Tetrafluorothiophene S, S-dioxide: A perfluorinated building block

Article abstract of DOI:10.1021/jo402373x

Tetrafluorothiophene S,S-dioxide, a highly reactive diene and dienophile, has been synthesized. A new route to 3,4-difluoro- and tetrafluorothiophene has been realized, and the previously unknown 2,3,4-trifluorothiophene has been obtained. The reactivity of tetrafluorothiophene S-oxide has been compared with that of the S,S-dioxide.

Full text of DOI:10.1021/jo402373x

Featured Article
pubs.acs.org/joc
Tetrafluorothiophene S,S‑Dioxide: A Perfluorinated Building Block
David M. Lemal,* Marc Akashi, Yan Lou, and Vivek Kumar
Department of Chemistry, Dartmouth College, Hanover, New Hampshire 03755, United States
S
* Supporting Information
ABSTRACT: Tetrafluorothiophene S,S-dioxide, a highly
reactive diene and dienophile, has been synthesized. A new
route to 3,4-difluoro- and tetrafluorothiophene has been
realized, and the previously unknown 2,3,4-trifluorothiophene
has been obtained. The reactivity of tetrafluorothiophene S-
oxide has been compared with that of the S,S-dioxide.
Scheme 3
INTRODUCTION
■
Decades ago, Raasch synthesized tetrachlorothiophene S,S-
dioxide (1) and showed it to be a very reactive and versatile
diene.1 Accompanied by extrusion of sulfur dioxide, its Diels−
Alder reactions result in incorporation of a tetrachlorobuta-
diene fragment (CCl)4 into the product (shown with alkenes in
Scheme 1). In the hope of being able to build a (CF)4 fragment
Scheme 1
RESULTS AND DISCUSSION
Synthesis of Fluorothiophenes. We carried out the four-
step synthesis outlined in Scheme 4, beginning with
■
Scheme 4
in analogous fashion into a diverse array of molecular
architectures, we set out to synthesize tetrafluorothiophene
S,S-dioxide (2).2 We envisioned simply oxidizing the known
tetrafluorothiophene (3) (Scheme 2). The first synthesis of this
Scheme 2
commercially available diol 4. Ditosylate 55 was cyclized to
3,3,4,4-tetrafluorothiolane (6) in the high boiling solvent
dimethylacetamide, from which it was isolated by vacuum
transfer. If sodium sulfide nonahydrate was used for this
reaction, a significant amount of 3,3,4,4-tetrafluorooxolane was
always formed as a byproduct, but the use of hydrate containing
60−63% of the sulfide (cheaper by far than the anhydrous salt)
gave both better yields of 6 and very little of the oxolane.
Fluorination of 6 with Selectfluor to afford hexafluorothiolane 7
molecule, reported in a patent, proceeded in three steps from
1,4-dichlorotetrafluorobutadiene (Scheme 3).3 No yields were
reported for any of the steps, and the starting material also
required synthesis. Later, Tatlow’s group prepared 3 in two
steps from the parent thiophene by fluorination followed by
dehydrofluorination (Scheme 3).4 The fluorination step
required a stirred bed reactor and gave a complex mixture;
both steps proceeded in low yield. For a practical route to the
target molecule 2, a more efficient synthesis of 3 was essential.
Received: October 24, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jo402373x | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
was accomplished via a tandem fluoro-Pummerer rearrange-
ment. With a single equivalent of Selectfluor the reaction
yielded pentafluorothiolane 9; the presumed mechanism for its
formation is shown in Scheme 5.6
chloride catalyst smoothly transformed the thiolane into its S,S-
dioxide 12 at rt (Scheme 7).12 The cis/trans ratio was 1.5:1,
respectively.
Scheme 7
Scheme 5
Unfortunately, treatment of 12 with strong, hindered bases
killed it without yielding a detectable amount of the thiophene
dioxide. The gentler base cesium carbonate in acetonitrile or
adiponitrile effected the first dehydrofluorination step at rt,
affording initially thiolene 13. This quickly isomerized almost
completely to 14, presumably via SN2′ attack by the liberated
fluoride ion (Scheme 8).13 Longer reaction time with the
The key transformation to 7 required much optimization.
The solvent had to be sufficiently polar to dissolve the salt
Selectfluor, yet non-nucleophilic so it would not compete with
fluoride ion for the extremely electrophilic intermediate 8.
Solvents we tried that met the first criterion failed the second,
with the exception of sulfolane. A further advantage of that
solvent was its very high boiling point (285 °C), which again
allowed isolation of the product by vacuum transfer. The cis
and trans isomers of 7 were formed in the ratio 1.5:1.7 Two-
fold dehydrofluorination of the isomer mixture with potassium
hydroxide in DMSO occurred smoothly at rt, and the resulting
tetrafluorothiophene (3) was obtained in high yield and quite
pure form by vacuum transfer. The overall yield of 3 was 33%.
Fluorothiophenes are of interest because of the important
role thiophenes play in conducting polymers8 and liquid crystal
displays.9 We have obtained two more starting from diol 4: 3,4-
difluorothiophene (10)10 and 2,3,4-trifluorothiophene (11).
The latter is the sole previously unknown fluorothiophene.11
Compounds 10 and 11 are available in high yield by 2-fold
dehydrofluorination of 6 and 9, respectively (Scheme 6).
Scheme 8
carbonate again resulted in decomposition. The role of fluoride
ion here suggested that 2 might actually have been formed
under these conditions but immediately attacked by fluoride
ion to regenerate a thiolene. It seemed that if there were a way
to sequester fluoride ion, the 2-fold dehydrofluorination of 12
might be accomplished successfully.
In light of the low solubility of lithium fluoride,
dehydrofluorination of 12 was attempted with 1 equiv of
lithium tert-butoxide in ether/hexane at 0 °C. In contrast to the
result with cesium carbonate, the product was 13 unadulterated
with 14 (Scheme 8). Thus, sequestration worked, so the
reaction was repeated with 2 equiv of the hindered base.
Gratifyingly, prominent signals for the desired sulfone 2
appeared in the 19F NMR spectrum of the product. Signals
for 13 and four new ones ascribed to an adduct of tert-butyl
alcohol with sulfone 2 were present as well, however. Because
attack of tert-butoxide on 2 was apparently competing
effectively with its dehydrofluorination of 13, we concluded
that a more hindered lithium base was needed. Accordingly, 2-
phenyl-2-propanol was converted to its lithium salt with
methyllithium in ether, but this new base gave results similar to
those of the tert-butoxide. No conditions were found that
delivered sulfone 2 in acceptable yield and purity.
