Copolymerization of ethylene, tetrafluoroethylene, and an olefin-containing fluorosulfonyl fluoride: Synthesis of high-proton-conductive membranes for fuel-cell applications

Article abstract of DOI:10.1002/anie.200461104

Excellent thermal stability and mechanical properties are exhibited by high-proton-conductive membranes synthesized by a facile copolymerization of tetrafluoro-ethylene, ethylene, and CH₂=CHCF₂CF ₂O-CF₂CF₂SO₂F (see scheme and AFM image). These membranes have been successfully used in a direct MeOH fuel cell.

Full text of DOI:10.1002/anie.200461104

Communications
Membranes
Copolymerization of Ethylene,
Tetrafluoroethylene, and an Olefin-Containing
Fluorosulfonyl Fluoride: Synthesis of High-
Proton-Conductive Membranes for Fuel-Cell
Applications**
Zhen-Yu Yang* and Raj G. Rajendran
Fuel cells, which are devices that convert the chemical energy
stored in the fuel directly into electricity, are one of the most
important technologies in the 21st century. They not only
provide pollution-free clean energy, but also give high-quality
and more-reliable power.[1] Of all the types known today, the
technology based on proton-exchange membrane fuel cells
(PEMFC) is the most attractive for various applications,
especially transportation and for stationary and portable
devices.[2] The most important component of the PEMFC is
the membrane electrode assembly (MEA), at the heart of
which is the proton-conductive membrane. The key functions
of the membrane are to transport protons from the anode to
the cathode of the cell and to separate the fuel and the
oxidant.
Significant resources have been devoted worldwide to the
development of high-performance and reliable membranes.
[*] Dr. Z.-Y. Yang, Dr. R. G. Rajendran
DuPont Central Research and Development
Experimental Station, Wilmington, DE 19880-0328 (USA)
Fax: (+1)302-695-9799
E-mail: Zhenyu.yang-1@usa.dupont.com
[**] Publication no.: 8560. The authors thank Mr. D. Irino for technical
assistance, the Pressure Lab of DuPont for running all polymer-
ization reactions, Prof. S. Zhu at the Shanghai Institute of Organic
Chemistry, China, for providing compound 1, and Dr. A. E. Feiring
and Dr. M. Hofmann for their comments on this manuscript.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
564
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200461104
Angew. Chem. Int. Ed. 2005, 44, 564 –567

