Trifluoromethylthiolation of aromatic substrates using thiophosgene - fluoride salt reagents, and formation of byproducts with multi-carbon chains

Article abstract of DOI:10.1016/S0022-1139(99)00063-9

Reaction of potassium fluoride or tetramethylammonium fluoride with thiophosgene leads to the formation of a nucleophilic source of trifluoromethanethiolate, suitable for the preparation of trifluoromethyl aryl sulfides from activated haloaromatics. Analy

Full text of DOI:10.1016/S0022-1139(99)00063-9

Journal of Fluorine Chemistry 95 (1999) 171±176
Tri¯uoromethylthiolation of aromatic substrates using
thiophosgeneЯuoride salt reagents, and formation
of byproducts with multi-carbon chains
*
Stewart J. Tavener, Dave J. Adams, James H. Clark
Department of Chemistry, University of York, Heslington, York YO10 5DD, UK
Received 15 January 1999; accepted 24 February 1999
Abstract
Reaction of potassium ¯uoride or tetramethylammonium ¯uoride with thiophosgene leads to the formation of a nucleophilic source of
tri¯uoromethanethiolate, suitable for the preparation of tri¯uoromethyl aryl sul®des from activated haloaromatics. Analysis of the by-
4
products in the system demonstrates that complex molecules with up to C chains may be formed by the reaction of ¯uoride salts with
thiophosgene. # 1999 Elsevier Science S.A. All rights reserved.
Keywords: Tri¯uoromethylthio; Tetramethylammonium ¯uoride (TMAF); Dithiine; Dithione
1
. Introduction
to introduce ±SCF groups to penta¯uoropyridine, tetra-
3
¯uoropyridazine and tetra¯uoropyrimidine, although large
Tri¯uoromethyl aryl sul®des are important target mole-
cules for the pharmaceutical and agrochemical industries
scale reactions gave a red-brown polymeric tar as a by-
product. We recently reported that the reaction between
thiophosgene (Cl CS) and KF provides a cheap and con-
venient source of tri¯uoromethanethiolate anion in solution,
which can be used to perform nucleophilic aromatic sub-
stitution reactions on activated aromatic substrates [16]
(Scheme 1). We have now investigated this reaction in more
detail, and have investigated the effects of using an organic
¯uoride source in place of KF, and studied the by-products
formed.
[
[
1,2], because of the lipophilicity [3] and electronic effects
4] imparted by the ±SCF group. These compounds have
2
3
traditionally been prepared using Swartz chemistry which
involves photochemical chlorination of a methyl aryl sul®de
followed by ¯uorination with either anhydrous HF at high
pressure and temperature, or SbF [1,2]. As these harsh
3
reaction conditions limit the variety of other functional
groups which may be attached to the ring, several milder
methods have been developed, including the use of tri¯uoro-
methanesulfenyl chloride and substituted aromatics [5], and
Grignard reagents [6]. Nucleophilic sources of tri¯uoro-
methanethiolate include Hg(SCF3)2 [7,8], and CuSCF3
2. Results and discussion
[
of aryl tri¯uoromethyl sulphides. AgSCF is only suitable
9±12], both of which have been used in the preparation
2.1. The reactions of thiophosgene with
tetramethylammonium fluoride (TMAF) and potassium
fluoride
3
for reactions with highly polarised C±X bonds, reacting, for
example, with benzotrichloride to form the trithiolated
product, PhC(SCF ) [13].
The reaction of haloaromatics with Cl CS and KF was
2
3
3
Dmowski and Haas [14,15] found that potassium and
cesium tri¯uoromethanethiolates may be formed at low
found to be limited to substrates which are highly activated
to nucleophilic substitution [16]. Although product yields
for the relatively inactive substrate, 2-chloro-5-nitrobenzo-
nitrile, could be improved by heating the reaction, this also
(
� 158C) temperatures from the reaction of thiocarbonyl
¯
uoride (F CS) with the alkali metal ¯uoride, and used this
2
decreased the selectivity of the reaction. The KSCF formed
3
in situ decomposes on warming, forming F CS and KF, and
2
*
Corresponding author. Tel.: +44-1904-432-559; fax: +44-1904-432-
5
0
16; e-mail: jhcl@york.ac.uk
this is likely to thwart attempts to extend the range of
022-1139/99/$ ± 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 6 3 - 9

1
72
S.J. Tavener et al. / Journal of Fluorine Chemistry 95 (1999) 171±176
Table 1
Reaction of haloaromatic substrates with thiophosgene and fluoride
a
c
Substrate
Product
Yield/% (GC area)
b
KF
TMAF
>90
Scheme 1.