Protection of Tetrafluorothiophene. Clearly, the highly
electrophilic sulfone had to be generated under conditions
where it was safe from nucleophilic attack. We speculated that
protection of the thiophene by bromine addition might be a
successful tactic, as that should make possible both oxidation at
sulfur and deprotection under mild conditions. Tetrafluor-
othiophene (3) was reported to react extremely slowly with
bromine,4b but we found that addition occurs readily under
Scheme 6
Early Attempts To Prepare Tetrafluorothiophene S,S-
Dioxide (2). Raasch successfully oxidized tetrachlorothiophene
to its S,S-dioxide (1) with m-chloroperbenzoic acid,1 but the
expectation that tetrafluorothiophene (3) could be similarly
oxidized to dioxide 2 was mistaken. A series of oxidizing agents
either obliterated the thiophene or failed to react with it: m-
chloroperbenzoic acid, peroxytrifluoroacetic acid, dimethyldiox-
irane, and sodium periodate/ruthenium(III) chloride. To
circumvent this problem, we decided to reverse the order of
the dehydrofluorination and oxidation steps.
Oxidation of hexafluorothiolane 7 with peroxytrifluoroacetic
acid in methylene chloride/trifluoroacetic acid at rt stopped at
the monoxide stage, but sodium periodate with ruthenium III
B
dx.doi.org/10.1021/jo402373x | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
irradiation with visible light, presumably a radical chain process.
Addition takes place exclusively at the 2- and 5-positions,
yielding a 1.4:1 mixture of trans (15) and cis (16) isomers
(Scheme 9). The basis for the configurational assignments is
discussed below.
computation. Treatment of the isomer mixture with sodium
bromide in DMF at rt interconverted the three, probably via a
series of SN2′ transformations with bromide ion. The least
abundant, present in very small amount initially, became the
dominant compound, and the other symmetric isomer nearly
vanished. Calculation of the free energies of the three left no
doubt that the most stable isomer had the cis,cis configuration
(19) (Figure 1).
Scheme 9
Interestingly, treatment of the mixture with sodium iodide in
acetone brought about highly selective reduction of the trans
isomer 15 back to the thiophene (Scheme 10). Only with the
Figure 1. Equilibration of the dibromosulfoxides and computed
Scheme 10
relative energies (kcal/mol).16
Kinetically controlled oxidation of the cis dibromide had
introduced the oxygen on the less hindered face of the
molecule to afford 18. Less repulsion between bond dipoles
may help explain why 19 is the more stable of the two forms.
Treatment of the sulfoxide mixture with sodium iodide in
acetone at 0 °C resulted again in very selective reduction of the
form with trans bromines (17) (Scheme 12). In the 19F NMR
trans isomer can reaction occur antarafacially on the CC π
bond, as shown. Thus, the observed selectivity supports the
view that a CC double bond has an inherent preference for
reacting in transoid fashion.14
Scheme 12
Synthesis and Reactivity of Tetrafluorothiophene S-
Oxide (20). Oxidation of the dibromide mixture with
peroxytrifluoroacetic acid at rt proceeded only to the sulfoxide
stage despite that fact that the first oxidation step for a sulfide is
usually slower than the second. Since the reagent contained
considerable trifluoroacetic acid, protonation on or hydrogen
bonding to the sulfoxide oxygen may have inhibited further
oxidation, as both Bronsted and Lewis acids have been shown
to stop oxidation of less electron-deficient thiophenes at the
sulfoxide stage.15 All three stereoisomeric sulfoxides were
obtained: cis, trans (17); trans, trans (18); and cis, cis (19), in
decreasing order of abundance (Scheme 11). The dominant
spectrum a new pair of signals appeared that represented
tetrafluorothiophene S-oxide (20), but they faded at 0 °C and
were replaced by eight signals of roughly equal intensity
signifying formation of its Diels−Alder dimer 21 (configuration
at sulfur discussed below). Zinc−copper couple in acetonitrile
was a better choice of reducing system because it smoothly
debrominated all three sulfoxides to give 20 at 0 °C, the cis,cis
isomer 19 most rapidly. We were surprised to find that the
dimer 21 spontaneously decomposed upon attempting to
isolate it.
Scheme 11
Sulfoxide 20 is a highly reactive diene that can be trapped by
some dienophiles before it can dimerize. Cyclopentene, for
example, yielded adduct 22, but this compound also
decomposed even at 0 °C (Scheme 13). In contrast, reaction
with an alkyne leads to a stable product. With phenylacetylene,
the resulting adduct spontaneously extrudes sulfur monoxide
with the driving force of aromatization, affording 2,3,4,5-
Scheme 13
isomer was easily recognized as cis, trans by 19F NMR because
of its lack of symmetry. Since it had to arise from the more
abundant of the two dibromide isomers, that one had to have
the trans configuration.
Assignment of stereochemistry to the two symmetric
sulfoxide isomers derived from the cis dibromide was not
obvious; it was done by a combination of equilibration and
C
dx.doi.org/10.1021/jo402373x | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
tetrafluorobiphenyl (23) (Scheme 14).17 It was apparent that
the SO group was somehow responsible for the lability of the
calibration the change from oxidizing the saturated analogue 29
to 30. Adduct 28 was found to lie 7.38 kcal/mol lower in
energy than 27, and the (negative) ΔE for 26 → 28 was 8.66
kcal/mol greater than that for 29 → 30 (Figure 3). Thus,
Scheme 14
dimer and alkene adducts, and to understand this it was
imperative that the configuration at sulfur in those compounds
be established.
Thiophene S-oxides in general are reactive Diels−Alder
dienes that show a marked preference for syn-facial addition,
i.e., addition that orients the oxygen on the same side as the
developing σ bonds.18−21 Interaction of the sulfur lone pair
with a σ* orbital of those bonds, an example of the Cieplak
effect,22 has been invoked by a number of authors to explain
this selectivity.15a,19,23 It has also been suggested that ground-
state distortion of the sulfoxide from planarity is a contributing
factor.20 Based on AM1 calculations, Werstiuk proposed that
relative product stability plays a role in determining π-facial
selectivity.24 For the reaction of 2,5-dimethylthiophene S-oxide
with maleic anhydride, he calculated a 4.15 kcal/mol lower ΔHf
for the syn-facial product than for the anti. No reason was given
for the difference he found.
