Synthesis and characterization of network type single ion conductors

Article abstract of DOI:10.1021/ma035690g

New single ion conductors were synthesized by grafting the allyl group-containing lithium salt, lithium bis(allylmalonato)borate (LiBAMB), onto allyl group-containing comb-branch polyacrylate or polymethacrylate ethers by means of hydrosilylation. The highest ambient temperature conductivity of 3.5 × 10^-7 S cm^-1 was obtained for a polyacrylate ether-based single ion conductor containing eight EO units in the side chain and five EO units in the cross-linking side chain, to which the anion was fixed with a salt concentration of EO/Li = 20. For polyacrylate ether-based single ion conductors, an increase of chain length in both side chains and cross-linking anion chains favors an increase of ionic conductivity. The addition of 50 wt % EC/DMC (1/1, wt/wt) increased the ionic conductivity by more than 2 orders of magnitude due to both the increase in ionic mobility from the liquid phase and the increase in the concentration of free ions from the high dielectric constant of the solvent. The preliminary Li/Li cycling profiles of dry polyacrylate- and polymethacrylate ether-based single ion conductors are encouraging as almost no concentration polarization or relaxation was observed. The observed increase in cell potential with cycling is apparently due to an increase in the interfacial impedance associated with the SEI layer, and the cell failure is accompanied by the decomposition of the ester bond of the polyacrylate backbone.

Full text of DOI:10.1021/ma035690g

Macromolecules 2004, 37, 2219-2227
2219
Synthesis and Characterization of Network Type Single Ion Conductors
Xia o-Gu a n g Su n , Cr a ig L. Reed er , a n d J oh n B. Ker r *
Lawrence Berkeley National Laboratory, MS 62-203, One Cyclotron Road, Berkeley, California 94720
Received November 10, 2003; Revised Manuscript Received J anuary 9, 2004
ABSTRACT: New single ion conductors were synthesized by grafting the allyl group-containing lithium
salt, lithium bis(allylmalonato)borate (LiBAMB), onto allyl group-containing comb-branch polyacrylate
or polymethacrylate ethers by means of hydrosilylation. The highest ambient temperature conductivity
of 3.5 × 10-7 S cm-1 was obtained for a polyacrylate ether-based single ion conductor containing eight
EO units in the side chain and five EO units in the cross-linking side chain, to which the anion was fixed
with a salt concentration of EO/Li ) 20. For polyacrylate ether-based single ion conductors, an increase
of chain length in both side chains and cross-linking anion chains favors an increase of ionic conductivity.
The addition of 50 wt % EC/DMC (1/1, wt/wt) increased the ionic conductivity by more than 2 orders of
magnitude due to both the increase in ionic mobility from the liquid phase and the increase in the
concentration of free ions from the high dielectric constant of the solvent. The preliminary Li/Li cycling
profiles of dry polyacrylate- and polymethacrylate ether-based single ion conductors are encouraging as
almost no concentration polarization or relaxation was observed. The observed increase in cell potential
with cycling is apparently due to an increase in the interfacial impedance associated with the SEI layer,
and the cell failure is accompanied by the decomposition of the ester bond of the polyacrylate backbone.
1. In tr od u ction
In this paper we report the synthesis of a new lithium
salt, lithium bis(allylmalonato)borate (LiBAMB), and its
incorporation into polymers to form single ion conduc-
tors. Synthesis of this new salt is based on two consid-
erations. One is that the new salt (LiBAMB) is similar
to the recently studied lithium bis(oxalato)borate (Li-
BOB)13 in which the lithium cation is well dissociated
from the anion. It is expected that the new salt should
be well dissociated in the polymer media so that higher
conductivities can be obtained. The other consideration
is that LiBAMB has allyl groups on both sides, and after
attachment to the polymer through tetramethyldisilox-
ane by means of hydrosilylation, it not only acts as a
lithium source but also acts as a cross-linking site to
enhance the mechanical properties of the resulting
single ion conductors. So far, examples of papers using
lithium salts both as the lithium source and the cross-
linker are rare.8,10 In this paper the synthesis, charac-
terization, and Li/Li cell performance of these new
network single ion conductors are reported.
