phosphine oxides.54 This is significantly more
wasteful, especially for expensive and sensitive phosphines that
would thus have to be used in superstoichiometric amounts (with
respect to palladium(II)). Moreover, the phosphine oxides are
much less Lewis basic than the strongly electron-rich phosphines
and are ineffective at maintaining catalyst life and promoting
facile oxidative additions to aryl halides.
Figure 1. General catalytic cycle for cross-coupling of silanolate salts.
throughout the course of the reaction. Additionally, the amount
of homocoupling is also negligible with the current catalyst
system. The origin of this result is likely the increased steric
bulk around the palladium center that decreases the opportunity
for either a second oxidative insertion of the aryl halide or
simply a bimolecular recombination from occurring.
Although these studies do not provide direct evidence for
the turnover-limiting step for the catalytic cycle, circumstantial
evidence supports the postulate that with potassium salts,
transmetalation is turnover limiting. Experimental evidence for
this notion comes from lower temperature required for the
coupling of sodium (2-benzofuranyl)dimethylsilanolate
(Na+13-) compared to other sodium arylsilanolates (60 °C
compared to 90 °C). Whereas productive cross-coupling for the
arylsilanolates is observed at lower temperatures, these reactions
quickly stall because of the slower rate of transmetalation. This
difference can be attributed to the key transmetalation event
that forms the Pd-sp2 carbon bond prior to reductive elimina-
tion. Previous mechanistic studies performed in these labora-
tories have demonstrated that displacement of an arylPd(X)Ln
with silanolate salts to form ii (without ligands) occurs readily
at lower temperatures.26b,52 Even though the exact molecular
events in transmetalation are currently unknown, the aromaticity
of the arene ring must be disrupted when palladium atom bonds
to the ipso carbon of the silanolate. Thus, fused heteroarylsil-
anolates, particularly the π-excessive types, are expected to
undergo more rapid transmetalation because of the lesser
energetic cost in interrupting aromaticity53 and greater electron
density at the ipso carbon.
4.5. Scope of Arylsilanolate Structure. The most significant
advance described in this work is the expansion of scope of the
arylsilanolate beyond the 4-anisylsilanolate, K+1-, reported
previously.25 For the silicon-based, cross-coupling reaction to
be competitive with other methods for the synthesis of biaryls,
a much wider range of both coupling partners was needed.
Fortunately, the combination of preformed silanolate salts and
(t-Bu3P)2Pd enabled that expansion of scope. The discussion
below follows the presentation of the different classes of
substrates in the Results section.
4.5.1. Electron-Rich Arylsilanolates. Although the cross-
coupling of K+1- was already well documented, the limited
scope of these studies prompted a full investigation of the
functional group compatibility under the newly developed
conditions.25 As expected, K+1- reacted smoothly with a
number of aryl bromides in good yields (78-88%). In addition,
K+4- and K+5- in which the position of the electron-donating
methoxy group was moved to the meta and ortho positions also
underwent cross-coupling in comparable yields. The reaction
times with K+5- were slightly longer than the other two
substrates and the coupling with the sterically demanding
2-bromotoluene was successful albeit in slightly diminished
yield.
4.4. Optimization of Reaction Conditions. The reaction
conditions developed herein have allowed the successful cross-
coupling of a broad range of arylsilanolate reagents containing
various functional groups. Previous reports from these labora-
tories described the use of arylsilanols in combination with
external activators such as fluoride or Brønsted bases.22,25
Unfortunately, the substrate scope using these protocols was
limited to electron-rich arylsilanols. The introduction of pre-
formed arylsilanolate salts described herein represents significant
operational and preparative advantages. The arylsilanolate
nucleophile can be charged directly into the reaction mixture
with an aryl electrophile in the presence of a palladium catalyst.
