A novel type of nucleophilic substitution reactions on nonactivated aromatic compounds and benzene itself with trimethylsiliconide anions.

Article abstract of DOI:10.1021/ol015666s

[reaction: see text]. The reaction of fluorobenzene with Me3Si- anion (1) in HMPA at room temperature surprisingly affords o- and p-fluorotrimethylsilylbenzenes (substitution of aromatic H for TMS, 76% yield) 7a and 7b and also 14% of trimethylsilylbenzene (2). Benzene itself reacts at 50 degrees C to furnish 4 in 45% yield. Pyridine affords p-trimethylsilylpyridine quantitatively. Mechanistic studies are presented.

Full text of DOI:10.1021/ol015666s

ORGANIC
LETTERS
2001
Vol. 3, No. 8
1197-1200
A Novel Type of Nucleophilic
Substitution Reactions on Nonactivated
Aromatic Compounds and Benzene
Itself with Trimethylsiliconide Anions
Al Postigo and Roberto A. Rossi*
INFIQC, Departamento de Qu´ımica Orga´nica, Facultad de Ciencias Qu´ımicas,
UniVersidad Nacional de Co´rdoba, Co´rdoba 5000, Argentina
rossi@dqo.fcq.unc.edu.ar
Received February 6, 2001
ABSTRACT
The reaction of fluorobenzene with Me3Si- anion (1) in HMPA at room temperature surprisingly affords o- and p-fluorotrimethylsilylbenzenes
(substitution of aromatic H for TMS, 76% yield) 7a and 7b and also 14% of trimethylsilylbenzene (2). Benzene itself reacts at 50 °C to furnish
4 in 45% yield. Pyridine affords p-trimethylsilylpyridine quantitatively. Mechanistic studies are presented.
The available mechanisms to effect aromatic nucleophilic
substitution vary greatly, depending on the aromatic moiety,
the nucleophile, and the reaction conditions. During the past
five decades, and after the landmark review of Bunnett and
Zahler,1 organic chemists have recognized that aromatic
compounds can undergo nucleophilic substitution just as
easily as they undergo electrophilic substitution.
Trimethylsiliconide anion (Me3Si- (1)), generated from
reaction of hexamethyldisilane with NaOMe, reacts with
p-halotoluenes (Cl, Br, and I) to afford mainly the tri-
methylsilyl-substituted products (63-92%) and toluene as
a minor product (4-26%) in HMPT as solvent.4 There is
30% deuterium incorporation when the reaction is quenched
with D2O. On the other hand, when NaOCH3 is replaced by
NaOCD3 in the reaction of p-iodotoluene and hexamethyl-
disilane, toluene is recovered with 64% of deuterium
incorporation. From these results there appear to be at least
two intermediates involved in this reaction.4
We prepared the highly basic and nucleophilic 1,5 in dry
and de-oxygenated HMPA as solvent and allowed it to react
with haloarenes under nitrogen. There is a fast dark (15 min)
reaction between PhI and 1 to afford PhSiMe3 (2) in 99%
yield. PhBr and PhCl react in the same fashion to afford 2
in 89 and 71% yields, respectively, in 30-min reactions
(Scheme 1).
Being interested in the reactions of triorganyltin ions as
nucleophiles by the SRN1 mechanism and their synthetic
applications,2 we decided to explore the triorganylsiliconide
analogues. Although the reactions of these anions with
haloarenes and haloheteroarenes to afford substitution prod-
ucts have previously been reported, mechanistic studies are
sparse and dubious.3,4
(1) Bunnett, J. F.; Zahler, R. E. Chem. ReV. 1951, 49, 275.
(2) (a) Yammal, C. C.; Podesta, J. C.; Rossi, R. A. J. Org. Chem. 1992,
57, 5720. (b) Yammal, C. C.; Podesta, J. C.; Rossi, R. A J. Organomet.
Chem. 1996, 509, 1. (c) Lockhart, M. T.; Chopa, A. B.; Rossi, R. A. J.
Organomet. Chem. 1999, 582, 229. (d) Co´rsico, E. F.; Rossi, R. A. Synlett
2000, 227. (e) Co´rsico, E. F.; Rossi, R. A Synlett 2000, 230.
(3) (a) Gilman, H.; Wittemberg, D. J. Org. Chem. 1958, 23, 2677. (b)
Benkeser, R. A.; Severson, R. G. J. Org. Chem. 1951, 16, 1424. (c) Brook,
A. G.; Wolfe, S. J. Am. Chem. Soc. 1957, 79, 1431
(5) Kondo, F.; Sakurai, H. J. Organomet. Chem. 1975, 92, C46. Buncel,
E.; Venkatachalam, T. K.; Edlund, U. J. Organomet. Chem. 1992, 437, 5.
Sakurai, H.; Okada, A.; Kira, M.; Yonezawa, K. Tetrahedron Lett. 1971,
19, 1511.
(4) Dervan, P. B.; Shippey, M. A. J. Org. Chem. 1977, 42, 2654.
10.1021/ol015666s CCC: $20.00 © 2001 American Chemical Society
Published on Web 03/22/2001

