Synthesis of a polymerizable fluorosurfactant for the construction of stable nanostructured proton-conducting membranes

Article abstract of DOI:10.1021/jo101196w

The synthesis of the polymerizable fluorinated surfactant sodium 1,1,2,2-tetrafluoro-2-(1,1,2,2-tetrafluoro-2-(4-vinylphenyl)ethoxy) ethanesulfonate (1) and a number of related fluorocarbon compounds is described. Compound 1 is synthesized using a copper-mediated cross-coupling reaction of 4-bromobenzaldehyde and sodium 5-iodooctafluoro-3-oxapentanesulfonate. The resulting benzaldehyde is converted to a styrene unit using a Wittig reaction with methyltriphenylphosphonium bromide in acetonitrile, using DBU as a base. This strategy for converting an iodo-functionalized fluorosurfactant to a styrene-containing fluorosurfactant is highly efficient because both reactions are performed in polar solvents and are compatible with the sulfonate moiety. In addition, the copper-mediated cross-coupling reaction is most efficient with electron-poor aryl bromides like 4-bromobenzaldehyde. We wish to employ 1 for the construction of nanostructured membranes by polymerization of 1 in a microemulsion or in lyotropic liquid crystalline phases.

Full text of DOI:10.1021/jo101196w

pubs.acs.org/joc
Synthesis of a Polymerizable Fluorosurfactant for the Construction
of Stable Nanostructured Proton-Conducting Membranes
†
Mohan N. Wadekar, Wolter F. Jager,* Ernst J. R. Sudholter, and Stephen J. Picken
,‡
‡
†
€
†Laboratory of NanoStructured Materials and ‡Laboratory of Nano-Organic Chemistry,
Department of Chemical Engineering DelftChemTech, Delft University of Technology, Julianalaan 136,
2628BL Delft, The Netherlands
w.f.jager@tudelft.nl
Received June 18, 2010
The synthesis of the polymerizable fluorinated surfactant sodium 1,1,2,2-tetrafluoro-2-(1,1,2,2-tetra-
fluoro-2-(4-vinylphenyl)ethoxy)ethanesulfonate (1) and a number of related fluorocarbon com-
pounds is described. Compound 1 is synthesized using a copper-mediated cross-coupling reaction of
4-bromobenzaldehyde and sodium 5-iodooctafluoro-3-oxapentanesulfonate. The resulting benzalde-
hyde is converted to a styrene unit using a Wittig reaction with methyltriphenylphosphonium
bromide in acetonitrile, using DBU as a base. This strategy for converting an iodo-functionalized
fluorosurfactant to a styrene-containing fluorosurfactant is highly efficient because both reactions
are performed in polar solvents and are compatible with the sulfonate moiety. In addition, the
copper-mediated cross-coupling reaction is most efficient with electron-poor aryl bromides like
4-bromobenzaldehyde. We wish to employ 1 for the construction of nanostructured membranes by
polymerization of 1 in a microemulsion or in lyotropic liquid crystalline phases.
Introduction
integrity, and the anionic moieties form proton-conducting
hydrophilic channels in the membrane.2 Ideally, the proton
conducting polymer must possess long-term chemical stabil-
ity, notably against negative pH, high temperatures, and
aggressive radicals like OH• and OOH•, which are present in
operating fuel cells.3 In practice, (partially) fluorinated and
aromatic polymers have4 the appropriateproperties, andthus,
they are suitable for making proton conducting membranes.
In the present generation PEMFCs, the sulfonated fluoro-
polymer Nafion, a statistical copolymer of tetrafluoroethylene
Nanostructured polymer membranes, composed of poly-
mers with a well-defined morphology and consisting of
at least two phases, find application in many areas.1 Well
known are membranes for filtration that are composed of
cross-linked polymers, which contain mesoscale pores. We
are interested in developing proton-conducting membranes
for polymer electrolyte membrane fuel cells (PEMFC). For
this application, apolar polymers decorated with anionic
hydrophilic proton-conducting moieties, usually -SOh3, are
employed.2 The apolar polymer phase provides structural
(3) LaConti, A. B.; Hamdan, M.; McDonald, R. C.; Vielstich, W. In
Handbook of Fuel Cells-Fundamentals, Technology and Applications;
Gasteiger, H. A., Lamm, A., Eds.; John Wiley and Sons: New York, 2003;
Vol. 3, Chapter 49.
(4) Souzy, R.; Ameduri, B.; Boutevin, B.; Gebel, G.; Capron, P. Solid
State Ionics 2005, 176, 2839–2848.
(1) Gin, D. L.; Bara, J. E.; Noble, R. D.; Elliott, B. J. Macromol. Rapid
Commun. 2008, 29, 367–389.
