Metallation reactions. XXV. A re-examination of the metallation reaction of (alkylthio)fluorobenzenes

Article abstract of DOI:10.1016/S0022-1139(99)00097-4

The metallation of (alkylthio)fluorobenzenes by organolithium compounds, lithium amides and butyllithium/potassium tert-butoxide superbasic mixture was re-examined. To avoid the formation of defluorinated compounds all the reactions must be carried out below -80°C. The para-substituted 1a and 1b showed a regiochemistry directed by the halogen; the ortho-derivative 1c underwent metallation ortho to the halogen when treated with lithium tetramethylpiperidide, an α-metallation with butyllithium while sec-butyllithium was less selective. Compounds 1a and 1c allowed the preparation of disubstituted products. At temperatures higher than -80°C increasing amounts of dehalogenated products are formed, whose formation can be explained through the intermediacy of arynes.

Full text of DOI:10.1016/S0022-1139(99)00097-4

Journal of Fluorine Chemistry 98 (1999) 97±106
Metallation reactions. XXV. A re-examination of the
metallation reaction of (alkylthio)¯uorobenzenes1
M.G. Cabiddu, S. Cabiddu*, E. Cadoni, C. Fattuoni, S. Melis
Á
Dipartimento di Scienze Chimiche, Universita di Cagliari, Via Ospedale, 72, Cagliari I-09124, Italy
Received 4 August 1998; accepted 9 April 1999
Abstract
The metallation of (alkylthio)¯uorobenzenes by organolithium compounds, lithium amides and butyllithium/potassium tert-butoxide
superbasic mixture was re-examined. To avoid the formation of de¯uorinated compounds all the reactions must be carried out below
808C. The para-substituted 1a and 1b showed a regiochemistry directed by the halogen; the ortho-derivative 1c underwent metallation
ortho to the halogen when treated with lithium tetramethylpiperidide, an ꢀ-metallation with butyllithium while sec-butyllithium was less
selective. Compounds 1a and 1c allowed the preparation of disubstituted products. At temperatures higher than 808C increasing amounts
of dehalogenated products are formed, whose formation can be explained through the intermediacy of arynes. # 1999 Elsevier Science
S.A. All rights reserved.
Keywords: Metallation; Organolithium; Superbases; Lithium amides; (Alkylthio)¯uoroarenes
1. Introduction
after quenching with iodomethane, the product 2 [2], while
the use of two equivalents of BuLi yielded the isomers 2 and
3, the last one arising from the substitution of a thiomethylic
hydrogen. The same reaction at 508C gave 2, 3, 4 and 5
whereas, at 308C, we found 8 and 9a and traces of 4, 5, 6
and 7.
In a previous paper [2] we reported the monometallation
reaction of (alkylthio)¯uorobenzenes by butyllithium show-
ing that the halogen atom is an ortho-directing group
stronger than the alkylthio. We also reported that our
attempts to bimetallate these molecules were unsuccessful,
leading only to de¯uorinated products. To complete our
study on these substrates we examined several metallating
reagents; butyllithium (BuLi), sec-butyllithium (sec-BuLi),
lithium diisopropylamide (LDA), lithium tetramethylpiper-
idide (LITMP), superbasic mixture obtained by mixing BuLi
and potassium tert-butoxide (LICKOR), in various working
conditions. All the metallated intermediates were quenched
by iodomethane and examined by GC/MS for identi®cation
of products.
The use of 2 molar equivalents of sec-BuLi in cyclohex-
ane, at 808C, gave 2 as main product and small amounts of
3, 10 and 11; using 4 molar equivalents of the same reagent
we obtained a larger amount of 10 (22%): the main product
was 2 with traces of 3, 11, 12 and 13 and the dehalogenated
9b and 14. Using one equivalent of the LICKOR in hexane at
1008C we obtained 2, while two equivalents of the same
reagent led to 10 as the main product (80%) and to small
amounts of 2 and 13. The use of LDA in THF gave
unsatisfactory results in the range of temperatures between
708C and 258C, or using several molar ratios of amide
(from 2 to 4 molar equivalents): we always recovered the
starting material 1a, a small amount of 2 (never more than
17%) and traces of 3. On the other hand, LITMP (1.5 molar
equivalents) in hexane and THF gave 2 and traces of 12 and
15, while using three equivalents of this reagent increased
the amount of 2 up to 58%.
2. Results and discussion
(a) 1-Fluoro-4-(methylthio)benzene (1a) (Scheme 1,
Table 1) in anhydrous tetrahydrofuran (THF) reacted with
1 molar equivalent of BuLi in hexane at 1008C to give,
(b) 1-Fluoro-4-(isopropylthio)benzene (1b) (Scheme 2,
Table 1) reaction with 1 molar equivalent of BuLi at
1008C con®rmed the already reported results [2]: we
obtained product 11 arising from the substitution of a
*Corresponding author. Tel.: +39-70-6758625; fax: +39-70-6758605;
e-mail: scabiddu@vaxcal.unica.it
1For Part XXIV, see [1].
0022-1139/99/$ ± see front matter # 1999 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 9 9 ) 0 0 0 9 7 - 4

98
M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
Scheme 1.
hydrogen ortho to the ¯uorine. Using two or more equiva-
lents of BuLi at 1008C or 858C we obtained 11 and a
small amount of 17. When the reaction was performed at
308C these products were present only in traces with 18,
19 and 20 in almost equal amounts. The use of one equiva-
lent of LICKOR led almost exclusively to 11, while with
two or more equivalents, 11 decreased and the main pro-
ducts were the de¯uorinated 19, 20 and 21.
(c) 1-Fluoro-2-(methylthio)benzene (1c) (Scheme 3,
Table 1) treated with one equivalent of BuLi at 1008C
gave the product 22 and a small amount of 23. The use of
two or more equivalents of BuLi, aimed at the preparation of
the product 24, yielded an increased amount of 23. When the
metallating reagent was sec-BuLi we obtained the same
compounds 22 and 23, with 23 the main product. Two
equivalents of the same reagent gave the disubstituted 24
in an acceptable yield (41%). When we tried to increase this
yield using three or four equivalents of sec-BuLi we noticed
a decrease of the amount of 24 and the formation of four new
products 25, 26b, 29 and 30. The LICKOR gave only
de¯uorinated products (26a, 27, 28, 31) while LDA did
not even react. On the other side, LITMP was shown to be a
good reagent to prepare 23 in good yield, but 24 was
obtained in small amounts and also disubstituted 32.