Conventional solid polymer electrolytes are formed by
the dissolution of salts in a suitable polymer host, and
thus both cation and anion are mobile.1 However, since
lithium batteries only involve the intercalation/deinter-
calation and plating/stripping of the lithium cations,
practical cells using these conventional binary salt
electrolytes will suffer from problems caused by con-
centration gradients of the salt and cell polarization,
which eventually results in battery failure. Therefore,
in practical applications, polymer electrolytes with high
Li+ transference numbers are highly preferred. So far,
there are two approaches of increasing Li+ ion transfer-
ence number in polymer electrolytes. One is to incor-
porate electronically deficient moieties into the polymer
chain, which can act as an anion-trapping site to limit
the free movement of anions.2,3 In this system, a Li+
transference number as high as 0.7 has been realized.
Another approach is to chemically attach the anions to
the polymer backbones, so that Li+ transference number
of one, a single ion conductor, can be realized.4-12 Most
of the single ion conductors have been synthesized by
fixing either alkyl sulfonate4,5 or carboxylate6,7 to the
polymer backbone. However, the ion dissociation in
polyether media is very limited, and as a result, the
ambient temperature conductivities are very low, usu-
ally in the range 10-8-10-7 S cm-1. To enhance the ionic
dissociation and improve room temperature ionic con-
ductivity, polyelectrolytes with different anions have
been synthesized,8-12 and the resulting conductivities
have increased to the range 10-7-10-5 S cm-1. Among
them, higher ionic conductivities are usually found for
oligomers, low molecular weight polymers with poor
mechanical properties, so they cannot be used for
commercial devices where thin films are preferred. The
synthesis of single ion conductors with both high ambi-
ent conductivity and high mechanical strength presents
a challenge.
2. Exp er im en ta l Section
2.1. Ma ter ia ls. Allyloxyethanol, 2-[2-chloroethoxy]ethanol,
3,4-dihydro-2H-pyran, acrylic acid, methacrylic acid, tetrabu-
tylamonium hydrogen sulfate, triethylene glycol monomethyl
ether, polyethylene glycol monomethyl ether acrylate (Mn
)
454, n ) 8), polyethylene glycol monomethyl ether methacryl-
ate (Mn ) 475, n ) 8), hydrochloric acid, diethyl malonate,
trimethyl borate, allyl bromide, trichloromethylsilane, 1,2-
dichloroethane, 1,1,3,3-tetramethyldisiloxane, sodium metal,
lithium ribbon, toluene, benzene, anhydrous methanol, anhy-
drous ethanol, anhydrous MgSO4, NaOH pellet, and KOH
pellet were all obtained from Sigma-Aldrich and were used
directly without further purification. Tetrahydrafuran (B&J ,
distilled in glass) was obtained from VWR and was refluxed
over CaH2 for several days before use. AIBN was recrystallized
from methanol twice. The platinum-divinyltetramethyldisi-
loxane complex in vinylsilicon was obtained from Gelest, Inc.
2.2. Syn th esis of Mon om er s (Sch em e 1). 2-[2-(2-Chloro-
ethoxy)ethoxy]tetrahydropyran was synthesized from the ad-
dition of 2-(2-chloroethoxy)ethanol to 3,4-dihydro-2H-pyran
with hydrochloric acid as catalyst. The crude addition product
was dissolved in ether and neutralized with dilute NaOH
* Corresponding author: e-mail jbkerr@lbl.gov; Tel 1-510-486-
6279; Fax 1-510-486-4995.
10.1021/ma035690g CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/26/2004

2220 Sun et al.
Macromolecules, Vol. 37, No. 6, 2004
Sch em e 1. Syn th esis of Mon om er s
aqueous solution, washed with distilled water, and dried with
anhydrous MgSO4. The solution was filtered and concentrated
on a rotary evaporator and distilled with a purity of 99% by
GC analysis. 1H NMR (CDCl3, ppm): 4.91 (s, 1H), 3.86 (t, 3H),
3.76 (t, 3H), 3.67 (t, 3H), 3.58 (t, 3H), 3.49 (t, 3H), 1.35-1.90
(m, 6H). 13C NMR (CDCl3, ppm): 99.25, 71.60, 70.89, 66.91,
62.5, 43.00, 30.82, 25.67, 19.74. Elemental Analysis: found C
51.87, H 8.48; calculated C 51.80, H 8.15.