To help assign the configuration at sulfur of the
tetrafluorothiophene S-oxide−cyclopentene adduct 22, we
examined the reaction of sulfoxide 20 with ethylene computa-
tionally. Finding the syn-facial adduct 25 a full 7.0 kcal/mol
lower in free energy than the anti product 24 gave us an
appreciation for Werstiuk’s surmise, as that amount could
certainly account for most of the 4.2 kcal/mol difference in
transition state free energy we found for the two reaction
pathways (Figure 2). Curious about whether the 7 kcal/mol
results from stabilization of the favored product or destabiliza-
tion of the other, we examined this question with a stripped-
down model system. We calculated the energy changes arising
from oxidizing sulfide 26 to oxides 27 and 28 and for
Figure 3. Computed energy differences for syn vs anti cycloaddition
and for both vs a saturated reference system.16
stabilization of the syn-facial product is responsible for the syn/
anti difference. Among the several highest occupied MOs of 28,
the CC π bond participates only in the HOMO-2 and to a
much smaller extent in the HOMO. Both lie below their anti
adduct counterparts, as does the HOMO-1, but it is not
obvious why. Thus, the absence of a detectable stabilizing
interaction between the π bond and the sulfur leaves unresolved
the interesting question of the origin of the product energy
difference.
In any event, it seems clear that the cyclopentene adduct 22
has the syn-facial configuration at sulfur, but the question
remains as to why the compound is so labile. A discovery made
in our laboratory in the 1970s may be relevant. We synthesized
Dewar thiophene S-oxide 31 and found that it undergoes
degenerate allylic rearrangement at an enormous rate (Scheme
15).25 There is a close correspondence in geometry between 22
Scheme 15
and 31, which have their sulfur atom similarly positioned to
react with the π bond, and both have their oxygen on the
opposite face of the sulfur. We suggest that the decomposition
of the dimer and alkene adducts of tetrafluorothiophene S-
oxide (20) begins with allylic rearrangement, as illustrated with
adduct 22 in Scheme 15.26
Synthesis of Tetrafluorothiophene S,S-Dioxide (2).
Tetrafluorothiophene S-oxide (20) is a very reactive Diels−
Figure 2. Comparison of computed free energy changes in the syn and
anti cycloadditions of sulfoxide 20 to ethylene (kcal/mol).16
D
dx.doi.org/10.1021/jo402373x | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
Alder diene, but its usefulness in synthesis is severely limited by
both its rapid dimerization and the self-destruction of its alkene
adducts. It was therefore essential to obtain the original
synthetic target, dioxide 2, and that meant completing oxidation
of trans- and cis-2,5-dibromotetrafluorothiol-3-ene (15 and 16)
to their dioxides (32 and 33). Sodium periodate with
ruthenium chloride as catalyst, a very effective combination
for oxidation of electron-deficient sulfides,12 accomplished this
transformation at rt in good yield (Scheme 16).
Scheme 16
Figure 4. Frontier orbital energies for tetrafluorothiophene and its
oxides (eV).
2,2,3,3-Tetrafluoro-1,4-di(tosyloxy)butane (4).5 Into a 1 L
round-bottom flask were placed 37.50 g (0.231 mol) of 2,2,3,3-
tetrafluorobutane-1,4-diol, 107 g (0.561 mol) of tosyl chloride, and
400 mL of pyridine. The flask was immersed in a bath at ∼55 °C for
24 h. Contents were poured into a 2 L Erlenmeyer flask cooled in ice,
and 400 mL of water was added with good swirling. Product was
collected by filtration, washed with water, and then dissolved in 550
mL of CH2Cl2. A wash with 360 mL of 5% H2SO4 in two portions
followed, and the organic phase was dried over Na2SO4. Solvent was
removed by rotary evaporation, leaving a white solid, 104.0 g (0.221
mol, 96% yield). Mp: 90.5−91.5 °C. 1H NMR (CDCl3): δ 7.79 (d, J =
7.8 Hz, 4H), 7.38 (d, J = 7.8 Hz, 4H), 4.36 (m, 4H), 2.47 (s, 6H). 19F
NMR (CDCl3): δ −121.1 (s, 4F). 13C NMR (CDCl3): δ 146.2, 131.8,
130.4, 128.2, 114.1 (tt, J = 256, 33 Hz), 64.0 (t, J = 28 Hz), 21.9. Anal.
Calcd for C18H18F4O6S2: C, 45.95; H, 3.86; S, 13.63. Found: C, 45.94;
H, 4.05; S, 13.60.
Zinc/copper couple in acetonitrile smoothly reduces this
mixture of isomeric sulfones at rt to the elusive 2. To avoid
aqueous workup of this highly reactive electrophile, a solvent
was needed from which 2 could be obtained directly. As the
sulfone is somewhat volatile, the need was for a high-boiling
solvent that, like acetonitrile, was dipolar aprotic but not so
nucleophilic that it would destroy the product. Adiponitrile met
these criteria nicely, with a dielectric constant of 30,27 a dipole
moment of 3.9,28 and a boiling point of 295 °C. The reduction
proceeded in good yield at 50 °C in this underutilized dipolar
aprotic solvent, and 2 was obtained as a mobile liquid in fairly
pure form simply by vacuum transfer (Scheme 17).
Scheme 17
3,3,4,4-Tetrafluorothiolane (6). A 1 L round-bottom flask was
charged with 50.00 g (0.106 mol) of the ditosylate, 20.7 g of sodium
sulfide hydrate (Acros, 60−63% Na2S, ∼0.16 mol,), and 225 mL of
dimethylacetamide. The mixture was purged (two aspirator/N2
cycles), placed in a bath at 70 °C, and magnetically stirred for 8 h.
The flask was fitted with a reflux condenser that was connected via a
large liquid N2-cooled U-trap to a mechanical pump. Pressure was
controlled with a nitrogen bleed and measured with a digital
manometer. Vacuum transfer was carried out, finally down to a
pressure of 12 Torr at 70 °C for 0.5 h. Product was pipetted out of the
trap into a 50 mL pear-shaped flask. To facilitate separation of the
thiolane from accompanying water, a drop of red food coloring was
added, then the colorless lower layer was carefully removed by pipet.