The above results completed those previously reported
and showed that the occurrence of de¯uorinated products
was minimised by working at temperatures between 708C
and 1008C. The use of iodomethane as electrophile
allowed us to detect even traces of those products through
GC/MS. Scheme 4 shows the possible paths leading to the
products formed from 1a.
The formation of 3 can be explained by a substitution of a
thiomethylic hydrogen (leading to the intermediate A). This
reaction is in competition with the main path involving the
hydrogen ortho to the ¯uorine atom leading to the inter-
mediate B, that can give 2 or, at higher temperatures, react to
the aryne C by elimination of alkali ¯uoride. C would
undergo addition of 1 mol of organolithium (BuLi) to yield

M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
99
Table 1
Metallation of (alkylthio)fluorobenzenes (1a±c)
Entry
RM
(equivalent)
T (8C)
t (min)
St. mat.
recov. (%)
a-S
(%)
o-F
(%)
a-S-o-F
(%)
o,-o0-F
(%)
Defluor.
prod. (%)
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1b
1b
1b
1b
1b
1c
1c
1c
1c
1c
1c
1c
1c
1c
1c
1c
1c
1c
1c
BuLi (1)
100
100
100
50
360
360
360
360
360
240
240
120
360
360
240
240
240
240
360
360
360
360
360
360
240
240
240
240
240
240
240
240
240
240
240
240
240
240
9
<1
<1
<1
<1
19
11
7
91
70
75
68
BuLi (2)
30
25
13
BuLi (4)
BuLi (2)
BuLi (2)
18a
30
ꢀ100b
sec-BuLi (2)
sec-BuLi (4)
sec-BuLi (4)
LICKOR (1)
LICKOR (2)
LDA (2)
80
4
6
53
37
50
93
5
20c
22
80
70
5d
12e
32e
11
100
100
25
7
5
80f
85
80
50
18
15
10
<1
<1
<1
2
3
13
17
42
58
85
82
90
5
LDA (4)
25
LITMP (1.5)
LITMP (3)
BuLi (1)
80
4g
11h
80
100
100
85
BuLi (2)
8
10
Tr
BuLi (2)
BuLi (2)
30
95i
90j
LICKOR (1)
LICKOR (2)
BuLi (1)
100
100
100
100
100
90
ꢀ100
10
15
30
40
42
34
25
12
10
8
73
60
52
30
13
14
BuLi (2)
BuLi (4)
sec-BuLi (1)
sec-BuLi (2)
sec-BuLi (4)
LICKOR (1)
LICKOR (2)
LDA (2)
28
12
12
90
41
30k
90
90
12l
ꢀ100
90
28
~99
~98
~98
31
1
Tr
Tr
Tr
3
1
9
61m
80
Tr
Tr
Tr
50
67
71
LDA (2)
LDA (4)
25
25
LITMP (1.5)
LITMP (3)
LITMP (6)
80
7n
5o
5o
80
80
13
9
4
4
a The defluorinated products were a mixture of 4 and 5 in the ratio 1 : 2.
b The defluorinated products were a mixture of 4, 5, 6, 7, 8, 9a in the ratio 1 : 3 : 3 : 3 : 45 : 45.
c About 4% of the product o-F-ꢀ,ꢀ0-S trisubstituted 11 was also detected.
d About 5% of the product o,o0-F-ꢀ-S trisubstituted 13 and 2% of 11 were also detected.
e The defluorinated products were a mixture of 9b and 14 in the ratio 1 : 1.
f About 10% of the product 13 was also detected.
g About 4% of the product 15 was also detected.
h About 10% of the product 15 and 3% of 16 were also detected.
i The defluorinated products were a mixture of 18, 19, 20 in the ratio 32 : 34 : 34.
j The defluorinated products were a mixture of 19, 20, 21 in the ratio 17 : 39 : 44.
k About 7% of the product o-F-ꢀ,ꢀ0-S trisubstituted 25 was also detected.
l The defluorinated products were a mixture of 26b, 29, 30 in the ratio 50 : 37 : 13.
m The defluorinated products were a mixture of 26a, 27, 28, 31 in the ratio 24 : 20 : 28 : 28.
n About 9% of the product 32 was also detected.
o About 11% of the product 32 was also detected.
D and E, which lead to 8 and 9a after quenching with
iodomethane, or 6 and 7 by abstraction of a hydrogen atom
from solvent (THF). The product 2 can give F by a metal±
halogen exchange, or 12 by further metallation/quenching
with iodomethane. From F it is possible to obtain either 5 or
the aryne G by elimination of alkali ¯uoride. The formation
of G was con®rmed by the presence of 14 through the
intermediacy of H, obtained by addition of 1 mol of orga-
nolithium. Another secondary reaction can be the metal±
halogen exchange of 1a leading to I and then 4. The
bimetallation of 1a can give either J, that leads to 10, or
the intermediate K, that leads to 15. This latter, with an
excess of metallating mixture, can afford L and then 16. The
product 10 can undergo a further metallation ortho to the
¯uorine atom to give M and then 13. 10 can be ꢀ-metallated
on the thiomethyl group to give N and 11.

100
M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
Similar considerations can be applied to the reaction
pathways of 1b and 1c. The reactions reveal the following
interesting features:
1. From 1a and 1 molar equivalent of organolithium or
superbase or LITMP (Scheme 5) it is possible to
substitute the hydrogen atom ortho to the ¯uorine, while
with 2 molar equivalents of superbase it is possible to
substitute both the hydrogen atom of the thiomethyl
group and that ortho to the halogen.
2. From 1c and organolithium compounds it is possible to
substitute a thiomethyl hydrogen (Scheme 6), while with
LITMP it is possible to substitute the aryl hydrogen ortho
to the fluorine. In this way these reagents allow the
achievement of complementary results. It is also possible
to obtain acceptable yields of the disubstituted products
by treating 1c with 2 molar equivalents of sec-BuLi.
