Liquid crystals based on hypervalent sulfur fluorides: Exploring the steric effects of ortho-fluorine substituents

Article abstract of DOI:10.1002/ejoc.200500125

Liquid crystals that have ortho-fluorinated pentafluoro- λ⁶-sulfanyl terminal groups are the most polar materials that are still compatible with current active matrix LCD technology. The synthesis of the first examples of this class of substance has been achieved by direct fluorination. The physico-chemical characterization, supported by DFT calculations, revealed some surprising features of these new compounds: the bulky SF₅ group is easily deformed and responds in a quite unusual way to the steric strain exerted by its ortho substituents. Wiley-VCH Verlag GmbH & Co. KGaA, 2005.

Full text of DOI:10.1002/ejoc.200500125

FULL PAPER
Liquid Crystals Based on Hypervalent Sulfur Fluorides: Exploring the Steric
Effects of ortho-Fluorine Substituents[‡]
Peer Kirsch*[a] and Alexander Hahn[a]
Keywords: Density functional calculations / Fluorination / Liquid crystals / Hypervalent sulfur fluorides / Mesogens /
Molecular modelling
Liquid crystals that have ortho-fluorinated pentafluoro-λ6-
sulfanyl terminal groups are the most polar materials that are
still compatible with current active matrix LCD technology.
The synthesis of the first examples of this class of substance
has been achieved by direct fluorination. The physico-chemi-
cal characterization, supported by DFT calculations, revealed
some surprising features of these new compounds: the bulky
SF5 group is easily deformed and responds in a quite unusual
way to the steric strain exerted by its ortho substituents.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
In our previous communication[7a] we pointed out that
the pentafluoro-λ6-sulfanyl group is easily deformed, re-
sulting in significant changes in its dipole moment. There-
fore, the effect of ortho-fluorination on the sterically de-
manding pentafluoro-λ6-sulfanyl group might be quantita-
tively as well as qualitatively different to the effect on the
smaller trifluoromethyl group for which the simple addition
of local dipole moments leads to an observed increase in
Δε of 8.5 units.
Introduction
More than 40 years after their initial synthesis and char-
acterization,[1] the chemistry of compounds that have a
pentafluoro-λ6-sulfanyl substituent is currently experienc-
ing a renaissance.[2,3] The combination of high polarity and
lipophilicity (“polar hydrophobicity”),[4] which is unique to
the SF5 group, renders it an attractive building block not
only for biomedical applications[5] but also for materials sci-
ences.[6] The use of a pentafluoro-λ6-sulfanyl group offers
some advantages,[7] particularly in liquid crystals, because
it is the most polar terminal group that is still compatible
with active matrix LCD (liquid crystal display) technol-
ogy.[8] For this application it is imperative that the polar
function does not coordinate to cations, thereby mobilizing
ubiquitous ionic impurities in the liquid crystal and other
peripheral display components.[8b] Nevertheless, there is a
need for even more polar liquid crystalline materials for
LCDs with lower power consumption particularly in mobile
devices, such as notebook PCs, cellular phones and PDAs
(personal digital assistants).
Results and Discussion
Synthesis and Characterization of Liquid Crystals
In order to obtain experimental evidence on how the
pentafluoro-λ6-sulfanyl group responds to steric pressure
from ortho substituents, the liquid crystal 4 was selected as
a target compound. The material 5 was also chosen because
the combination of an ortho-fluoro(pentafluoro-λ6-sulfanyl)-
phenyl group with a difluorooxymethylene bridge in the
mesogenic core structure promised some attractive features,
such as high dielectric anisotropy (Δε) and also a broad
nematic phase range.[7c] Liquid crystals are a particularly
convenient tool for the experimental determination of the
mode of sterically induced deformation of the SF5 group
because even relatively small changes in the molecular di-
pole moment translate into noticeable changes in dielectric
anisotropy.[9]
The molecular dipole moment (μ) and the dielectric an-
isotropy (Δε) of liquid crystals is usually increased by flank-
ing the polar terminal group by one or two ortho-fluorine
substituents on the aromatic moiety.[7] Thus, for the liquid
crystal 1, addition of two ortho-fluorine atoms (2) increases
the dielectric anisotropy by 8.5 units (Scheme 1 and
Table 1).
In his first paper on pentafluoro-λ6-sulfanyl com-
pounds[1a] Sheppard pointed out that the synthesis of ortho-
substituted aromatic SF5 derivatives would be problematic
as a result of strong steric interactions with the bulky SF5
group. More than 40 years later Thrasher and co-workers
succeeded in the first synthesis of ortho-fluoro(pentafluoro-
[‡] Liquid Crystals Based on Hypervalent Sulfur Fluorides, 2. Part
1: Ref.[7a]
[a] Merck KGaA, Liquid Crystals Division,
64293 Darmstadt, Germany
Fax: +49-6151-72-2593
E-mail: peer.kirsch@merck.de
Eur. J. Org. Chem. 2005, 3095–3100
DOI: 10.1002/ejoc.200500125
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3095

P. Kirsch, A. Hahn
FULL PAPER
Scheme 1. Increase in the dielectric anisotropies (Δε) of liquid crystals induced by the replacement of CF3 by the SF5 group (1 Ǟ 3) and
by difluorination of the ortho positions of the polar terminal group (1 Ǟ 2).
