Uncatalyzed CO2Li-Mediated SNAr Reaction of Unprotected Benzoic Acids via Silicon Trickery

Article abstract of DOI:10.1071/CH15398

The alkyl and aryllithium nucleophilic aromatic substitution (S_NAr) displacement of a fluoro or methoxy group from unprotected 2-fluoro/methoxybenzoic acids is discussed. It was discovered that a TMS group located at the C6-position ortho to the carboxyl group shields effectively the carboxylate against nucleophilic attack, thus reducing dramatically ketone formation, and reorients nucleophilic substitution to the C2-position. The reactions with fluoro-substituted substrate 7 proceed efficiently; in contrast, the use of methoxy-functionalized benzoic acid 8 only affords moderate yields with s-BuLi and PhLi. This uncatalyzed coupling reaction, which provides a direct access to biaryl compounds, presumably proceeds by an addition-elimination sequence via intermediate formation of a resonance-stabilized pentavalent silalactone-Meisenheimer complex.

Full text of DOI:10.1071/CH15398

CSIRO PUBLISHING
Aust. J. Chem. 2016, 69, 307–313
Full Paper
http://dx.doi.org/10.1071/CH15398
Uncatalyzed CO2Li-Mediated SNAr Reaction of
Unprotected Benzoic Acids via Silicon Trickery
A
A
A
Mickael Belaud-Rotureau, Anne-Sophie Castanet, Thi Huu Nguyen,
and Jacques Mortier
A B
,
A
Universite´ du Maine and CNRS UMR 6283, Institut des Mole´cules et Mate´riaux
du Mans, Faculte´ des Sciences et Techniques, Avenue Olivier Messiaen,
72085 Le Mans Cedex 9, France.
B
Corresponding author. Email: jacques.mortier@univ-lemans.fr
The alkyl and aryllithium nucleophilic aromatic substitution (SNAr) displacement of a fluoro or methoxy group from
unprotected 2-fluoro/methoxybenzoic acids is discussed. It was discovered that a TMS group located at the C6-position
ortho to the carboxyl group shields effectively the carboxylate against nucleophilic attack, thus reducing dramatically
ketone formation, and reorients nucleophilic substitution to the C2-position. The reactions with fluoro-substituted
substrate 7 proceed efficiently; in contrast, the use of methoxy-functionalized benzoic acid 8 only affords moderate yields
with s-BuLi and PhLi. This uncatalyzed coupling reaction, which provides a direct access to biaryl compounds,
presumably proceeds by an addition–elimination sequence via intermediate formation of a resonance-stabilized
pentavalent silalactone-Meisenheimer complex.
Manuscript received: 2 July 2015.
Manuscript accepted: 31 July 2015.
Published online: 15 September 2015.
Introduction
group at the C6-position to shield the carboxylate in the adjacent
position, thus reorienting the nucleophilic attack at the C2-
position. The developed strategy relies on the following obser-
vations: (1) based on sparse information, the steric congestion
resulting from the silyl group is able to reduce the electrophili-
city of a carboxylate;[5] (2) arylsilanes are stable to alkyllithium
reagents;[6] (3) arylsilanes bearing an electron-withdrawing
function tolerate ipso-desilylation processes;[7] and (4) intro-
duction of a silyl group at C6 should provide useful precursors
for cross-coupling reactions.[8]
The unprotected carboxyl-directing metalation group exhibits a
remarkable degree of regiocontrol between non-equivalent
ortho centres in directed ortho metalation chemistry (DoM,
Fig. 1).[1] As deprotonation of 2-anisic acid (2) occurs exclu-
sively at the C6-position adjacent to the carboxylate with the
1 : 1 complex s-BuLi/N,N,N0,N0-tetramethylethylenediamine
(TMEDA) at ꢀ788C, a complete reversal of regioselectivity is
observed with the Schlosser–Lochmann superbase (n-BuLi/
t-BuOK).[2] Exposing 2-fluorobenzoic acid (1) to lithium
2,2,6,6-tetramethylpiperidide (LTMP) at ꢀ788C deprotonates
the C3-position ortho to the fluorine, whereas 2-chloro- and
2-bromobenzoic acids ((3) and (4)) are exclusively metalated
in the vicinity of the carboxylate.[3]
Results and Discussion
The original principle of regiocontrol established by the ‘silyl
trick’[9] reported herein is illustrated by the transformations of
2-fluoro- and 2-methoxy-6-(trimethylsilyl)benzoic acids ((7)
and (8)). Compound 8 was prepared by metalation of
2-methoxybenzoic acid (2) followed by chlorotrimethylsilane
(TMSCl) trapping using the in situ quench (ISQ) technique[10]
(89 %, Scheme 1). Because 2-fluorobenzoic acid (1) cannot be
metalated at the position C6 adjacent to the carboxylate (see
Fig. 1), an access to 2-fluoro-6-(trimethylsilyl)benzoic acid (7)
was envisioned through the acid 10, which was easily preparedby
lithiationandcarboxylation of1-bromo-3-fluorobenzene (9). The
halogen-lithium exchange/TMSCl-quench sequence (10 - 7)
was studied under several conditions (Table 1). Under external
quench (EQ) conditions at ꢀ788C, besides the expected product
7, 2-fluorobenzoic acid (1) was obtained in 28% yield (entry 1).
At ꢀ908C, an 82 % yield of the expected product 7 was attained,
and the amount of 1 produced decreased to 10 % (entry 2).