(2) Mauritz, K. A.; Moore, R. B. Chem. Rev. 2004, 104, 4535–4585.
6814 J. Org. Chem. 2010, 75, 6814–6819
Published on Web 09/16/2010
DOI: 10.1021/jo101196w
r
2010 American Chemical Society

Wadekar et al.
JOCArticle
and a sulfonate containing fluorinated comonomer, is the
most commonly used proton-conducting material.2 From an
emulsion of this polymer, proton-exchange membranes are
produced by various pressing, stamping, and ink deposi-
tion techniques. These membranes are cocontinuous and
highly proton conducting with channel sizes in the order of
1-4 nm.
For membranes prepared from the statistical copolymers
like Nafion, little control over the microstructure is obtained
because of the low mobility of the polymer during membrane
preparation. In addition, the statistical nature of the poly-
mer does not facilitate a discrete phase separation and the
formation of regular and well-developed morphologies.
Another drawback associated with the use of polymers is
that the degree of sulfonation is limited to fairly low values
because highly sulfonated polymers are water-soluble. It
is a known fact that the proton conductivity, water uptake,
and subsequent swelling are strongly influenced by chemi-
cal structure, morphology, and the percentage of proton-
carrying groups in the polymer.5,6 Thus, it is a challenge to
enhance the conductivity by increasing the degree of sub-
stitution with proton carrying groups without increasing the
water uptake and swelling. This problem can partly be solved
by synthesizing block or graft copolymers, in which one
of the blocks is apolar in nature and the other is hydrophi-
lic and carries anionic functionalities. These polymers can
form phase-separated systems. For example, Chung and co-
workers6 have shown that the conductivity of phase sepa-
rated membranes of poly(vinylidenefluoride)-g-sulfonated
polystyrene (PVDF-g-SPS) graft copolymers is better than
that of Nafion 112 due to a higher degree of sulfonation.7 In
this polymer, the water swelling was controlled by the graft
copolymer morphology. However, the rate of membrane
manufacturing is still limited by the slow dynamics of poly-
mer chains.
In an alternative approach, nanostructured membranes
can be assembled by polymerizable surfactants or surfmers
in microemulsions8 or lyotropic liquid crystalline (LLC)
phases.9 Now the mobility of the surfactant molecules is no
longer a limiting factor, and equilibrium states with well-
defined morphologies may be obtained quickly. Mechani-
cally stable membranes are obtained by a subsequent poly-
merization of the polymerizable surfactants. By employing
multifunctional monomers in these systems, cross-linked
polymer membranes are obtained. For such membranes
the degree of sulfonation, and therefore the conductivity,
can be increased, since the cross-linking prevents the polymer
from becoming water-soluble and will also control the degree
of swelling.
FIGURE 1. Polymerizable fluorosurfactant 1 and nonpolymeriza-
ble fluorosurfactant 2.
The approach of using surfmers was pioneered by Luzzati
and co-workers.10 The preparation of proton-conducting
membranes by polymerization of a bicontinuous microemul-
sion of a zwitterionic surfmer, 3-((11-acryloyloxyundecyl)-
imidazoyl)propyl sulfonate, has been reported.11 However,
the chemical stability of this polymer is expected to be low
due to the presence of an easily hydrolyzable acrylate group.
Polymerizable surfactants with hydrocarbon backbones12
and (semi)fluorinated backbones13 have been synthesized
and studied for various objectives like micellar polymeriza-
tion, drug delivery systems, and applications in paints. Most
polymerizable fluorosurfactants are cationic and nonionic
in nature. In one report,14 an anionic (-COOh-based) fluori-
nated surfmer, monofluoroalkyl maleate, has been used by
Pich et al. as a stabilizer in the miniemulsion polymerization
of styrene and n-butyl methacrylate.
Here, we report on the synthesis of the polymerizable
fluorosurfactant 1 and its nonpolymerizable analogue 2
(Figure 1). In the design of 1, we chose the sulfonate group
as the proton- conducting moiety because of its high stability
and high proton conductivity. In addition, many fluoro-
carbons containing fluorosulfonyl groups, from which the
sulfonate group is readily obtained by basic hydrolysis, are
available.13 The styryl functionality was selected because
aromatic moieties can directly be attached to fluorocarbon
chains by a copper-mediated coupling reaction.15 In this
fashion, hydrolyzable ester and amide functionalities, that
are present in acrylates and acrylamides, are avoided.
Although polymerizable surfactants have been synthesized
by direct coupling of styryl moieties using methods like
esterification16 or etherification,17,18 the harsh conditions
of the copper-mediated coupling prevented such a direct
approach. Therefore, 4-bromobenzaldehyde was coupled to
the fluorosurfactant molecule and subsequently converted
into a functional styrene group.