Scheme 1
When an HMPA solution of p-chlorotoluene is allowed
to react in the dark with 1, only p-trimethylsilyltoluene is
obtained in 90% yield. No meta isomer is observed in the
reaction mixture. Minor amounts of toluene are also found
(<2%). The above reaction performed with an excess of
NaOMe (an excellent hydrogen atom donor to aryl rad-
icals)6 does not lead to an increase in reduction product
toluene.
Figure 1. Relative yields of the disappearance of 3 and the
formation of 4 and 5 in HMPA vs time.
2-Bromo-1,3,5-trimethylbenzene (bromomesitylene) reacts
with 1 in HMPA, furnishing 2-trimethylsilyl-1,3,5-tri-
methylbenzene7 in 48% yield, along with reduction product
mesitylene (30% yield). These facts preclude a benzyne
mechanism to be operating in the substitution reaction of
halobenzenes with 1 in excess, in HMPA as solvent.
Upon reaction of p-dichlorobenzene (3) with 1 in HMPA,
p-chlorotrimethylsilyl benzene7 (4) (15%), p-bis(trimethyl-
silyl)benzene7 (5) (71%), and traces of 2 (3%) are obtained
(Scheme 2). This indicates that the reaction proceeds in a
p-fluorotrimethylsilylbenzenes (7)9 are obtained in 76%
overall yield, as well as some 2 (14%) (experiment 1, Table
1) (Scheme 3).10
Table 1. Reactions of Aromatic Compounds with 1 in HMPA
time
products
(yield, %)
expt
substr (mM)
1, mM
(min)
1
2
3
4
6 (106)
C6D5F (107)
8 (515)c
8 (618)c
8 (494)e
benzene (447)
benzene (5634)
benzene (223)
C6D6 (223)
272
272
927
30
30
120
15
7 (76),a 2 (14)
7b (15), 2b (9)
9 (99)
Scheme 2
742
10 (99)d
5
1480
1480
1878
1790
15
11 (87),f 9 (9)
2 (20
2 (45)
2 (11)
6g
7h
8
2280
2304
2280
C6D5SiMe3 (6.4)
a 7a and 7b elute together in our VPC conditions. A ratio of ca. 60:40
is found for the two isomers according to 1H NMR spectroscopy.
b Deuterated 7a, 7b, and 2; C6D5F was recovered in 78% yield. c Quenched
with water. d Isolated as 9 (92%). e Quenched with ethyl chloroformate.
f Isolated in 78% yield. g Duplicate experiments, ratio C6H6:1 ) 0.3. Yield
of 2 based on benzene. h At 50 °C, ratio C6H6:1 ) 3. Yield of 2 based on
1.
consecutive fashion (i.e., 4 being the precursor of 5) in
contrast to some nucleophilic disubstitutions via the SRN
1
mechanism that do not proceed through the monosubstitution
intermediates.8 There is a fast depletion of 3 concomitant
with the appearance of 4, which then decreases as 5 increases,
indicating a stepwise mechanism (Figure 1).
In all these examples, the substrates suffer an ipso
substitution. Contrasting results are obtained with PhF (6).
When 6 is allowed to react with 1 in HMPA, o- and
A plot of product formation vs time indicates that the
production of 7 and 2 occurs simultaneously, as two
competitive reactions (Figure 2): a fast nucleophilic substitu-
tion reaction by 1 on the benzene ring to yield 7 and a
fluorine-atom replacement by 1 to furnish 2. At prolonged
reaction times (6 h) there is a decrease in the yield of product
7, and small amounts of o- and p-bis(trimethylsilyl)benzenes
(6) Tomaselli, G. A.; Bunnett, J. F. J. Org. Chem. 1992, 57, 2710.
Tomaselli, G. A.; Cui, J. J.; Chen, Q. F.; Bunnett, J. F. J. Chem. Soc.,
Perkin Trans. 2 1992, 9. Bunnett, J. F. Acc. Chem. Res. 1992, 25, 2.
(7) Spectroscopic data for these compounds are in agreement with
published data: Haubold, W.; Herdtle, J.; Gollinger, W.; Einholz, W. J.
Organomet. Chem. 1986, 315, 1.
(8) Bunnett, J. F.; Creary, X. J. Org. Chem. 1974, 39, 3611. Bunnett, J.
F.; Shafer, S. J. J. Org. Chem. 1978, 43, 1873.
Scheme 3
(9) The spectroscopic evidence agrees well with the published spectra,
see: Freeburger, M. E.; Hughes, M.; Buell, G. R.; Tierman, T. O.; Splatter,
L. J. Org. Chem. 1971, 36, 933. See also: Nishimura, J.; Fukurawa, J.;
Kawabata, N. J. Organomet. Chem. 1971, 29, 237. Cartledge, F. K.; Riedel,
K. H. J. Organomet. Chem. 1972, 34, 11. Moerlin, S. M. J. Organomet.
Chem. 1987, 319, 29. Eaborn, C.; Jaura, K. L.; Walton, D. R. M. J. Chem.
Soc. 1964, 1198.
(10) Products 3a and 3b were quantified together. There is 3a:3b ratio
of ca. 60:40.
1198
Org. Lett., Vol. 3, No. 8, 2001