3. On the contrary, attempts to prepare disubstituted pro-
ducts starting from 1b were unsuccessful.
All the results can be explained by considering that the
regiocontrol of metallation reactions depends on the coor-
dinating properties, on the electron density and on the
inductive and mesomeric effects exerted by the substituents
on the aromatic ring [2±11]. The reaction of para-deriva-
tives 1a and 1b involves mainly the position ortho to the
Scheme 2.
Scheme 3.

M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
101
Scheme 4.
¯uorine atom because its strong inductive effect prevails
over the weak coordinating power of the sulphur atom of the
alkylthio group. On the other hand, the ortho-derivative 1c
shows a different regioselectivity depending on the metal-
lating reagent used: this can be justi®ed by noting that BuLi
acts in a coordinative way through a cyclic intermediate
involving the ¯uorine and the sulphur atom [2] leading to ꢀ-
substitution. On the contrary, the more basic LITMP attacks
the aromatic hydrogen inductively activated by the adjacent
Scheme 5.

102
M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
‡
some
‡
fragmentations
C2H5 ± C2H5) contradicting the even-electron rule,
(M
C2H5 ± CH3
and
M
which can be justi®ed only when much conjugation is
present [14,15]. This could suggest that 9b, 14, 26b, 29
and 30 bear a sec-butyl and a SR group ortho or para each
other. The comparison of such spectra with those of other
de¯uorinated products obtained only in traces (and not
reported in this paper) reveals that 9b, 14, 26b, 29 and
30 have discrete peaks at m/e ˆ 104 and 105 or at m/
e ˆ 119, 117 and 115 due to the loss of SCH3 or SC2H5.
These data suggest that 9b, 14, 26b, 29 and 30 have a lower
double bond character in the Carom±S bond. Finally, the
formation of odd-electron ions from even-electron ones,
present in similar quantity in these compounds, can be
explained from the fact that the ion formed by the loss of
C2H5 from sec-butyl can transpose to tropylium [16] caus-
ing an increase of conjugation even when the sec-butyl and
the sulphur are meta to each other. These results, in accord
with the de¯uorination proposed, are also supported by the
fact that sec-butyl, bulkier than the n-butyl, would enter the
less hindered position.
Scheme 6.
¯uorine. With sec-BuLi, which is even more basic than
BuLi, competition between the positions ortho to the ¯uor-
ine and alpha to the sulphur is brought into play.
All products were identi®ed by comparison with authen-
tic samples (3, 4, 5 and 18) or by analysis of their mass
spectra, assuming that metallations lead to carbanions ortho
to the heteroatom or alpha in the chain bonded to the
heteroatom [4±12]. In the de¯uorinated products the butyl
and the methyl groups obviously occupy the same position
as the ¯uorine or the site ortho to this. The value of the
molecular ion allows determination of many low methyl
groups are in the substrate, the presence of the halogen atom
or the butyl group. The occurrence of ethyl or isopropyl
groups linked to the sulphur is proved by the loss of ethyl,
ethylene or propylene, respectively, in 10, 13, 22, 24, 26a±b,
28, 31, 11, 18, 19, 20, 21, 25. The presence of the methyl
3. Experimental
3.1. General
1H and 13C NMR spectra were recorded on a Varian
VXR-300 spectrometer with tetramethylsilane as internal
reference. IR spectra were recorded on a Perkin-Elmer 1310
grating spectrophotometer. The GC±MS analyses were
performed at 70 eV with a Hewlett-Packard 5989A GC±
MS system with HP 5890 GC ®tted with a capillary column
(50 mÂ0.2 mm) packed with DH 50.2 Petrocol (0.50 mm
®lm thickness). All ¯ash chromatography was on silica G60
(Merck) columns. Microanalyses were carried out with a
Carlo-Erba 1106 elemental analyser. Melting points were
obtained on a Ko¯er hot stage microscope and are uncor-
rected.
group ortho to the SCH3 or to the SC2H5 is shown by the
‡
high abundance of the peak (M
‡
CH3) for the ®rst and
‡
(M
C2H5) more abundant than (M
second [13]. The value of the abundance ratio of peaks
C2H4) for the
‡
‡
(M
C3H7) and (M
C3H6) allows to determine the
Commercially available reagent-grade starting materials
and solvents were used. Solutions of BuLi in hexane were
obtained from Aldrich and were analysed by the Gilman
double titration method before use [17].
position of the butyl group relative to the SR group in 6, 7, 8,
9a, 26a, 27, 28, 31. A high value of this ratio shows that the
two groups are ortho or para to each other, while a low value
suggests a meta relation. This can be explained considering
that the loss of a propyl moiety in ortho or para isomers
3.2. Starting materials
leads to ions more stabilised than meta. Moreover, 26a but
‡
not 27, 28, 31, shows three peaks (M
SR), 147, 105 and
1-(Methylthio)- (1a) and 1-[(1-methylethyl)thio]-4-¯uor-
obenzene (1b) and 1-(methylthio)-2-¯uorobenzene (1c)
were prepared by published methods [18±20].
91 with a high abundance con®rming the poor double bond
character of the Carom±S bond.
Compounds 19, 20 and 21 show a similar behaviour
3.3. Authentic samples
though the presence of the isopropyl group linked to the
‡
sulphur leads to an increase of the peak M C3H6, as in
18. Compound 17 is identi®ed through the loss of C(CH3)3
1-(Ethylthio)-4-¯uorobenzene (3), 1-methyl- (4), and
1,2-dimethyl-4-(methylthio)benzene (5) were prepared by
published methods [21±23]. The following compounds were
obtained as previously described.
‡
and the presence of the peak CꢁCH3†3.
The attribution of structure to 9b, 14, 26b, 29 and 30
needs a more accurate analysis. These substances show

M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
103
‡
‡
‡
3.3.1. 1-(Ethylthio)-2-fluorobenzene (1c)
C4H3S ), 69 (10.5%, C5H9 ), 57 (15.9%, C4H9 ), 45
‡
Yield 83%; n3D0 1.5400; 1H NMR (CDCl3) ꢁ: 1.27 (t, 3H,
CH3CH2), 2.89 (q, 2H, CH2CH3), 7.03 (m, 2H, Ar±H), 7.15
(m, 1H, Ar±H), 7.33 (m, 1H, Ar±H); EI±MS: m/e ˆ 156
(40.6%, CHS ).