In the absence of aprotic Lewis base additive, the reaction
follows a different pathway. In the benzene series, direct
nucleophilic aromatic substitution (SNAr) coupling of 1 and 2
is observed with common organolithium reagents (n-BuLi,
s-BuLi, t-BuLi, and PhLi) leading to 2-alkyl and arylbenzoic
acids. However, the process is hampered by the simultaneous
susceptibility of the polar species to react with the lithium
carboxylate (nucleophilic addition (NA)), providing ketone
by-products. In the naphthalene series, substitution of the
ortho-fluoro and methoxy groups of 5 and 6 by alkyl, vinyl, or
aryllithiums, lithioamides, or Grignard reagents occurs exclu-
sively, furnishing the corresponding substitution products in
good-to-excellent yields.[4]
We are now able to propose a more rigorous and convenient
approach to circumvent the issue encountered with 1 and 2. This
new approach is based on the introduction of a trialkylsilyl
Journal compilation Ó CSIRO 2016
www.publish.csiro.au/journals/ajc

308
M. Belaud-Rotureau et al.
RLi NA
CO2H
RLi NA
CO2H
RM SNAr
X
DoM
s-BuLi/
TMEDA
DoM
LTMP
RM
SNAr
RLi
RLi
SNAr
CO2H
CO2H
SNAr
X
2
2
OMe
F
X
CO2H
6
6
3
3
n-BuLi/
t-BuOK
DoM
LTMP
DoM
X ꢂ Cl
3
4 X ꢂ Br
1
2
5
6
R ꢂ alkyl, aryl, vinyl
X ꢂ OMe, F
R ꢂ alkyl, aryl, vinyl, NR؅
2
M ꢂ Li, Mg
NA: Nucleophilic addition
DoM: Directed ortho metalation
SNAr: Nucleophilic aromatic substitution
Fig. 1. Reactivity of unprotected benzoic acids towards strong bases.
CO2H
OMe
TMS
LTMP/TMSCI
ISQ
2
8 (89 %)
CO2H
CO2H
2
n-BuLi (n equiv.)
THF, T (ꢀC)
6
TMS
Br
F
F
Br
F
1) LDA (2 equiv.)
ꢁ 1
2) CO2
3) H3Oꢁ
TMSCI
see Table 1
7
9
10 (90 %)
Scheme 1. Preparation of 2-fluoro- and 2-methoxy-6-(trimethylsilyl)benzoic acids (7) and (8).
Accordingly, we were able to obtain the required silylated
compound 7 in pure form by dissolving TMSCl and 10 in
sodium-dried THF, cooling to ꢀ908C, and then adding n-BuLi
(entry 3). In the literature, similar results have been obtained for
other aromatic compounds.[12]
Table 1. Br-Li exchange/(TMSCl or D2O) quench sequence (10 - 7
or 1-6d1)
Entry T [8C]
AddA
nB
EXC
7, 1-6dD1
1
10E
1
2
3
4
5
ꢀ78
ꢀ90
EQ
EQ
2
2
TMSCl 65 (7)
28
0
0
A rather interesting and significant event occurred during this
study. Deuterium was incorporated at the C6-position at a level of
30% by adding n-BuLi to a flask containing 10 dissolved in
sodium dried-THF, subsequently wetted with D2O (entry 4).
A maximal amount of deuterium incorporation of 52% was
obtained when a larger excess of n-BuLi (6equiv.) and D2O
(8 equiv.) were used (entry 5). It follows that the Br-Li exchange/
D2O trapping sequence can compete successfully with the direct
attackofD2O byn-BuLi. Infact, thissituation isnotunique. Forty
years ago, Taylor reported a convenient approach to synthesize
9-tritiated anthracene by addition of n-BuLi to a mixture of
9-bromoanthracene and T2O in ether at ꢀ708C, which implies
that n-BuLi reacts with the bromoarene faster than with tritiated
water.[12] More recently, Capriati et al. showed that directed
ortho-lithiation of diaryltetrahydrofurans can compete success-
fully with protonolysis in protic eutectic solvent mixtures.[13,14]
The existence of a cluster or mixture of aggregates held
together by non-covalent bonds (‘supramolecules’) could
explain why n-BuLi does not react more effectively with water.
Under certain conditions, water can act as an effective ligand for
lithium, and the resulting ‘coordinated’ water exhibits a differ-
ent reactivity from ‘free’ water. The organolithium species RLi
and 10 being involved within aggregates by weak electrostatic
interactions, the topology of the pocket might also be responsi-
ble for the reaction pathway observed after the electrophile
penetrates the cluster.[15] The reluctance of water and its
isotopes (D2O and T2O) to react with n-BuLi at low tempera-
tures (,ꢀ808C) casts doubt on reports regularly quoted in the
scientific literature.[16]
TMSCl 82 (47*) (7) 10
ꢀ90 - rt ISQ
ꢀ90 - rt ISQ
ꢀ90 - rt ISQ
2.5 TMSCl 70 (54*) (7)
0
0
0
0
2.5 D2O
D2O
30 (1-6d1)
52 (1-6d1)
65
48
6
AAddition mode – EQ: external quench (n-BuLi was allowed to react with 10
at T (8C) before addition of the electrophile); ISQ: internal quench (10 and
the electrophile were pre-mixed before addition of n-BuLi).
Bn-BuLi (n equiv.).
CThe amount of electrophile (EX) used was 4 equiv., except for entry 5
(8 equiv.).
DCrude yields (%), estimated by 1H NMR. Isolated yields (recrystallized or
chromatographed) are denoted by an asterisk (*).
EAmount (%) of unreacted starting acid 10.
These results are reproducible; however, 1 and 7 proved difficult
to separate by recrystallization or chromatographic means.
Two reasons can be postulated to explain the formation of 1:
(1) the organometallic intermediate reacts with water traces
contained in the solvent and (2) the intermediate abstracts
hydrogen from the solvent.
Formation of 1 was shown not to arise from the presence of
water traces by changing the order of reagent addition. n-BuLi
was added to sodium-dried THF (this would remove any
moisture traces), and neat 2-bromo-6-fluorobenzoic acid (10)
and TMSCl were added successively. 2-Fluorobenzoic acid (1)
was again found to be present in similar quantities.