7
8
0
5
substrates simply by heating the reaction. We thought that a
bigger cation might help stabilise the relatively large tri-
96
78
¯
uoromethanethiolate anion. Since the reaction needs to be
performed under anhydrous conditions, tetramethylammo-
nium ¯uoride (TMAF) was chosen as this may be dried
ef®ciently [17].
d
3
8
When Cl CS was added to a suspension of dry TMAF
2
(
three equivalents) in acetonitrile at � 158C, the reaction
mixture rapidly turned pink, indicating the formation of
F CS) CS (which has been reported elsewhere as being red
(
3
2
e
when undiluted [18,19]): the presence of this compound was
80
5
0
1
9
con®rmed by F NMR spectroscopy and GCMS. By the
time the reaction had warmed to � 108C, it had taken on a
dark brown colour, and showed a large number of peaks in
0
1
9
the F NMR spectrum. By contrast, the reaction between
Cl CS and KF does not turn pink until it reaches 08C. TMAF
2
a
The area not accounted for is starting material, unless otherwise noted.
Conditions: KF (78 mmol), substrate (13 mmol) and thiophosgene
is a more active source of ¯uoride than KF, in terms of both
nucleophilicity and basicity [20], and ¯uoride is known to
catalyse the condensation of thiocarbonyl ¯uoride [18]. The
lower temperature for formation of condensation products
demonstrates that the greater activity of the TMAF out-
weighs any stabilisation effects imparted by use of the larger
b
(13 mmol). The reaction was kept at � 158C for 4 h, and then stirred at
room temperature overnight.
c
Conditions: TMAF (2.4 mmol), substrate (0.79 mmol) and thiophosgene
(
0.79 mmol). The reaction was kept at � 488C for 4 h, and then stirred at
room temperature overnight.
d
1
4% 4-fluoro-3,5-dinitrobenzotrifluoride also observed.
e
cation. Addition of dichloromethane to the TMAF/Cl CS
2
A complex mixture of products was formed, many of which were
solution causes precipitation of tetramethylammonium
chloride, (the identity of which was con®rmed by infrared
spectroscopy and AgNO3 titration), indicating that salt
formation is the fate of the thiophosgene's chlorine atoms.
Like the KF/Cl CS system, TMAF/Cl CS may be used as
unidentified.
F CSC(S)F, (F CS) CS, and heavier condensation products
3
3
2
of formula (CSF ) which were not identi®ed. We found that
2 n
2
2
a nucleophilic source of tri¯uoromethanethiolate, but better
results are obtained with the latter system if the reactions are
started at � 408C (rather than � 158C used for KF). Several
the KF/Cl CS reactions always contained F CSSCF ,
2
3
3
F CSSSCF and (F CS) CS as by-products, along with a
3
3
3
2
compound with a molecular weight of 516 and an MS
fragmentation pattern consistent with a compound of for-
mula C F S (1).
substrates were studied, and compared to the KF/Cl CS
2
system (Table 1). Good conversions to the aryl tri¯uoro-
methyl sulphide products were observed for the simple
activated substrates penta¯uoropyridine, 2,4-dinitro¯uoro-
benzene and 4-chloro-3,5-dinitrobenzotri¯uoride. However,
the use of TMAF led to destruction of the furazan ring in 4-
chloro-7-nitrobenzofurazan (which was left intact when KF
was used), and a large number of unidenti®ed products were
seen in the GC trace for this reaction. It appears that, whilst
dry TMAF may be used as a ¯uoride source for this reaction,
it offers no advantages over KF, and in some situations its
increased activity can be a drawback.