Table 1. The physical properties of the liquid crystals 1–5 and 14– quite harsh and strongly acidic conditions. The monofluoro
17 for comparison.[a]
compound 10a was converted into the corresponding phe-
nol 11, although we did not succeed in obtaining an analyti-
cally pure sample. Nevertheless, subjecting the crude phenol
to oxidative phenoxydifluorodesulfuration in the presence
of the ketene dithioketal 12[7c] yielded 7% of the liquid crys-
tal 5 which was isolated and purified by chromatography
and subsequent crystallization.
An attempt to convert the bromide 10b into the corre-
sponding phenol failed completely to furnish even an im-
pure product. Also, Suzuki coupling of 10b with the bo-
ronic acid 13 produced an untypically[7a] low yield (16%)
Com-
pound
Phase sequence
Δεvirt
Δnvirt
TNI,virt
of the liquid crystal 4.
1
These new compounds were all characterized by H and
19F NMR spectroscopy and by EI mass spectrometry. Some
of the intermediates contained minor by-products resulting
from over-fluorination. In these cases the compounds were
used without further purification in the subsequent reaction
steps. The purities were determined by HPLC and are noted
in the Exp. Sect.
1
2
C 134 I
C 86 I
C 109 N (87.8) I
C 103 I
C 50 N 101.9 I
C 66 N 94.1 I
C 133 I
13.0
21.5
14.3
21.4
14.3
9.7
9.5
11.7
11.8
0.165
0.149
0.154
0.132
0.080
0.075
0.091
0.093
0.080
109.5
47.1
94.6
49.6
87.6
74.7
112.2
95.5
108.2
3
4
5
14
15
16
17
C 121 I
C 67 N 116.5 I
In conclusion, it seems that in some cases the yields of
the synthetic conversions of ortho-fluoro(pentafluoro-λ6-
sulfanyl)arenes decrease with an increasing number of flu-
orine substituents. On the other hand, none of the isolated
and purified reaction products shows any sign of instability.
The “virtual” parameters,[12] which were used to charac-
terize the new liquid crystals 4 and 5 (Table 1), reflect the
contribution of a given compound to the anisotropic char-
acteristics of a liquid crystal mixture. They play an impor-
tant role in the application-oriented evaluation of new com-
pounds as they allow electrooptic, viscoelastic and mesog-
enic properties to be compared even if they do not form a
thermodynamically stable mesophase.
[a] The virtual electrooptical parameters (Δεvirt, Δnvirt) and clearing
points (TNI,virt) were extrapolated from the mixture ZLI-4792
(Merck)[12] (C = crystalline, N = nematic, I = isotropic; the phase
transition temperatures are given in °C, numbers in parentheses
denote monotropic phase transitions.).
λ6-sulfanyl)benzene derivatives[3] and were also able to per-
form a variety of nucleophilic replacement reactions on the
ortho-fluorine atom with even more sterically demanding
substituents. However, the resulting derivatives did not have
the right substitution pattern for the synthesis of liquid
crystals.
Like its trifluoromethyl analogue 2, compound 4 does
As the starting point for our syntheses (Scheme 2), we not exhibit any mesophase. It shows a slightly higher virtual
chose the direct fluorination[10] of the suitably fluorinated clearing point which may be attributed to the larger steric
bis(4-nitrophenyl) disulfides 7a and 7b, which were synthe- bulk of the SF5 group. Compared with the simple penta-
sized from 6a and 6b, respectively.[11] The ortho-fluorinated fluoro-λ6-sulfanyl derivative 3 the dielectric anisotropy
(pentafluoro-λ6-sulfanyl)benzenes 8a and 8b were obtained (Δεvirt) of its ortho-difluoro analogue is increased by only
in rather low but reproducible yields (9% and 4.6%, respec- about 7 units and has nearly the same dielectric anisotropy
tively). These nitro compounds were catalytically hydroge- as its trifluoromethyl analogue 2. This means that the con-
nated to the amines 9a and 9b in high yields without com- tribution of the SF5 group to the overall dipole moment
plications. Also the subsequent Sandmeyer reaction fur- and dielectric anisotropy of 4 is the same as that of the CF3
nished the bromides 10a and 10b with no indication of in- group in 2, whereas usually (e.g., for the pair 15 and 16) Δε
stability in any of the reaction intermediates, even under is about 2 units higher for the SF5 compound.