The ability to employ other electron-rich silanolates contain-
ing sensitive functional groups illustrates the mildness and
generality of the method. For example, the successful couplings
in the presence of the TBS-protected phenol in K+3-, and the
dimethylamino group in K+8- further demonstrated the mildness
of this protocol. Obviously, the use of fluoride sources to
promote the coupling precludes the use of silicon protecting
groups. The coupling of the stilbene-derived silanolate K+6-
(54) (a) Mason, M. R.; Verkade, J. G. Organometallics 1992, 11, 2212–
2220. (b) Amatore, C.; Jutand, A.; Barki’, M. A. M. Organometallics
1992, 11, 3009–3013. (c) Ozawa, F.; Kubo, A.; Hayashi, T. Chem.
Lett. 1992, 2177–2180. (d) Amatore, C.; Carre, E.; Jutand, A.; Barki’,
M. A. M. Organometallics 1995, 14, 1818–1826. (e) Amatore, C.;
Jutand, A. J. Organomet. Chem. 1999, 576, 254–278.
(53) Clar, E. The Aromatic Sextet; John Wiley and Sons: London, 1972.
9
3114 J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009

Cross-Coupling Reactions of Silanolates with Halides
A R T I C L E S
illustrated that an alkene is also compatible. The double bond
did not adversely affect the desired cross-coupling and the
products were obtained with similar yields compared to other
electron-rich arylsilanolates. Additionally, no products resulting
from a Heck-type process were observed.
to proceed as the concentration of the silanolate decreased. In
addition, the more rapid displacement prevented protiodesilyl-
ation from occurring.
4.5.3. Heterocyclic Silanolates. The importance of biaryls
containing one or more heterocyclic nucleus in medicinal
chemistry and material science cannot be overstated. As part
of this ongoing program we have already reported the ability
to synthesize and successfully couple π-excessive heterocyclic
silanolates without significant protiodesilylation.38 Two hetero-
cyclic silanolates K+12- and Na+13- were investigated here.56
The cross-coupling of K+12- proceeded smoothly with a variety
of aryl bromides in excellent yields without any modification
of the standard conditions. This result is not surprising because
the dibenzothiophene has no serious liabilities.
On the other hand, the cross-coupling of 2-benzofurylsilano-
late Na+13- failed under standard conditions and a significant
amount of protiodesilylation was detected. As was also the case
in the cross-coupling of 2-indolylsilanolate, carrying out the
reactions at lower temperatures led to the smooth cross-coupling
of Na+13- while maintaining the stability of the silanolate
throughout the course of the reaction.38 Attempts to cross couple
Na+13- with electron-deficient aryl bromides failed using (t-
Bu3P)2Pd in toluene, but simply replacing toluene with ethereal
solvents successfully promoted the desired cross-coupling. The
use of ethereal solvents must serve to improve the nucleophi-
licity of sodium silanolate by increasing the charge separation.
Presumably, the consequence of this transformation improves
the rate of displacement of the palladium halide, thereby creating
a more efficient catalytic cycle.
This hypothesis became more apparent in attempts to cross
couple Na+13- with aryl chlorides. Under conditions that were
successful with aryl bromides, both 16a′ and 16c′ failed to
provide the desired products. However, conditions recently
developed for the cross-coupling of aryl chlorides with alk-
enylsilanolates (APC, SPhos, THF, 60 °C) were successful.40
Good yields of the biaryl products were achieved in short
reaction times. These results were surprising because the
mechanism of cross-coupling reaction for bromides and chlo-
rides is believed to be similar. Both bromides and chlorides are
known to undergo oxidative addition with (t-Bu3P)2Pd.42a
Additionally, once the displacement has occurred, the pathways
converge for the different halide families and transmetalation
followed by a facile reductive elimination must proceed for both.