It is expected that the styrene moiety in 1, in contrast to
most fully fluorinated monomers,19,20 is readily polymerized
(11) Lim, T. H.; Tham, M. P.; Liu, Z.; Hong, L.; Guo, B. J. Membr. Sci.
2007, 290, 146–152.
(12) Holmberg, K. Novel Surfactants Preparation, Applications, and
Biodegradability; Marcel Dekker: New York, 1998; Chapter 10.
(13) Kissa, E. Fluorinated Surfactants and Repellents; Marcel Dekker:
New York, 2001; Chapters 1-2.
(5) Ding, J.; Chuy, C.; Holdcroft, S. Macromolecules 2002, 35, 1348–
1355.
(6) Zhang, Z.; Chalkova, E.; Fedkin, M.; Wang, C.; Lvov, S. N.;
Komarneni, S.; Chung, T. C. M. Macromolecules 2008, 41, 9130–9139.
(7) The degree of sulfonation is best represented by the equivalent weight
(14) Pich, A.; Datta, S.; Musyanovych, A.; Adler, H.-J. P.; Engelbrecht,
L. Polymer 2005, 46, 1323–1330.
(15) (a) McLoughlin, V. C. R.; Thrower, J. Tetrahedron 1969, 25, 5921–
5940. (b) Chen, G. J.; Tamborski, C. J. Fluorine Chem. 1989, 43, 207–228.
(16) Yamaguchi, K.; Watanabe, S.; Nakahama, S. Makromol. Chem.
1989, 190, 1195–1205.
-
(EW), which represents the mass of the polymer per SO3 unit. The
equivalent weights for Nafion 112 and the (PVDF-g-SPS) graft copolymer
reported by Chung are 1100 and 454, respectively.
(8) Gan, M. G.; Chow, P. Y.; Liu, Z.; Hana, M.; Queka, C. H. Chem.
Commun. 2005, 35, 4459–4461.
(17) Reppy, M. A.; Gray, D. H.; Pindzola, B. A.; Smithers, J. L.; Gin,
D. L. J. Am. Chem. Soc. 2001, 123, 363–371.
(9) Gin, D. L.; Lu, X.; Nemade, P. R.; Pecinovsky, C. S.; Xu, Y.; Zhou,
M. Adv. Funct. Mater. 2006, 16, 865–878.
(10) Herz, J.; Reiss-Husson, F.; Rempp, P.; Luzzati, V. J. Polym. Sci. Part
C Polym. Symp. Ed. 1963, 4, 1275–1290.
(18) Soula, O.; Guyot, A. Langmuir 1999, 15, 7956–7962.
(19) Souzy, R.; Ameduri, B. Prog. Polym. Sci. 2005, 30, 644–687.
(20) Kirsh, Y. E.; Smirnov, S. A.; Popkov, Y. M.; Timashev, S. F. Russ.
Chem. Rev. 1990, 59, 560–574.
J. Org. Chem. Vol. 75, No. 20, 2010 6815

Wadekar et al.
JOCArticle
with a wide variety of comonomers under anionic, cationic,
coordination, and controlled radical conditions.21 It is also
expected that the partial fluorination of 1, by the attachment
of a fluorocarbon tail, increases the stability of Poly-1.22 To
the best of our knowledge, compound 1 is the only fluori-
nated surfactant that contains a sulfonate and a readily
polymerizable styrene that can be synthesized using standard
synthetic tools. Special equipment for handling volatile and
highly toxic halogenated compounds is not required. We
wish to employ 1 for the construction of nanostructured
membranes by polymerization of 1 in a microemulsion or in
an LLC phase. Compound 2 is expected to be very similar to
1 regarding its physical properties but without the ability to
be polymerized. Therefore, 2 is regarded as a nonpolymer-
izable model surfactant that is well suited for investigating
the morphology formation in microemulsions and LLC
phases.
SCHEME 1. Synthesis of Compounds 5a-c
SCHEME 2. Synthesis of Non-polymerizable
Fluorosurfactant 2
Results and Discussion
The synthesis of compounds 1 and 2 is based upon sodium
5-iodooctafluoro-3-oxapentanesulfonate (6), which contains
an iodo and a sulfonate group at opposite ends of a fluoro-
carbon chain.13 For the substitution of the iodo moiety by an
aromatic substituent, a copper-mediated cross-coupling is
employed. In this reaction, an iodine to copper exchange
takes place that forms a copper(I) fluoroalkyl adduct, which
subsequently reacts with an aromatic iodide or bromide.15a,b
This reaction has been described using a wide range of
aromatic iodides and bromides and appears to be tolerant
to the presence of functional groups like esters, carboxylic
acids, nitro groups, and amines.15a
We decided to investigate the copper-mediated cross-
coupling reactions in the synthesis of model compounds 5,
which did not contain sulfonate groups (Scheme 1). In this
manner, potential interference of the sulfonate with the
reaction and tedious surfactant purifications are avoided.