‡
4: EI±MS: m/e ˆ 138 (100%, M ), 137 (14.7%,
‡
‡
‡
M
M
H), 123 (27.8%, M
‡
CH3), 105 (6.5%,
‡
‡
‡
‡
(100%, M ), 141 (48.9%, M
CH3), 128 (99.4%,
HS) 91 (83.6%, C7H7 ), 65 (10.6%, C5H5 ).
‡
‡
M C2H4), 127 (15.2%, M
‡
C6H5S ), 108 (51.9%, C6H4S ), 84 (47.4%, C5H5F ), 83
C2H5), 109 (13.1%,
‡
5: EI±MS: m/e ˆ 152 (100%, M ), 137 (34.1%,
‡
‡
‡
M CH3), 105 (81%, M
C6H5 ), 45 (45.1%, CHS ).
SCH3), 77 (28.1%,
‡
‡
‡
‡
(52.5%, C5H4F ), 75 (14.4%, C2FS ), 69 (34.5%,
‡
‡
‡
‡
C4H2F ), 57 (33.3%, C3H2F ), 45 (91.9%, CHS ). Ele-
mental analysis: Found: C, 61.40; H, 5.74; S, 20.35. C8H9FS
(156.2); calc.: C, 61.51; H, 5.81; S, 20.52%.
6: EI±MS: m/e ˆ 180 (24.3%, M ), 137 (100%,
‡
‡
‡
M
C3H7), 91 (7.6%, C7H7 ), 45 (4.4%, CHS ).
‡
7: EI±MS: m/e ˆ 180 (48.7%, M ), 138 (57.1%,
‡
‡
M C3H6), 137 (25.8%, M
C7H7 ), 45 (24.5%, CHS ).
C3H7), 91 (100%,
‡
‡
3.3.2. 3,4-Dimethyl-1-[(methylethyl)thio]benzene (19)
Yield 74%; n3D0 1.5340; 1H NMR (CDCl3) ꢁ: 1.29
(d, 6H, (CH3)2CH), 2.55 (s, 6H, Ar±CH3), 3.32 (m, 1H,
SCH), 7.08 (d, 1H, Ar±H), 7.18 (d, 1H, Ar±H), 7.23 (s, 1H,
‡
8: EI±MS: m/e ˆ 194 (25.8%, M ), 152 (9.7%,
‡
‡
M
C8H9 ), 91 (1.2%, C7H7 ).
C3H6) 151 (100%, M
‡
C3H7), 105 (1%,
‡
‡
9a: EI±MS: m/e ˆ 194 (82.5%, M ), 152 (37.3%,
‡
Ar±H); EI±MS: m/e ˆ 180 (79.8%, M ), 165 (4.4%,
‡
‡
M
C8H9 ), 91 (19.4%, C7H7 ), 45 (16.9%, CHS ).
C3H6), 151 (83.2, M
‡
C3H7), 105 (100%,
‡
‡
‡
M
M
CH3), 138 (100%, M
CH3CHCH2 CH3), 105 (92.3, C8H9 ), 91 (15.3%,
CH3CHCH2), 123 (8.0%,
‡
‡
‡
‡
‡
‡
C7H7 ), 77 (19.9%, C6H5 ), 41 (19.8%, C3H5 ). Elemental
analysis: Found: C, 73.19; H, 8.91; S, 17.69. C11H16S
(180.31); calc.: C, 73.28; H, 8.94; S, 17.78%.
3.4.2. Method B
To a vigorously stirred solution of 1a (2 mmol) and
anhydrous THF (5 ml) cooled to 808C, a 1.2 M solution
of sec-BuLi in cyclohexane (4.8 mmol, 4 ml) was gradually
added under argon. After 4 h, an excess of iodomethane
(4 mmol) was added, the cooling bath removed and the
reaction completed by stirring overnight at room tempera-
ture. The reaction mixture was poured into water, worked up
as described above and analysed. The GC/MS analyses (see
Table 1) exhibited four products 2, 3, 10 and 11 in the ratio
of 65 : 5 : 25 : 5. The starting material remaining was 19%.
If the reaction was performed at the same temperature
with 4 molar equivalents of organolithium eight products (2,
3, 10, 11, 12, 13, 9b, 14) were detected in the ratio
40 : 7 : 25 : 6 : 6 : 2 : 7 : 7. The remaining starting material
was 11%.
3.4. Metallation of 1a
3.4.1. Method A
To a vigorously stirred solution of 1a (2.5 mmol) and
anhydrous THF (5 ml) cooled to 1008C, a 1.2 M solution
of BuLi in hexane (2.6 mmol) was gradually added under
argon, and stirring was continued at the same temperature
for ca. 3 h. The resulting mixture was then treated with
iodomethane (2.6 mmol), left at room temperature for 3 h
and poured into water. The organic layer was separated, the
aqueous layer extracted with diethyl ether and the organic
solutions combined and dried (Na2SO4) were analysed by
GC/MS. These (see Table 1), exhibited the presence of
product 2 (91%). If the reaction was performed with 2
molar equivalents of BuLi, the GC/MS analyses exhibited
two products (2, 3) in the ratio of 7 : 3, respectively. If the
reaction was performed with 4 molar equivalents of BuLi,
the GC/MS analyses exhibited two products (2, 3) in the
ratio 7.5 : 2.5.
When the reaction was performed at 708C, the GC/MS
exhibited four products (2, 10, 9b, 14) in the ratio of
54 : 12 : 17 : 17. The remaining starting material was 7%.
‡
9b: EI±MS: m/e ˆ 194 (33.3%, M ), 165 (100%,
‡
‡
M
(5.2%, M
C2H5), 150 (21.1%, M
‡
C2H5 CH3), 131
‡
C2H5 H2S), 118 (11.6%, C9H10), 117
‡
‡
‡
If the reaction was performed at 508C, four products (2,
3, 4, 5) were detected in the ratio 68 : 13 : 6 : 12.