2-{2-[2-(2-Allyloxyethoxy)ethoxy]ethoxy}tetrahydropyran. 2-Al-
lyoxyethanol (153 g, 1.5 mol), 2-[2-(2-chloroethoxy)ethoxy]-
tetrahydropyran (208 g, 1 mol), and tetrabutylammonium
hydrogen sulfate (55 g, 0.162 mol) were mixed with 50%
sodium hydroxide aqueous solution (400 mL) and toluene (350
mL). The mixture was heated to 70 °C with mechanical stirring
for 16 h. After cooling to room temperature the mixture was
poured into 200 g of crushed ice and mixed well. The upper
organic layer was isolated, and the aqueous layer was ex-
tracted with ether (100 mL × 3). The organic layer was then
combined and washed with H2O (75 mL × 2) and dried over
MgSO4. The solution was concentrated on a rotary evaporator.
The volatiles were removed under vacuum of 0.2 Torr in 130
°C oil. The residual brownish oil (180 g) in the pot was the
crude product. Yield: 66%, 90% pure by GC analysis. This
crude product was directly used for the next synthesis step.
HCl in 50 mL of distilled water was added. The solution was
refluxed overnight, neutralized with NaHCO3, and concen-
trated under reduced pressure. The residual oil was vacuum-
distilled, and the fraction at 85-87 °C/0.15 T was collected to
1
yield 105 g product; yield: 84%, 98% pure by GC analysis. H
NMR (CDCl3, ppm): 5.80 (m, 1H), 5.16 (d, 1H), 5.07 (d, 1H),
3.91 (d, 2H), 3.4-3.7 (m, 12H), 2.80 (s, 1H). 13C NMR (CDCl3,
ppm): 134.90, 117.43, 72.77, 72.47, 70.87, 70.81, 70.56, 69.58,
61.89. Elemental Analysis: found C 56.45, H 9.24; calculated
C 54.84, H 9.47.
2-[2-(2-{2-[2-(2-Allyloxyethoxy)ethoxy]ethoxy}ethoxy)ethoxy]-
tetrahydropyran. The crude product was obtained as described
for 2-{2-[2-(2-allyloxyethoxy)ethoxy]ethoxy}tetrahydropyran as
a brown oil; yield: 54%, 91% pure by GC analysis. This crude
product was directly used for the next synthesis step.
2-(2-{2-[2-(2-Allyloxyethoxy)ethoxy]ethoxy}ethoxy)ethanol. The
crude product was obtained as described for 2-[2-(2-allyloxy-
ethoxy)ethoxy]ethanol as a brownish oil. It was then distilled
at 0.15 Torr, and the fraction that boiled at 124-126 °C was
collected; yield: 82%, 99% pure by GC analysis. 1H NMR
(CDCl3, ppm): 5.87 (m, 1H), 5.24 (d, 1H), 5.15 (d, 1H), 3.99
(d, 2H), 3.5-3.8 (m, 20H), 2.72 (s, 1H). 13C NMR (CDCl3,
ppm): 135.01, 117.38, 72.79, 72.49, 70.87, 70.83, 70.59, 69.65,
61.98. Elemental Analysis: found C 55.60, H 9.58; calculated
C 56.12, H 9.35.
2-[2-(2-Allyloxyethoxy)ethoxy]ethanol. 180 g of crude product
of 2-{2-[2-(2-allyloxyethoxy)ethoxy]ethoxy}tetrahydropyran was
dissolved in 300 mL of ethanol, and 10 mL of concentrated
2-[2-(2-{2-[2-(2-Methoxyethoxy)ethoxy]ethoxy}ethoxy)ethoxy]-
tetrahydropyran. Following the same procedure as for 2-{2-

Macromolecules, Vol. 37, No. 6, 2004
Network Type Single Ion Conductors 2221
Sch em e 2. Syn th esis of P r ep olym er s
Sch em e 3. Syn th esis of LiBAMB
[2-(2-allyloxyethoxy)ethoxy]ethoxy}tetrahydropyran triethyl-
ene glycol monomethyl ether (300 g, 1.83 mol), 2-[2-(2-
chloroethoxy)ethoxy]tetrahydropyran (208 g, 1 mol), and
tetrabutylamonium hydrogen sulfate (55 g, 0.162 mol), 50%
sodium hydroxide aqueous solution (400 mL) and toluene (350
mL) were used. After work-up, removal of volatiles by rotary
evaporation, and further purification by distillation under high
vacuum of 0.2 Torr in a 130 °C oil bath, viscous liquid was
left; 200 g, yield 59%, 82.5% pure by GC analysis. This crude
product was used directly for the next synthesis step.