In order to suppress this side reaction, it is necessary for the
coupling reagent to be in situ before addition of n-BuLi.[11]

SNAr Reaction of Unprotected Benzoic Acids
309
Table 2. SNAr reactions of 2-fluoro and 2-methoxy-6-(trimethylsilyl)benzoic acids (7) and (8)A,B
COR
CO2H
CO2H
Y
Y
R
X
Y
X
1) RM, THF
2) H3Oꢁ
M ꢂ Li, Mg
1 X ꢂ F, Y ꢂ H
11 Y ꢂ H
13 X ꢂ F, Y ꢂ H
7 X ꢂ F, Y ꢂ TMS
12
Y ꢂ TMS
14 X ꢂ F, Y ꢂ TMS
15 X ꢂ OMe, Y ꢂ H
16 X ꢂ OMe, Y ꢂ TMS
X ꢂ OMe, Y ꢂ H
2
8
X ꢂ OMe, Y ꢂ TMS
a R ꢂ n-Bu
b R ꢂ s-Bu
c
R ꢂ t-Bu
R ꢂ Ph
R ꢂ 4-MeOC6H4
R ꢂ 4-Me2NC6H4
d
e
f
Entry
Acid
RM (2.2 equiv.), conditionsC
11, 12
13–16
1
1
1
1
1
7
7
7
7
7
7
7
7
2
2
8
8
8
n-BuLi, ꢀ788C
0 (11a)
38 (13a)
2
s-BuLi, ꢀ788C
58 (47*) (11b)
68 (63*) (11c)
17 (11d)
39 (13b)
3
t-BuLi, ꢀ788C
0
4
PhLi, ꢀ308C - rt
n-BuLi, ꢀ788C
75 (13d)
5
70 (65*) (12a)
88 (69*) (12b)
80 (72*) (12c)
60 (57*) (12d)
69 (50*) (12e)
60 (43*) (12f)
,5 (12a)
20 (14a)
6
s-BuLi, ꢀ788C
10 (7*) (14b)
7
t-BuLi, ꢀ788C
0
8
PhLi, ꢀ308C - rt
4-MeOC6H4Li, ꢀ308C - rt
4-Me2NC6H4Li, ꢀ308C - rt
n-BuMgBr, 308C
PhMgBr, 308C
0
9
0
10
11
12
13
14
15
16
17
0
17 (14a)
,5 (12d)
25 (11b)
17 (14d)
s-BuLi, ꢀ658C
45 (15b)
t-BuLi, ꢀ788C - 08C
n-BuLi, ꢀ788C
39 (11c)
0
0
44 (14a)
s-BuLi, ꢀ788C
36 (12b)
0
0
PhLi, ꢀ308C - rt
38 (12d)
ACrude yields (%) estimated by 1H NMR. Isolated yields are denoted by an asterisk (*).
BEntries 1–4, 13, 14: see refs.[2,3]
CRM denotes the nucleophile.
6-Silylated benzoic acids 7 and 8 were subjected to a series
of organolithium and Grignard reagents (Table 2). The results
obtained with the unsilylated benzoic acids 1 and 2 are also
provided for comparison purposes. n-BuLi has the stronger
nucleophilic character and the weaker complexing ability
towards the carboxylate. The SNAr/NA selectivity follows
the order n-Bu , s-Bu , t-Bu when the corresponding alkyl-
lithiums are allowed to react with 2-fluorobenzoic acid (1)
(entries 1–3). Introduction of a TMS group in C6 shields
effectively the carboxylate and suppresses in most cases the
undesired nucleophilic addition (entries 5–7). In all cases
studied, the SNAr reaction proceeds smoothly and the aromatic
ring is not lithiated at all. The method described provides
excellent latitude with respect to the synthesis of biaryl com-
pounds (compare entries 8–10 with entry 4).[17] The 2-methoxy
derivative 8 is less reactive than its fluoro counterpart 7.[18]
Alkyl and aryl Grignard reagents are not effective for these
transformations. Removal of the ‘traceless’ trialkylsilyl group
can be accomplished with various sources of fluoride.[6]
that 2,6-disubstituted benzoate esters tetrahedral hydrolysis
mechanisms are disfavoured over SN1 or SN2 processes,[20]
lithium carboxylate would be resistant to NA, provided the
carboxylate is orthogonal to the arene and attack is then
sterically prevented. However, the carboxylate must be coplanar
to the arene for the SNAr. Consequently, the electronic effect
associated with the silicon atom most likely plays the most
important role. If it is assumed that the reaction proceeds via an
addition–elimination sequence,[21] the carboxylate and the fluo-
rine atom could initially provide a residence site for the lithium
cation (complex-induced proximity effect process, CIPE)
(Fig. 2).[22,23] The resulting complex A9 allows the carboxylate
to orientate itself with the aromatic ring in a coplanar fashion.
The transition state leading to B may be envisioned as forming
from A9, where the R group enters from the side almost perpen-
dicular to the aromatic ring (to the p cloud). The ability of the
silicon 3p orbitals to engage in electron-rich three-centre four-
electron (3c-4e) bonding allows Si to expand its coordination
sphere to a pentacoordinate silicon complex, and formation at
a lower energy cost of a silalactone scaffold[24] in the resonance-
stabilized Meisenheimer complex intermediate B must be
envisaged.[25] Hypervalency could also originate from vacant
d orbitals at silicon combined with the potential influence of s*
(Si–C/O) orbitals. However, these two theories are still a matter
of debate.[25] The intramolecular coordination presumably
accelerates the nucleophilic substitution rate. It is noteworthy
It is by no means obvious to rationalize how a bulky
substituent can shield a neighbouring carboxylate against nucle-
ophilic attack so effectively. Although details are yet to be
clarified, we propose the following reaction mechanism. At first
sight, the obstruction of the nucleophilic attack at the carboxyl-
ate can be attributed to trivial steric hindrance caused by the
bulky trialkylsilyl substituent.[6,19] Taking into account the fact

310
M. Belaud-Rotureau et al.
R
Li
F
Me
Me
RLi
Si
Me
ꢃ O
ꢃ
Liꢁ
O
R
O
R
F
ꢃLiF
then Hꢁ
F
A؅ (CIPE)
Me
Me
ꢃ
Li
Si
Me
Si
Me
Me
Me
CO2H
Me
Me
O
Liꢁ
Si
Me
ꢃ
R
RLi
O
O
ꢁ
Li
F
B (Meisenheimer)
pentavalent Si
Li
Me
Me
Si
Me
O
ꢁ
O
Li
A؆ (CIPE)
pentavalent Si
Fig. 2. Plausible reaction mechanism.
that cyclization leading to the silalactone complex A00 might
precede complexation with RLi.
thus reducing dramatically the formation of ketones, and
reorients the nucleophilic substitution to the 2-position.