When the reaction was performed at 308C, the GC/MS
exhibited six products (4, 5, 6, 7, 8, 9a) in the ratio
1 : 3 : 3 : 3 : 45 : 45.
(16.6%, C9H9 ), 105 (4.9%, C8H9 ), 91 (7.3%, C7H7 ),
‡
65 (4.6%, C7H7 ).
10: EI±MS: m/e ˆ 170 (100%, M ), 155 (52.4%,
‡
‡
‡
‡
M
C7H6F ), 83 (11.4%, C4H3S ), 45 (96.7%, CHS ).
CH3), 142 (37.7%, M
‡
C2H4), 109 (78.6%,
‡
‡
13: EI±MS: m/e ˆ 184 (100%, M ), 169 (44.4%,
‡
‡
‡
‡
2: EI±MS: m/e ˆ 156 (100%, M ), 141 (44%,
M
M
CH3), 156 (31.7%, M
C2H5S), 109 (23.8%, C6H5S ), 45 (85.7%,
C2H4), 123 (87.3%,
‡
‡
‡
‡
M
C7H7F ), 109 (21.5%, C7H6F ), 97 (11.9%, C5H5S ),
CH3), 123 (12.1%, M
‡
HS), 110 (16.2%,
‡
‡
CHS ).
11: EI±MS: m/e ˆ 184 (51.1%, M ), 142 (100%,
‡
‡
45 (34.2%, CHS ).
‡
‡
‡
3: EI±MS: m/e ˆ 156 (100%, M ), 141 (51.2%,
M
C7H6F ), 97 (8.1%, C5H5S ), 83 (16.1%, C5H4F ), 45
C3H6), 141 (11.5%, M
‡
C3H7), 109 (81.9%,
‡
‡
‡
‡
‡
M
M
CH3), 128 (61.1%, M
C2H4 HF), 84 (13.7%, C4H4S ), 83 (33.5%,
C2H4), 108 (27%,
‡
‡
‡
(27.8%, CHS ), 43 (2.7%, C3H7 ).

104
M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
‡
‡
14: EI±MS: m/e ˆ 208 (33.8%, M ), 179 (100%,
12: EI±MS: m/e ˆ 170 (100%, M ), 155 (43.2%,
‡
‡
‡
‡
‡
M
(6.6%, M
C2H5), 164 (9.8%, M
‡
C2H5 CH3), 149
‡
M
M
CH3), 137 (21.5%, M
CH2S), 123 (17.9%, M
HS), 124 (18.0%,
SCH3), 109 (34.5%,
‡
C2H5 C2H6), 132 (9.3%, M
‡
‡
‡
‡
C2H5 SCH3), 117 (8.4%, M
‡
C2H5 SCH3
‡
C7H6F ), 77 (9.7%, C6H5 ), 45 (44.1%, CHS ).
‡
CH3), 91 (5.4%, C7H5 ), 77 (6.1%, C6H5 ).
15: EI±MS: m/e ˆ 170 (100%, M ), 155 (83.2%,
‡
‡
‡
M
M
CH3), 154 (6.4%, M
CH4 H), 140 (8.1%, M
CH4), 153 (4.8%,
‡
C2H6), 139
‡
3.4.3. Method C
‡
(7.1%, M
122 (12.9%, M
C2H6 H), 123 (29.0%, M
‡
SCH3),
‡
A solution of BuLi in hexane (2.7 mmol) was cooled to
1008C under argon and a solution of 1a (2.5 mmol) in
hexane (10 ml) was added. Finely powdered potassium tert-
butoxide (2.7 mmol) was added. After 4 h, an excess of
iodomethane (4 mmol) was added, the cooling bath
removed and the reaction completed by stirring overnight
at room temperature. The reaction mixture was poured into
water, worked up as described above and analysed. The GC/
MS analyses (see Table 1) exhibited only the product 2
(93%).
CH3SH), 109 (32.2%, C7H6F ),
‡
‡
103 (16.1%, C8H7 ), 96 (19.3%, C5H4S ), 77 (11.3%,
‡
‡
C6H5 ), 45 (35.5%, CHS ).
‡
16: EI±MS: m/e ˆ 184 (100%, M ), 169 (74.2,
‡
‡
‡
M
M
CH3), 168 (4.8%, M
C2H6), 153 (7.5%, M
CH4), 154 (6.7%,
C2H6 H), 151
‡
‡
‡
(6.4%, M
(19.7%, M
96 (6.2%, C5H4S ), 45 (25.5%, CHS ).
SH), 137 (20.9%, M
CH3 CH2S), 109 (17.1%, C7H6F ),
CH3S), 123
‡
‡
‡
‡
When 2 molar equivalents of superbase were used, the
reaction mixture exhibited three products (2, 10, 13) in the
ratio 5 : 80 : 10. Analogous results were obtained perform-
ing the reaction with 3 or more molar equivalents of the
same reactant.
3.5. Metallation of 1b
3.5.1. Method A
The same procedure described above was followed.
When 1 molar equivalent of BuLi was used GC/MS analysis
showed, beside the starting material (15%), the product
11 in 85% yield. When 2 molar equivalents of organo-
lithium were used at 1008C, the reaction mixture showed
11 and 17 in the ratio of 91 : 9 with 10% of starting material.
The same two products 11 and 17 in a 90 : 10 ratio were
obtained performing the reaction at 858C; only traces of
the starting material were detected. When the reaction was
performed at 308C, the reaction mixture showed 11, 17,
18, 19, 20 in the ratio of 5:1:30:32:32; the starting material
was <1%.
3.4.4. Method D
A solution of diisopropylamine (8.8 mmol) in THF
(10 ml) was cooled to 708C under argon and a solution
of BuLi in hexane (8.8 mmol, 6.2 ml), was added dropwise.
After 15 min a solution of 1a (3.5 mmol) in THF (5 ml)
was added. After 4 h, an excess of iodomethane (9 mmol)
was added, the cooling bath removed and the reaction
completed by stirring overnight at room temperature. The
reaction mixture was poured into water, worked up as
described above and analysed. GC/MS analyses (see
Table 1) exhibited small amounts (15%) of two products
(2, 3 in the ratio of 6 : 1, respectively); the remainder was
starting material.