Wt.: 13.54 g. Concentrated H2SO4 (4 mL) was added to sequester the
very small amount of water present, and the thiolane was short-path
distilled to free it of dimethylacetamide. The colorless distillate
weighed 12.60 g (74% yield). It contained 1.3 mol % of 3,3,4,4-
tetrafluorooxolane30 [19F NMR (CDCl3): δ −123.8 (s, 4F)]. Thiolane
bp: 100 °C. 19F NMR (CDCl3): δ −120.3 (s, 4F). 1H NMR (CDCl3):
δ 3.22 (m, 4H). 13C NMR (CDCl3): δ 119.4 (tt, J = 262, 29 Hz), 30.0
(m). A sample was prepared for microanalysis by GC: inj 120 °C, col
90 °C, det 160 °C. Anal. Calcd for C4H4F4S: C, 30.00; H, 2.52; F,
47.46. Found: C, 29.75; H, 2.39; F, 47.19.
Synthesis of 2 has been completed in seven steps in 18%
overall yield: tosylation of diol 4, cyclization to thiolane 6,
tandem fluoro-Pummerer rearrangement giving 7, dehydro-
fluorination to thiophene 3, bromination affording 15 and 16,
oxidation to 32 and 33, and debromination to yield 2. No
chromatography or recrystallization is required.
In sharp contrast to the monoxide 20 that dimerizes readily
at 0 °C, dioxide 2 survives in solution for many hours at 100
°C.29 Neither frontier orbital energies (Figure 4) nor their
shapes contribute much to an understanding of this striking
difference, a problem for the future. The dioxide’s reluctance to
dimerize is most fortunate, as that allows it to function as a
highly versatile diene and dienophile, capable of reacting with
alkenes and alkynes across the polarity spectrum. A foray into
its cycloaddition chemistry will be reported elsewhere.
3,4-Difluorothiophene (10).3 Into a 25 mL round-bottom flask
were placed 905 mg (5.66 mmol) of 3,3,4,4-tetrafluorothiolane and 7
mL of DMSO. Powdered 85% KOH, 1.8 g, (27 mmol) was added, and
the mixture was stirred at rt for 1 h. The thiophene was isolated by
vacuum transfer to a liquid N2-cooled U-trap, finally to ∼3 Torr and
70 °C for 10 min. The liquid in the trap after thawing was pipetted
into a 5 mL round-bottom flask, and the water droplet on the bottom
was dyed with a drop of red food coloring to facilitate its separation
from the product. Careful pipetting removed the thiophene: 560 mg
EXPERIMENTAL SECTION
■
NMR spectra were measured on 300 and 500 MHz spectrometers. 19
F
NMR spectra were referenced to internal chlorotrifluoromethane, as
1H and 13C NMR spectra were referenced to TMS. Electrospray mass
spectra were obtained with a q-TOF detector (Harvard FAS System).
Preparative GC was carried out on a 1/4″ x 10’ column packed with
10% OV-101 on Chromosorb W AW-DMCS.
1
(82% yield). A little tetrafluorooxolane (1.8 mol %) was present. H
NMR (CDCl3): δ 6.71 (s, 2H). 19F NMR (CDCl3): δ −139.1 (s, 2F).
13C NMR (CDCl3): δ 145.8 (1JCF = 261 Hz), 103.4.
E
dx.doi.org/10.1021/jo402373x | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
2,3,3,4,4-Pentafluorothiolane (9). To a 50 mL round-bottom
flask were added 1.50 g (9.38 mmol) of 3,3,4,4-tetrafluorothiolane, 15
mL of sulfolane, and 3.70 g (10.4 mmol) of powdered Selectfluor. The
mixture was stirred in a bath at ∼60 °C for 2.5 h. Vacuum transfer was
carried out, finally to 70 °C for several minutes at full oil pump
vacuum. After thawing, the cloudy material in the trap was transferred
into a 5 mL round-bottom flask by pipet, with warming using a heat
gun. It set to a cloudy, colorless, waxy solid, 1.238 g (74% yield). 19F
NMR (CDCl3): δ −111.8 (dm, J = ∼243 Hz, 1F), −121.8 (d, J =
∼243 Hz, 1F), −124.9 (d, J = ∼252 Hz, 1F), −127.3 (dm, J = ∼252
Hz, 1F), −162.8 (dm, J = 56 Hz, 1F). Two mol % of
tetrafluorooxolane was present plus an additional impurity signal of
about the same area. A sample was purified by GC: inj 120 °C, col 97
°C, det 160 °C. The waxy solid had an ill-defined melting range, but it
(large excess) was introduced. The resulting dark brown slurry was
stirred for 2.5 h at rt and then subjected to vacuum transfer to a liquid
N2-cooled trap at pressures down to 3 Torr at 45 °C. After thawing,
the trap contained colorless, mobile liquid that was removed by pipet,
leaving a very small amount of water as droplets on the walls of the
trap (3.24 g, 20.8 mmol, 81% yield). 19F NMR (CDCl3): δ −155.6 (t, J
= 12.0 Hz, 2F), −164.5 (t, J = 12.0 Hz, 2F). 13C NMR (CDCl3): δ
134.2 (1JCF = 291 Hz), 127.0 (1JCF = 263 Hz).
cis- and trans-2,3,3,4,4,5-Hexafluorothiolane S,S-Dioxide
(12). A 200 mL round-bottom flask was charged with hexafluor-
othiolane (4.42 g, 22.6 mmol), CCl4 (45 mL), CH3CN (45 mL), H2O
(90 mL), and NaIO4 (13.0 g, 60.8 mmol). Several milligrams of RuCl3·
xH2O was added, and the mixture was vigorously stirred at rt for 22 h.
The reaction mixture was partitioned between 160 mL of ether and
160 mL of water; the ether layer was washed with 20 mL of saturated
NaHCO3 solution and then 50 mL of brine. After drying over Na2SO4,
the nearly colorless ether solution was passed through a silica gel pad
several mm thick on a 30 mL sintered glass funnel, and followed with
an ether wash. Rotary evaporation left partially crystalline, pale yellow
sulfone as a 1.5: 1 (cis: trans) mixture of isomers, 4.00 g, 17.5 mmol
(78% yield). Product from another run was short-path distilled, bp
∼146 °C, to give a white solid. A sample of this sublimed at 55 °C and
23 Torr afforded a waxy solid. 19F NMR (CDCl3): cis, δ −120.9 (d, J =
280 Hz, 2F), −130.0 (d, J = 280 Hz, 2F), −190.2 (d, JHF = 50 Hz, 2F);
trans, −119.3 (d, J = 274 Hz, 2F), −132.5 (d, J = 274 Hz, 2F), −182.7
(d, JHF = 50 Hz, 2F). 1H NMR (CDCl3): cis, δ 5.59 (dm, JHF = 50 Hz,
2H); trans, δ 5.52 (dm, JHF = 50 Hz, 2H). 13C NMR (CDCl3): cis, δ
1
became a clear liquid at 28 °C. H NMR (CDCl3): δ 5.92 (d, J = 56
1
Hz, 1H), 3.40 (m, 2H). 13C NMR (CDCl3): δ 118.7 (tm, JCF = 263
1
Hz), 116.1 (tm, JCF = 267 Hz), 93.0 (dddd, J = 236, 39, 22, 2.3 Hz),
31.2 (t, J = 27 Hz). HRMS, APCI: calcd for C4H4F5S+ 178.9948, found
178.9946.