3096
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 3095–3100

Liquid Crystals Based on Hypervalent Sulfur Fluorides
FULL PAPER
Scheme 2. Synthesis of the liquid crystals 4 and 5; reagents and conditions: a) Na2S·9H2O, DMSO, room temp. Ǟ 50 °C; b) NaBO4·3H2O,
HOAc, room temp. (7a: 63%, 7b: 77%); c) 10% F2/N2, CH3CN, –3 Ǟ 0 °C (8a: 9%, 8b: 4.6%); d) cat. Raney Ni, H2, THF, room temp.,
1 bar (9a: 95%, 9b: 95%); e) 1. NaNO2, 47% HBr, 0–5 °C; 2. CuBr, 85 °C (10a: 51%, 10b: 49%). f) 1. tBuLi, Et2O, –70 °C; 2. B(OMe)3;
3. HOAc, H2SO4, 30% H2O2, –20 Ǟ 35 °C (crude product used for next step); g) 1. 12, CF3SO3H, CH2Cl2, 0 °C (5 min) Ǟ room temp.
(30 min) Ǟ –70 °C; 2. 11, NEt3, CH2Cl2, –70 °C; 3. NEt3·3HF, –70 °C; 4. Br2, CH2Cl2, –70 Ǟ –10 °C (7%); h) 13, NaBO2·8H2O, H2O,
THF, cat. Pd(PPh3)4, reflux, 18 h (16%).
The general tendency of SF5-substituted liquid crystals requiring less than 1 kcalmol–1 in terms of steric strain. In
not to form thermodynamically stable mesophases is shared contrast to the experimental results, this would mean that
by their CF3 analogues. These materials have found some the gain in the molecular dipole moment (μ) and the dielec-
limited practical applications[8a] owing to their high po- tric anisotropy (Δε) for SF5-substituted liquid crystals
larity, but in general, liquid crystals with a broad nematic should be much higher than for their otherwise similar CF3
phase, such as 14, are much more versatile and have a wide analogues. Alternatively, the steric strain might be relieved
range of applications. In our previous communication[7c] we by the elongation of the carbon–sulfur bond.
showed that the ability of liquid crystals to form a nematic
phase can be effectively promoted by the insertion of a di-
fluorooxymethylene bridge into the mesogenic core struc-
ture. This design principle also works for the material 17 in
which a broad nematic phase (relative to 16) is induced by
the CF2O bridge, along with an increase in the virtual clear-
ing point of about 13 °C. The ortho-fluorine substituent in
5 leads to a further increase in Δε of 2.5 units (compared
with 17) and retains a broad nematic phase range. The ob-
served drop in the virtual clearing points (5 vs. 17, as well
as 4 vs. 3) is a normal effect of ortho-fluorination of any
polar terminal group[8] and can be seen, in terms of practi-
cal applications, as the price to pay for increased polarity.
Nevertheless, quantum chemical calculations [geometry
optimization at the B3LYP/6-31+G(d) level of theory][9]
indicate that the SF5 group copes with the repulsive steric
interactions with ortho-fluoro substituents (Table 2) in a dif-
ferent way: the equatorial fluorine atoms are neither pushed
“forward” by the ortho substituents (A) nor is the carbon–
sulfur bond elongated (B); instead they are pushed apart
allowing the ortho-fluorine atom to “squeeze” in between
(C). The C–S–Feq angle (for the monofluorinated model
system the averaged value of 92.34° for the two relevant
angles has to be considered) is even reduced by the ortho-
fluorination. Of these three deformation modes, only for
A could an extraordinary increase in the molecular dipole
moment or the dielectric anisotropy be expected to occur.
Together with the experimental evidence, the molecular
modelling results indicate that ortho-difluorination (4 vs. 3)
distorts the SF5 group by mode C (Table 2), that is, by en-
larging the Feq–S–Feq angle. A concomitant reduction of
the C–S–Feq angle leads to the observed decrease in the
polarity of the SF5 group in the liquid crystal 4.
Quantum Chemical Calculations
Intuitively, it may be expected that the repulsive interac-
tions of the ortho-fluorine atoms “push” the equatorial flu-
orine atoms away from the aromatic moiety rendering the
pentafluoro-λ6-sulfanyl group as a whole more polar,
thereby exceeding the effect of the simple addition of lateral
carbon–fluorine dipole moments. In our previous com-
For ortho-monofluorinated SF5 derivatives this effect is
munication[7a] we showed that the molecular dipole mo- not observed experimentally: compound 5 shows the ex-
ment is increased by about 0.3 D per degree of deformation, pected increase in Δε induced by its ortho-fluorine substitu-
Eur. J. Org. Chem. 2005, 3095–3100
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3097

P. Kirsch, A. Hahn
FULL PAPER
Table 2. Results of the quantum chemical calculations (geometry 7.8 kcalmol–1 for 20 (two ortho-fluorine atoms). During the
optimization) performed at the B3LYP/6-31+G(d) level of theory.[a]
transition, the “passing” equatorial fluorine atoms are
pushed “forward” by about 0.8° by the aromatic ortho-hy-
drogen atoms and by about 1.5° by the ortho-fluorine
atoms.
Conclusions
18
19
20
For the first time we have been able to synthesize liquid
crystals with a polar pentafluoro-λ6-sulfanyl group and ad-
ditional ortho-fluorine substituents. These materials were
used in combination with quantum chemical calculations
to determine how the bulky SF5 group responds to steric
pressure, and how the dynamics of its rotation are affected.