Angewandte
Chemie
Academic researchers are working on the development of
new membrane materials, while government and industry are
driving the commercialization of fuel-cell technology.[3]
Although many proton-exchange membranes have been
developed, the perfluorinated-ionomer
Although copolymerization of 2 and TFE was unsuccess-
ful, and only a small amount of 2 was incorporated into the
polymers, we found that copolymerization of TFE, ethylene,
and 2 was more effective. Copolymerization in the presence of
membrane nafion is still the best known
today.[4] Nafion, which was invented and
introduced by DuPont in the early 1960s,
revolutionized fuel-cell technology, and is
synthesized by copolymerization of tetra-
fluoroethylene (TFE) and a perfluorovinyl
ether (PSEPVE) containing a sulfonyl fluo-
ride.[5] Although nafion is the most widely
used membrane in fuel cells and provides
the best overall performance at temper-
an organic peroxide, such as lupersol 11, as an initiator in
100 mL of 1,1,2-trichloro-1,2,2-trifluoroethane (CFC113) in a
240-mL shaker tube at 608C for seven hours provided the
terpolymers shown in Table 1.
All polymerizations proceeded smoothly to give polymers
in good yields. The remarkably high conversion of termon-
omer 2 makes this polymerization system particularly attrac-
tive for practical applications. In most cases, conversions of 2
atures below 1008C, it is relatively expensive. The high cost of
nafion is primarily attributed to the difficult polymerization
and the need for an expensive monomer
(PSEPVE). In addition, perfluorinated
nafion is difficult to further modify to
improve its desired properties. We report
here a new, potentially lower cost and high-
performance proton-conductive membrane
for fuel-cell applications.
Table 1: Coolymerization of TFE, ethylene (E), and 2.
Entry Monomers
[g, mol]
(TFE+E)/2 DP[a]
[mol]
Yield Analysis[b]
EW
Conductivity[c]
[mScmÀ1
[psi/7h] [g]
(TFE + E)/2
]
1
2
3
4
5
6
7
TFE (22, 0.22) 15.2
E (7, 0.25)
2 (10, 0.031)
TFE (22, 0.22) 11.8
E (7, 0.25)
2 (13, 0.040)
TFE (22, 0.22) 10.9
E (7, 0.25)
2 (14, 0.043)
TFE (22, 0.22) 10.2
E (7, 0.25)
2 (15, 0.046)
316
247
282
222
244
186
152
35.7
33.5
35.8
30.2
35.5
35.6
30.6
(E)7.7(TFE)6.5(2) 1192 40
14.2/1
Although perfluorinated polymers pro-
vide the best combination of both thermal
and chemical properties, selective introduc-
tion of H atoms into polymers does not
affect the stability dramatically.[6] In fact, a
copolymer of ethylene and tetrafluoroethy-
lene (ETFE), tefzelꢀ, made by DuPont, still
has excellent thermal and chemical stability
and has better mechanical properties than
perfluorinated polymers.[7] In addition, the
ETFE portion can be cross-linked readily to
further improve its properties, whereas
perfluorinated polymers are extremely dif-
ficult to modify.[8] More importantly, TFE
and ethylene copolymerize mostly alter-
nately, which may help in the incorporation
of olefins other than expensive perfluoro-
vinyl ethers.[9] If this is the case, we might be
able to design a low-cost, fluorinated mono-
mer containing a sulfonyl fluoride group.
Monomer 2 was readily synthesized from
(E)7.3(TFE)5.5(2) 1080 75.7
12.8/1
(E)6.4(TFE)5.5(2) 1055 77.8
11.9/1
(E)6.2(TFE)3.7(2) 870
9.9/1
94.6
99.6
102
122
TFE (22, 0.22)
E (7, 0.25)
2 (16, 0.049)
TFE (22, 0.22)
E (7, 0.25)
2 (20, 0.061)
TFE (22, 0.22)
E (7, 0.25)
9.6
7.7
6.1
(E)5.7(TFE)4.0(2) 886
9.7/1
(E)5.4(TFE)3.3(2) 807
8.7/1
(E)3.6(TFE)3.0(2) 727
6.6/1
2 (25, 0.077)
[a] Pressure change between beginning and end of polymerization in 7 h. [b] Based on elemental
analysis. [c] At room temperature.
ethylene and 1, which is commercially available, or by
were up to 90%. The polymerization pressure drops with
polymerization time, while the rate of polymerization
decreases with an increase of monomer 2 in the
presence of given amounts of TFE and ethylene.
Incorporation of termonomer 2 increases in the
polymers with an increase of the pre-charged
amount of 2. It is interesting to note that the ratio
[10]
À
reaction of TFE, FSO2CF2COF, and I Cl in one step.
Angew. Chem. Int. Ed. 2005, 44, 564 –567
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
565