2,3,4-Trifluorothiophene (11). Into a 25 mL round-bottom flask
were placed 944 mg (5.30 mmol) of 2,3,3,4,4-pentafluorothiolane and
6 mL of DMSO. Powdered 85% KOH, 1.9 g (29 mmol), was
introduced, and the mixture warmed considerably. It was stirred for 45
min and then subjected to vacuum transfer. The pressure was lowered
to ∼3 Torr and the temperature raised to 60 °C for 10 min. Product
was pipetted from the U-trap into a small pear-shaped flask, and the
large drop of water on the bottom was dyed by addition of a bit of red
food coloring. Careful pipetting isolated the colorless thiophene layer.
Weight: 582 mg (80% yield). There was 2.3 mol % of
tetrafluorooxolane in the product. A sample was purified by GC: inj
120 °C, col 84 °C, det 160 °C. 19F NMR (CDCl3): δ −133.1 (s, 1F),
1
1
111.5 (tm, JCF ∼ 273 Hz), 96.1 (dm, JCF = 251 Hz); trans, δ 111.5
1
1
(tm, JCF ∼ 273 Hz), 95.3 (dm, JCF = 251 Hz). Anal. Calcd for
C4H2F6O2S: C, 21.06; H, 0.88; S, 14.06. Found: C, 21.16; H, 0.92; S,
13.90.
−148.3 (s, 1F), −156.0 (s, 1F). 1H NMR (CDCl3): δ 6.10 (s, 1H). 13
C
2,3,4,4,5-Pentafluorothiol-2-ene S,S-Dioxide (13). Into a small
side arm flask fitted with septum were placed hexafluorothiolane
dioxide (300 mg, 1.32 mmol) and 3 mL of ether. The flask was cooled
in an ice bath under N2, and LiOt-Bu, 1 M in hexane (1.5 mL, 1.5
mmol) was added via syringe with vigorous stirring during 10 min.
The bath was removed after another 5 min; 15 min later a 19F
spectrum showed that reaction was complete, giving the title sulfone
plus tetrafluorothiophene S,S-dioxide in a 5: 1 ratio. Water (5 mL) was
added, and after shaking and layer separation the aqueous phase was
extracted with ether (2 × 5 mL). The combined ether solution was
dried over Na2SO4 and then evaporated. The residue was chromato-
graphed on silica gel (4 g) with 30% CH2Cl2 as eluent, giving 41 mg of
colorless oil (15% yield). 19F NMR (CDCl3): δ −101.3 (dm, J = 266
Hz, 1F), −113.5 (dm, J = 266 Hz, 1F), −147.5 (s, 1F), −150.2
NMR (CDCl3): δ 145.1 (ddd, J = 287, 10.6, 6.9 Hz), 142.8 (dd, J =
262, 16.6 Hz), 129.8 (ddd, J = 260, 22, 7.6 Hz), 88.6 (dd, J = 17.0, 2.8
Hz). MS: m/z 138 (M+), 118 (M+ − HF), 107 (M+ − CF, base), 93
(M+ − CHS). HRMS, APCI: calcd for C4HF3S+ 137.9746, found
137.9741.
cis- and trans-2,3,3,4,4,5-Hexafluorothiolane (7). Into a 200
mL round-bottom flask were placed 8.00 g (50.0 mmol) of 3,3,4,4-
tetrafluorothiolane and 65 mL of sulfolane. Selectfluor (15.0 g, 42.3
mmol) was added as a powder in portions through flexible tubing from
a 50 mL round-bottom flask while the reaction flask was contained in a
cool water bath and vigorously stirred. Stirring was continued for 25
min after addition was complete, and then another 24.0 g (67.7 mmol)
of Selectfluor was added all at once (total, 39.0 g, 110 mmol). The
flask was mounted in a bath at 65 °C and stirred for 4 h. 19F NMR
revealed that reaction was complete, so the mixture was subjected to
vacuum transfer into a liquid N2-cooled U-trap at ∼0.1 Torr. The bath
temperature was gradually raised to 120 °C and maintained there for
several minutes. After thawing, the product was transferred by pipet
with the help of a heat gun into a 10 mL round-bottom flask (7.92 g)
and short-path distilled without water in the condenser. Fraction 1:
80−97 °C, 0.522 g of mobile, pale yellow oil; fraction 2: 97−105 °C,
5.793 g that set to a nearly colorless, waxy solid. The first fraction was
mostly the desired thiolane but contaminated with impurities. Fraction
2 constituted a 59% yield of thiolane isomers in the ratio 1.3: 1 (cis:
trans). In other runs the ratio was nearly 1.5:1. A sample dissolved in
ether was purified by GC: inj 140 °C, col 75 °C, det 160 °C), and a
portion of it was sublimed at 46 °C and 3−4 Torr. 19F NMR (CDCl3):
cis, δ −120.2, −123.0 (ABq, J = 255 Hz, 4F); −163.6 (d, JHF = 51 Hz,
2F); trans, δ −121.3 (d, J = 255 Hz, 2F), −134.1 (d, J = 255 Hz, 2F),
−161.1 (d, JHF = 50 Hz, 2F). 1H NMR (CDCl3): both isomers, δ 6.09
(dm, JHF (apparent) = 56 Hz due to nonidentical δs, 2H). 13C NMR
(CDCl3): cis, δ 115.4 (tm, 1JCF = 271 Hz), 94.3 (dm, 1JCF = 244 Hz);
1
(narrow m, 1F), −185.0 (d, JHF = 50 Hz, 1F). H NMR (CDCl3): δ
5.66 (dm, JHF = 50 HZ, 1H). 13C NMR (CDCl3): δ 145.0 (1JCF = 327
Hz), 138.5 (1JCF = 308 Hz), 109.6 (1JCF = 259 Hz), 95.9 (1JCF = 251
Hz). Since a little water was detected in the NMR solution, it was
dried over Na2SO4, evaporated, and sent for microanalysis. Anal. Calcd
for C4HF5O2S: C, 23.09; H, 0.48; S, 15.41. Found: C, 22.99; H, 0.50;
S, 15.14.