In the literature, coupling reactions between polar organo-
metallic reagents and halo or alkoxybenzenes are most fre-
quently carried out using transition metal catalysts and aryl
halides.[26] In the pharmaceutical industry, these reactions have
major shortcomings. Because the limits set by health authorities
are very low,[27] the effective removal of transition metals
(e.g. Pd, Cu) in active pharmaceutical ingredients (API) con-
tinues to be a challenge. When the synthetic scheme of an
industrial process requires the use of a metal of significant safety
concern and that the standards of metal content permitted in the
API are exceeded, it is generally necessary to find empirically a
disposal method such as nanofiltration or use of scavenging
agents, which is costly in time and money.[28]
Experimental
Materials and Measurements
All experiments were carried out under argon with anhydrous
THF in dried glassware, using syringe-septum cap techniques.
For standard working practice, see ref.[34] THF was purchased
from Aldrich and dried using the drying station GT S100 (Glass
Technology). n-BuLi and s-BuLi were purchased from Acros
Chemicals and Aldrich Chemical Co. as solutions in hexane and
cyclohexane, respectively, and were titrated periodically against
N-benzylbenzamide.[35] Flash column chromatography was
carried out using Merck Kieselgel 60 silica gel (particle size:
32–63). Analytical TLC was performed using Merck pre-coated
silica gel 60 F-254 sheets. 1H 200 MHz, 1H 400 MHz,
13C 50 MHz, 13C 100 MHz, and 19F 376 MHz NMR spectra were
recorded on a Bruker DPX 200 spectrometer or a Bruker
AC-400 spectrometer. Infrared (IR) spectra were recorded neat
or as thin films using a Nicolet Avatar 370 DTGS Fourier
transform infrared (FT-IR) spectrometer. High-resolution mass
spectroscopy (HRMS) was conducted on a Micromass GCT
Premier. Melting points were measured on a Bu¨chi Melting
Point B-540 apparatus and are uncorrected. Elemental analyses
were performed by the Service de Microanalyse, CNRS ICSN,
Gif-sur-Yvette (France).
In contrast, silicon is considered to be an element of a
relatively low order of toxicity.[29]
Direct (uncatalyzed) coupling of Grignard or lithium
reagents is rarely successful although exceptions are known.
In the Meyers reaction,[30] 2-fluoro- and 2-methoxyaryl oxazo-
lines react with Grignard, organolithium reagents, and lithio
amides usually around ꢀ208C, furnishing the substituted aro-
matics. This reaction has been used for the synthesis of a variety
of natural products and in some cases, this direct approach has
proved superior to transition metal-catalyzed approaches.[31]
Miyano et al. reported that esters and sulfonyl substituents can
take the place of the oxazolinyl group to give similar chelation-
assisted SNAr reactions.[19] Nevertheless, these transformations
require laborious protection and deprotection steps to restore
the carboxylic acid moiety. Deprotection of aryloxazolines
requires the use of acidic conditions (3 M HCl, 12–24 h, reflux),
which are not compatible with delicate structures.[30] Further-
more, 2,6-disubstituted aryloxazolines, benzoate esters, and
benzamides are inert to hydrolysis except in cases where
anchimeric assistance by ortho-introduced electrophiles can
give five- or six-membered ring tetrahedral intermediates,
which greatly enhances hydrolytic rates.[32]
Preparation of 2-Fluoro-6-trimethylsilylbenzoic Acid (7)
6-Bromo-2-fluorobenzoic Acid (10)
Adapted from a literature procedure.[36] To a flask containing
anhydrous diisopropylamine (9.24 mL, 66 mmol) in dry THF
(300 mL) at ꢀ308C, n-BuLi (1.6 M in hexane, 41.3 mL,
66 mmol) was added dropwise. The mixture was then cooled
to ꢀ788C, and 3-bromofluorobenzene (9) (6.6 mL, 60 mmol)
was slowly added. After 2 h of reaction at this temperature, the
reaction mixture was quenched with dry ice. Stirring was
maintained for 3 h, and the mixture was then slowly allowed
to warm to room temperature (rt), and hydrolyzed with water
(300 mL). The aqueous layer was washed with ether
(2 ꢁ 150 mL), acidified with 2 M HCl (pH 1–2), and extracted
with ether (3 ꢁ 150 mL). The combined organic layers were
dried over MgSO4, filtered and concentrated under vacuum.
Purification of the residue by recrystallization (cyclohexane/
ethylacetate) afforded 6-bromo-2-fluorobenzoic acid (10) as a
white solid (11.8 g, 90 %), mp 154–1558C (lit. 152–1548C[37]).
Conclusion
To the best of our knowledge, alkyl and aryl organolithium
displacement of an ortho-fluoro group from unprotected ben-
zoic acids is unprecedented. This facile process, which occurs
under unexpectedly mild conditions (ꢀ78–408C, THF), pos-
sesses considerable potential and opens a route to asymmetri-
cally substituted biaryls.[33] Unprecedented is the finding that a
TMS group shields the carboxylate against nucleophilic attack,

SNAr Reaction of Unprotected Benzoic Acids
311
nmax (neat)/cmꢀ1 2872, 2665, 1696, 1600, 1449, 1297.
dH (CDCl3, 400 MHz) 7.45 (1H, d, J 8.1), 7.32 (1H, dt, J 8.4,
5.8), 7.14 (1H, dd, J 9.4, 1.1). dC (100 MHz, [D6]DMSO) 164.6,
158.2 (d, J 250.0), 132.1 (d, J 8.9), 128.6, 125.7 (d, J 22.0), 118.6
(d, J 5.1), 115.1 (d, J 21.2).