8
12 6
In order to investigate the thermal stability of the KSCF3
formed in situ, a reaction between KF and Cl C=S without
2
substrate was sampled from � 408C (MeCN±CO bath) to
2
208C at 58C intervals. These samples were ®ltered to
19
remove solid KF, and analysed by F NMR spectroscopy.
There was no peak in any of the spectra which could be
assigned to KSCF itself: presumably the active source of
3
tri¯uoromethanethiolate is formed at the surface of the KF,
and is either insoluble, or unstable in solution. At � 408C,
the solution showed no ¯uorine containing compounds, but
at � 58C the spectrum showed the presence of F CSSCF
3
3
2
.2. By-product formation
and traces of F CSC(S)F (3) and (F CS) CS (4). By the
3 3 2
time the reaction reached room temperature the major
component was 4, and on stirring for 4 h most of this
was converted to 1, with some disulphide, F CSSCF .
Haas and Klug [18] reported that thiocarbonyl ¯uoride
reacts in the presence of alkali-metal ¯uorides to form
3
3

S.J. Tavener et al. / Journal of Fluorine Chemistry 95 (1999) 171±176
173
The ®nal product distribution was not affected either by
light or by the presence of oxygen, although increasing the
initial ratio of KF to Cl CS was found to increase the ®nal
2
1
ratio of 1 to the disulphide, as determined by F NMR
9
spectroscopy.
A large-scale (50 g KF, 20 g Cl CS) reaction was run
2
without substrate, from which compound 1 (8% yield) was
isolated after removal of volatiles on a rotary evaporator and
puri®cation by column chromatography. 1 was obtained as
an orange oil with an odour of garlic, which crystallised
1
9
slowly, and showed two peaks in the F NMR spectrum, at
ˆ � 40.9 and � 42.6 ppm. This was identi®ed as 3,4,5,6-
tetrakis(tri¯uoromethylthio)-1,2-dithiine, with melting
ꢀ
19
point, F NMR and IR spectra in good agreement with
values available in the literature [21]. The data could also be
assigned to a but-2-ene-1,4-dithione structure (2) [22], but
such compounds are known to be thermodynamically less
stable, and spontaneously cyclise to dithiines [23], hence the
dithiine structure is more likely. Unfortunately, attempts to
obtain a crystal of 1 of suf®cient quality for an X-ray
structure were unsuccessful.
Scheme 2.
four equivalents of triphenylphosphite [P(OPh) ] to this
3
solution caused conversion of the trithiocarbonate (4) to
tetrakis(tri¯uoromethanethio)ethene (6), which again co-
distilled with acetonitrile (70±808C). The alkene (6) has
been reported to react with trithiocarbonate (4) to form the
dithione (2, which presumably cyclises to 1) in the presence
of HgCl at 1708C [22]. However, mixing 4 and 6 together
2
Compound 1 is not a simple oligomer of formula (CSF ) ,
2
under less forcing conditions, in the presence of KF (with or
without UV irradiation) did not lead to any detectable
n
being de®cient in sulphur: the missing sulphur atoms are
accounted for by the presence of the trisulphide,
F CSSSCF . Remarkably, it contains a 4-carbon chain,
19
products (by GCMS, F NMR), although a colour change
from pink to orange was observed. Photolysis of 6 in the
presence of carbon disulphide led to the formation of 1 in
5% yield (GC areas), along with a roughly equal quantity of
tris(tri¯uoromethanethio)ethene, (F CS) C=CH(SCF ), 7.
3
3
which has been generated from a C1 molecule under
relatively mild conditions. Although it may seem surprising
that a molecule as complex as 1 is formed simply from
KF and thiophosgene, there is literature precedent for a
number of reaction steps which could lead to the formation
of 1, most of which may occur in the presence of ¯uoride or
light: only the ®nal step was reported as proceeding under
more forcing conditions (Scheme 2). Compounds 4, 5 and 6
were all detected as by-products in reactions involving KF/
3
2
3
Cl CS, although they were typically found at low levels (1±
2
5
% of 1). Whilst investigating the reactions of silver(I)
tri¯uoromethanethiolate with halosilanes, we observed a
different cyclic condensation product, of formula
(
detected in any of the reactions involving KF/Cl CS or
CS) (SCF ) [24]. However, that compound was not
3 3 2
2
TMAF/Cl CS.