Because the isolated yield was so low in this experiment, owing
perhaps to volatility, the reaction was repeated as above to determine
the yield by NMR. Carefully measured hexafluorobenzene was
introduced as an area standard after a little dilute HCl had been
added to ensure that no base remained to attack it. Integration of the
19F spectrum of the ether solution, taken with a 6 s delay between
pulses to prevent differential relaxation, revealed that the thiol-2-ene
dioxide was present in 70% yield. There was also 11% of
tetrafluorothiophene S,S-dioxide.
2,2,3,4,5-Pentafluorothiol-3-ene S,S-Dioxide (14). Into a 10
mL round-bottom flask were placed hexafluorothiolane dioxide (440
mg, 1.93 mmol), adiponitrile (4 mL), and Cs2CO3. (695 mg, 2.13
mmol). Mixture was stirred vigorously at RT for ∼75 min, then
subjected to vacuum transfer into a small liquid N2-cooled U-trap. At
full oil pump vacuum, the temperature was raised to 100 °C during
∼20 min; solvent was refluxing by the end. After thawing, the trap was
washed down with a little CH2Cl2 which was transferred to a
graduated test tube (∼0.3 mL). The solution was diluted to 1 mL with
1
1
trans, δ 114.7 (tm, JCF ∼260 Hz), 91.3 (dm, JCF = 239 Hz). Anal.
Calcd for C4H2F6S: C, 24.49; H, 1.03; S, 16.35. Found: C, 24.31; H,
0.91; S, 16.13.
Tetrafluorothiophene (3).4 To a 200 mL round-bottom flask was
added 5.02 g (25.6 mmol) of hexafluorothiolane and 55 mL of DMSO.
The flask was placed in a water bath at rt, and 7.5 g of powdered KOH
F
dx.doi.org/10.1021/jo402373x | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
pentane and chromatographed on 5 g of silica gel with 30% CH2Cl2 in
pentane as eluent. The yield of thiol-3-ene dioxide was ∼120 mg
(30%). 19F NMR (CDCl3): δ −95.8 (d, J = 218 Hz, 1F), −111.6 (dd, J
= 218, 12.6 Hz, 1F), −134.5 (narrow m, 1F), −148.7 (d, J = 37 Hz,
1F), −169.2 (dd, J = 56, 25 Hz, 1F). 1H NMR (CDCl3): δ 5.97 (dm, J
= 56 Hz, 1H). 13C NMR (CDCl3): δ 140.7 (1JCF = 295 Hz), 137.3
(1JCF = 300 Hz), 113.8 (1JCF = 290 Hz), 92.3 (1JCF = 239 Hz). Anal.
Calcd for C4HF5O2S: C, 23.09; H, 0.48; S, 15.41. Found: C, 22.99; H,
0.48; S, 15.19.
2,3,4,5-Tetrafluorobiphenyl (23). Into a 10 mL round-bottom flask
were placed 158 mg (0.476 mmol) of the dibromothiolene oxides, 143
mg (1.40 mmol) of phenylacetylene, and 3 mL of CH3CN. Purged as
above, the mixture was cooled in an ice bath, treated with 237 mg of
Zn(Cu), purged again, and stirred for 6 h. The reaction was very
sluggish, so additional Zn(Cu) was added in two portions during the
next 10 h (total 732 mg), still at 0 °C. After another 14 h at RT
reaction was essentially complete. The reaction mixture was filtered
through Celite, which was then washed with 5 mL of CH2Cl2. Filtrate
was evaporated to a wet brown solid that was chromatographed on 3 g
of silica gel with hexanes as eluent, giving 44 mg of the biphenyl (41%
yield). Mp: 65−66 °C (lit.17a mp 61−62 °C). 19F NMR (1: 1 CCl4/
CDCl3 for lit. comparison): 4.8, 6.8, 18.1, and 22.2 ppm relative to
hexafluorobenzene (lit.17a 4.7, 6.7, 18.3, and 22.4 ppm).
trans- and cis-2,5-Dibromo-2,3,4,5-tetrafluorothiol-3-ene
S,S-Dioxide (32, 33). A 250 mL round-bottom flask was charged
with 7.10 g (22.5 mmol) of dibromothiolenes, 35 mL of CCl4, 35 mL
of CH3CN, 70 mL of H2O, and 12.5 g (58.5 mmol) of NaIO4. A few
milligrams of RuCl3·xH2O was added, and the mixture was vigorously
stirred for 23 h at rt. 19F NMR showed that reaction was complete.
The mixture was partitioned between 270 mL of ether and an equal
amount of water. The ether phase was washed with saturated
NaHCO3 solution (30 mL) and 80 mL of brine; it was dried over
Na2SO4. After filtration, it was concentrated to ∼35 mL and then
passed through a silica gel plug followed by an ether wash. Evaporation
left 6.01 g of light brown oil (77% yield), with a trans/cis ratio of 1.4:
1. A sample was chromatographed on silica gel with 3% CH2Cl2 in
pentane to give the isomer mixture as a colorless oil. 19F NMR
(CDCl3): trans, δ −110.0 (d, J = 25 Hz, 2F), −142.6 (d, J = 25 Hz,
2F); cis, δ −91.6 (d, J = 25 Hz, 2F), −139.6 (d, J = 25 Hz, 2F). 13C
NMR (CDCl3): trans, δ 137.9 (1JCF = 296 Hz), 97.1 (1JCF = 304 Hz);
cis, δ 138.5 (1JCF = 293 Hz), 98.0 (1JCF = 304 Hz). Anal. Calcd for
C4Br2F4O2S: C, 13.81; H, 0.00; F, 21.84. Found: C, 13.88; H, 0.00; F,
21.66.
trans- and cis-2,5-Dibromo-2,3,4,5-tetrafluorothiol-3-ene
(15 and 16). Into a 250 mL round-bottom flask were placed
tetrafluorothiophene (3.08 g, 19.7 mmol), CH2Cl2 (55 mL), and Br2
(3.5 g, 21.9 mmol). The flask was mounted in a large crystallizing dish
which was filled continuously to overflowing with cold water that was
caught in a basin with a drain. The solution was irradiated for 5.5 h
with a 120 W tungsten spot lamp that was mounted directly
underneath the water bath. Evaporation of the solvent left 5.88 g of the
dibromides as oil containing some crystals (94% yield). The trans/cis
ratio was 1.4:1. A sample was chromatographed with pentane as eluent
on silica gel to obtain the isomer mixture as analytically pure, colorless
oil. 19F NMR (CDCl3): trans, δ −75.8 (m, 2F), −140.7 (m, 2F); cis, δ
−64.0 (m, 2F), −140.4 (m, 2F). 13C NMR (CDCl3): trans, δ 137.8
(1JCF ∼ 290 Hz), 93.6 (1JCF ∼ 290 Hz); cis, δ 138.0 (1JCF ∼ 290 Hz),
93.1 (1JCF ∼ 290 Hz). Anal. Calcd for C4Br2F4S: C, 15.21; H, 0.00; Br,
50.59; F, 24.06. Found: C, 15.22; H, 0.00; Br, 50.30; F, 23.80.