2-(2-{2-[2-(2-Methoxyethoxy)ethoxy]ethoxy}ethoxy)ethanol. Fol-
lowing the same procedure used for 2-[2-(2-allyloxyethoxy)-
ethoxy]ethanol, 200 g of 2-[2-(2-{2-[2-(2-methoxyethoxy)ethoxy]-
ethoxy}ethoxy)ethoxy]tetrahydropyran was hydrolysized. After
work-up and distillation, the fraction at 118-120 °C/0.2 Torr
was collected, and 110 g of product was obtained; yield 74.6%,
1
96.7% pure by GC analysis. H NMR (CDCl3, ppm): 3.3-3.6
(m, 20H), 3.25 (s, 3H), 3.08 (t, 1H). 13C NMR (CDCl3, ppm):
72.74, 72.03, 70.71, 70.66, 70.60, 70.46, 61.69, 59.12. Elemental
Analysis: found C 51.89, H 9.47; calculated C 52.38, H 9.52.
Synthesis of Acrylate or Methacrylate Monomers. Following
the reported procedure,14 monomethyloligoethylene glycols and
1.5-fold excess of acrylic acid or methacrylic acid, hydroquinone
inhibitor, and concentrated sulfuric acid catalyst were dis-
solved in benzene. The solution was heated to reflux, and water
was removed by azeotropic distillation. The reaction was
continued for 16 h, after which the solution was cooled to room
temperature and washed with 5 wt % NaOH aqueous solution
until it is basic. The solution was then washed with distilled
water until it is neutral. The solution was dried over anhy-
drous MgSO4 and filtered. The pure product with a purity of
99% was obtained after removing the solvent on a rotary-
evaporator.
135.00, 131.29, 128.53, 117.38, 72.49, 70.87, 70.83, 69.66,
69.36, 63.96. Elemental Analysis: found C 57.62, H 8.50;
calculated C 57.83, H 8.43.
2-[2-(2-Methoxyethoxy)ethoxy]ethyl Acrylate. 1H NMR (CDCl3,
ppm): 6.33 (d, 1H), 6.06 (d, 1H), 5.74 (d, 1H), 4.22 (t, 2H),
3.4-3.8 (m, 10H), 3.29 (s, 3H). 13C NMR (CDCl3, ppm): 166.38,
131.24, 128.49, 72.11, 70.65, 69.31, 63.91, 59.22. Elemental
Analysis: found C 53.99, H 8.40; calculated C 55.05, H 8.26.
2-(2-{2-[2-(2-Methoxyethoxy)ethoxy]ethoxy}ethoxy)ethyl Ac-
1
rylate. H NMR (CDCl3, ppm): 6.36 (d, 1H), 6.08 (d, 1H), 5.77
2-Allyloxyethyl Acrylate. 1H NMR (CDCl3, ppm): 6.44 (d,
1H), 6.13 (d, 1H), 5.90 (d, 1H), 5.80 (d, 1H), 5.26 (d, 1H), 5.17
(d, 1H), 4.30 (t, 2H), 4.02 (d, 2H), 3.66 (t, 2H). 13C NMR (CDCl3,
ppm): 166.44, 134.63, 131.29, 128.51, 117.68, 72.39, 68.09,
64.01. Elemental Analysis: found C 59.97, H 7.96; calculated
C 61.54, H 7.69.
(d, 1H), 4.25 (t, 2H), 3.4-3.8 (m, 18H), 3.31 (s, 3H). 13C NMR
(CDCl3, ppm): 166.38, 131.25, 128.51, 72.15, 70.80, 69.33,
66.85, 63.93, 59.26. Elemental Analysis: found C 55.33, H 8.63;
calculated C 54.90, H 8.50.