2
In order to con®rm the identities of some of the by-
products, alternative methods were devised for their pre-
paration (Scheme 3). Copper(I) tri¯uoromethanethiolate
reacted readily with thiophosgene in acetonitrile to form
bis(tri¯uoromethyl)trithiocarbonate (4) in quantitative
yield. This was puri®ed by co-distillation with the acetoni-
trile: the two compounds formed a low-boiling azeotrope at
6
0±708C, which was collected as a red solution. Addition of
Scheme 3.

1
74
S.J. Tavener et al. / Journal of Fluorine Chemistry 95 (1999) 171±176
Exactly what role the carbon disulphide plays in this reac-
tion is, as yet, unclear, but it should be noted that carbon
disulphide is formed on photolysis of 4 [25], and is therefore
� 428C (acetonitrile/CO bath). The substrate (0.79 mmol)
2
was dissolved in acetonitrile (2 ml) and injected into the
reaction ¯ask. Thiophosgene (0.091 g, 0.79 mmol, 60 ml)
was then injected into the cold, stirring reaction mixture.
The reaction was kept at � 428C for 4 h, and then allowed to
warm to room temperature. The reaction mixture was
separated using diethylether and washed with aqueous
sodium hydrogen carbonate solution, to neutralise any
residual thiophosgene. The ether layer was dried over
magnesium sulfate, ®ltered, and reduced on a rotory eva-
likely to be present in the KF/Cl CS reactions, and so may
2
be involved in the formation of 1 in this system.
3
. Conclusions
Tri¯uoromethanethiolate salts of potassium and tetra-
19
methylammonium ¯uoride cations may be formed in
acetonitrile at low temperatures by reaction of the corre-
sponding ¯uoride with Cl CS. Either of these may be used
porator. Products were identi®ed by GCMS and F NMR
spectroscopy.
2
as a source of nucleophilic tri¯uoromethanethiolate for the
preparation of tri¯uoromethyl aryl sulphides. The tetra-
methylammonium salt is less thermally stable than the
potassium salt, but both decompose even at 08C to form
4.4. Analytical data for reaction products
2,3,5,6-Tetrafluoro-4-(trifluoromethylthio)pyridine
F NMR (CD CN): ꢀ ˆ � 41.5 (s, 3F, SCF ), � 90.2 (m,
19
3
3
a range of SCF -containing by-products. Reaction of thio-
2F, ortho to pyridine N), � 133.6 (m, 2F, meta to
3
phosgene with ¯uoride salts may result in multiple carbon±
carbon bond formation: the major by-product, 3,4,5,6-tetra-
kis(tri¯uoromethylthio)-1,2-dithiine, contains a C back-
pyridine N) ppm
MS: m/z ˆ 251 (53%), 69 (100), 138 (27), 87 (18), 93
(12), 163 (12), 232 (12), 182 (9)
2,4-Dinitro(trifluoromethylthio)benzene
4
bone.
1
9
F NMR (CD CN): ꢀ ˆ � 42.0 (s) ppm
3
MS: m/z ˆ 268 (23%), 199 (100), 69 (88), 63 (87), 95
4
. Experimental
(45), 183 (26), 79 (24), 137 (20).