Scheme 4a
a Reagents and conditions:( a) (i) 1 (2 h); (ii) H2O; (b) (i) 1 (15
min); (ii) H2O; (c) (i) 1 (15 min); (ii) ClCO2Et.
Figure 2. Relative yields for the disappearance of 2 and formation
of 3 and 4 in HMPA vs time.
The intrinsic reactivity of 1 toward 6 and 8 led us to
consider benzene itself as a candidate to undergo substitution
by 1. When benzene is allowed to react with 1 under similar
experimental conditions, 2 is obtained in 20% yield (48 h)
(Table 1, experiment 6). When the reaction is performed at
50 °C the yield of 2 rises to 45% (experiment 7, Table 1).
A primary kinetic isotope effect,12 1.7 ( 0.1, is observed
when benzene-d6 is placed in competition with ordinary
benzene (Table 1, experiment 8).
In other competition experiments we determined the
relative reactivity pattern of halobenzenes (toward 1) and
found values for PhI (3.5), PhBr (1.9), and PhF (0.08)15 vs
p-chlorotoluene as unity. These results show that the span
in reactivity between the four halogens toward 1 is, from
PhF to PhI, only about 44.
are obtained (5%).11 o- and p-bis(trimethylsilyl)benzenes
arise from further reaction of o- and p-fluorotrimethylsilyl-
benzenes 7 with 1.
In competition experiments of 6 (426 mM) and per-
deuteriofluorobenzene (C6D5F, 426 mM), toward 1 (426 mM,
2 h), a primary deuterium kinetic isotope effect,12 DKIE
(kH/kD ) 3.2 ( 0.1), is found for the formation of compound
7. This is in agreement with loss of hydride as the
rate-limiting step from a σH adduct. No DKIE is observed
for the formation of product 2 (fluorine as leaving group)
(kH/kD ) 1.0 ( 0.1).
Formation of compound 7 entails loss of hydrogen as
leaving group, whereas to form product 2 the formal leaving
group is fluorine. With loss of hydrogen being a rate-
determining step in the formation of 7 (but not in the
formation of 2), this should also be reflected in the relative
yields of products 7:2 when C6D5F and 6 are used as
substrates, respectively. When C6D5F is allowed to react with
1, a 7:2 product ratio of ca. 1.7 (experiment 2, Table 1) is
observed, in contrast to a 7:2 ratio of 5.4, (experiment 1,
Table 1) when 6 is utilized as substrate.
The formation of substitution products can be rationalized
as arising from attack of 1 on the aromatic moiety to yield
the σH adduct 12 (Scheme 5). This adduct could proceed to
Scheme 5
Given the remarkable reactivity of 1, we estimated that
an aromatic moiety even without a formal leaving group such
as pyridine (8) would be able to react with 1. Upon reaction
of 1 with 8 (2 h) and subsequent quenching with water,
4-trimethylsilylpyridine (9) is obtained in quantitative yield
(99% yield).13 At shorter reaction times (15 min), dihydro
derivative 10 is obtained (99%, isolated as 9 in 92% yield).
When the above reaction (15 min) is quenched with ethyl
chloroformate, urethane 11 is obtained in 87% isolated yield,
along with minor amounts of 9 (9% yield) (Table 1,
experiments 3-5) (Scheme 4).14
(11) Identified by GC-MS and compared with the authentic samples.
(12) Determination of the relative concentrations of deuterated and
nondeuterated products were carried out by GC/MS analysis (Shimadzu
GC-17 A Series, DB5-MS 30 m × 0.2 mm × 0.18 mm film column,
integrated to a Shimadzu GCMS-QP 5050A mass selective detector) by
mass selective integration.
(13) The spectroscopic evidence agrees well with the published spectra,
see: Sulzbach, R. A. J. Organomet. Chem. 1970, 24, 307.
(14) For the methyl chloroformate adduct, see: Fowler, F. J. Org. Chem.
1972, 37, 1321.
products through hydride ion elimination.16 A more plausible
route to products could be hydride shift to form the
pentacoordinate silicon species 13,17 which can persist in an
anionic form until quenching of the reaction, with consequent
(15) Taking only into account the ipso substitution product on substrate
2, i.e., 4, the relative reactivity is ca. 0.016.
Org. Lett., Vol. 3, No. 8, 2001
1199