2.3. Syn th esis of P r ep olym er s (Sch em e 2). The following
is a specific example of the polymerization of 2-[2-(2-methoxy-
ethoxy)ethoxy]ethyl acrylate-co-2-allyloxyethyl acrylate to il-
lustrate the general polymerization procedure of acrylates or
methacrylates. 2-[2-(2-Methoxyethoxy)ethoxy]ethyl acrylate
(21.8 g, 0.1 mol) and 2-allyloxyethyl acrylate (2.46 g, 0.016 mol)
were mixed in 250 mL of methanol, and 0.16 g of AIBN (1 mol
% relative to double bonds) was added as the initiator. The
solution was degassed first and then flushed with dry argon.
The above process was repeated three times, and finally the
solution was heated to 60 °C for 3 days. The solution was
concentrated and precipitated into dry diethyl ether twice. The
prepolymer was dried under high vacuum over P2O5 for 2 days.
15.5 g of viscous liquid was obtained; yield 61.0%. Mw 5.13 ×
1
2-Allyloxyethyl Methacrylate. H NMR (CDCl3, ppm): 6.15
(d, 1H), 5.88 (m, 1H), 5.52 (d, 1H), 5.21 (d, 1H), 5.16 (d, 1H),
4.32 (t, 2H), 3.98 (d, 2H), 3.66 (t, 2H), 1.90 (s, 1H). 13C NMR
(CDCl3, ppm): 167.63, 136.37, 134.70, 126.01, 117.50, 72.37,
68.11, 64.63, 18.56. Elemental Analysis: found C 63.35, H 8.29;
calculated C 63.53, H 8.24.
2-[2-(2-Allyloxyethoxy)ethoxy]ethyl Acrylate. 1H NMR (CDCl3,
ppm): 6.43 (d, 1H), 6.12 (d, 1H), 5.90 (m, 1H), 5.80 (d, 1H),
5.21 (d, 1H), 5.16 (d, 1H), 4.29 (t, 2H), 3.99 (d, 2H), 3.5-3.8
(m, 10H). 13C NMR (CDCl3, ppm): 166.42, 134.94, 131.28,
128.51, 117.40, 72.47, 70.90, 70.86, 69.63, 69.35, 63.94. El-
emental Analysis: found C 59.02, H 8.20; calculated C 59.98,
H 8.01.
1
104 g/mol, PDI ) 1.26. H NMR (CDCl3, ppm): 5.85 (m, 1H),
2-(2-{2-[2-(2-Allyloxyethoxy)ethoxy]ethoxy}ethoxy)ethyl Ac-
5.23 (d, 1H), 5.17 (d, 1H), 4.15 (t, 14H), 3.98 (d, 2H), 3.4-3.8
(m, 66H), 3.34 (s, 19H), 2.30 (t, 7H), 1.63 (d, 14H). 13C NMR
(CDCl3, ppm): 174.50, 135.61, 115.54, 72.18, 70.85, 70.78,
69.16, 63.72, 59.28, 36.60, 31.55.
1
rylate. H NMR (CDCl3, ppm): 6.39 (d, 1H), 6.14 (d, 1H), 5.90
(m, 1H), 5.80 (d, 1H), 5.21 (d, 1H), 5.16 (d, 1H), 4.27 (t, 2H),
3.98 (d, 2H), 3.5-3.8 (m, 18H). 13C NMR (CDCl3, ppm): 166.50,

2222 Sun et al.
Sch em e 4. Syn th esis of Netw or k Sin gle Ion
Macromolecules, Vol. 37, No. 6, 2004
spectrum showed that double bond signal from LiBAMB had
disappeared, indicating the addition reaction was completed.
The solvent was removed, and the residual solid was dried in
Con d u ctor s
1
the antechamber of a glovebox; 8.9 g of solid was isolated. H
NMR (CDCl3, ppm): 4.64 (m, 1H), 3.42 (t, 1H), 1.80 (t, 2H),
1.37 (m, 2H), 0.49 (t, 2H), 0.15 (m, 12H). 13C NMR (DMSO,
ppm): 168.90, 48.34, 31.31, 20.90, 18.43, 1.46, 1.25. Elemental
Analysis: found C 41.75, H 7.39; calculated C 42.11; H 7.02.
Single Ion Conductors. Equivalent amounts of intermediate
I and prepolymers were dissolved in dry THF and refluxed
for 5 days under the Pt catalyst.The apparent cross-linking
by the lithium salt grafting was observed by the precipitation
of the product from the solution. The solvent was removed,
and the crude single ion conductors were dried in a vacuum
oven at 50 °C overnight, after which they were extracted with
dry THF to remove unreacted intermediate I. Finally, the
single ion conductors were dried in a vacuum oven at 85 °C
over P2O5 for 2 days.