cis,trans-; trans,trans-; and cis,cis-2,5-Dibromo-2,3,4,5-tetra-
fluorothiol-3-ene S-Oxide (17−19). Into a 25 mL round-bottom
flask was placed 1.808 g (5.72 mmol) of the dibromothiolene mixture,
and 9 mL (9 mmol) of 1 M CF3CO3H in CF3CO2H/CH2Cl2 was
added. The solution became warm, and after 6.5 h at rt the oxidation
was complete. The mixture was transferred with 10 mL of CH2Cl2 to a
125 mL Erlenmeyer flask, cooled in ice, and made alkaline with ice-
cold 10% Na2CO3 solution (35 mL). Layers were separated, and the
aqueous phase was extracted with CH2Cl2 (3 × 10 mL). The
combined extract was dried over Na2SO4, filtered, and evaporated to
leave 1.804 g of colorless oil (95% yield). The order of abundance of
the S-oxide isomers was cis, trans > trans, trans > cis, cis. A sample of
the mixture was chromatographed on silica gel with 5% CH2Cl2/
hexanes as eluent. 19F NMR (CDCl3): cis, trans, δ −106.1 (m, 1F),
−109.8 (m, 1F), −135.6 (m, 1F), −137.2 (dd, J = 23, 10 Hz, 1F);
trans, trans, δ −99.3 (d, J = 20 Hz, 2F), −140.2 (d, J = 20 Hz, 2F); cis
cis, δ −112.6 (d, J = 20 Hz, 2F), −139.0 (d, J = 20 Hz, 2F). 13C NMR
(CDCl3, 19F decoupled): cis, trans, δ 139.5, 137.9, 110.4, 95.6 or 95.3;
trans, trans, δ 140.1, 95.6 or 95.3; cis, cis, δ 137.4, 109.4. Anal. Calcd
for C4Br2F4OS: C, 14.47; H, 0.00; Br, 48.15. Found: C, 14.42; H, 0.00;
Br, 48.31.
Tetrafluorothiophene S,S-Dioxide (2). To a 50 mL round-
bottom flask were added 1.128 g (3.24 mmol) of the dibromothiolene
dioxides and 5 mL of adiponitrile. After a triple purge (aspirator, N2),
1.0 g of zinc−copper couple was introduced and purging was repeated.
Flask was placed in a bath at 50 °C and stirred for 75 min. Product was
collected by vacuum transfer to a liquid N2-cooled U-trap, finally at full
oil pump vacuum and ∼80 °C. After thawing, the trap afforded 443 mg
of colorless, mobile liquid (73% yield). The sulfone decomposed on
silica gel, but could be distilled via kugelrohr (aspirator, < 70 °C). 19F
NMR (CDCl3): δ −151.7 (s, 2F), −162.7 (s, 2F). 13C NMR (CDCl3):
1
1
δ 134.1 (dm, JCF = 298 Hz), 133.4 (dm, JCF = 331 Hz). HRMS,
APCI: calcd for C4F4O2S+ 188.9628, found 188.9632.
Tetrafluorothiophene S-Oxide (20): Generation and Trap-
ping. Thiophene Oxide and Its Dimer. Into a 25 mL round-bottom
flask were placed 326 mg (0.982 mmol) of the dibromothiolene oxides
and 5 mL of CH3CN. The solution was purged three times (aspirator,
N2) and cooled in an ice bath. Purging was repeated after 468 mg of
zinc−copper couple was added. The slurry was vigorously stirred at 0
°C for 4 h, whereupon the 19F spectrum revealed that reaction was
complete, yielding the S-oxide, a larger quantity of its Diels−Alder
dimer (21), and a small amount of the over-reduction product
tetrafluorothiophene. The dimer decomposed during an attempt to
isolate it. 19F NMR (CH3CN): S-oxide, δ −144.5 (s, 2F), −160.4 (s,
2F); dimer, δ −129.4 (s, 1F), −135.1 (d, J = 25 Hz), −144.0 (s, 1F),
−153.8 (s, 1F), −177.3 (s, 1F), −182.4 (s, 1F), −195.9 (s, 1F),
−200.5 (s, 1F); thiophene, δ −156.5 (t, J = 12.4 Hz, 2F), −164.7 (t, J
= 12.4 Hz, 2F).
ASSOCIATED CONTENT
■
S
* Supporting Information
1H, 19F, and 13C NMR spectra; energies and Cartesian
coordinates for structures in Figures 1−4. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: david.m.lemal@dartmouth.edu.
Notes
The authors declare no competing financial interest.
Cyclopentene Adduct (22). Into a 5 mL round-bottom flask were
placed 101 mg (0.304 mmol) of the dibromothiolene oxides, 40 μL
(31 mg, 0.39 mmol) of cyclopentene, and 2 mL of CH3CN. Purged as
above, the solution was cooled in ice, treated with 145 mg of Zn(Cu),
purged again, and stirred for 1.5 h. NMR showed that reaction was
complete and the product was quite clean cyclopentene Diels−Alder
adduct. 19F NMR (CH3CN): δ −154.7 (s, 2F), −184.5 (s, 2F). The
adduct gradually decomposed spontaneously at 0 °C.
ACKNOWLEDGMENTS
■
We thank the National Science Foundation for support of this
work (Grant No. CHE-0653935). We are grateful for mass
spectra obtained by Sunya Trauger of the Harvard Small
Molecule Mass Spectrometry Facility. D.M.L. is pleased to
acknowledge a helpful discussion with Eric Block (SUNY
Albany).
G
dx.doi.org/10.1021/jo402373x | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
(20) Furukawa, N.; Zhang, S.-Z.; Horn, E.; Takahashi, O.; Sato, S.
Heterocycles 1998, 47, 793.
(21) Nakayama, J.; Sugihara, Y. Sulfur Rep. 1997, 19, 349.
(22) (a) Cieplak, A. S. J. Am. Chem. Soc. 1981, 103, 4540.
(b) Cieplak, A. S.; Tait, B. D.; Johnson, C. R. J. Am. Chem. Soc. 1989,
111, 8447.
(23) Naperstkow, A. M.; Macaulay, J. B.; Newlands, M. J.; Fallis, A.
G. Tetrahedron Lett. 1989, 30, 5077.
(24) Werstiuk, N. H.; Ma, J. Can. J. Chem. 1994, 72, 2493.