This analysis leaves the displacement as the only problematic
step in the catalytic cycle for the aryl chlorides. Although no
direct evidence has been secured for the differences under these
conditions, these results can be explained by a combination of
factors that influence the rate of displacement on monoligated,
T-shaped arylpalladium halide complexes. First, the displace-
ment of a bromide should be more rapid compared to displace-
ment of a chloride on the basis of the affinities of the halides
for Pd(II).57 Second, the sodium silanolate is a poorer nucleo-
phile compared to the potassium silanolates that coupled
successfully with chlorides using (t-Bu3P)2Pd.25b,40 Finally, the
lesser steric bulk around the phosphorus atom in SPhos
compared to (t-Bu3P)2Pd also contributes to the more facile
attack of the silanolate on the palladium center. To understand
the effect of these ligands, various derivatives of SPhos were
employed in the cross-coupling reaction of 4-chloroanisole in
THF at 60 °C (Scheme 1). The use of biphenyl ligand L1
4.5.2. Electron-Deficient Arylsilanolates. The cross-coupling
of electron-deficient arylsilanols represented one of the greatest
challenges to this enterprise. Only one example of this class of
arylsilanols had been reported previously (9), and the results
were far from optimal (low yields with CsOH as the activator
at 110 °C in toluene). The forcing conditions limited the scope
of aryl electrophiles that could participate. Presumably, the
problem is associated with the poor equilibrium concentration
of active silanolate because of the ready and favorable formation
of the disiloxane. In addition, the electron-deficient nature of
the silanolate leads to slower displacement and transmetalation
(see Section 4.3). However, these problems were effectively
overcome by the use of the preformed potassium silanolate
K+9-. Silanolate K+9- was resistant to disiloxane formation
and underwent efficient cross-coupling with a wide range of
electrophiles in moderate to good yields. Similarly, 4-fluoro-
phenylsilanolate K+11- afforded biaryls in very good yields
from aryl bromides representing a wide range of electronic and
steric character. In addition, 4-chlorophenylsilanolate K+10- also
underwent cross-coupling with range of aryl bromides albeit in
lower yield compared to K+11-. The reason for the attenuated
yields is not clear, but no products arising form cross-coupling
at the chlorine-bearing carbon were detected.
Arylsilanolate K+15- is not only electron deficient, but it also
contains the reactive carboalkoxy functional group. The ester
survived the reaction conditions to provide biaryls in good
yields. The decreased nucleophilicity of this arene was reflected
in the increased reaction times (6-7 h) required to reach
completion and lack of cleavage of the tert-butyl ester compared
to other silanolates.55 In addition, the decreased nucleophilicity
of K+15- allowed for the successful formation of 28n by
combination with 16n. Attempts to prepare 28n from the
alternative pair wise combination of K+3- with 16l led to
increased amounts of phenol-based products resulting from
deprotection of 16l.
Although silanolate K+7- containing a dioxolane protective
group is not rigorously an electron-deficient arene, it nonetheless
could be prepared easily and underwent smooth cross-coupling
with a number of aryl bromides in excellent yields. The parent
aldehyde was not compatible with these reaction conditions.
Silanolate Na+14- also proved to be an extremely interesting
example both preparatively and mechanistically. Attempts to
directly translate the optimized conditions to this substrate were
ineffective. Although the product was detected, the reaction
quickly stalled with the formation of palladium black. Moreover,
significant amounts of protiodesilylation were observed by GC/
MS leading to an overall decrease in the concentration of active
silanolate. These two observations lead to the hypothesis that
the sodium silanolate along with the steric environment were
dramatically slowing down the displacement step. To improve
the nucleophilicity, ethereal solvents were surveyed and
were found to improve the lifetime of the sodium salt with only
trace amount of protiodesilylation observed. Presumably, by
increasing the charge separation of the two ions, the nucleo-
philicity of Na+14- was improved and allowed for the reaction
(55) Langanis, E. D.; Chenard, B. L. Tetrahedron Lett. 1984, 25, 5831–
(56) The preparation and coupling of p-deficient heterocyclic silanolates
will be the subject of a forthcoming report.
5834.