2.6. Mea su r em en ts. 1H and 13C NMR (TMS as internal
reference) spectra were collected on a Bruker 300 NMR
spectrometer. The GC experiment was performed on a 5890
series II plus gas chromatograph with a SPB5 column (Su-
pelcowax 30 m, 0.32 mm i.d., film thickness 0.25 µm) and
carrier gas of helium (3 mL/min) with a temperature program
of 5 min at 40 °C, 10 °C/min to 230 °C, and 5 min at 230 °C.
Elemental analysis was performed on a Perkin-Elmer series
II CHNS/O analyzer 2400.
Glass transition temperatures Tg were measured using a
Perkin-Elmer-7 DSC instrument. The heating rate was kept
at 10 °C/min. The molecular weights of the prepolymers were
determined by using Agilent 1100 series HPLC equipped with
a refractive index detector. A PLgel 10 µm miniMIX-B 250 ×
4.6 mm column was employed with THF eluent and calibrated
by using polystyrene standard samples in THF with molecular
weights ranging from 500 to 3 000 000.
2.4. Syn th esis of Sa lt (Sch em e 3). LiB(OCH3)4 was
synthesized according to the literature15 with a yield of 93.6%.
1H NMR (DMSO-d6, ppm): 3.16 (s). 13C NMR (DMSO-d6, δ,
ppm): 48.56.
2-Allyldiethyl malonate was synthesized following the lit-
erature procedure16 with a yield of 30%. 92% pure by GC
analysis. 1H NMR (CDCl3, ppm): 5.70 (m, 1H), 5.2(d, 1H), 5.17
(d, 1H), 4.14 (q, 4H), 3.42 (t, 1H), 2.61 (t, 2H), 1.29 (t, 6H). 13
C
NMR (CDCl3, ppm): 169.50, 134.47, 117.60, 61.49, 53.28,
33.23, 14.33.
2-Allylmalonic Acid. To a hot solution of 26 g of KOH in 60
mL of water, 33 g of 2-allyldiethyl malonate was added slowly
under vigorous stirring. When the reaction solution became
clear, it was cooled and poured carefully into 200 mL of
hydrochloric acid solution. The resulting organic solution was
extracted with ether (150 mL × 4). The combined ether was
dried over MgSO4 and filtered. Ether was removed, and 19.6
g of product was obtained with a yield of 82.5%, 95% by GC
The measurement of ac conductivity was carried out using
a
Swagelok cells that have been described before.18,19
A
Solartron SI 1254 four-channel frequency response analyzer
and a 1286 electrochemical interface were used to measure
the impedance of the electrolyte films of known thickness in
constant volume cells with blocking electrodes. Li/Li cells were
constructed inside a helium-filled drybox, and the cell cycling
was carried out at 85 °C using an Arbin BT 4020 mutichannel
cycler.
1
analysis. H NMR (CDCl3, ppm): 5.78 (m, 1H), 5.16 (d, 1H),
5.10 (d, 1H), 3.51 (t, 1H), 2.68 (t, 2H). 13C NMR (CDCl3, ppm):
169.21, 134.43, 117.51, 53.21, 33.20.
2-Allyldi(trichloromethylsilyl) Malonate (AllylDTMSM). Fol-
lowing the literature procedure,17 23 g of 2-allylmalonic acid
was mixed with 50 mL of chlorotrimethylsilane and 50 mL of
1,2-dichloroethane. The solution was refluxed in 80 °C oil bath
for 4 days. Solvent was removed on a rotary evaporator. The
residual was distilled, and the fraction at 73-74 °C/0.25 T was
collected. 26 g of product (AllylDTMSM) was obtained; yield:
35%, 96.7% pure by GC analysis. 1H NMR (CDCl3, ppm): 5.79
(m, 1H), 5.03 (d, 1H), 4.97(d, 1H), 3.38 (t, 1H), 2.57 (t, 3H),
0.27 (s, 18H). 13C NMR (CDCl3, ppm): 167.35, 135.90, 117.80,
45.13, 31.10, -0.15.