6-Deutero-2-fluorobenzoic Acid (1-6d1) (Table 1, Entry 5)
To a solution of 2-fluoro-6-bromobenzoic acid (10) (0.423 g,
3 mmol) and D2O (0.43 mL, 24 mmol) in dry THF (18 mL) at
ꢀ908C, n-BuLi (1.6 M in hexane, 18 mmol) was added drop-
wise. After stirring of the reaction mixture at this temperature for
2 h, the reaction mixture was allowed to warm to rt over a period
of 6 h, and hydrolyzed with water (30 mL). The aqueous layer
was washed with ether (2 ꢁ 30 mL), acidified with 2 M HCl
(pH 1–2), and extracted with ether (3ꢁ 30mL). The organic
phase was dried (MgSO4), filtered, and concentrated under vac-
2-Fluoro-6-trimethylsilylbenzoic Acid (7)
The synthesis of 2-fluoro-6-trimethylsilylbenzoic acid (7)
under EQ conditions (Table 1, entry 2) is described as follows.
n-BuLi (1.6 M in hexane, 26 mL, 41 mmol) was added
dropwise to a solution of 2-fluoro-6-bromobenzoic acid (10)
(4.38 g, 20 mmol) in dry THF (120 mL) at ꢀ908C. The mixture
was stirred for 2 h, and TMSCl (9.2 mL, 70 mmol) was added.
The solution was hydrolyzed with 1 M NaOH (pH 10–11), and
the aqueous layer was washed with ether (2 ꢁ 50 mL) and
acidified with 2 M HCl (pH 1–2). After extraction of the aqueous
layer with ether (3 ꢁ 60 mL), the combined organic layers were
dried over MgSO4, filtered, and concentrated under vacuum to
give a sticky residue containing 2-fluoro-6-trimethylsilyl-
benzoic acid (7) (82 % yield) and 2-fluorobenzoic acid
1
uum. The crude yield (52 %) was estimated by H NMR spec-
troscopy. dH (CDCl3, 200 MHz) 7.61(1H, m), 7.13–7.31 (2H, m).
Nucleophilic Aromatic Substitution Reactions
of 2-Fluoro-6-trimethylsilylbenzoic Acid (7)
and 2-Methoxy-6-trimethylsilylbenzoic Acid (8)
General Procedure Involving Alkyllithium Nucleophiles
The alkyllithium (10.5 mol) was added dropwise to a solution
of 7 or 8 (5 mmol) in dry THF (20 mL) at ꢀ788C under argon.
After 2 h of stirring, the reaction mixture was allowed to warm
to rt, and hydrolyzed with water (30 mL). The aqueous layer was
washed with ether (2 ꢁ 30 mL). The aqueous layer was acidified
with 1 M HCl (pH 1–2), and extracted with ether (3 ꢁ 30 mL).
The organic phase was dried over MgSO4, filtered, and concen-
trated under vacuum to give the ipso-substitution product, which
was purified by chromatography.
1
(10 % yield) (the yields were estimated by H NMR analysis).
2-Fluoro-6-trimethylsilylbenzoic acid (7) was separated by
chromatography on silica gel (hexane/ether 90 : 10) as a white
solid (2.0 g, 47 %), mp 88.4–898C. nmax (neat)/cmꢀ1 2957, 2654,
1686, 1464, 1291, 1236. dH (CDCl3, 400 MHz) 7.47–7.52 (2H,
m), 7.14 (1H, m), 0.34 (9H, s). dC (CDCl3, 50 MHz) 172.8, 161.0
(d, J 258.0), 144.1, 132.6 (d, J 8.1), 130.6, 124.1 (d, J 10.6),
117.0 (d, J 22.3), 0.54.
2-n-Butyl-6-trimethylsilylbenzoic Acid (12a)
(Table 2, Entry 5)
The synthesis of 2-fluoro-6-trimethylsilylbenzoic acid (7)
under ISQ conditions (Table 1, entry 3) is described as follows.
To a flask containing a stirred solution of 2-fluoro-6-bromo-
benzoic acid (10) (0.876 g, 4 mmol) in dry THF (24 mL) at
ꢀ908C under argon, TMSCl (2.24 mL, 16 mmol) and n-BuLi
(1.6 M in hexane, 6.25 mL, 10 mmol) were added successively
The mixture was slowly allowed to warm to rt over a period of
6 h, and hydrolyzed with water (20 mL). Then, 1 M NaOH was
added (pH 9–10). The aqueous layer was washed with ether
(2 ꢁ 30 mL), acidified with 1 M HCl (pH 1–2), and the aqueous
phase was extracted with ether (3 ꢁ 30 mL). The organic layers
were combined and driedover MgSO4, filtered, and concentrated
under vacuum. Compound 7 was purified by filtration on silica
gel and recrystallized (hexane/toluene) as a white solid (0.46 g,
54 %), mp 88.4–898C.
Compound 12a was obtained according to the general
procedure from 2-fluoro-6-trimethylsilylbenzoic acid (7)
(0.424 g, 2 mmol) and n-BuLi (2 M in hexane, 2.2 mL,
4.4 mmol). Chromatography on silica gel (cyclohexane/diethy-
lether 90 : 10) gave 12a as a white solid (0.331 g, 65 %), mp 76–
77.88C. dH (CDCl3, 400 MHz) 7.46 (1H, dd, J 7.2, 1.6), 7.36
(1H, t, J 7.2), 7.28 (1H, dd, J 7.6, J 1.5), 2.76 (2H, m), 1.73–1.56
(2H, m), 1.49–1.31 (2H, m), 0.94 (3H, t, J 7.2), 0.34 (9H, s).
2-s-Butyl-6-trimethylsilylbenzoic Acid (12b)
(Table 2, Entry 6)
Compound 12b was obtained according to the general
procedure from 2-fluoro-6-trimethylsilylbenzoic acid (7)
(1.19 g, 4.77 mmol) and s-BuLi (1.3 M in hexane, 8.1 mL,
10.5 mmol). Chromatography on silica gel (cyclohexane/
ethylacetate 90 : 10) afforded 12b as a colourless oil (69 %
yield). nmax (neat)/cmꢀ1 2961, 2650, 1686, 1457, 1285, 1246.
dH (CDCl3, 400 MHz) 7.44 (1H, dd, J 7.4, 1.5), 7.40 (1H, t, J
7.6), 7.33 (1H, dd, J 7.6, 1.5), 2.93 (1H, m), 1.76–1.56 (2H, m),
1.27 (3H, d, J 6.8), 0.85 (3H, t, J 7.6), 0.35 (9H, s). dC (CDCl3,
100 MHz) 177.9, 174.5, 144.3, 137.6, 132.2, 129.5, 126.7,
38.1, 31.2, 22.2, 12.3, ꢀ0.6. HRMS m/z 250.1395; calcd for
C14H22O2Si ([M]þ) 250.1389.