Communications
of TFE plus ethylene to 2, (TFE + E)/2, in the polymers is
similar to that of the precharged monomers.
Terpolymers 3 show excellent thermal stability. As
observed in the TGA studies, the polymers start to decom-
pose at about 3708C and a 10% weight loss occurs up to
4258C. DSC analysis indicated that the polymers have no
melting points, except for the polymer of entry 1, which has a
broad melting point at 2108C, although X-ray diffraction
studies showed that the polymers have some degree of
crystallinity. The degree of crystallinity decreases with
increasing amounts of incorporated 2 in the polymers. These
polymers are insoluble in common organic solvents at room
temperature, so it is difficult to obtain their molecular
weights.
Figure 1. AFM images: a) Tapping-force height: 50 nm scale; b) phase:
750 nm scale.
All terpolymers 3 were readily pressed to transparent,
strong and tough films at 220–2508C, thus indicating their
high molecular weight. The films were hydrolyzed with 10%
KOH in MeOH/H2O/DMSO (4:5:1, v/v/v) at 608C to give the
potassium salt of the ionomer membrane. The films became
light yellow in color when hydrolysis was carried out above
1008C. This is probably a consequence of the attack of the
hydroxy anion on the backbone of the polymers. The
potassium ionomers were immersed in 10% HNO3 at 608C
for two hours twice, and then washed and boiled with
deionized water several times to obtain membranes 4.
The proton conductivity of 4 is, in general, similar to, or
higher than that of nafion, and depends on the amount of
incorporated termonomer 1. It was found that the conductiv-
ity follows a linear relationship with the ratio of units of
(TFE + ethylene)/2, or with equivalent weight (EW), as
indicated in Table 1. Interestingly, the conductivity of the
new membranes is higher than that of nafion samples
containing the same mol percent of PSEPVE. For example,
membranes containing less than 10 mol% of 2 (entries 4 or 5
of Table 1) have the same conductivity as nafion 117, which
contains 15 mol% PSEPVE. Membranes containing
15 mol% of 1 (entry 7) have a conductivity of 122 mScmÀ1,
which is 22% higher than that of nafion 117.
and the cathode with oxygen (3 L minÀ1). The cell was heated
to 808C, and the fuel-cell performance data were recorded.
The fuel-cell performance of the membrane 4 is compared
with the performance of commercial nafion 115 under the
same conditions in Figure 2. The new 5-mil (127 mm) mem-
Figure 2. MeOH/O2 fuel-cell performance with membrane 4 (entry 5)
at 808C by monitoring cell voltage and power density versus current
density. The anode was fed with 1m methanol at a rate of 25 mLminÀ1
and the cathode was fed with oxygen at a rate of 3 LminÀ1
.
Although the exact reason for the high conductivity of
membrane 4 is not fully understood, the more-polar and
flexible backbones of the polymer and lower equivalent
weight may, in part, be responsible for the proton conduction,
as a result of higher water uptake.[11] In fact, the membrane in
entry 5 takes-up 49% (by weight) water compared with 39%
water for nafion under the same conditions. A clear phase-
separation in 4 (entry 5) between the hydrophilic sulfonic acid
and the hydrophobic hydrofluorocarbon backbone was
observed by atomic-force microscopy (AFM; Figure 1),
which is very similar to the images generated from nafion.[12]
The experimental membrane 4 from entry 5 in Table 1
without any optimization was tested in a direct MeOH fuel-
cell. The membrane and electrode assembly required for the
brane exhibits a similar power output to a 5-mil commercial
nafion membrane, although it has 9% higher methanol
crossover (10.9 ꢂ 10À4 gminÀ1 cmÀ2 for membrane 4 and
9.9 ꢂ 10À4 gminÀ1 cmÀ2 for 5-mil nafion). This difference is
partly a consequence of the variation in the thickness of the
new membrane. We anticipate that higher EW samples may
reduce the methanol crossover significantly. Of course, the
high EW also affects the fuel-cell performance, although
cross-linking may reduce MeOH crossover and further
improve the fuel-cell performance.
In conclusion, we have discovered an effective copoly-
merization of tetrafluoroethylene and ethylene with 2 to give
melt-processable terpolymers, which can be readily hydro-
lyzed and acidified to give membranes. The new membranes
exhibit excellent conductivity and stability. The fuel-cell
performance is comparable to, or slightly better than, that
of nafion, although it is still to be optimized. Their polymeric
lithium salts also exhibit excellent lithium-ion conductivity
and are attractive candidates for lithium-battery applications.
The new membranes could be produced more cheaply
fuel-cell evaluation was fabricated using
a standard
method.[13] The anode and cathode catalysts were Pt–Ru
and Pt black, respectively, and the loadings were kept close to
4 mgcmÀ2. Commercial carbon cloths—E-Tekꢁs ELAT and
Zoltekꢁs plain carbon cloth—were used as the cathode and
anode gas diffusion backing, respectively. The anode was fed
with a 1m methanol/water mixture at a rate of 25 mLminÀ1
,
566
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 564 –567