‡
‡
‡
17: EI±MS: m/e ˆ 184 (25.8%, M ), 155 (1.3%,
‡
M
C2H6 H), 147 (3.8%, M
‡
ꢁCH3†3), 57
(100%, CꢁCH3†3), 41 (61.3%, C3H5 ).
18: Identified by comparison with an authentic
sample.
The same results were obtained performing the reaction
with 4 molar equivalents of LDA.
‡
19: EI±MS: m/e ˆ 222 (27.5%, M ), 180 (20.0%,
‡
‡
‡
M
M
C3H6), 179 (24.8%, M
C3H6 C3H6), 137 (100%,
C3H7), 138 (9.6%,
‡
3.4.5. Method E
M
C3H7
‡
A solution of BuLi in hexane (15 mmol, 10 ml) and THF
(10 ml) was cooled to 808C and 2,2,6,6-tetramethylpiper-
idine (15 mmol) under argon was added dropwise. Then a
solution of 1a (10 mmol) in THF (5 ml) was added. After
4 h, an excess of iodomethane (15 mmol) was added, the
cooling bath removed and the reaction completed by stirring
overnight at room temperature. The reaction mixture was
poured into water, worked up as described above and
analysed by GC/MS that exhibited (see Table 1) 2 (42%)
with traces of 12 and 15; the remainder was starting
material.
‡
C3H6), 105 (4.8%, C8H9 ), 91 (7.2%, C7H7 ), 45
‡
‡
(10.2%, CHS ), 43 (13.1%, C3H7 ).
‡
20: EI±MS: m/e ˆ 222 (30.1%, M ), 180 (35.2%,
‡
‡
‡
M
M
C3H6), 179 (8.5%, M
C3H6 C3H6), 137 (100%,
C3H7), 138 (10.1%,
‡
M
C3H7
‡
‡
C3H6), 105 (8.2%, C8H9 ), 91 (8.1%, C7H7 ), 45
‡
‡
(8.3%, CHS ), 43 (10.1%, C3H7 ).
3.5.2. Method C
The same procedure described above was followed oper-
ating at 1008C. When 1 molar equivalent of superbasic
mixture was used GC/MS analysis showed product 11 at
about 100%. When 2 molar equivalents of the same reactant
were used, the reaction mixture showed 11, 19, 20, 21 in the
When the reaction was performed with 3 molar equiva-
lents of metallating mixture, 2 was obtained in 58% yield.
The GC/MS analyses exhibited the presence of minor
amounts of 12, 15 and 16.

M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
105
‡
ratio of 10 : 15 : 35 : 40; the starting material was <1%.
Similar results were obtained performing the reaction with
more equivalents of the same superbasic mixture.
26b: EI±MS: m/e ˆ 208 (62.7%, M ), 193 (6.7%,
‡
‡
‡
M
M
CH3), 179 (100%, M
C2H5 C2H4), 150 (13.9%, M
C2H5), 151 (26.8%,
‡
C2H5 CH5)
C2H5 CH5 H), 119 (14.2%,
‡
149 (10.8%, M
‡
‡
21: EI±MS: m/e ˆ 208 (28.4%, M ), 193 (1.6%,
‡
‡
C9H11), 117 (13.5%, C9H9 ) 115 (15.4%, C9H7 ), 105
‡
‡
‡
‡
‡
M
M
CH3), 166 (32.2%, M
C3H6), 165 (9.7%,
C3H6 C3H7), 109
‡
‡
(9.7%, C8H9 ), 91 (15.4%, C7H7 ), 77 (6.5%, C6H5 ),
‡
C3H7), 123 (100%, M
45 (8.4%, CHS ).
‡
‡
‡
(2.3%, C6H5S ), 91 (7.1%, C7H7 ), 45 (8.2%, CHS ),
‡
‡
29: EI±MS: m/e ˆ 194 (56.8%, M ), 179 (2.8%,
43 (12.9%, C3H7 ).
‡
‡
‡
M
M
CH3), 165 (100%, M
C2H5 CH3), 149 (4.5%, M
C2H5), 150 (46.9%,
‡
C2H5 CH4),
‡
‡
3.6. Metallation of 1c
119 (9.1%, C9H11), 118 (12.0%, C9H10) 117 (17.3%,
‡
‡
‡
C9H9 ) 115 (15.9%, C9H7 ), 91 (14.0%, C7H7 ), 77
‡
‡
3.6.1. Method A
(6.5%, C6H5 ), 45 (9.9%, CHS ).
30: EI±MS: m/e ˆ 180 (63.3%, M ), 151 (100%,
‡
The same procedure described above was followed.
When 1 molar equivalent of BuLi was used at 1008C,
GC/MS analysis showed, beside the starting material (12%),
the products 22 and 23 in the ratio of 83 : 17. When 2 molar
equivalents of the same reactant were used, the reaction
mixture showed the same products 22 and 23 in the ratio of
67 : 33; the starting material was 10%. 22 and 23 were also
obtained in the ratio of 56 : 44 operating with 4 molar
equivalents of BuLi; the starting material was 8%.
‡
‡
M
(23.9%, M
C2H5), 136 (56.7%, M
‡
C2H5 CH3), 135
‡
C2H5 CH4), 117 (7.8%, C9H9 ), 105
‡
‡
(9.1%, C8H9 ),104 (14.4%, C8H8 ), 91 (19.1%,
‡
‡
‡
C7H7 ),77 (10.4%, C6H5 ) 45 (7.3%, CHS ).
3.6.3. Method C
The same procedure described above was followed.
Using 2 molar equivalents of superbase, the GC/MS ana-
lysis showed 22, 23, 24, 26a, 27, 28, 31 in the ratio
1 : 1 : 9 : 15 : 12 : 17 : 17; the starting material was 28%.
22: Identified by comparison with an authentic
sample.