,6-Dinitro-4-(trifluoromethylthio)benzotrifluoride
2
19
4
.1. Instrumentation
F NMR (CD CN): ꢀ ˆ � 39.2 (s, 3F, SCF ); � 62.4 (s,
3
3
3
F, CF ) ppm
3
1
9
F NMR spectra were recorded at 208C in CH CN with a
MS: m/z ˆ 336 (4%), 267 (100), 251 (86), 69 (22), 175
3
few drops of C D added as a signal lock and referenced to
CFCl on a Jeol EX270 spectrometer operating at 254 MHz,
(18), 106 (15), 317 (12), 81 (11)
4-Nitro-7-(trifluoromethylthio)benzofurazan
6
6
3
1
unless otherwise stated. The F and C NMR spectra of 1
9
13
19
F NMR (CD CN): ꢀ ˆ � 39.6 (s) ppm
3
were obtained in C D on a Bruker AMD 400 operating at
6
MS: m/z ˆ 265 (92%), 69 (100), 196 (68), 120 (28), 180
6
1
76 MHz for F and 100 MHz for C. Connectivity was
9
13
3
(27), 80 (23), 207 (19), 136 (16)
5-Nitro-2-(trifluoromethylthio)benzonitrile
1
9
13
proved by F± C 2D gradient assisted HMQC technique
26], running an experiment optimised for long-range cou-
19
[
F NMR (CD CN): ꢀ ˆ � 40.8 (s) ppm
3
pling to adjacent carbons. Mass spectra were obtained on a
VG analytical Autospec instrument or on a Finnegan MAT
Magnum GC-MS instrument. IR spectra were recorded on a
Bruker Equinox 55 FT-IR spectrometer.
MS: m/z ˆ 248 (100%), 69 (98), 133 (50), 152 (22), 218
(20), 82 (15), 202 (13), 229 (10).
4.5. Reactions with KF and thiophosgene
4
.2. Chemicals
Typical reaction conditions: KF (spray dried, 4.53 g,
8 mmol) was oven-dried at 3008C for at least 4 h. This
7
Anhydrous solvents stored under nitrogen were obtained
from Aldrich and used without further puri®cation, unless
otherwise stated. Copper(I) tri¯uoromethanethiolate
was placed in a ¯ask with a magnetic stirrer bar, the
substrate (13 mmol) and acetonitrile (30 ml, freshly distilled
from calcium hydride), and ®tted with a rubber septum. The
reaction was ¯ushed with argon and cooled to � 158C in an
(
method [10].
CuSCF ÁCH CN) was prepared by a previously reported
3
3
ethylene glycol±dry ice bath, at which point Cl CS (1.50 g,
2
1
3 mmol, 1.0 ml) was injected through the septum. The
4
.3. Reactions with TMAF and thiophosgene
reaction was kept at � 158C for 4 h, and then allowed to
warm to room temperature. The reaction mixture was
separated using diethylether and washed with aqueous
sodium hydrogen carbonate solution, to neutralise any
residual thiophosgene. The ether layer was dried over
magnesium sulfate, ®ltered, and reduced on a rotory eva-
Typical reaction conditions: TMAF was dried by the
method of Christe [17]. TMAF (0.22 g, 2.4 mmol) and a
stirrer bar were transferred in an inert-atmosphere glovebox
to a reaction ¯ask, which was ®tted with a rubber septum.
Acetonitrile (8 ml) was added and the reaction cooled to
19
porator. Products were identi®ed by GCMS and F NMR.

S.J. Tavener et al. / Journal of Fluorine Chemistry 95 (1999) 171±176
175
For the isolation of compound 1, a large scale reaction
was performed without an aromatic substrate, under con-
ditions similar to those mentioned, but using 50 g of KF and
4.8. Reaction of bis(trifluoromethyl)trithiocarbonate with
tetrakis(trifluoromethylthio)ethene
2
0 g of thiophosgene. Isolation was performed by removal
The solution of 6 obtained from the distillation described
above was combined with equimolar quantities of 4 and KF.
The reaction was allowed to stand in a sample tube over-
night, during which time no products formed. This was then
irradiated with a 300-W UV lamp for 1 h. Again, no
products were observed by GCMS. Finally, two drops of
carbon disulphide were added to the reaction, which was re-
irradiated. After an hour, a small quantity of 1 (5%) could be
detected by GCMS, along with a compound with a mass
spectrum consistent with tris(tri¯uoromethylthio)ethene
of volatiles on a rotary evaporator, followed by column
chromatography on Kieselgel 60 silica gel using cyclohex-
ane as the eluent.
3,4,5,6-Tetrakis(trifluoromethylthio)-1,2-dithiine,
C F S , 1
8
12 6
Elemental analysis: expected for C F S : 18.60% C;
8
12 6
found 19.14%; m.p. 408C
1
9
F NMR (C D ): ꢀ ˆ � 40.9 (s, 6F, C±C±SCF ), � 42.6
6
6
3
(s, 6F, S±C±SCF ) ppm
3
1
(
5%), 7.