9
J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009 3115

A R T I C L E S
Denmark et al.
provided no trace of biaryl product although the silanolate was
unproductively consumed to benzofuran. Interestingly, the more
bulky tert-butyl substituted ligand L2 promoted the cross-
coupling, albeit more slowly compared to SPhos which bears
two cyclohexyl substituents. The small change in the steric
environment around the palladium center is critical to the success
of this reaction but a concrete conclusion cannot be provided
at this time.58
prolonged heating was required (>29 and >19 h) compared to
the reactions of the corresponding potassium silanolates (3.5
and 3.5 h). The difference in reaction time implies that
displacement becomes competetive as the turnover-limiting step
in the catalytic cycle. Significant amounts of disiloxane and
unconsumed bromo ester 16l were observed in the reaction with
Na+10-. From these findings, the use of the sodium arylsilano-
lates (Na+3- and Na+10-) did not offer any preparative
advantage over the potassium analogues.
Scheme 1
4.6. Scope of Electrophile Structure. 4.6.1. Electron-Rich
Electrophiles. A wide range of substituted, electron-rich aryl
bromides were surveyed in the cross-coupling with arylsilano-
lates. The position of the functional group on the electrophile
had little influence on the overall process. For example, the
cross-coupling of K+1- with 4- and 2-bromotoluene afforded
the biaryl compounds 20b and 20c in the same yield and reaction
time. This result demonstrates the insensitivity the catalyst feels
due to steric effects. However, this conclusion must be tempered
by the results from the reaction of K+9- with 2- and 4-bro-
moanisole, which afforded the biaryl products 17a and 17j in
92% and 78% yield, respectively. From these results it is clear
that the heteroatom plays a role in the overall effectiveness of
this process. Coordination of the Lewis basic aryl oxygen to an
open coordination site at the palladium center lowers the reaction
rate. Subsequently, partial inhibition leads to lower yields as
compared with the methyl-substituted systems. However, unlike
other systems reported previously, the coordination retards, but
does not completely inhibit, the reaction.27
4.6.2. Electron-Deficient Electrophiles. Generally, the cross-
coupling of arylsilanolates with electron-poor electrophiles
afforded good to excellent yields of the biaryl products.
Reactions of electron-rich silanolates K+2- and K+8- with
electron-poor bromides yielded the biaryl products in good
yields (71-88%). Similarly, cross-coupling of electron-poor
silanolates K+9- and K+11- with tert-butyl 4-bromobenzoate
gave the biaryl products in almost identical yields (68% and
70%) albeit with slightly higher catalyst loading for the less
nucleophilic silanolate K+9-. The mildness of this protocol is
illustrated by the competency of esters, ketones, and acetals to
be present without any appreciable reaction occurring at these
sensitive functional groups.
4.6.3. Heterocyclic Bromides. Previously, the cross-coupling
of arylsilanols with heteroaryl bromides was only moderately
successful. However, the reaction conditions described herein
successfully promote the cross-coupling without any decrease
in reaction rate or yield. Reactions of 3-bromoquinoline,
3-bromopyridine, and 3-bromobenzothiophene with a variety
of arylsilanolates afforded biaryl compounds in good yields
(61-85%) with no significant differences in rate. The success
of these reactions is attributed to facile catalyst turnover without
interference of the heteroatom binding to the palladium center
that would result in stalling.
4.6.4. Aryl Chlorides. The cross-coupling of aryl chlorides
is more challenging because oxidative addition of C-Cl bonds
is more difficult compared to C-Br bonds. Furthermore, the
halide displacement from Aryl-Pd(X)Ln is further attenuated by
the greater affinity of chloride for palladium(II).58 This obstacle
has been overcome and arylsilanolates smoothly cross-couple
with aryl chlorides, as illustrated by the successful reaction of
K+1- with 2- and 4-chlorotoluene to afford 20b and 20c.