The polarity of a terminal SF5 group can be enhanced sig-
nificantly by ortho-monofluorination (5). Introduction of a
second ortho-fluorine atom (4) leads to a sterically induced
deformation of the SF5 group which reduces its polarity to
about that of the more “conventional” CF3 group. In spite
of this limitation, use of the pentafluoro-λ6-sulfanyl group
has allowed the design of liquid crystals with the most polar
terminal functions that are still compatible with active ma-
trix LCD technology. The unique combination of high po-
larity, low polarizability and chemical stability renders the
SF5 group an interesting structural feature not only for
polymers and functional materials in molecular electronics,
but also in medicinal chemistry.
C–S [Å]
S–Fax [Å]
1.838
1.633
1.642
–
92.57
–
89.82
–
4.65
1.835
1.631
1.642
1.632
91.97
92.70
89.39
90.74
5.16
1.837
1.630
–
1.632
–
92.11
–
90.32
5.44
S–Feq (o-H) [Å]
S–Feq (o-F) [Å]
C–S–Feq (o-H) [°]
C–S–Feq (o-F) [°]
Feq–S–Feq (o-H) [°]
Feq–S–Feq (o-F) [°]
μ [D]
[a] The calculations (geometry optimization at the B3LYP/6-
31+G(d) level of theory)[13] indicate that the pentafluoro-λ6-sul-
fanyl group responds to the repulsive interactions with the ortho
substituents X1 and X2 by widening the Feq–S–Feq angle as a result
of the partial insertion of the ortho-fluorine atom. Note that, in
contrast to the calculations reported here, the only reported experi-
mental value for the dipole moment of (pentafluoro-λ6-sulfanyl)-
benzene (18) is 3.44 D.[1]
ent relative to 17. The quantum chemical calculations on
the model system 19 indicate that only the equatorial fluor-
ine atoms on the side of the ortho-fluorine atom are pushed
apart. At the same time they are also pushed forward by
about 0.1°. On the side of the ortho-hydrogen atom, the C–
S–Feq angle is slightly reduced, resulting in an overall tilt
of the plane formed by the equatorial fluorine atoms. The
effects on the two sides of the SF5 group seem to cancel
each other and so the group dipole moment is not signifi-
cantly changed.
Experimental Section
General: The starting materials 6a and 6b are commercially avail-
able from Apollo Scientific. They were converted into the corre-
sponding disulfides 7a and 7b according to methods previously re-
ported.[11] Caution: Working with elemental fluorine is potentially
hazardous! The fluorination reactions were carried out under a
well-ventilated hood in a 1-L PFA bottle with PTFE and PFA tub-
ing. Even traces of non-fluorine-resistant organic compounds, such
In addition the dynamics of the molecular rotation of the
SF5 group around the central carbon–sulfur bond is affec-
ted by the steric hindrance of the ortho-fluorine atoms. In as lubricants, can combust spontaneously. Before and after opera-
tion the apparatus was purged with dry nitrogen. Excess fluorine
and hydrofluoric acid were scrubbed in a column filled with a mix-
ture of granulated charcoal and aluminium oxide pearls. The aceto-
nitrile used as the reaction solvent (Merck KGaA, DNA synthesis
grade) contained less than 10 ppm of water. Interruption of the
fluorination for several hours overnight did not reduce the reaction
yields.
the transition state of this rotation (Scheme 3) two of the
equatorial fluorine atoms have to “pass” the aromatic ortho
substituents. Calculations at the B3LYP/6-311+G(2d,p)//
B3LYP/6-31+G(d) level of theory indicate that the relative
energies of these transition states are elevated by ortho-fluo-
rination. Whereas the rotational barrier of (pentafluoro-λ6-
sulfanyl)benzene (18) is only 1.8 kcalmol–1, it is
4.4 kcalmol–1 for 19 (one ortho-fluorine atom) and
8a: A solution of 7a (35.0 g, 101.6 mmol) in dry acetonitrile