evolution of molecular hydrogen. Hydrogen gas can also be
envisaged as evolving from a species such as 12 through
deprotonation of the ancillary methyl group from the
trimethylsilyl substituent, to furnish anion 14 (Scheme 5).
However, when the reaction is quenched with methyl iodide,
ion 14 was not trapped. Other mechanistic possibilities may
also be considered.
With respect to halobenzenes, the reactivity order found
(PhI > PhBr > PhCl > PhF) toward 1 is in agreement with
both HME and ET reactions, except that the span in reactivity
is only of about 44. The span in reactivity in competition
experiments of the four halobenzenes toward acetone enolate
anion in SRN1 reactions is over 100 000.18 On the other hand,
PhI reacts with Me3Sn- ions by HME in a fast reaction,
whereas PhCl does not by this mechanism.2a,c
tion of a σX complex that evolves to a pentacoordinate
intermediate, from which the substitution product ensues.19
The reactivity order is also opposed to that found in SNAr
reactions. However, in the mechanism proposed a transfer
of the leaving group (X ) halogen or H in Scheme 5) to
silicon is envisaged, to form a pentacoordinate species and
not to free halide ions, as occurs in ordinary SNAr reactions.
As far as we know, there is no precedent for silylation of
aromatic compounds to replace hydrogen through the action
of trialkylsiliconide ions. The operation of a σ adduct-type
reaction is due to the powerful nucleophilicity of 1 in HMPA
that can effect substitution on substrates otherwise unreactive
with nucleophiles, including benzene.
Further work is in progress to shed more light on the
mechanistic aspects of these reactions, as well as reactions
with other substituted benzene rings and heteroarenes.
Another mechanistic possibility for the ipso trimethylsilyl-
sustitution products from halobenzenes could be the forma-
Acknowledgment. We thank Professor Joseph F. Bunnett
for his comments and valuable suggestions. This work was
supported in part by the SECYT, National University of
Co´rdoba, FONCYT, CONICOR, and CONICET, Argentina.
(16) (a) It is known that under certain conditions the hydrogen atom
can indeed be substituted (vicarious nucleophilic substitution, VNS). This
elimination is assisted by a leaving group (generally X, ArO, RO, ArS)
and therefore does not imply hydride ion elimination. The VNS requires
the presence of an EWG such as nitro or cyano, which stabilizes the σH
adduct.16b Lately, alkyl-hydroperoxide anions have been employed for the
introduction of the hydroxy functionality on the aromatic moiety,16c and
more recently it has been reported that the reaction of nitroarenes with a
large excess of potassium methoxide leads to removal of the ipso hydrogen
to form a dianion.16d (b) Makosza, M.; Miniarski, J. Acc. Chem. Res. 1987,
20, 143. (c) Makosza, M.; Sienkiewicz, K. J. Org. Chem. 1998, 63, 4199.
(d) Kawakami, T.; Suzuki, H. J. Chem. Soc., Perkin Trans. 1 2000, 1259.
(17) Pentacoordinated and other hypervalent states of silicon are well-
known, see: Holmes, R. R. Chem. ReV. 1996, 96, 927. Chuit, C.; Corriu,
J. P.; Reye, C.; Colin Young, J. Chem. ReV. 1993, 93, 1371. Anglada, J.
M.; Bo, C.; Bofill, J. M.; Crehuet, R.; Poblet, J. M. Organometallics 1999,
18, 5584 and references therein.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL015666S
(19) By semiempirical methods (AM1) the energy of the pentacoordinate
complex is much lower than that of the σ adduct (for PhF and PhI). The
intermediate σ adduct does not exist (AM1) for PhCl and PhBr. Calculations
at a higher level of theory are in progress for all the compounds in this
system.
(18) Bunnett, J. F. Acc. Chem. Res. 1978, 11, 413.
1200
Org. Lett., Vol. 3, No. 8, 2001