Similarly, reaction of electron-deficient silanolate K+9- with
various aryl chlorides (e.g., TBS-protected 4-chlorophenol, tert-
butyl-4-chlorobenzoate, and 4-chlorobenzophenone) afforded the
4.5.4. Nucleophilicity and Basicity of Aryl Silanolates. Previ-
ous studies in these laboratories demonstrated that the counterion
of an in situ generated silanolate has a dramatic effect on the
rate of reaction.25b Activators such as Na2CO3 and K2CO3 did
not promote the cross-coupling of 1 with ethyl 4-iodobenzoate.
In contrast, 8% and 45% conversion of the iodide was observed
using Rb2CO3 and Cs2CO3 in 24 h, respectively. Furthermore,
the dependence of the alkali metal was also seen in the
alkenyl-aryl cross-coupling reactions in which the potassium
salt of (E)-heptenylsilanolate reacted significantly faster than
did the sodium salt.40 Thus, the expected correlation between
the size of the alkali metal and the nucleophilicity of the
silanolate is seen.59 Larger alkali metals typically form more
ionic metal-oxygen bonds, and consequently, these alkali metal
silanolates are more nucleophilic than the silanolates of smaller
alkali metals.
With this observation in mind, Na+3- and Na+10- were
prepared and the reactivity of these nucleophiles was evaluated
in (t-Bu3P)2Pd-catalyzed cross-coupling reactions. Electrophiles
bearing either carboxylic esters or silicon protecting groups were
surveyed. These substrates were chosen because lower yields
of the biaryl products 24a, 24c, 30n, 30r, and 30l were obtained
using the corresponding potassium salts K+3- and K+10-. It
was hypothesized that decreasing the nucleophilicity of the
aryldimethylsilanolate could potentially minimize the undesired
side reactions such as cleavage of the TBS-protected phenol
and tert-butyl ester. Additionally, the formation of the phenol
from 30l via SNAr pathyway may be avoided. The cross-
coupling of Na+3- and Na+10- with 16a, and 16n provided
the desired biaryls 24a and 30n by GC analysis. However,
(57) (a) Flemming, J. P.; Pilon, M. C.; Borbulevitch, O. Y.; Antipin, M. Y.;
Grushin, V. V. Inorg. Chim. Acta 1998, 87–98. (b) Stambuli, J. P.;
Weng, Z.; Incarvito, C. D.; Hartwig, J. F. Angew. Chem., Int. Ed.
2007, 46, 7674–7677.
(58) The primary influence of the ligand was also illustrated by carrying
out the cross-coupling of Na+13- and 16a′ with APC and SPhos
in toluene at 90 °C. After 5 h, complete conversion of 16a′ to 33a
was observed with a negligible amount of protiodesilylation
detected (GC).
(59) Kurts, A. K.; Sakembaeva, S. M.; Beletskaya, I. P.; Reutov, O. A.
Zh. Org. Chim. 1974, 10, 1572–1580.
9
3116 J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009

Cross-Coupling Reactions of Silanolates with Halides
A R T I C L E S
(15 mL) and dried (MgSO4, 5 g). The volatiles were removed in
vacuo (30 °C, 10 mmHg) to afford a viscous, brown oil. Purification
by MPLC (24 g SiO2, 30 mL/min, hexane/ethyl acetate 9:1 (three
column volumes), gradient to hexane/ethyl acetate 3:2 (five column
volumes), hexane/ethyl acetate 3:2 (six column volumes)) followed
by sublimation (80 °C, 0.1 mmHg) afforded 198 mg (80%) of 23e
as analytically pure yellow plates. Data for 23e: mp ) 127-129
°C; 1H NMR: (500 MHz, CDCl3) 9.19 (d, 1 H, J ) 2.2 Hz, HC(2)),