(500 mL) was cooled to –3 °C. A stream of 10% F2/N2 was bubbled
into the vigorously stirred solution at a rate of 200 mLmin–1 at a
temperature between –3 and 0 °C for approximately 32 h under
GC-MS control (the samples were injected into the GC apparatus
without prior aqueous work-up, in order to avoid extraction of
water-soluble intermediates). After completion of the reaction, the
apparatus was purged with nitrogen and the solution was concen-
trated to dryness. The residue was dissolved in CH2Cl2 (400 mL)
and stirred with ice/water (500 mL). The organic layer was sepa-
rated and washed three times with 10% aqueous NaOH and twice
with water. The solution was dried with Na2SO4 and the solvent
evaporated to dryness. The residue (25.1 g) was distilled in a Kugel-
rohr apparatus (b.p. 120 °C/0.1 mbar) to furnish 13.4 g of a crude
Scheme 3. The activation barriers, EA, for the rotation of the SF5
group around the central C–S bond: 18: 1.8 kcalmol–1; 19:
4.4 kcalmol–1; 20: 7.8 kcalmol–1.[13]
3098
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 3095–3100

Liquid Crystals Based on Hypervalent Sulfur Fluorides
FULL PAPER
product, which was purified by chromatography (silica gel; toluene/
n-heptane, 3:2) and distilled again (b.p. 102 °C/0.1 mbar) to yield
the product 8a (4.9 g, 9%) as a yellow oil, m.p. 11 °C (92% purity
by GC, over-fluorinated by-product). 1H NMR (300 MHz, CDCl3,
DMSO, 303 K): δ = 7.67 (d, J = 8.8 Hz, 1 H, Ar-H), 7.92–7.98 (m,
2 H, Ar-H) ppm. 19F NMR (235 MHz, CDCl3, 300 K; standard
CFCl3): δ = –104.9 to –104.4 (mc, 1 F, Ar-F), 66.8 (ddd, J = 151.9,
J = 24.6, J = 3.5 Hz, 4 F, SFeq), 86.1 (quint, J = 151.9 Hz, 1 F,
303 K): δ = 8.00 (dd, J = 9.8, J = 6.9 Hz, 1 H), 8.08–8.15 (m, 2 H) SFax; an additional, insufficiently resolved fine structure is underly-
ppm. 19F NMR (235 MHz, CDCl3, 300 K; standard CFCl3): δ = –
103.1 (mc, 1 F, Ar-F), 66.8 (dd, J = 150.9, J = 25.2 Hz, 4 F, SFeq),
77.2 (quint, J = 150.9 Hz, 1 F, SFax; an additional, insufficiently
resolved fine structure is underlying the quint) ppm. MS (EI,
70 eV): m/z (%) = 267 [M+] (100), 248 (30), 221 (24), 113 (58), 101
(16), 94 (31), 93 (31), 89 (45). HRMS for [M+] (C6H3F6NO2S):
calcd. 266.9789; found 266.9796.
ing the quint) ppm. MS (EI, 70 eV): m/z (%) = 300/302 (100) [M+],
281/283 (18), 192/194 (40), 173/175 (18), 113 (78), 94 (60), 93 (25),
89 (34), 74 (18).
10b: Starting from 9b (1.5 g, 3.2 mmol) the same procedure as that
in the synthesis of 10a was applied. The yield of 10b was 510 mg
(49%) as a colorless oil (70% purity by GC, containing 14% of
over-fluorinated by-product). 1H NMR (300 MHz, CDCl3, 303 K):
8b: Starting from 7b (35.0 g, 92.0 mmol) the same procedure as that δ = 7.25 (d, J = 7.6 Hz, 2 H, overlap with solvent signal) ppm. 19
F
in the synthesis of 8a was applied. The final Kugelrohr distillation
(b.p. 85 °C/0.1 mbar) furnished 8b (2.4 g, 4.6%) as a yellow oil
NMR (235 MHz, CDCl3, 300 K; standard CFCl3): δ = –105.0 to
–104.5 (m, 2 F, Ar-F), 71.0–72.1 (m, 4 F, SFeq), 73.0–76.8 (m, 1 F,
(79% purity by GC, over-fluorinated by-product). 1H NMR SFax) ppm. MS (EI, 70 eV): m/z (%) = 318/320 (100) [M+], 299/301
(300 MHz, CDCl3, 303 K): δ = 7.92 (d, J = 8.3 Hz, 2 H) ppm. 19F
NMR (235 MHz, CDCl3, 300 K; standard CFCl3): δ = –98.2 to
–97.7 (m, 2 F, Ar-F), 71.6–75.7 (m, 5 F, SF5; the SFax quint with
an additional, insufficiently resolved fine structure is overlapping a
complex SFeq signal group) ppm. MS (EI, 70 eV): m/z (%) = 285
[M+] (100), 266 (38), 239 (13), 131 (39), 119 (25), 112 (34), 99 (16),
89 (49), 83 (23), 81 (42), 69 (18), 65 (90), 62 (22). HRMS for [M+]
(C6H2F7NO2S): calcd. 284.9694; found 284.9690.
(18), 228/230 (13), 210/212 (53), 191/193 (15), 149 (16), 131 (65),
130 (20), 112 (68), 89 (40), 81 (36).