The reaction of 1-iodofluorohexane (3) with bromo-4-ethyl-
benzene (4a) yielded 1-ethyl-4-(perfluorohexyl)benzene (5a)
in a moderate 60% yield. A direct coupling of 4-bromostyr-
ene (4b) with 3 was attempted, but instead of the desired
1-(perfluorohexyl)-4-vinylbenzene (5b), a highly viscous poly-
meric mass was obtained. Clearly, the reaction conditions
were too harsh for the styryl moiety, and premature styrene
polymerization had taken place. In order to obtain 5b, we
decided to couple 4-bromobenzaldehyde (4c) to the fluoro-
carbon iodide 3, and 4-(perfluorohexyl)benzaldehyde (5c)23
was obtained in 88% yield. Compound 5c appeared to be
very sensitive to oxidation by air (even in the solid state) and
had to be handled under an argon atmosphere to avoid
oxidation to carboxylic acid. A subsequent Wittig reaction
with methyl triphenylphosphonium bromide in DCM,24a
using 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) as a base,
yielded the desired styrene compound 5b in 77% yield; see
Scheme 1. The introduction of a styryl functionality by
various methods, like dehydration of an alcohol,25 dehydro-
halogenation,26 or a Wittig reaction with an aldehyde,24a,b
has been reported. We have chosen for the attachment an
aldehyde followed by a Wittig olefination, because this
method is efficient, is performed under mild conditions,
and can be employed in polar solvents in which fluorosulfo-
nate compounds are soluble.24a
The synthesis of the model surfactant 2 is outlined in
Scheme 2. After the cross-coupling of sodium 5-iodoocta-
fluoro-3-oxapentanesulfonate (6) and 4a, a mixture of 20,
which may contain Cuþ or Naþ as the counterion, and the
hydrogen-substituted byproduct 7 was obtained.27 The
1
crude yield of this reaction, based on H NMR spectra,
was typically 70%. For standard reactions, in which the ratio
between the reactants 6 and 4a was 1:1, the 20:7 ratio was
typically 8:1. The substitution of iodide by hydrogen in this
cross-coupling has been reported and is ascribed to reac-
tion of the Cu(I)-fluorocarbon adduct with water during
workup.15a
Isolation of 2 from this reaction mixture turned out to be
difficult and very sensitive to the exact workup procedure. It
turned out that only if the starting material contains DMSO
can 2 and 7 be separated by column chromatography. Based
on these findings, we performed the synthesis of 2 using
(21) (a) Higginson, W. C. E.; Wooding, N. S. J. Chem. Soc. 1952, 760–
774. (b) Overberger, C. G.; Arnold, L. H.; Tanner, D.; Taylor, J. J.; Alfrey,
T., Jr. J. Am. Chem. Soc. 1952, 74, 4848–53. (c) Gao, H.; Pei, L.; Song, K.;
Wu, Q. Eur. Polym. J. 2007, 43, 908–914. (d) Chang, C.-C.; Studer, A.
Macromolecules 2006, 39, 4062–4068.
(22) Tayouo, R.; David, G.; Ameduri, B. Eur. Polym. J. 2010, 46, 1111–
1118.
(23) The synthesis of 5c via ethyl 4-bromobenzoate has been reported
previously, but a reaction step using LiAlH4 in this synthetic scheme, which
may reduce the sulfonate group, made this procedure unsuitable for our
work; see: Reichardt, C.; Eschner, M.; Schafer, G. J. Phys. Org. Chem. 2001,
14 (11), 737–751.
(25) Overberger, C. G.; Saunders, J. H. Org. Synth. 1948, 28, 31–33.
(26) Halpern, M.; Zahalka, H. A.; Sasson, Y.; Rabinovitz, M. J. Org.
Chem. 1985, 50, 5088–5092.
(24) (a) Okuma, K.; Sakai, O.; Shioji, K. Bull. Chem. Soc. Jpn. 2003, 76,
1675–1676. (b) Marsh, G. P.; Parsons, P. J.; McCarthy, C.; Corniquet, X. G.
Org. Lett. 2007, 9, 2613–2616.
(27) It is noteworthy that we did not find a similar hydrogenated product
while preparing 5a. Presumably, this product has been formed, but since it is
highly volatile, it has not been isolated.
6816 J. Org. Chem. Vol. 75, No. 20, 2010

Wadekar et al.
JOCArticle
SCHEME 3. Synthesis of the Polymerizable Fluorosurfactant 1
2 equiv of 4a, which reduces the contents of 7 in the crude
product to a mere 4%. For the removal of copper salts from
the reaction, an ethyl acetate-brine extraction was used.