8.22 (d, 1 H, J ) 2.2 Hz, HC(7)), 8.11 (d, 1 H, J ) 8.4 Hz, HC(3)),
7.84 (d, 1 H, J ) 8.4 Hz, HC(6)), 7.67 (ddd, 1 H, J1 ) 1.4 Hz, J2
) 7.0 Hz, J3 ) 8.3 Hz, HC(4)), 7.64 (d, 2 H, J ) 8.9 Hz, HC(2′)),
7.54 (ddd, 1 H, J1 ) 1.4 Hz, J2 ) 7.0 Hz, J3 ) 8.3 Hz, HC(5)),
6.86 (d, 2 H, J ) 8.9 Hz, HC(3′)), 3.03 (s, 6 H, H3C(5′)); 13C NMR:
(126 MHz, CDCl3) 150.4 (C(1′)), 149.9 (C(2)), 146.7 (C(8)), 133.8
(C(4′)), 131.3 (C(7)), 129.1 (C(3)), 128.6 (C(4)), 128.3 (C(1)), 128.0
(C(2′)), 127.7 (C(6)), 126.7 (C(5)), 125.4 (C(9)), 112.8 (C(3′)), 40.5
(C(5′)), 40.4 (C(5′)); IR: (KBr) 2952 (w), 2849 (w), 2798 (w), 1602
(s), 1525 (s), 1355 (m), 1292 (w), 1212 (m), 1061 (w), 950 (w),
813 (s); MS: (EI, 70 eV) 248 (M+, 100), 232 (12), 218 (1), 204
(12), 176 (4), 151 (2), 124 (10), 102 (6), 88 (4); HRMS: Calcd for
C17H16N2 (248.1314); Found: 248.1312; TLC: Rf ) 0.26 (hexanes/
ethyl acetate, 3:2); Anal. Calcd for C18H20O3: C, 82.22; H, 6.49;
N, 11.28, Found: C, 81.88; H, 6.46; N, 11.23.
Preparation of tert-Butyl-4′-(trifluoromethyl)biphenyl-4-carb-
oxylate (17l) from tert-Butyl-4-chlorocarboyxlate Using K+9-.
To an oven-dried, 5-mL, round-bottomed flask equipped with a
magnetic stir bar, reflux condenser, and three-way argon adapter
was charged (t-Bu3P)2Pd (12.8 mg, 0.025 mmol, 2.5 mol %) as a
colorless solid in a drybox. The flask was sealed away from the
atmosphere and removed to a hood. Dry toluene (2 mL) was added
via syringe through the three-way adapter resulting in a colorless
solution upon stirring. tert-Butyl-4-chlorocarboxylate (213 mg, 1.0
mmol) and K+9- (388 mg, 1.5 mmol, 1.5 equiv) were added
sequentially by removal of the adapter, adding the reagents as liquid
and solid, respectively, and replacing the adapter as quickly as
possible. The flask was placed into a preheated 90 °C oil bath and
stirred at 90 °C under a static pressure of argon. After 5 h, the
reaction was cooled to room temperature and poured onto water
(10 mL). The aqueous layer was extracted with ethyl acetate (3 ×
20 mL) and the organics washed with brine (15 mL) and dried
(MgSO4, 5 g). The volatiles were removed in vacuo (30 °C, 10
mmHg) to afford a viscous, brown oil. Purification by MPLC (12
g SiO2, 30 mL/min, hexanes (3 column volumes), ramp to hexane/
ethyl acetate 3:1 (20 column volumes), hexane/ethyl acetate 3:1 (5
column volumes)) followed by recrystallization (EtOH) afforded
200 mg (62%) of 17l as analytically pure, colorless plates. Data
for 17l: mp ) 116-117 °C; 1H NMR: (500 MHz, CDCl3) 8.09 (d,
2 H, J ) 8.5 Hz, HC(3)), 7.72 (app. S, 4 H, HC(2′) and HC(3′)),
7.64 (d, 2 H, J ) 8.5 Hz, HC(2)), 1.63 (s, 9 H, H3C(2′′)); 13C NMR:
(126 MHz, CDCl3) 165.4 C(5), 143.7 C(1), 143.5 C(1′), 131.7 C(4),
130.1 C(2′), 130.0 (q, J ) 32 Hz, C(4′)), 127.6 C(2), 127.1 C(3),
125.8 (q, J ) 3.9 Hz, C(3′)), 124.1 (q, J ) 271 Hz, C(5′)), 81.3
C(1′′), 28.2 C(2′′); 19F NMR: (470 MHz, CDCl3) -62.9; IR: (KBr)
2983 (w), 2935 (w), 1711 (s), 1613 (w), 1395 (m), 1375 (w), 1330
(m), 1300 (m), 1255 (w), 1171 (m), 1123 (m), 1075 (m), 1001
(w), 834 (s), 777 (s), 736 (m), 702 (w); MS: (EI, 70 eV) 322 (M+,
10), 303 (4), 266 (100), 249 (44), 221 (4), 201 (17), 183 (8), 152
(20), 57 (24); TLC: Rf ) 0.53 (hexanes/ethyl acetate, 9:1). Anal.