11: A solution of 10a (2.0 g, 6.6 mmol) in diethyl ether (50 mL) was
cooled to –70 °C, and tert-butyllithium (9.0 mL of a 15% solution
in n-pentane, 13.3 mmol; Caution: ignites spontaneously on contact
with air!) was added dropwise. After stirring at –70 °C for 30 min,
trimethyl borate (0.8 mL, 7.2 mmol) was added, and after 30 min
the temperature was allowed to rise to –20 °C. Then, acetic acid
(0.4 mL, 7.0 mmol), 96% H2SO4 (0.3 mL, 5.6 mmol), 30% hydro-
gen peroxide (1.3 mL, 13 mmol) and water (5 mL) were added. Af-
ter warming the mixture to 35 °C over 2 h, it was cooled to room
temp. and the organic phase was separated. The aqueous phase was
9a: A solution of 8a (4.5 g, 15.5 mmol) in THF (30 mL) was hydro-
genated in the presence of Raney nickel (0.2 g, THF wet) at room
temp. and ambient pressure. The mixture was filtered through Ce-
lite and the solvents were evaporated to dryness to furnish 9a (3.5 g,
95%) as an off-white solid, m.p. 56.5 °C (93% purity by HPLC). extracted a second time with diethyl ether. The combined organic
1H NMR (300 MHz, CDCl3, 303 K): δ = 4.10 (br. s, 2 H, NH2), phases were concentrated to dryness, and the residue was co-evapo-
6.32–6.41 (m, 2 H), 7.47 (t, J = 8.5 Hz, 1 H) ppm. 19F NMR rated three times with methanol and subsequently dried in vacuo.
(235 MHz, CDCl3, 300 K; standard CFCl3): δ = –107.8 to –108.3
(mc, 1 F, Ar-F), 69.9 (dd, J = 150.9, J = 23.5 Hz, 4 F, SFeq), 85.1
(quint, J = 150.9 Hz, 1 F, SFax; an additional, insufficiently re-
solved fine structure is underlying the quint) ppm. MS (EI, 70 eV):
m/z (%) = 237 (100) [M+], 128 (27), 129 (70), 110 (51), 102 (13),
101 (14), 90 (12), 89 (14), 83 (48). HRMS for [M+] (C6H5F6NS):
calcd. 237.0047; found 237.0044.
The crude phenol 11 (1.3 g, 81%) was used in the next reaction
step without further purification.
4: A mixture of 10b (450 mg, 1.41 mmol), NaBO2·8H2O (300 mg),
water (5.0 mL), PdCl2(PPh3)2 (50 mg) and hydrazine hydrate
(50 μL) was stirred at room temp. for 15 min. Then, 13 (400 mg,
1.63 mmol) was added, and the mixture was heated at reflux under
nitrogen for 18 h. After cooling, the solution was diluted with tert-
9b: Starting from 8b (2.0 g, 7.0 mmol) the same procedure as that butyl methyl ether (100 mL), and water (20 mL) was added. The
in the synthesis of 9a was applied. The yield of 9b was 1.7 g (95%).
organic layer was separated, washed with brine, dried with Na2SO4
and the solvents were evaporated to dryness. The residue was sub-
The crude product was used without further purification in the
1
next step. H NMR (300 MHz, CDCl3, 303 K): δ = 3.22 (br. s, 2 jected to chromatography through a short silica gel column (n-hep-
H, NH2), 6.19 (d, J = 11.7 Hz, 2 H) ppm. 19F NMR (235 MHz, tane). According to HPLC analysis, the purity of the eluate was
CDCl3, 300 K; standard CFCl3): δ = –107.3 to –106.7 (m, 2 F, Ar-
F), 73.3–74.5 (m, 4 F, SFeq), 77.8–82.0 (m, 1 F, SFax) ppm. MS
(EI, 70 eV): m/z (%) = 255 (100) [M+], 236 (14), 205 (12), 165 (13),
147 (55), 128 (51), 119 (17), 101 (61), 89 (15).
90.4% with 9.3% of over-fluorinated by-product. The product frac-
tion was purified by preparative HPLC (RP-18 phase; acetonitrile)
and crystallized from n-heptane at –70 °C to yield 4 (110 mg, 16%)
as colorless crystals (purity 99.9% by HPLC), m.p. 103 °C. 1H
NMR (300 MHz, CDCl3, 303 K): δ = 0.91 (t, J = 7.0 Hz, 3 H),
0.99–1.56 (m, 9 H), 1.87–1.94 (m, 4 H), 2.53 (mc, 1 H), 7.22 (d, J
= 10.8 Hz, 2 H, Ar-H; overlap with solvent signal), 7.33 (d, J =
8.4 Hz, 2 H, Ar-H), 7.48 (d, J = 8.4 Hz, 2 H, Ar-H) ppm. 19F
NMR (235 MHz, CDCl3, 300 K; standard CFCl3): δ = –105.6 to
–105.0 (m, 2 F, Ar-F), 72.7–73.7 (m, 4 F, SFeq), 77.3–79.8 (m, 1 F,
SFax) ppm. MS (EI, 70 eV): m/z (%) = 440 (100) [M+], 355 (29),
342 (95), 329 (50), 228 (14), 215 (13), 202 (17), 86 (16).