‡
23: EI±MS: m/e ˆ 156 (100%, M ), 141 (23.9%,
‡
26a: EI±MS: m/e ˆ 208 (100%, M ), 207 (50.0%,
‡
‡
‡
M
C7H7F ), 109 (35.9%, C7H6F ), 97 (11.1%, C5H4S ),
CH3), 123 (30.4%, M
‡
SH), 110 (18.2%,
‡
‡
‡
‡
‡
‡
M
M
M
H), 179 (46.7%, M
C3H6), 165 (58.0%, M
C2H5), 166 (32.2%,
C3H7), 151 (25.8%,
‡
‡
96 (7.6%, C5H3S ), 83 (8.1%, C5H4F ), 57 (7.3%,
‡
‡
C2H5 C2H4), 147 (90.3%, M
‡
SC2H5), 138
C3H6 C2H4) , 1 3 7 ( 4 5 . 1 % ,
‡
C3H2F ), 45 (40.4%, CHS ).
( 2 5 . 8 % ,
‡
M
C3H7 C2H4), 123 (22.6%, C7H7S ), 109
‡
M
(31.9%, C6H5S ), 105 (82.2%, C8H9 ), 91 (25.7%,
‡
‡
3.6.2. Method B
‡
‡
‡
The same procedure described above was followed.
When 1 molar equivalent of sec-BuLi at 908C was used,
GC/MS analysis showed, beside the starting material (28%),
the products 22 and 23 in the ratio of 42 : 58. When 2 molar
equivalents of the same reactant were used, the reaction
mixture showed three products, 22, 23 and 24, in the ratio of
15 : 39 : 46; the starting material was 12%. When 4 molar
equivalents of sec-BuLi were used, the reaction mixture
showed seven products, 22, 23, 24, 25, 26b, 29 and 30 in the
ratio 16 : 28 : 34 : 4 : 4 : 3 : 1; the starting material was
13%.
C7H7 ), 77 (32.2%, C6H5 ), 71 (27.4%, C3H3S ), 65
‡
‡
(27.7%, C5H5 ), 45 (58.1%, CHS ), 43 (32.4%,
‡
‡
C3H7 ), 42 (71.6%, C3H6 ).
27: EI±MS: m/e ˆ 194 (19.5%, M ), 151 (100%,
‡
‡
‡
M
(3.2%, M
C3H7), 136 (5.4%, M
‡
C3H7 CH3), 135
‡
C3H7 CH4), 105 (7.0%, C8H9 ), 91
‡
‡
(12.9%, C7H7 ), 71 (35.4%, C3H3S ), 45 (16.1%,
‡
‡
CHS ), 43 (11.2%, C3H7 ).
28: EI±MS: m/e ˆ 194 (29.0%, M ), 151 (100%,
‡
‡
‡
M
(1.2%, C6H5S ), 91 (10.3%, C7H7 ), 45 (17.7%,
C3H7), 123 (29.6%, M
‡
C3H7 C2H4), 109
‡
‡
‡
‡
CHS ), 43 (4.2%, C3H7 ).
24: EI±MS: m/e ˆ 170 (94.2%, M ), 155 (43.8%,
‡
‡
‡
‡
31: EI±MS: m/e ˆ 208 (25.8%, M ), 179 (1.0%,
M
M
CH3), 142 (59.3%, M
C2H5), 121 (15.6%, M
C2H4), 141 (12.7%,
C2H5 HF), 109
‡
‡
‡
‡
M
M
C2H5), 165 (100%, M
C3H7 C2H4), 105 (3.5%, C8H9 ), 91 (9.3%,
C3H7), 137 (19.3%,
‡
‡
‡
(100%, M
(10.7%, C5H3S ), 83 (12.5%, C4H5F , 69 (9.3%,
SC2H5), 97 (11.6%, C5H4S ), 96
‡
‡
‡
‡
‡
C7H7 ), 77 (8.4%, C6H5 ), 45 (12.9%, CHS ).
‡
‡
‡
C4H2F ), 57 (10.9%, C3H2F ), 45 (96.6%, CHS ).
‡
25: EI±MS: m/e ˆ 184 (43.5, M ), 169 (3.2%,
3.6.4. Method D
‡
‡
‡
M
M
CH3), 142 (100%, M
C3H7), 121 (9.2%, M
C3H6), 141 (9.7%,
C3H7 HF), 109
The same procedure described above was followed.
When 2 molar equivalents of LDA at 808C were used,
the GC/MS analyses showed traces of two products (22, 23);
the remainder was starting material. The same results were
obtained performing the reaction at 258C with 2 and 4
molar equivalents of the same reactant.
‡
‡
‡
(62.9%, C7H6F ), 97 (5.6%, C5H4S ), 96 (4.0%,
‡
‡
‡
C5H3S ), 83 (6.4%, C5H4F ), 77 (8.0%, C6H5 ), 69
‡
‡
‡
(4.1%, C4H2F ), 57 (3.9%, C3H2F ), 45 (17.7 CHS ),
‡
43 (24.2%, C3H7 ).

106
M.G. Cabiddu et al. / Journal of Fluorine Chemistry 98 (1999) 97±106
3.6.5. Method E
Acknowledgements
The same procedure described above was followed.
When 1.5 molar equivalents of LITMP at 808C were
used, the GC/MS analyses showed four products, 22, 23,
24, 32 in the ratio 73 : 4 : 13 : 10; the starting material was
31%. When 3 molar equivalents of the same reactant were
used, the reaction mixture showed the same products in the
ratio of 77 : 4 : 13 : 6; the starting material was 13%. When
6 molar equivalents of LITMP were used, the reaction
mixture showed the same products in the ratio of
78 : 4 : 12 : 6; the starting material was 9%.
Á
Financial support from the Ministero dell'Universita e
della Ricerca Scienti®ca e Tecnologica, Rome and by the
University of Cagliari (National Project ``Stereoselezione in
Sintesi Organica. Metodologie ed Applicazioni'') and from
the C.N.R. (Italy) is gratefully acknowledged. One of us
(M.G.C.) gratefully acknowledges ®nancial support from
Regione Autonoma of Sardinia (Grant no. 206, art. 37 L.R.
2/94).