2-Methoxy-6-trimethylsilylbenzoic Acid (8)
To a solution of LTMP (9 mmol) in THF (20 mL) at ꢀ788C,
TMSCl (1.38 mL, 10.5 mmol) in THF (5 mL) and 2-methoxy-
benzoic acid (2) (0.46 g, 3 mmol) in THF (3 mL) were added
successively. After gradual warming of the reaction mixture to rt
over a period of 4 h, aqueous 2 M NaOH was added until the pH
reached 10. The aqueous layer was washed with ether, acidified
with aqueous 4 M HCl, and extracted with ether. The combined
organic layers were dried over MgSO4, filtered, and concen-
trated under vacuum. The residue was purified by column
chromatography on silica gel (cyclohexane/ether 90 : 10) to give
2-methoxy-6-trimethylsilylbenzoic acid (8) as a white solid
(0.60 g, 89 %), mp 85.5–87.08C.[2] dH (CDCl3, 200 MHz) 0.33
(9H, s), 3.98 (3H, s), 7.05 (1H, d, J 8.4), 7.37 (1H, d, J 7.8), 7.51
(1H, dd, J 8.4, 7.8). dC (CDCl3, 100 MHz) 0.6 (3C), 56.0, 112.1,
124.2, 127.9, 131.8, 144.4, 157.2, 170.3. nmax (neat)/cmꢀ1 2840,
1681, 1571, 1445, 1243, 1126, 950. Anal. Calcd. for
C11H16O3Si: C 58.89, H 7.19. Found: C 58.94, H 7.12 %.
2-t-Butyl-6-trimethylsilylbenzoic Acid (12c)
(Table 2, Entry 7)
Compound 12c was obtained according to the general
procedure from 2-fluoro-6-trimethylsilylbenzoic acid (7)
(1.19 g, 4.77 mmol) and t-BuLi (1.6 M in hexane, 6.6 mL,
10.5 mmol). Chromatography on silica gel (cyclohexane/
ethylacetate 90 : 10) afforded 12c as a white solid (72 % yield),
mp 163–1658C. nmax (neat)/cmꢀ1 2960, 2637, 1688, 1287, 1244,
1103. dH (CDCl3, 400 MHz) 7.54 (1H, dd, J 8.0, 1.3), 7.47 (1H,

312
M. Belaud-Rotureau et al.
dd, J 7.3, 1.1), 7.36 (1H, dd, J 7.8, 7.6), 1.48 (9H, s), 0.35 (9H, s).
dC (CDCl3, 100 MHz) 179.8, 146.2, 137.8, 136.9, 132.4, 128.9,
127.5, 36.2, 31.5, ꢀ0.6. HRMS m/z 250.1387; calcd for
C14H22O2Si ([M]þ) 250.1389. Anal. Calc. for C14H22O2Si:
C 67.15, H 8.86. Found: C 66.86, H 8.58 %.
(6H, s), 0.38 (9H, s). dC (CDCl3, 100 MHz) 176.8, 149.7, 139.8,
138.3, 137.5, 132.7, 131.0, 129.1, 113.4, 41.0, ꢀ0.6.
Acknowledgements
This research was supported by the CNRS and Universite´ du Maine.
MBR and THN gratefully acknowledge the Soprema Group and the French
Ministry of Research and Technologies for doctoral fellowships. Reviewers
are cordially acknowledged for useful comments and suggestions.
General Procedure Involving Aryllithium Nucleophiles
At ꢀ788C, n-BuLi (4.13 mL, 6.6 mmol, 1.6 M in hexane) was
added slowly to a solution of arylbromide (6.6 mmol) in dry
THF (12 mL). After 1 h of stirring, the mixture was allowed to
warm to ꢀ508C, and the acid (7 or 8) (3 mmol) in THF (4 mL)
was added. The mixture was stirred at rt for 24 h (for 7) or
refluxed for the same period (for 8). Water (30 mL) was added,
and the aqueous layer was washed with ether (2 ꢁ 30 mL) and
acidified with 1 M HCl (pH 1–2). After extraction of the aqueous
layer with ether (3 ꢁ 30 mL), the combined organic layers were
dried over MgSO4, filtered and concentrated under vacuum to
give the ipso-substitution products 12d–f, which were purified
by chromatography or recrystallization.
References
[1] F. Leroux, J. Mortier, in Arene Chemistry: Reaction Mechanism and
Methods for Aromatic Compounds (Ed. J. Mortier) 2015, Ch. 26,
pp. 743–776 (John Wiley & Sons: Hoboken, NJ).
[2] T. H. Nguyen, A.-S. Castanet, J. Mortier, Org. Lett. 2006, 8, 765.
doi:10.1021/OL0530427
[3] F. Gohier, A.-S. Castanet, J. Mortier, Org. Lett. 2003, 5, 1919.
doi:10.1021/OL034491E
[4] (a) R. Aissaoui, A. Nourry, A. Coquel, T. T. H. Dao, A. Derdour,
J.-J. Helesbeux, O. Duval, A.-S. Castanet, J. Mortier, J. Org. Chem.
2012, 77, 718. doi:10.1021/JO202017Z
(b) M. Belaud-Rotureau, T. T. Le, T. H. T. Phan, T. H. Nguyen,
R. Aissaoui, F. Gohier, A. Derdour, A. Nourry, A.-S. Castanet, K. P. P.
Nguyen, J. Mortier, Org. Lett. 2010, 12, 2406. doi:10.1021/
OL1007586
(c) J. Mortier, A.-S. Castanet, A. Nourry, M. Belaud-Rotureau, WO
Patent 2011101604 A1 2011.