3
C NMR (C D ): ꢀ ˆ 127.5 (q, J ˆ 313.8 Hz, 2C,
6
6
C±C±SCF ), 129.1 (q, J ˆ 313.7 Hz, 2C, S±C±SCF ),
Tris(trifluoromethylthio)ethene, (F CS) C=CHSCF , 7
3
3
3
3
2
1
30.5 (s, 2C, C±C=C), 85.5 (s, 2C, S±C=C) ppm
MS: m/z ˆ 328 (98%), 69 (100), 158 (74), 89 (48), 215
MS: m/z ˆ 516 (42%), 69 (100), 447 (81), 145 (70), 346
63), 88 (55), 277 (26), 189 (23)
IR: 1500 (m), 1449 (s), 1153 (vs, broad), 1097 (vs), 1016
(40), 259 (28), 145 (18), 227 (12).
(
4
.9. Analytical data for other observed by-products
�
1
(
m), 975 (m), 911 (m), 876 (m), 755 (s), 477 (s) cm
.
Trifluoromethylfluorodithionoformate, (F CS)C(S)F, 3
3
1
9
4
.6. Reaction of CuSCF with thiophosgene
3
F NMR (CH CN): ꢀ ˆ ‡43.2 (s, 1F, C±F), � 40.2 (s,
3
3
F, SCF ) ppm
3
CuSCF ÁCH CN (2.0 g, 9.2 mmol) was dissolved in acet-
Hexakis(trifluoromethylthio)ethane, (F CS) CC(SCF ) ,
3 3 3 3
3
3
onitrile (10 ml). The solution was degassed with argon and
5
then Cl CS (0.52 g, 4.5 mmol, 345 ml) was added with a
MS: m/z ˆ 630 (10%), 459 (100), 145 (42), 461 (25),
2
microsyringe. The reaction was stirred for 10 min, during
which time it developed a deep red colour and CuCl
precipitated. The reaction mixture was the ®ltered and
distilled at ambient pressure. A red fraction was collected
between 608 and 808C, which contained a mixture of
acetonitrile and product 4.
390 (4), 321 (4)
Bis(trifluoromethyl)disulphide, F CSSCF
3
3
1
9
F NMR (CH CN): ꢀ ˆ � 46.1 (s) ppm
3
MS: m/z ˆ 202 (69%), 183 (100), 69 (65), 64 (48), 114
(32), 133 (15), 63 (23), 95 (14)
Bis(trifluoromethyl)trisulphide, F CSSSCF
3
3
19
F NMR (C D ): ꢀ ˆ � 44.5 (s) ppm
6
6
Copper(I) trifluoromethanethiolate, CuSCF
1
3
9
MS: m/z ˆ 234 (72%), 83 (100), 64 (34), 115 (29), 133
F NMR (CD CN): ꢀ ˆ � 25 (s) ppm
3
�
1
(27), 165 (18), 69 (17), 215 (4).
IR: 1084 (s, C±F), 751 (s, C±S) cm
Bis(trifluoromethyl)trithiocarbonate, (F CS) CS, 4
F NMR (CH CN): ꢀ ˆ � 41.8 (s) ppm
3
2
1
9
3
Acknowledgements
MS: m/z ˆ 246 (5%), 145 (100), 69 (55), 72 (35).
We thank the EPSRC Clean Technology Unit for funding
to S.J. Tavener) and the Royal Academy of Engineering/
4
.7. Reaction of bistrifluoromethyltrithiocarbonate with
triphenylphosphite
(
EPSRC for a fellowship (to J.H. Clark). Thanks also to Dr.
19
13
Simon Duckett for running the HMQC F± C NMR
A quarter of the solution (ꢀ1 mmol) collected from the
reaction of CuSCF with Cl CS was reacted with P(OPh)
3
spectrum of compound 1.