After addition of DMSO to the mixture of 20 and 7 a single
column chromatography separation followed by cation ex-
change using an ion-exchange column was sufficient to
obtain 2 in pure form.28 The isolated yield after purification
was 60%.29
surfactants 1 and 2, however, were markedly lower than
those of 5a and 5b, but this was due to losses during
purification of the surfactants. For the surfactant purifica-
tion, ion-exchange and column chromatography were re-
quired. In particular, the isolation of 2 from mixtures with
the side product 7 was very challenging. Attempts to avoid
the formation of sulfonate containing surfactants by revers-
ing the sequence of reaction steps, and making sulfonyl
fluorides instead, were not successful. We have found that
sulfonyl fluorides are not compatible with the copper-
mediated cross-coupling conditions.
The physical properties of 1 and 2, with regard to their
phase behavior in aqueous systems and the polymerization
of 1 are currently under investigation. In view of the en-
visaged application of 1, however, it is appropriate to show
the most important results here. Compound 1 is readily
polymerized under standard free-radical polymerization
conditions31 in polar solvents like water and DMF, and 1H
and 19F NMR spectra of the resulting polymer are shown in
Figures S36 and S37 (Supporting Information). The phase
behavior of 1 and 2 is very similar, and various lyotropic
phases have been identified by cross-polarized optical micro-
scopy pictures of binary mixtures of 2 in water, as illustrated
in Figure S38 (Supporting Information). These findings
imply that compound 1 is capable of forming various nano-
structured phases in water and can be readily polymerized in
this solvent. On the basis of these observations, it is reason-
able to assume that the polymerizable fluorosurfactant 1 is
well suited for the construction of nanostructured proton-
conducting membranes.
The synthesis of the polymerizable surfactant 1 is depicted
in Scheme 3. The cross-coupling of 4c and 6 produced com-
pound 8. After ethyl acetate and brine extraction, a proce-
dure by which ionic contaminants and DMSO are removed
and eventual Cuþ ions are exchanged by Naþ ions, the yield
of 8 was 85%. Despite precautions to exclude air, traces of
the carboxylic acid were observed in the NMR spectra of 8.30
Remarkably, the hydrogenated product 7, a compound that
1
would be easily identified by H and 19F NMR, was not
detected. The subsequent Wittig reaction, performed with
2-3 equiv of Wittig reagent, yielded the styrene 9, in which
the triphenylmethylphosphonium ion acts as the counterion.
After ion exchange using a sodium-loaded Dowex column,
followed by chromatography on a silica column, 1 was
obtained. Recrystallization from toluene gave crystalline 1
in 50% yield.
For the synthesis of compounds 1, 2, 5a, and 5b, the cross-
coupling reactions between an aromatic moiety ArBr 4 and
the fluorocarbon iodides 3 and 6 were employed. We found
that this reaction performed much better with 4-bromoben-
zaldehyde than with 1-bromo-4-ethylbenzene. With 4-bro-
mobenzaldehyde both the yields were higher and hydro-
genated side-products were absent. These observations are
best explained by assuming that the reactivity of the aromatic
bromide in this reaction is enhanced by electron-withdraw-
ing substituents at the aromatic core. This conclusion is
consistent with reported yields for this reaction.15a For the
synthesis of fluoroalkyl-substituted styrenes a direct coupl-
ing of 4-bromostyrene was not possible. We have found that
the coupling of benzaldehyde followed by a Wittig reaction
was an efficient method for the synthesis of these com-
pounds. Both the copper-mediated cross coupling and the
Wittig reactions were not affected by the presence of a sul-
fonate moiety. The isolated yields of the sulfonate containing
Conclusions
Compounds 1 and 2 were synthesized using the readily
accessible sodium 5-iodooctafluoro-3-oxapentanesulfonate
6 as a starting compound. Coupling of aromatic molecules
ArBr (4) to 6 was achieved by a copper-mediated cross-
coupling. The presence of a sulfonate moiety did not inter-
fere with this reaction. The reaction of 6 with 1-bromo-4-
ethylbenzene 4a was slow and produced the hydrogenated
side product 7. The formation of 7 during the synthesis of 2
resulted in a challenging purification procedure. Attachment
of a styryl functionality to 6 was achieved in two steps. The
attachment of 4-fluorobenzaldehyde to 8 was a fast, efficient,
and clean reaction. The conversion of the aldehyde 8 to the
polymerizable surfactant by a Wittig reaction proceeded
(28) After two columns, traces of 7 could be detected by 1H and 19F
NMR. The percentage of 7 in the product was below 1%.