Lithium Bis(alllylmalonato)borate (LiBAMB). 6.0 g of LiB-
(OCH3)4 (0.042 mol) and 24.2 g of AllylDTMSM (0.084 mol)
were mixed in 200 mL of anhydrous acetonitrile and heated
in a 45-50 °C oil bath for 24 h. After the reaction was
completed, the solvent was removed by rotary evaporator at
50 °C and by drying in a vacuum oven at 75 °C for 24 h. It is
difficult to recrystallize from acetonitrile/benzene solution, so
the solid was dissolved in acetonitrile and filtered, the solvent
was removed under reduced pressure, and 11.0 g of product
was obtained after drying under vacuum at 85 °C for 2 days;
yield 86.7%. 1H NMR (DMSO, ppm): 5.75 (m, 1H), 5.02 (d,
1H), 4.98 (d, 1H), 3.54 (t, 1H), 2.56 (t, 2H). 13C NMR (DMSO,
ppm): 168.20, 135.90, 117.78, 48.12, 31.16. Elemental Analy-
sis, found C 47.55, H 4.33; calculated C 47.68, H 3.97.
2.5. Syn th esis of Netw or k Sin gle Ion Con d u ctor s
(Sch em e 4). Intermediate I. Tetramethyldisiloxane (36 g, 0.27
mol) was dissolved in 100 mL of dry THF, and 0.34 g of Pt
catalyst was added. The solution was stirred at 70 °C, and
4.0 g of LiBAMB (0.0132 mol) dissolved in 50 mL of THF was
added to this solution slowly under the purge of dry N2. The
reaction was continued for 3 days, after which the 1H NMR
3. Resu lts a n d Discu ssion
3.1. P r ep a r a tion of Mon om er s a n d Netw or k
Sin gle Ion Con d u ctor s. The allyl-terminated and
methoxy-terminated oligoethylene glycols were synthe-
sized by stepwise chain extension and hydrolysis as
shown in Scheme 1. Gas chromatography (GC) was used
to monitor the process and the identity of the distilled
product was confirmed by 1H NMR. The final termi-
nated glycols were esterified with acrylic acid or meth-
acrylic acid using catalytic amounts of concentrated
sulfuric acid rather than with acryloyl chloride or
methacryloyl chloride via the catalysis of triethylamine,
since the product obtained through the former process
is of higher purity, and also the byproduct of triethyl-
amine hydrogen chloride of the latter process is difficult
to remove completely.
The prepolymers were obtained by free radical po-
lymerization in methanol at 65 °C with AIBN as the
initiator. They were purified by repeated precipitations
from methanol into dry ether.
The synthesis of network single ion conductors pro-
ceeded through two steps. Intermediate compounds
were synthesized by reacting LiBAMB with excess
tetramethyldisiloxane, and their identities were con-
1
firmed in the H NMR spectra through the disappear-
ance of the double bond signal from the salt and the

Macromolecules, Vol. 37, No. 6, 2004
Network Type Single Ion Conductors 2223
Ta ble 1. P h ysica l P r op er ties of P r ep olym er s
prepolymers
polyacrylates
R
m
n
x
Mn
Mw
PDI
Tg/°C
appearance
H
H
H
H
H
H
H
H
3
5
5
5
5
8
8
8
8
8
1
1
1
1
1
1
1
3
5
1
6.33
7.80
3.80
1.80
0.80
2.27
1.08
0.84
0.60
1.06
4.07 × 104
9.32 × 103
1.17 × 104
1.35 × 104
6.58 × 104
5.92 × 103
9.89 × 106
6.93 × 103
5.00 × 103
1.77 × 106
5.13 × 104
1.48 × 104
2.20 × 104
3.12 × 104
1.17 × 105
8.25 × 103
1.54 × 107
8.97 × 103
7.86 × 103
2.73 × 106
1.26
1.59
1.88
2.31
1.78
1.39
1.56
1.30
1.57
1.54
-56.0
-64.7
-63.0
-59.5
-44.8
-61.3
-55.7
-58.7
-61.1
-55.9
viscous liquid
viscous liquid
viscous liquid
viscous liquid
glassy
viscous liquid
rubber
viscous liquid
viscous liquid
rubber
H
CH3
polymethacrylates
presence of a silane signal. The obtained intermediates
were then reacted with prepolymers using a platinum
catalyst in dry THF. To ensure complete reaction, a
higher catalyst concentration and extended reaction
time was used. The residual unreacted intermediates
were removed from the single ion conductors by extrac-
tion with THF, in which the intermediates were well
dissolved.