(25) (a) Ross, J. A.; Seiders, R. P.; Lemal, D. M. J. Am. Chem. Soc.
1976, 98, 4325. (b) Bushweller, C. H.; Ross, J. A.; Lemal, D. M. J. Am.
Chem. Soc. 1977, 99, 629.
(26) Apparently, the fluorine substituents are responsible for the
lability of alkene adducts of 20, as other thiophene S-oxide adducts
with alkenes are stable (refs 15, 18−21, and 23).
(27) Abu-Lebdeh, Y.; Davidson, I. J. Electrochem. Soc. 2009, 156, A60.
(28) Sengupta, K.; Roy, S. K.; Roy, S. B. Indian J. Phys. 1977, 51B, 34.
(29) Tetrachlorothiophene S,S-dioxide dimerizes very slowly in
refluxing 1,2- dichloroethane.
REFERENCES
■
(1) Raasch, M. S. J. Org. Chem. 1980, 45, 856.
(2) Thrasher’s group has synthesized 3,4-difluorothiophene S,S-
dioxide: Nowak, I.; Rogers, L. M.; Rogers, R. D.; Thrasher, J. S. J.
Fluorine Chem. 1999, 93, 27.
(3) Ilgenfritz, E. M.; Ruh, R. P., U.S. Pat. 2,932,651, 1960.
(4) (a) Burdon, J.; Campbell, J. G.; Parsons, I. W.; Tatlow, J. C.
Chem. Commun. 1969, 27. (b) Burdon, J.; Campbell, J. G.; Parsons, I.
W.; Tatlow, J. C. J. Chem. Soc. C 1971, 352.
(5) Ball, T. F.; Henrie, R. N., II. Org. Prep. Proc. Int. 1989, 21, 245.
(6) Review: Hugenberg, V.; Haufe, G. J. Fluorine Chem. 2012, 143,
238.
(7) The major isomer had a somewhat higher boiling point and
longer GC retention time than the minor isomer. It was assigned the
cis configuration because the calculated dipole moment of that isomer
(2.90 D) is significantly higher than that of the trans form (2.02 D).
19F NMR provided evidence supporting this assignment. Calculated
structures for both isomers have strongly twisted rings, relieving
eclipsing between C3 and C4 (ref 16). The twist renders inequivalent
the members of each pair of geminal fluorines at those carbons. Time
averaging eliminates this effect in the cis isomer, but not the trans, in
which one of the two twisted conformers is preferred over the other.
Therefore, a larger chemical shift difference between the geminal
fluorines is expected for the trans isomer (barring a strongly
countervailing influence of substituents at C2 and C5). That difference
for the minor isomer is 12.8 ppm, but only 2.8 for the major one,
consistent with the conclusion that the latter is cis.
(30) Henne, A. L.; Richter, S. B. J. Am. Chem. Soc. 1952, 74, 5420.
(8) (a) Jameh-Bozorghi, S.; Il Beigi, H. S. J. Fluorine Chem. 2011, 132,
190. (b) Sakamoto, Y.; Komatsu, S.; Suzuki, T. J. Am. Chem. Soc. 2001,
123, 4643.
(9) (a) Kawabata, K.; Goto, H. Materials 2009, 2, 22. (b) Funahashi,
M.; Hanna, J.-I. Adv. Mater. 2005, 17, 594. (c) Zhang, H.; Shiino, S.;
Shishido, A.; Kanazawa, A.; Tsutsumi, O.; Shiono, T.; Ikeda, T. Adv.
Mater. 2000, 12, 1336. (d) Knawby, D. M.; Swager, T. M. Chem.
Mater. 1997, 9, 535.
(10) (a) Christiansen, H.; Gronowitz, S.; Rodmar, B.; Rodmar, S.;
́
Rosen, U.; Sharma, M. K. Ark. Kemi 1969, 30, 561. (b) See also ref 8b.
(11) The other partially fluorinated thiophenes are discussed in ref
10a and: Rodmar, S.; Rodmar, B.; Sharma, M. K.; Gronowitz, S.;
Christiansen, H.; Rosen, U. Acta Chem. Scand. 1968, 22, 907.
́
(12) Su, W. Tetrahedron Lett. 1994, 35, 4955.
(13) Calculations (ref 16) found 14 to be more stable than 13 in
both the gas phase and polar solution and even more stable than
unobserved isomer 2,3,4,5,5- pentafluorothiol-2-ene S,S-dioxide that
could arise by prototropic rearrangement of 14 in the basic medium.
(14) The SN2′ reaction is a good testing ground for the question of
facial preference of the CC π bond. In a recent computational study
of this reaction, Streitwieser et al. found syn (suprafacial) stereo-
chemistry in cases where the attacking reagent was an ion pair and
therefore able to react via a cyclic transition state. However, when the
attacking species was a lone nucleophile, reaction took place in anti
(antarafacial) mode, consistent with an inherent preference for that
stereochemistry. Streitwieser, A.; Jayasree, E. G.; Hasanayn, F.; Leung,
S. S.-H. J. Org. Chem. 2008, 73, 9426.
(15) (a) Li, Y.; Thiemann, T.; Sawada, T.; Mataka, S.; Tashiro, M. J.
Org. Chem. 1997, 62, 7926. (b) Li, Y.; Matsuda, M.; Thiemann, T.;
Sawada, T.; Mataka, S.; Tashiro, M. Synlett 1996, 461. (c) Pouzet, P.;
Erdelmeier, I.; Ginderow, D.; Mornon, J.-P.; Dansette, P. M.; Mansuy,
D. J. Chem. Soc., Chem. Commun. 1995, 473. (d) Pouzet, P.;
Erdelmeier, I.; Ginderow, D.; Mornon, J.-P.; Dansette, P. M.;
Mansuy, D. J. Heterocycl. Chem. 1997, 34, 1567.
(16) All calculations were carried out at the B3LYP/6-311**+ level
of theory. Jaguar, version 7.0, Schrodinger, LLC, New York, 2007.
(17) (a) Bogachev, A. A.; Kobrina, L. S. Russ. J. Org. Chem. 1997, 33,
681. (b) Wei, Y.; Kan, J.; Wang, M.; Su, W.; Hong, M. Org. Lett. 2009,
11, 3346.
(18) Nakayama, J. Bull. Chem. Soc. Jpn. 2000, 73, 1.
(19) Nakayama, J. Sulfur Rep. 2000, 22, 123.
H
dx.doi.org/10.1021/jo402373x | J. Org. Chem. XXXX, XXX, XXX−XXX