3
2
(
1
4 mmol, 1.2 g) in a sample tube at room temperature for
h. The reaction mixture was then distilled at ambient
References
pressure. The colourless product, 6, formed a low-
boiling azeotrope with the acetonitrile and was collected
at 708C.
[
[
1] M.A. McClinton, D.A. McClinton, Tetrahedron 48 (1992) 6555.
2] J.H. Clark, D. Wails, T.W. Bastock, Aromatic Fluorination, CRC
Press, 1996, pp. 119.
Tetrakis(trifluoromethylthio)ethene,
C(SCF ) , 6
(F CS) C=
3 2
[
3] T. Fujita, Prog. Phys. Org. Chem. 14 (1983) 75.
3
2
1
9
[4] W.A. Sheppard, J. Am. Chem. Soc. 87 (1965) 2410.
5] S. Andreades, J.F. Harris Jr., W.A. Sheppard, J. Org. Chem. 29
1964) 898.
[6] W.A. Sheppard, J. Org. Chem. 29 (1964) 895.
F NMR (C D ): ꢀ ˆ � 39.5 (s) ppm
6
6
[
MS: m/z ˆ 428 (100%), 145 (82), 69 (62), 258 (57), 88
55), 189 (48), 359 (42), 327 (35).
(
(

1
76
S.J. Tavener et al. / Journal of Fluorine Chemistry 95 (1999) 171±176
[
7] E.H. Man, D.D. Coffman, E.L. Muetterties, J. Am. Chem. Soc. 81
1959) 3575.
[18] A. Haas, W. Klug, Chem. Ber. 101 (1968) 2609.
(
[19] R.N. Haszeldine, J.M. Kidd, J. Chem. Soc. (1955) 3871.
[20] J.H. Clark, Chem. Rev. 80 (1980) 429.
[21] A. Haas, C. Klare, H.-U. Kr aÈ chter, J. Heterocyclic Chem. 22 (1985)
[
[
8] J.F. Harris, J. Org. Chem. 32 (1967) 2063.
9] N.V. Kondratenko, A.A. Kolomeytsev, V.I. Popov, L.M. Yagupolskii,
Synthesis (1985) 667.
1631.
[
[
[
10] J.H. Clark, C.W. Jones, A.P. Kybett, M.A. McClinton, J.M. Miller,
D. Bishop, R.J. Blade, J. Fluorine Chem. 48 (1990) 249.
11] S. Munavalli, A. Hassner, D.I. Rossman, S. Singh, D.K. Rohrbaugh,
C.P. Ferguson, J. Fluorine Chem. 73 (1995) 7.
[22] G. Didderrich, A. Haas, M. Yazdanbakhsch, Chem. Ber. 110 (1977)
916.
[23] G. Mazzanti, E. van Helvoirt, L.A. van Vliet, R. Ruinard, S. Maiero,
B.F. Bonini, B. Zwanenburg, J. Chem. Soc., Perkin Trans. 1 (1994)
12] S. Munavalli, D.I. Rossman, D.K. Rohrbaugh, C.P. Ferguson, H.D.
Hurst, J. Fluorine Chem. 76 (1996) 7.
3299.
[24] D.J. Adams, S.J. Tavener, J.H. Clark, J. Fluorine Chem. 90 (1998)
[
[
[
[
[
13] M. Yagupolskii, Y.L. Yagupolskii, J. Fluorine Chem. 72 (1995) 225.
14] W. Dmowski, A. Haas, J. Chem. Soc., Perkin Trans. 1 (1987) 2119.
15] W. Dmowski, A. Haas, J. Chem. Soc., Perkin Trans. 1 (1988) 1179.
16] J.H. Clark, S.J. Tavener, J. Fluorine Chem. 85 (1997) 169.
17] K.O. Christe, W.W. Wilson, R.O. Wilson, R. Bau, J. Feng, J. Am.
Chem. Soc. 112 (1990) 7619.
87.
[25] K. Schlosser, Z. Naturforsch. B. Anorg. Chem. Org. Chem. 37 (1982)
172.
[26] O.W. Sorensen, G.W. Eich, M.H. Levitt, G. BodenHausen, R.R.
Ernst, Prog. NMR Spect. 16 (1983) 163.