‡
32: EI±MS: m/e ˆ 170 (100%, M ), 155 (93.8%,
References
‡
‡
M
( 9 . 0 % ,
CH3), 140 (8.3%, M
‡
CH3 CH3), 139
CH3 CH3 H) , 1 2 3 ( 3 0 . 4 % ,
M
SCH3), 109 (18.3%, C7H6F ), 103 (16.6%,
[1] M.G. Cabiddu, S. Cabiddu, E. Cadoni, R. Cannas, C. Fattuoni, S.
Melis, Tetrahedron, 54 (1998) 14095.
‡
‡
M
C8H7 ), 96 (20.3%, C5H3S ), 91 (10.8%, C7H7 ), 83
‡
‡
‡
[2] S. Cabiddu, C. Fattuoni, C. Floris, G. Gelli, S. Melis, J. Organomet.
Chem. 419 (1991) 1.
‡
‡
(9.5%, C4H5F ), 69 (8.2%, C4H2F ), 57 (8.1%,
‡
‡
[3] S. Cabiddu, C. Floris, A. Lai, S. Melis, M. Monduzzi, Gazz. Chim.
Ital. 117 (1987) 759.
Â
[4] W. Bauer, P. von Rague Schleyer, J. Am. Chem. Soc. 111 (1989)
C3H2F ), 45 (26.3%, CHS ).
The metallated mixture, obtained by reaction of 1c with 3
molar equivalents of LITMP, was poured onto ca. 100 g of
crushed solid carbon dioxide. After 24 h the residue was
treated successively with 10% aqueous sodium bicarbonate
and then with ether. The alkali layer was separated, washed
with ether, and then acidi®ed with cold concentrated hydro-
chloric acid, extracted with chloroform and dried (Na2SO4).
The solvent was evaporated and the residue, which was
identi®ed as 2-¯uoro-3-(methylthio)benzoic acid (33), was
7191.
[5] S. Takagishi, G. Katsoulos, M. Schlosser, Synlett. (1992) 360.
[6] A. Mordini, in: V. Snieckus (Ed.), Advances in Carbanion Chemistry,
Chapter 1, Jai Press, Greenwich, CT, 1992.
[7] M. Schlosser in: R. Scheffold (Ed.), Modern Synthetic Methods, vol.
6, VHC, Basel, 1992, p. 227.
[8] M. Schlosser, Organometallic in Synthesis, Wiley, New York,
1994.
Â
[9] A.M. Sapse, P. von Rague Schleyer, Lithium Chemistry, Wiley, New
York, 1995.
crystallised from 1 : 1 aqueous ethanol. Yield 70%, m.p.
1
[10] F. Mongin, M. Schlosser, Tetrahedron Lett. 37 (1996) 6551.
[11] R. Maggi, M. Schlosser, J. Org. Chem. 61 (1996) 5430.
[12] R.D. Clark, A. Jahangir, Org. React. 47 (1995) 1.
[13] J.H. Bowie, S.O. Lawesson, J.é. Madson, G. Schroll, D.H. Williams,
J. Chem. Soc. B (1966) 951.
148±1508C. IR (nujol): 2900 (OH), 1689 cm
(C=O);
1H NMR (CD3COCD3) ꢁ: 2.52 (s, 3H, SCH3), 7.27 (t,
1H, H-5), 7.55 (dt, 1H, H-6), 7.71 (dt, 1H, H-4);
13C NMR
(CD3COCD3)
ꢁ:
14.12
2
(d,
SCH3,
[14] M. Karni, A. Maldenbaum, Org. Mass Spectrom. 15 (1980) 53.
[15] Hewlett-Packard HP 59943 Wiley 1 Database, Palo Alto, 1990.
[16] H. Budzikiewicz, C. Djerassi, D.H. Williams, Mass Spectrometry
of Organic Compounds, Holden-Day, San Francisco, CA, 1967.
[17] H. Gilman, A.H. Haubein, J. Am. Chem. Soc. 66 (1944) 1515.
[18] H. Zahn, H. Zuber, Chem. Ber. 86 (1953) 172.
[19] P. Cogolli, F. Maiolo, L. Testaferri, M. Tingoli, M. Tiecco, J. Org.
Chem. 44 (1979) 2642.
4JCF ˆ 2.17 Hz), 119.27 (d, C-1, JCF ˆ 10.95 Hz),
3
124.63 (d, C-6, JCF ˆ 4.35 Hz), 128.22 (d, C-3,
2JCF ˆ 18.0 Hz), 128.55 (s, C-5), 131.63 (d, C-4,
3JCF ˆ 3.82 Hz), 158.66 (d, C-2, JCF ˆ 267.52 Hz),
1
3
164.52 (d, CO2H, JCF ˆ 2.70 Hz); EI±MS: m/e ˆ 186
‡
‡
(100%, M ), 171 (5.3%, M
CH3), 169 (15.8%,
SH), 141 (6.8%,
CH2S), 139 (3.1%,
‡
‡
‡
‡
‡
[20] C.H. Yoder, F.K. Sheffy, R. Howell, R.E. Hess, L. Pacala, C.D.
Schaffer, J.J. Zukermann, J. Org. Chem. 41 (1976) 1511.
[21] T. Schaefer, G.H. Penner, Can. J. Chem. 66 (1988) 1641.
[22] E. Wedeking, D. Schenk, Chem. Ber. 54 (1921) 1604.
[23] G. Baccolini, E. Mezzina, P.E. Todesco, J. Chem. Soc., Perkin Trans.
I (1988) 3281.
M
M
M
OH), 153 (10.1%,
CO2H), 140 (3.6%, M
M
‡
‡
CH3S), 133 (3.9%, M
‡
HS-HF), 123 (6.6%,
‡
C7H7S ), 109 (3.2%, C6H5S ), 105 (2.9%, C8H9 ), 96
‡
‡
(1.74%, C6H5F ), 95 (2.9%, C6H4F ), 83 (8.1%,
‡
‡
‡
C5H4F ), 69 (2.8%, C4H2F ), 57 (3.1%. C3H2F ). Ele-
mental analysis. Found: C, 51.49; H, 3.74; S, 17.11.
C8H7FO2S (186.01); calc.; C, 51.60; H, 3.79; S, 17.22%.