3-(Trimethylsilyl)biphenyl-2-carboxylic Acid (12d)
(Table 2, Entry 8)
Compound 12d was obtained according to the general
procedure from bromobenzene (1.04 g, 0.7 mL, 6.6 mmol),
n-BuLi (4.13 mL, 6.6 mmol, 1.6 M in hexane), and 2-fluoro-6-
(trimethylsilyl)benzoic acid (7) (0.64 g, 3 mmol). Recrystalliza-
tion (cyclohexane) gave 12d as a white solid (0.46 g, 57 %), mp
146–1488C. This compound can also be prepared by metalation/
silylation of biphenyl-2-carboxylic acid.[38] nmax (neat)/cmꢀ1
2957, 2558, 1685, 1427, 1299, 1249, 1135. dH (CDCl3,
400 MHz) 7.60 (1H, dd, J 7.4, 1.3), 7.47 (1H, t, J 7.6), 7.40–
7.36 (5H, m), 0.38 (9H, s). dC (CDCl3, 100 MHz) 177.6, 140.9,
140.1, 138.7, 133.6, 130.8, 129.4, 128.5, 128.4, 127.5,
ꢀ0.6. HRMS m/z 270.1078; calcd for C16H18O2Si ([M]þ)
270.1076. Anal. Calc. for C16H18O2Si: C 71.07, H 6.71. Found:
C 71.11, H 6.81 %.
(d) J. Mortier, A.-S. Castanet, M. Belaud-Rotureau, WO Patent
2011101599 A1 2011.
[5] s-BuLi/TMEDA lithiation of 3-methoxy-2-trimethylsilylbenzoic acid
occurs regiospecifically, ortho to the carboxylate (C6) at ꢀ308C. The
ketone is formed with less than 5 % yield: T. H. Nguyen, N. T. T. Chau,
A.-S. Castanet, K. P. P. Nguyen, J. Mortier, Org. Lett. 2005, 7, 2445.
doi:10.1021/OL050761C
[6] (a) See Vol. 2, pp. 431–494 in: Z. Rappoport, Y. Apeloig, The
Chemistry of Organic Silicon Compounds 1998 (Wiley-Interscience:
New York, NY).
(b) M. A. Brook, Silicon in Organic, Organometallic and Polymer
Chemistry 2000 (Wiley: New York, NY).
[7] (a) B. Bennetau, J. Dunogue`s, Synlett 1993, 171. doi:10.1055/S-1993-
22392
2-(4-Methoxyphenyl)-6-trimethylsilylbenzoic Acid (12e)
(Table 2, Entry 9)
(b) Z. Zhao, V. Snieckus, Org. Lett. 2005, 7, 2523. doi:10.1021/
OL0506563
[8] H. F. Sore, W. R. J. D. Galloway, D. R. Spring, Chem. Soc. Rev. 2012,
Compound 12e was obtained according to the general
procedure from 4-methoxy-1-bromobenzene (1.24 g, 0.83 mL,
6.6 mmol), n-BuLi (4.13 mL, 6.6 mmol, 1.6 M in hexane), and
2-fluoro-6-(trimethylsilyl)benzoic acid (7) (0.64 g, 3 mmol).
Chromatography (hexane/ethylacetate 90 : 10) gave 12e as a
white solid (0.45 g, 50 %), mp 164–1668C. nmax (neat)/cmꢀ1
2958, 2639, 1683, 1513, 1299, 1240. dH (CDCl3, 400 MHz) 7.56
(1H, dd, J 7.6, 1.3), 7.43 (1H, t, J 7.6), 7.35 (1H, dd, J 7.8, 1.3),
7.31 (2H, d, J 8.8), 6.90 (2H, d, J 8.8), 3.83 (3H, s), 0.38 (9H, s).
dC (CDCl3, 100 MHz) 177.6, 159.2, 139.7, 138.6, 137.1, 133.3,
130.8, 129.5, 113.9, 55.3, ꢀ0.6.
41, 1845. doi:10.1039/C1CS15181A
[9] (a) C. Metallinos, H. Szillat, N. J. Taylor, V. Snieckus, Adv. Synth.
Catal. 2003, 345, 370. doi:10.1002/ADSC.200390042
(b) M. Schlosser, R. Ruzziconi, Synthesis 2010, 2111. doi:10.1055/
S-0029-1218810
[10] T. Krizan, J. Martin, J. Am. Chem. Soc. 1983, 105, 6155. doi:10.1021/
JA00357A034
[11] Halogen-lithium exchange reactions under in situ quench conditions:
(a) S. El Sheikh, H.-G. Schmalz, Curr. Opin. Drug Discovery Dev.
2004, 7, 882.
(b) S. Goto, J. Velder, S. El Sheikh, Y. Sakamoto, M. Mitani, S. Elmas,
A. Alder, A. Becker, J.-M. Neudo¨rfl, J. Lex, H.-G. Schmalz, Synlett
2008, 1361.
2-(4-Dimethylaminophenyl)-6-trimethylsilylbenzoic
Acid (12f) (Table 2, Entry 10)
[12] R. Taylor, Tetrahedron Lett. 1975, 16, 435. doi:10.1016/S0040-4039
(00)71886-7
[13] V. Mallardo, R. Rizzi, F. C. Sassone, R. Mansueto, F. M. Perna,
A. Salomone, V. Capriati, Chem. Commun. 2014, 50, 8655.
doi:10.1039/C4CC03149K
[14] See also R. Levine, M. J. Karten, W. M. Kadunce, J. Org. Chem. 1975,
40, 1770. doi:10.1021/JO00900A020
[15] D. Seebach, Angew. Chem., Int. Ed. Engl. 1988, 27, 1624. doi:10.1002/
ANIE.198816241
[16] See: J. Mortier, T. H. Nguyen, T. Tilly, A.-S. Castanet, ARKIVOC
2007, vi, 47 and references cited therein.
[17] I. Cepanec, Synthesis of Biaryls 2004 (Elsevier Ltd.: Oxford).
Compound 12f was obtained according to the general
procedure from 4-dimethylamino-1-bromobenzene (1.32 g,
6.6 mmol), n-BuLi (4.13 mL, 6.6 mmol, 1.6 M in hexane), and
2-fluoro-6-(trimethylsilyl)benzoic acid (7) (0.64 g, 3 mmol).