Calcd for C18H17F3O2: C, 67.07; H, 5.32. Found: C, 67.08; H, 5.54.
respective biaryl adducts in moderate yields. The cross-coupling
of 4-chlorobenzophenone is a further illustration of the mildness
of the reaction as the ketone function survived the reaction
conditions.
4.7. Other Cross-Coupling Reactions Using Arylsilanolates.
During the course of these studies, a number of pair wise
combinations were found to reach high conversions, but the
products could not be obtained in high yields because of
difficulties in isolation. Namely, those substrates that lacked
polar functional groups were not easily separated from the
polysiloxane byproducts. For example, K+2- smoothly coupled
with 3-bromopyridine, reaching 100% conversion within 5 h.
However, the low polarity of 25m complicated the removal of
the polysiloxanes and a clean product could not be obtained.
Another disappointing coupling resulted from the combination
of Na+14- with 2-bromotoluene. The desired product was
identified by GC/MS analysis but the isolated yield was lower
than expected (<50%).
A second class of substrates that were difficult to isolate
derived from those couplings involving K+6-. Cross-coupling
of K+6- with both 16a and 16i reached 100% conversion as
determined by GC analysis. However, the products could not
be separated easily from byproducts that contained the stilbene
moiety. Even recrystallization did not provide pure products as
compounds containing the stilbene unit would co-crystallize.
Presumably, the purification of the product could be simplified
using less K+6-.
Conclusion
A broadly applicable protocol for the cross-coupling of alkali-
metal, aryl-, and heteroarylsilanolates with aromatic bromides
and chlorides has been developed. Under catalysis by (t-
Bu3P)2Pd, a wide range of electron-rich, electron-poor, and
sterically hindered aryldimethylsilanolates underwent smooth
coupling with a wide range of aryl halides. The advantages of
using the preformed silanolate salts include their ease of
synthesis from inexpensive precursors, stability to storage,
resistance to disiloxane formation, and self-activating properties.
The broad substrate scope, functional group compatibility, and
anhydrous conditions for cross-coupling bode well for the
adoption of this method particularly in cases where the tradi-
tional boron- or tin-based reagents are problematic. The
development and application of silanolates derived from both
π-deficient and π-excessive heteroaromatic compounds are
currently under investigation.
Experimental Section
General Experimental. See the Supporting Information.
Preparation of 3-(4-N,N-Dimethylaminophenyl)quinoline (23e)
from 3-Bromoquinoline Using K+8-. To an oven-dried, 5-mL,
round-bottomed flask equipped with a magnetic stir bar, reflux
condenser, and three-way argon adapter was charged (t-Bu3P)2Pd
(12.8 mg, 0.025 mmol, 2.5 mol %) as a colorless solid in a drybox.