Angewandte
Chemie
[6] a) Fluoropolymers, Vol. XXV (Ed.: L. A. Wall), John Wiley &
because of the simple polymerization process and the low
monomer cost.
Sons, New York, 1972; b) Modern Fluoropolymers (Ed.: J.
Scheirs); John Wiley & Sons, New York, 1997.
[7] Ullmannꢀs Encyclopedia of Industrial Chemistry, Vol A11 (Eds.:
D. P. Carlson, W. Schmiegel), VCH, Weinheim, 1988, p. 395.
[8] J. S. Forsythe, D. J. T. Hill, Prog. Polym. Sci. 2000, 25, 101.
Experimental Section
General preparation of membranes: A 210-mL, stainless-steel tube
was charged with F113 (100 mL), 2 (15 g), and lupersol 11 (0.5 g), and
attached to a gas manifold. The tube was cooled in dry ice and the
contents degassed by several cycles of evacuation and repressuriza-
tion with nitrogen gas. After the final evacuation step, the tube was
pressurized with ethylene (7 g) and TFE (22 g). The tube was then
sealed and heated to 608C and held for 7 h to effect the polymeri-
zation. After completion of the polymerization, any remaining
ethylene and TFE were removed by venting, and the remaining
white solid was washed with MeOH and dried in a partial vacuum
oven at 1008C to give 30.2 g of polymer. IR(KBr): n˜ = 1464 cmÀ1
(SO2F). Elemental analysis of the polymer indicated that its
composition was (CF2CF2)3.7(CH2CH2)6.2(CH2CHCF2CF2OCF2CF2-
SO2F) on a molar basis, based on C 34.68%, H 2.64%, F 50.65%,
and S 3.72%. DSC showed that the polymer had no melting point. A
10% weight loss was noted up to 4308C under N2 by TGA. A
colorless, transparent, and tough film was pressed by placing a sample
of the polymer between the platens of a hydraulic press and heated to
2258C with a ram force 20000 lbs. This film was then immersed in a
suspension of 10% KOH in H2O/MeOH/DMSO (5:4:1 v/v/v) at 608C
for 6 h. The film was removed from the solution, washed with water
many times, and then treated with 10% HNO3 at 608C for 2 h twice. It
was then washed with deionized water until neutral, and then boiled
in deionized water for 1 h.
=
[9] TFE copolymerizes poorly with perfluoroalkenes such as CF2
CFRFSO2F or perfluoroalkylethylenes. Only small amounts of
co-monomers can be incorporated into the polymers; see:
a) C. G. Krespan, US Patent 4275255; b) A. W. Anderson, S. J.
Fritschel, H. E. Holmquist, US Patent 4522995.
[10] a) Z. Zhang, J. Lu, P. Song, C. Xu, N. Chen, Huaxue Shijie 1990,
31, 272; b) G. A. Bargigia, G. Caporiccio, M. Pianca, J. Fluorine
Chem. 1982, 19, 403.
[11] a) W. Y. Hsu, T. D. Gierke, J. Membr. Sci. 1983, 13, 307; b) C. A.
Edmondson, P. E. Stallworth, M. C. Wintersgill, J. J. Fontanella,
Y. Dai, S. G. Greenbaum, Electrochim. Acta 1998, 43, 1295.
[12] P. J. James, J. A. Elliott, T. J. McMaster, J. M. Newton, A. M. S.
Elliott, S. Hanna, M. J. Miles, J. Mater. Sci. 2000, 35, 5111;
b) R. S. McLean, M. Doyle, B. B. Sauer, Macromolecules 2000,
33, 6541; c) K. D. Kreuer, Solid State Ionics 2000, 136–137, 149.
[13] a) X. Ren, M. S. Wilson, S. Gottesfeld, J. Electrochem. Soc. 1996,
143, L12; b) X. Ren, T. E. Springer, S. Gottesfeld, J. Electro-
chem. Soc. 2000, 147, 92.
Received: June 28, 2004
Revised: September 1, 2004
Published online: December 13, 2004
Keywords: fluorine · fuel cells · membranes · polymerization
.
[1] a) M. S. Dresselhaus, I. L. Thomas, Nature 2001, 414, 337; b) G. J.
Schaeffer, R. Uyterlinde, J. Power Sources 1998, 71, 256; c) K. D.
William, W. K. D. Borthwick, J. Power Sources 2000, 86, 52.
[2] a) B. D. McNicol, D. A. J. Rand, K. R. Williams, J. Power
Sources 2001, 100, 47; b) S. Kawatsu, J. Power Sources 1998,
71, 150; c) R. Nolt, J. Power Sources 2001, 4573, 1; d) N. M.
Smammes, R. Boersma, J. Power Sources 2000, 86, 98; e) D. Chu,
R. Jiang, K. Gardner, R. Jacobs, J. Schmidt, T. Quakenbush, J. R.
Stephens, J. Power Sources 2001, 96,174.
[3] a) B. C. H. Steele, A. Heinzel, Nature 2001, 414, 345;
b) G. J. K. R. Acres, J. Power Sources 2001, 100, 60; c) M.
Conte, A. Iacobazzi, M. Ronchetti, R. R. Vellone, J. Power
Sources 2001, 100, 171; d) P. Costamagna, S. Srinivasan, J. Power
Sources 2001, 102, 242; e) H. L. Maynard, J. P. Meyers, J. Vac.
Sci. Technol. B. 2002, 20, 1287; f) X. Ren, P. Zelenay, S. Thomas,
J. Davey, S. Gottesfeld, J. Power Sources 2000, 86, 111.
[4] a) S. M. Haile, D. A. Boysen, C. R. I. Chisholm, R. Merle,
Nature 2001, 410, 910; b) G. Alberti, M. Casciola, Solid State
Ionics 2001, 145, 3; c) J. A. Kerres, J. Membr. Sci. 2001, 185, 3;
d) K. D. Kreuer, J. Membr. Sci. 2001, 185, 29; e) D. J. Jones, J.
Roziere, J. Membr. Sci. 2001, 185, 41; f) C. Heitner-Wirguin, J.
Membr. Sci. 1996, 120, 1.
[5] a) W. G. Grot, Macromol. Symp. 1994, 77–88, 161; b) A. Eisen-
berg, F. Bailey, ACS Symp. Ser. 1986, 302; c) A. Eisenberg, M.
King, Ion-Containing Polymers, Academic Press, New York,
1977; d) A. Eisenberg, H. Yeager, ACS Symp. Ser. 1982, 180;
e) D. Lloyd, ACS Symp. Ser. 1985, 269.
Angew. Chem. Int. Ed. 2005, 44, 564 –567
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
567