10a: A mixture of 9a (3.2 g, 13.5 mmol) and 23% HBr (40 mL) was
cooled to 0 °C, and a solution of NaNO2 (1.2 g, 17.4 mmol) in
water (7 mL) was added dropwise, keeping the temperature between
0 and 2 °C. After the completion of the diazotization reaction
(30 min), urea (50 mg) was added in order to destroy excess nitrous
acid. Then, a solution of CuBr (4.3 g, 30 mmol) in 47% HBr
(15 mL) was added at 0–5 °C. The strongly foaming mixture was
heated to 85 °C for 1 h. When the evolution of nitrogen had ceased,
it was cooled to room temp., poured into ice/water and extracted
with n-pentane. The aqueous phase was extracted twice with n-
pentane, and the combined organic phases were washed with water
and saturated aqueous NaHCO3 solution, dried with Na2SO4 and
the solvents evaporated to dryness. The residue was purified by
chromatography (silica gel; n-pentane) to yield 10a (2.1 g, 51%) as
a colorless oil (94% purity by GC). 1H NMR (300 MHz, [D6]
5: A solution of 12 (1.2 g, 3.6 mmol) in CH2Cl2 (30 mL) was treated
dropwise at 0 °C with triflic acid (320 μL, 3.65 mmol). After
10 min, the mixture was allowed to warm to room temp., stirred
for 30 min and then cooled to –70 °C. A solution of 11 (1.3 g of
crude product, ca. 5.5 mmol) and NEt3 (800 μL, 5.77 mmol) in
CH2Cl2 (10 mL) was added dropwise, followed after 30 min at
–70 °C first by NEt3·3HF (3.0 mL, 18.6 mmol) and then by bro-
Eur. J. Org. Chem. 2005, 3095–3100
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3099

P. Kirsch, A. Hahn
FULL PAPER
Mol. Cryst. Liq. Cryst. 2000, 346, 29–33; c) P. Kirsch, M.
Bremer, A. Taugerbeck, T. Wallmichrath, Angew. Chem. 2001,
113, 1528–1532; Angew. Chem. Int. Ed. 2001, 40, 1480–1484.
[8] a) P. Kirsch, M. Bremer, Angew. Chem. 2000, 112, 4384–4405;
Angew. Chem. Int. Ed. 2000, 39, 4216–4235; b) M. Bremer, S.
Naemura, K. Tarumi, Jpn. J. Appl. Phys. 1998, 37, L88–L90.
[9] Dielectric anisotropy is defined as Δε = ε|| – εЌ and birefrin-
gence as Δn = n|| – nЌ, where || represents the orientation paral-
lel and Ќ perpendicular to the nematic phase director, which
are approximated by the molecular orientation axis and the
long molecular axis, respectively. The correlation between Δε,
the molecular dipole moment μ and the angle β between the
dipole and the orientation axis is given by: Δε ∝ Δα –
F(μ2/2kBT)(1 – 3cos2β)S, where Δα is the anisotropy of the po-
larizability, F the reaction field factor and S the order parame-
ter: a) W. Maier, G. Meier, Z. Naturforsch. 1961, 16A, 262–267;
b) D. Demus, G. Pelzl, Z. Chem. 1982, 21, 1; c) J. Michl, E. W.
Thulstrup, Spectroscopy with Polarized Light: Solute Alignment
by Photoselection, in Liquid Crystals, Polymers, and Membranes,
VCH, Weinheim, 1995, pp. 171–221.
mine (1.0 mL, 19.5 mmol) in CH2Cl2 (10 mL). The orange mixture
was stirred at –70 °C for 1 h, and then it was allowed to warm to
–10 °C and poured into ice-cold 2 n NaOH. The organic phase was
separated and the aqueous phase was extracted again with CH2Cl2.
The combined organic phases were dried with Na2SO4, filtered and
the solvents evaporated to dryness. The tar-like crude product was
twice subjected to chromatography through a short silica gel col-
umn using n-heptane as eluent and crystallized from n-heptane at
–20 °C to yield 5 (120 mg, 7%) as colorless crystals, m.p. 50 °C,
1
nematic to 101.9 °C, isotropic (purity 99.0% by HPLC). H NMR
(300 MHz, CDCl3, 303 K): δ = 0.87 (t, J = 7.2 Hz, 3 H), 0.90–1.42
(m, 15 H), 1.69–1.87 (m, 6 H), 1.95–2.08 (m, 3 H), 7.00–7.06 (m,
2 H, Ar-H), 7.70 (t, J = 8.3 Hz, 1 H, Ar-H) ppm. 19F NMR
(235 MHz, CDCl3, 300 K; standard CFCl3): δ = –105.6 to –105.1
(m, 1 F, Ar-F), –79.2 (d, J = 8.6 Hz, 2 F, CF2O), 68.4 (ddd, J =
150.8, J = 24.6, J = 3.4 Hz, 4 F, SFeq), 81.5 (quint, J = 150.8 Hz,
1 F, SFax; an additional, insufficiently resolved fine structure is
underlying the quint) ppm. MS (EI, 70 eV): m/z (%) = 494 (22)
[M+], 475 (9), 256 (34), 125 (45), 111 (26), 83 (67), 69 (100).
[10] a) M. P. Greenhall, 15th International Symposium on Fluorine
Chemistry, Vancouver, Canada, Aug. 2–7, 1997, presentation
FRx C-2; b) R. D. Bowden, M. P. Greenhall, J. S. Moillet, J.