Recrystallization (cyclohexane/ethylacetate 90 : 10) gave 12f
as a white solid (0.404 g, 43 %), mp 182–1848C. nmax (neat)/
cmꢀ1 2955, 2558, 1692, 1609, 1521, 1285, 1250. dH (CDCl3,
400 MHz) 7.53 (1H, dd, J 7.6, 1.5), 7.42 (1H, t, J 7.6), 7.39 (1H,
dd, J 7.8, 1.3), 7.30 (2H, d, J 8.8), 6.71 (2H, d, J 8.8), 2.83

SNAr Reaction of Unprotected Benzoic Acids
313
[18] The fluoride often allows coupling at more sterically congested sites.
See: D. J. Cram, H. E. Katz, I. B. Dicker, J. Am. Chem. Soc. 1984, 106,
4987. doi:10.1021/JA00329A059
21 February 2008, http://www.ema.europa.eu/docs/en_GB/document_
library/Scientific_guideline/2009/09/WC500003586.pdf (accessed
30 June 2015).
[19] (a) T. Suzuki, H. Hotta, T. Hattori, S. Miyano, Chem. Lett. 1990, 19,
[28] (a) K. Ko¨nigsberger, G.-P. Chen, R. R. Wu, M. J. Girgis, K. Prasad,
O. Repic, T. J. Blacklock, Org. Process Res. Dev. 2003, 7, 733.
doi:10.1021/OP034072X
(b) C. J. Pink, H. T. Wong, F. C. Ferreira, A. G. Livingston, Org.
Process Res. Dev. 2008, 12, 589. doi:10.1021/OP800039G
[29] E. M. Carlisle, in 1984 Biochemistry of the Essential Ultratrace
Elements (Ed. E. Frieden) 1984, pp. 257–291 (Plenum Press:
New York, NY).
807. doi:10.1246/CL.1990.807
(b) T. Hattori, M. Suzuki, N. Tomita, A. Takeda, S. Miyano, J. Chem.
Soc., Perkin Trans. 1 1997, 1117. doi:10.1039/A607663G
[20] March’s Advanced Organic Chemistry: Reactions, Mechanisms, and
Structure, 7th edn (Ed. M. B. Smith) 2013, Ch. 16, pp. 1190–1197
(John Wiley & Sons: Hoboken, NJ).
[21] F. Terrier, Modern Nucleophilic Aromatic Substitution 2013 (Wiley-
VCH: Weinheim).
[30] J. Mortier, Curr. Org. Chem. 2011, 15, 2413. doi:10.2174/
138527211796150561
[22] M. N. Whisler, S. MacNeil, V. Snieckus, P. Beak, Angew. Chem., Int.
Ed. 2004, 43, 2206. doi:10.1002/ANIE.200300590
[23] V. Capriati, S. Florio, F. M. Perna, A. Salomone, Chem. – Eur. J. 2010,
16, 9778. doi:10.1002/CHEM.201000897
[24] Shindo disclosed the effect of a carboxyl group cis to a trimethylsilyl
group, achieving fluoride-free cross-coupling reaction: M. Shindo,
K. Matsumoto, K. Shishido, Angew. Chem., Int. Ed. 2004, 43, 104.
doi:10.1002/ANIE.200352705
[31] The direct displacement of alkoxy groups from the b-position of
aromatic and unsaturated esters and ketones by Grignard reagents was
recently reported: A. J. Brockway, M. Gonzalez-Lopez, J. C. Fettinger,
J. T. Shaw, J. Org. Chem. 2011, 76, 3515. doi:10.1021/JO102537N
[32] (a) V. Snieckus, Chem. Rev. 1990, 90, 879. doi:10.1021/CR00104A001
(b) M. I. Page, Angew. Chem., Int. Ed. Engl. 1977, 16, 449.
doi:10.1002/ANIE.197704491
[25] The formation of penta-coordinated silicon compounds with five
carbon ligands is well documented: (a) S. N. Tandura, M. G. Voronkov,
N. V. Alekseev, Top. Curr. Chem. 1986, 131, 99.
[33] J. Mortier, A.-S. Castanet, A. Boussonnie`re, in Arene Chemistry: Reac-
tion Mechanism and Methods for Aromatic Compounds (Ed. J. Mortier)
2015, Ch. 8, pp. 195–217 (John Wiley & Sons: Hoboken, NJ).
[34] M. Schlosser, Organometallics in Synthesis. A Manual 2nd Edn 2002
(Wiley: Chichester).
(b) S. Rendler, M. Oestreich, Synthesis 2005, 1727.
(c) S. E. Denmark, G. L. Beutner, Angew. Chem., Int. Ed. 2008, 47,
1560. doi:10.1002/ANIE.200604943
[35] A. F. Burchat, J. M. Chong, N. Nielsen, J. Organomet. Chem. 1997,
542, 281. doi:10.1016/S0022-328X(97)00143-5
[36] A. J. Bridges, A. Lee, E. C. Maduakor, C. E. Schwartz, Tetrahedron
Lett. 1992, 33, 7499. doi:10.1016/S0040-4039(00)60806-7
[37] F. Mongin, M. Schlosser, Tetrahedron Lett. 1996, 37, 6551.
doi:10.1016/0040-4039(96)01398-6
[26] (a) Metal Catalyzed Cross-Coupling Reactions and More (Eds A. De
Meijere, S. Bra¨se, M. Oestreich) 2014 (Wiley-VCH: Weinheim)
(b) Applied Cross-Coupling Reactions (Lecture Notes in Chemistry
Vol. 80) (Ed. Y. Nishihara) 2013 (Springer: Berlin).
[27] EMEA/CHMP/SWP/4446/2000 (final), Guideline on the Specifica-
tion Limits for Residues of Metal Catalysts or Metal Reagents:
Committee for Medicinal Products for Human Use (CHMP), London,
[38] D. Tilly, S. S. Samanta, A.-S. Castanet, A. De, J. Mortier, Eur. J. Org.
Chem. 2006, 174. doi:10.1002/EJOC.200500469