(29) Compound 2 can be recrystallized from toluene, but this procedure
does not improve the purity of 2.
(30) The coupled product 8 is very prone to oxidation. Fortunately, the
carboxylic acid that is formed is removed in the ethyl acetate and brine
extraction.
(31) For an example, see: Jager, W. F.; Sarker, A. M.; Neckers, D. C.
Macromolecules 1999, 32, 8791–8799.
J. Org. Chem. Vol. 75, No. 20, 2010 6817

Wadekar et al.
JOCArticle
smoothly. Extensive purification of this surfactant, including
column chromatography and ion-exchange, however, seri-
ously reduced the isolated yields. Along with 1 and 2, the
monomer 5b was synthesized by the cross-coupling of 4-bro-
mobenzaldehyde and 1-iodofluorohexane followed by a
Wittig reaction with methyltriphenylphosphonium bromide.
In our current research, the phase behavior of the surfac-
tants 1 and 2 in water are being investigated along with the
polymerization of 1 and 5b.
Na2SO4. The solvent was removed using vacuum at 40 °C, and
purification was done by column chromatography (pentane/
DCM = 0.9:0.1). Compound 5b was obtained as a colorless
1
liquid (2.05 g, 77%). H NMR, 19F NMR, and IR data were
consistent with those reported in ref 24a.
Sodium 2-(2-(4-Ethylphenyl)-1,1,2,2-tetrafluoroethoxy)-
1,1,2,2-tetrafluoroethanesulfonate (2). Copper powder (19.9 mmol,
1.25 g) was heated above 150 °C under flow of argon in a 25 mL
round-bottom flask. Sodium 5-iodooctafluoro-3-oxapentane-
sulfonate (4.31 mmol, 2.00 g) and 4-bromoethylbenzene (4.6
mmol, 0.85 g) were stirred in DMSO (3.6 mL) for 2 days at
110 °C under an argon atmosphere. The reaction mixture was
cooled, and 2-propanol (10 mL) was added. The solution was
filtered, and 2-propanol was removed by vacuum. The remain-
ing mixture was extracted using ethyl acetate (3 ꢀ 20 mL), and
all the ethyl acetate layers were washed using saturated brine to
remove the DMSO and inorganic impurities. The ethyl acetate
layer was dried using anhydrous Na2SO4. Around 50% DMSO
(w/w) was added to the dry powder obtained after evaporation
of ethyl acetate and column chromatography was performed
(DCM/2-propanol = 0.85:0.15) to give white 2 (1.09 g, 60%):
mp 241-245 °C; 1H NMR (300 MHz, DMSO-d6) 1.26 (t, J =
7.8 Hz, 3H), 2.71 (q, J = 7.5 Hz, 2H), 7.36 (d, J = 6.9 Hz, 2H),
7.56 (d, J = 6.9 Hz, 2H); 19F NMR (282.3 MHz, DMSO-d6),
-82.09 (t, J = 14.1 Hz, 2F), -87.10 (t, J = 13.0 Hz, 2F),
-113.37 (s, 2F), -117.53 (s, 2F); 13C NMR (75.4 MHz, DMSO-
d6) 148.83, 128.73, 127.15, 125.64 (aromatic-C), 121.92-108.01
(CF2), 28.43, 15.54 (ethyl-C); IR ν cm-1 3044, 2968, 2929, 2874,
1517, 1421, 1254, 1027, 1155, 1093, 1078, 1008; HRMS calcd for
C12H9F8Na2O4Sþ [M þ Na]þ 446.9889, found 446.9889. Anal.
Calcd for C12H9F8NaO4S: C, 33.97; H, 2.14; F, 35.83; Na, 5.42;
S, 7.56. Found: C, 33.92; H, 2.27; F, 35.20; Na, 5.38; S, 7.45.
Sodium 1,1,2,2-Tetrafluoro-2-(1,1,2,2-tetrafluoro-2-(4-formyl-
phenyl)ethoxy)ethanesulfonate (8). Copper powder (62.0 mmol,
3.84 g) was strongly heated above 150 °C in argon atmosphere
and cooled to room temperature. Added to it were compounds 6
(12.9 mmol, 6.00 g) and 4c (13.0 mmol, 2.40 g). DMSO (11 mL)
was added, and the reaction mixture was stirred for 22 h at
110 °C in inert conditions. 2-Propanol (20 mL) was added to the
reaction mixture, and it was filtered. The filtrate was evaporated
to remove 2-propanol and extracted with ethyl acetate (3 ꢀ
20 mL), and ethyl acetate layers were washed with brine. The
ethyl acetate layer was dried over Na2SO4 under argon. The
solvent was removed to give white solid 8 (4.77 g, 87%) and
stored under an inert atmosphere: 1H NMR (300 MHz, DMSO-
d6) 7.95 (d, J = 8.1 Hz, 2H), 8.09 (d, J = 7.8 Hz, 2H), 10.12 (s,
1H); 19F NMR (282.3 MHz, DMSO-d6) -81.86 (t, J = 12.98
Hz, 2F), -86.49 (broad s, 2F), -113.21 (s, 2F), -117.51 (s, 2F);
13C NMR (75.4 MHz, DMSO-d6) 193.47, 139.38, 133.32,
130.32, 128.40 (aromatic-C and -CHO), 122.09-108.18, (CF2);
IR ν cm-1 3063, 2865, 1710, 1587, 1513, 1400, 1287, 1178, 1024,
1005; HRMS calcd for C11H5F8Na2O5Sþ [M þ Na]þ 446.9525,
found 446.9521.