The flask was sealed away from the atmosphere and removed to a
hood. Dry toluene (2 mL) was added via syringe through the three-
way adapter resulting in a colorless solution upon stirring. 3-Bro-
moquinoline (136 µL, 1.0 mmol) and K+8- (350 mg, 1.5 mmol,
1.5 equiv) were added sequentially by removal of the adapter,
adding the reagents as liquid and solid, respectively, and replacing
the adapter as quickly as possible. The flask was placed into a
preheated 90 °C oil bath and stirred at 90 °C under a static pressure
of argon. After 3 h, the reaction was cooled to room temperature
and poured onto water (10 mL). The aqueous layer was extracted
with ethyl acetate (3 × 20 mL) and the organics washed with brine
Preparation of 2-(4-Methoxyphenyl)benzofuran (33a) from
4-Chloroanisole Using Na+13-. To an oven-dried, 5-mL, round-
bottomed flask containing a magnetic stir bar equipped with a reflux
condenser and a 3-way adaptor fitted with a septum, APC (9.1 mg,
0.025, mmol, 2.5 mol %) and 2-dicyclohexylphosphino-2′,6′-
dimethoxybiphenyl, SPhos (20.5 mg, 0.05 mmol, 5 mol %) were
combined as solids. The flask was sequentially evacuated and
purged with argon twice. THF (1.0 mL) was added via a syringe
resulting in a faint yellow solution upon stirring. 4-Chloroanisole
(125 µL, 1.0 mmol) was added followed by Na+13- (322 mg, 1.5
9
J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009 3117

A R T I C L E S
Denmark et al.
mmol, 1.5 equiv). The reaction vessel was placed into a preheated
60 °C oil bath and the mixture was stirred at 60 °C under a static
pressure of argon. After 3.5 h, the mixture was cooled to room
temperature, filtered through SiO2 (11 g, 20 mm diameter), and
the column was eluted with Et2O (100 mL). The filtrate was
concentrated and dissolved in a minimal amount of CH2Cl2 to give
a brown solution. Filtration through a plug of SiO2 (20 mm × 20
mm) eluting with CH2Cl2 (50 mL) gave a yellow solution.
Purification by recrystallization (EtOH) afforded 177 mg (77%) of
33a as colorless flakes. The spectroscopic data matched those from
the literature and was free of major impurities.60 Data for 33a: mp
NMR: (126 MHz, CDCl3) 159.5, 156.0, 154.7, 129.5, 126.4, 123.7,
123.3, 122.8, 120.5, 114.2, 111.0, 99.6, 55.4; MS (EI, 70 eV) 224
(M+, 100), 209 (61), 181 (32), 152 (15), 127 (3), 112 (7), 76 (3),
63 (4); Rf ) 0.22 (hexanes/CH2Cl2, 8:1) [silica gel, UV, CAM].
Acknowledgement. We are grateful to the National Institutes
of Health for generous financial support (GM63167) and Johnson-
Matthey for a gift of Pd2(dba)3. R.C.S. thanks Amgen for a graduate
fellowship and Sigma-Aldrich for an innovation award. W.T.C.
acknowledges the University of Illinois for a graduate fellowship.
J.M.M. acknowledges the NIH for a postdoctoral fellowship.
1
) 148-149 °C (EtOH); H NMR: (500 MHz, CDCl3) 7.81 (d, 2
H, J ) 9.0 Hz), 7.57 (dd, 1 H, J1 ) 7.5 Hz, J2 ) 1.0 Hz), 7.52 (dd,
1 H, J1 ) 7.5 Hz, J2 ) 1.0 Hz), 7.27 (ddd, 1 H, J1 ) J2 ) 7.5 Hz,
J3 ) 1.5 Hz), 7.23 (ddd, 1 H, J1 ) J2 ) 7.5 Hz, J3 ) 1.5 Hz), 6.99
Supporting Information Available: Detailed experimental
procedures, optimization studies, full characterization of all
products, and complete ref 48. This material is available free
of charge via the Internet at http://pubs.acs.org.
(d, 2 H, J ) 9.0 Hz), 6.90 (d, 1 H, J ) 1.0 Hz), 3.87 (s, 3 H); 13
C
(60) Kabalka, G. W.; Wang, L.; Pagni, R. M. Tetrahedron 2001, 57, 8107–
8128.
JA8091449
9
3118 J. AM. CHEM. SOC. VOL. 131, NO. 8, 2009