Thomson, WO 97/05106, 1997 [Chem. Abstr. 1997, 126,
199340].
Acknowledgments
[11] Disulfides 7a and 7b were prepared by procedures analogous
to those described in: a) E. A. Kuo, P. T. Hambleton, D. P. Kay,
P. L. Evans, S. S. Matharu, E. Little, N. McDowall, C. B. Jones,
C. J. R. Hedgecock, C. M. Yea, A. W. E. Chan, P. W. Hairsine,
I. R. Ager, W. R. Tully, R. A. Williamson, R. Westwood, J.
Med. Chem. 1996, 39, 4608–4621; b) A. McKillop, D.
Koyunçu, A. Krief, W. Dumont, P. Renier, M. Trabelsi, Tetra-
hedron Lett. 1990, 31, 5007–5010.
We thank J. Haas, H. Heldmann and K. Altenburg for the physical
characterization of the new compounds, and Dr. M. Bremer for
valuable advice on the quantum chemical calculations.
[1] a) W. A. Sheppard, J. Am. Chem. Soc. 1960, 82, 4751–4752; b)
W. A. Sheppard, J. Am. Chem. Soc. 1962, 84, 3064–3071; c)
W. A. Sheppard, J. Am. Chem. Soc. 1962, 84, 3072–3076.
[2] For reviews on pentafluoro-λ6-sulfanyl derivatives, see: a) D.
Lentz, K. Seppelt in Chemistry of Hypervalent Compounds
(Ed.: K. Akiba), Wiley, New York, 1999, chapter 10, p. 295; b)
R. Winter, G. L. Gard in Inorganic Fluorine Chemistry:
Towards the 21st Century (Eds.: J. S. Thrasher, S. H. Strauss),
ACS Symposium Series 555, Washington DC, 1994, p. 128.
[3] a) A. M. Sipyagin, C. P. Bateman, Y.-T. Tan, J. S. Thrasher, J.
[12] The “virtual” parameters TNI,virt, Δεvirt and Δnvirt were deter-
mined by linear extrapolation from a 10% w/w solution in the
commercially available Merck mixture ZLI-4792 (TNI
=
92.8 °C, Δε = 5.3, Δn = 0.0964). The extrapolated values were
corrected empirically for differences in the order parameter
which are induced by the analyte. For the pure substances, the
melting points were identified by differential scanning calorim-
etry (DSC) and the phases by optical polarization microscopy.
Fluorine Chem. 2001, 112, 287–295; b) A. M. Sipyagin, S. V. [13] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
Enshov, S. A. Kashtanov, C. P. Bateman, B. D. Mullen, Y.-T.
Tan, J. S. Thrasher, J. Fluorine Chem. 2004, 125, 1305–1316.
[4] a) P. Kirsch, Modern Fluoroorganic Chemistry: Synthesis Reac-
tivity, Applications, Wiley-VCH, Weinheim, 2004, p. 146; b)
H. W. Kim, P. Rossi, R. K. Shoemaker, S. G. DiMagno, J. Am.
Chem. Soc. 1998, 120, 9082–9083.
[5] a) J. E. D. Barton, G. Mitchell (Zeneca Ltd), GB 2276379, 1994
[Chem. Abstr. 1995, 123, 82948]; b) P. J. Crowley, G. Mitchell,
R. Salmon, P. A. Worthington, Chimia 2004, 58, 138–142.
[6] a) R. W. Winter, R. Dodean, L. Holmes, G. L. Gard, J. Fluor-
ine Chem. 2004, 125, 37–41; b) R. Winter, P. G. Nixon, R. J.
Terjeson, J. Mohtasham, N. R. Holcomb, D. W. Grainger, D.
Graham, D. G. Caster, G. L. Gard, J. Fluorine Chem. 2002,
115, 107; c) R. Winter, P. G. Nixon, G. L. Gard, D. G. Castner,
N. R. Holcomb, Y.-H. Hu, D. W. Grainger, Chem. Mater. 1999,
11, 3044; d) J. C. Hansen, P. M. Savu, US 5159105, 1992
[Chem. Abstr. 1992, 116, 58727].
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Mont-
gomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas,
J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson,
P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rab-
uck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Or-
tiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komar-
omi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M.
Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Re-
plogle, J. A. Pople, Gaussian98 (Revision A.6), Gaussian, Inc.,
Pittsburgh, PA, 1998. The geometries were optimized at the
B3LYP/6-31+G(d) level of theory, the minima only have posi-
tive eigenfrequencies, the transition states have one negative
eigenfrequency. The energies were calculated at the B3LYP/6-
311+G(2d,p) level of theory using the B3LYP/6-31+G(d)-de-
rived geometries and zero-point energies.
[7] a) P. Kirsch, M. Bremer, M. Heckmeier, K. Tarumi, Angew.
Chem. 1999, 111, 2174–2178; Angew. Chem. Int. Ed. 1999, 38,
1989–1992; b) P. Kirsch, M. Bremer, M. Heckmeier, K. Tarumi,
Received: February 18, 2005
3100
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 3095–3100