Experimental Section
Instrumentation. TLC analysis was performed on silica gel,
and 1H, 19F, and 13C NMR spectra were measured in CDCl3 or
CD3OD at 300 or 400 MHz and 282.3 and 75.4 MHz, respec-
tively. Chemical shifts are given in ppm (δ). Tetramethylsilane
1
(TMS) was used as internal standard for H and 13C NMR,
whereas for 19F NMR we used trifluoroacetic acid as the
standard. LC-MS data were collected with a liquid chromato-
graph mass spectrometer, with a diode-array detector. The col-
umn used was the Xbridge Shield RP 18.5 mm (4.6 ꢀ 150 mm)
on MeOH. Fast atom bombardment (FAB) mass spectrometry
was carried out with a four sector mass spectrometer, coupled
to a system program. Samples were loaded in a matrix solution
(3-nitrobenzyl alcohol) onto a stainless steel probe and bom-
barded with Xenon atoms with energy of 3 keV. During the
high-resolution FAB-MS measurements a resolving power of
10000 (10% valley definition) was used. IR spectra were from
KBr pellets with an FTIR spectrophotometer.
1-Ethyl-4-(perfluorohexyl)benzene (5a). A 25 mL round-bot-
tom flask was charged with copper powder (33 mmol, 3.07 g)
and heated above 150 °C under an argon flow using a hot gun.
The flask was cooled to ambient temperature, and perfluoro-
hexyl iodide (6.73 mmol, 3.00 g), 4-bromoethylbenzene (6.74
mmol, 1.27 g), and DMSO (6 mL) were added to the flask and
heated to 110 °C for 2 days using an oil bath. The reaction
mixture was cooled, 20 mL of water was added, and then the
mixture was filtered. The filtrate was further extracted with (3 ꢀ
20 mL) DCM. All of the DCM layers were collected, washed
with 20 mL of water two times, and dried over anhydrous
Na2SO4. Column chromatography with pentane gave 5a as a
colorless liquid (1.75 g, 61%): 1H NMR (400 MHz, CDCl3), 7.49
(d, J = 8.2 Hz, 2H), 7.32 (d, J = 8.16 Hz, 2H), 2.72 (q, J = 7.56
Hz, 2H), 1.27 (t, J = 7.6 Hz, H); 19F NMR (282.3 MHz, CDCl3),
-81.12 (t, J =9.03 Hz, 3F), -110.64 (t, J = 13.55 Hz, 2F),
-121.76 (s, 2F), -122.173 (s, 2F), -123.08 (s, 2F), -126.41 (s,
2F); 13C NMR (75.4 MHz, CDCl3), 148.73, 128.28, 127.02,
126.5 (aromatic-C), 120.00-105.00 (CF2 and CF3), 28.90, 15.25
(ethyl-C); IR ν cm-1 3045, 2974, 2941, 2881, 1517, 1464, 1420,
1363, 1239, 1091, 1040, 1015. Anal. Calcd for C14H9F13: C,
39.64; H, 2.14; F, 58.22. Found: C, 39.44; H, 2.13; F, 58.1.
4-(Perfluorohexyl)benzaldehyde (5c). Reaction conditions and
purification procedures were similar to those for 5a, except the
purification was done under an argon atmosphere. The reaction
was performed with copper powder (50 mmol, 3.15 g), perfluoro-
hexyl iodide (11.2 mmol, 5.00 g), and 4-bromobenzaldehyde
(10.0 mmol, 2.07 g) in DMSO (9 mL). The reaction was per-
formed for 22 h. Column chromatography (pentane/DCM =
0.7:0.3) gave a white solid 5c (4.5 g, 94%). 1H NMR, 19F NMR,
and IR data were consistent with those reported in ref 23.
1-(Perfluorohexyl)-4-vinylbenzene (5b). Methyltriphenylpho-
sphonium bromide (18.7 mmol, 6.69 g) and DBU (19 mmol,
2.89 g) were dissolved in freshly distilled DCM (65 mL). The
solution was refluxed for 30 min, and 5c (6.25 mmol, 2.65 g)
dissolved in 15 mL of freshly distilled DCM was added to the
refluxing solution at once. The mixture was refluxed for 2 h and
subsequently washed with water (3 ꢀ 20 mL) and dried over
Sodium 1,1,2,2-Tetrafluoro-2-(1,1,2,2-tetrafluoro-2-(4-vinyl-
phenyl)ethoxy)ethanesulfonate (1). Methyltriphenylphospho-
nium bromide (16.0 mmol, 5.71 g) and DBU (16.5 mmol,
2.51 g) were dissolved in 75 mL of freshly distilled acetonitrile.
The solution was refluxed for 30 min. Compound 8 (8.0 mmol,
3.85 g), dissolved in acetonitrile (20 mL), was added to the
refluxing solution at once, and the reaction mixture was refluxed
for 2 h. After the mixture was cooled to room temperature,
10 mL of water was added and the mixture filtered. The solvent
was removed using vacuum at 40 °C. The crude material was
extracted with ethyl acetate (3 ꢀ 20 mL) and washed with 0.001
N HCl (3 ꢀ 20 mL). All of the ethyl acetate layers were collected
and dried over Na2SO4. The solvent was removed, and the
purification was done by column chromatography (DCM/
2-propanol=0.8:0.2) to give a waxy white material 9. The yield
6818 J. Org. Chem. Vol. 75, No. 20, 2010

Wadekar et al.
JOCArticle
was 3.9 g (70%) with ∼75% purity:32 1H NMR (400 MHz,
CDCl3) 2.92 (d, J = 13.6 Hz, 3H), 5.33 (d, J = 11.2 Hz, 1H),
5.81 (d, J = 17.6 Hz, 1H), 6.70 (double d, J = 11.1 and 8.0 Hz,
1H), 7.43-7.68 (m, 19H); 19F NMR (282.3 MHz, CDCl3)
-82.62 (d, J = 12.70 Hz, 2F), -87.62 (s, 2F), -114.20 (s, 2F),
-117.96 (s, 2F); LRMS calcd for C12H7F8O4ShPþPh3CH3 neg
ion, C12H7F8O4Sh [Mh] 398.99, found 398.85; pos ion, PþPh3CH3
[M0]þ 277.11, found 276.95.
(75.4 MHz, DMSO-d6) 141.49, 136.13, 127.71, 127.48, 127.13,
118.02 (Ar-C and vinyl-C), 127.00-113.00 (CF2-C). IR ν cm-1
3087, 3007, 1651, 1571, 1515, 1407, 1293, 1255, 1220, 1170,
1100, 1080, 1008, 915, 841; HRMS calcd for C12H7F8Na2O4Sþ
[M þ Na]þ 444.9727, found 444.9721. Anal. Calcd for
C12H7F8NaO4S: C, 34.14; H, 1.67; F, 36.00; Na, 5.44; S, 7.59.
Found: C, 33.87; H, 1.67; F, 35.00; Na, 5.44; S, 7.59.
The solution of 9 in acetone/water mixture (1:1) was passed
through a Dowex cation exchange column loaded with Naþ
ions. Column chromatography (DCM/2-propanol = 0.8:0.2)
followed by recrystallization from toluene were performed to
Acknowledgment. The Ministry of Economic Affairs of
The Netherlands is gratefully acknowledged for funding this
research project via the EOS LT framework (EOSLT 02025).
We thank Kristina Djanashvili for performing 19F NMR
and 13C NMR measurements.
1
give 1, a white crystalline solid 2 g (50%): mp 250 °C dec; H
NMR (300 MHz, MeOD) 5.3 (d, J = 11.1 Hz, 1H), 5.8 (d, J =
16.8 Hz, 1H), 6.67 (double d, J = 11.1 and 6.6 Hz, 1H), 7.42 (d,
J = 8.9 Hz, 2H), 7.51 (d, J = 8.4 Hz, 2H); 19F NMR (282.3
MHz, MeOD) -83.12 (t, J = 13.54 Hz, 2F), -87.76 (t, J =
12.98 Hz, 2F), -114.80 (s, 2F), -118.20 (s, 2F); 13C NMR
Supporting Information Available: 1H NMR, 19F NMR,
and IR spectra of all synthesized compounds; 13C NMR spectra
for compounds 1, 2, 5a, and 8; HRMS for compounds 1, 2,
and 8; LRMS for 9; elemental analysis of 5a, 1, and 2. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(32) The remaining impurities were triphenylmethylphosphonium salts
that were easily removed by the subsequent ion exchange.
J. Org. Chem. Vol. 75, No. 20, 2010 6819