3.2. P h ysica l P r op er ties of P r ep olym er s a n d
Sin gle Ion Con d u ctor s. The obtained polyacrylate or
polymethacrylate polymers varied in mechanical prop-
erties from viscous liquids to rubbers, depending on the
molar ratio of the two monomers used. The general
physical properties of these prepolymers are listed in
Table 1.
The glass transition temperatures of prepolymers
depend on both the ratios of two monomers used and
also upon the number of EO units in the two monomers.
For the prepolymers from the same starting monomers
of pentaethylene glycol monomethyl ether acrylate and
allyloxyethyl acrylate, the more allyloxyethyl acrylate
that the polymer contained (the smaller value of x in
Table 1 (see also Scheme 2)) the higher were the
molecular weights and glass transition temperatures.
At similar molar ratios of the two monomers used for
copolymerization, it was observed that the glass transi-
tion temperature of the resulting prepolymers was
higher with fewer EO units in the side chains of the
two monomers. For example, the glass transition tem-
perature of the prepolymer from triethylene glycol
monomethyl ether acrylate and alloyoxyethyl acrylate
(x ) 6.33, Tg ) -56 °C) is higher than that of the
prepolymer from pentaethylene glycol monomethyl ether
acrylate and allyloxyethyl acrylate (x ) 7.80, Tg ) -64.7
°C). This also holds true for the prepolymers from the
same monomer of poly(ethylene glycol) monomethyl
ether acrylate (n ) 8) and different comonomers; i.e.,
the glass transition temperature of the prepolymer from
the comonomer of allyloxyethyl acrylate (1 EO unit) (x
) 1.08, Tg ) -55.7 °C) is higher than that from the
comonomer of 2-[2-(2-allyloxyethoxy)ethoxy]ethyl acryl-
ate (3 EO unit) (x ) 0.84, Tg ) -58.7 °C), which is
higher than that from the comonomer of 2-(2-{2-[2-(2-
allyloxyethoxy)ethoxy]ethoxy}ethoxy)ethyl acrylate (5
EO unit) (x ) 0.60, Tg ) -61.1 °C).
F igu r e 1. Arrhenius plots of ionic conductivities of single ion
conductors with general formula of PAEmXE1 (one EO unit in
cross-linker and different EO units in the side chains) at the
salt concentration of EO/Li ) 40.
chain length and different side chains at the same salt
concentration of EO/Li ) 40. As can be seen from the
Figure 1, the conductivity systematically increases with
increasing the side chain length. The best ambient
temperature conductivity of 1.8 × 10-8 S cm-1 is
obtained for the single ion conductor with a side chain
length of three ethylene oxide (EO) units; it increases
to 2.7 × 10-8 S cm-1 when side chain length is five EO
units and reaches the highest room temperature con-
ductivity of 1.9 × 10-7 S cm-1 when side chain length
is eight EO units. This is consistent with previously
reported conductivities of binary salt (LiTFSI) solutions
of polyacrylate ethers where it was observed that the
maximum values were obtained with side chains that
were at least six EO units long.20 In general, a difference
in ionic conductivity is either caused by a difference in
the number of charge carriers or by the ionic mobility.
As suggested by Nishimoto et al. in polyether-based
network polymer electrolytes with hyperbranched side
chains,21 it is quite unlikely that the dissociation of the
lithium salt differs enough to cause the changes in
conductivity reported above because the chemical struc-
ture of the single ion conductor is so similar. The
conductivity difference must be accounted for by the
difference of ionic mobility, which is due to the difference
in the mobility of the side chains within the network.
As reflected in Table 2, the glass transition temperature
of the single ion conductor decreases with an increase
in the side chain length, which is consistent with an
increase in side chain mobility.
After cross-linking with LiBAMB by hydrosilylation,
the glass transition temperature of the resulting dry
single ion conductor increases with respect to that of
the corresponding prepolymer. Here the increase of
glass transition temperature is due to not only the cross-
linking by the bridging lithium salt but also a result of
the complexation of lithium cation by carboxylate groups
in the anion and by the oligoether side chains.
3.3. Ion ic Con du ctivity. Effect of Side Chain Length-
PAEmXE1-40. Figure 1 shows the ionic conductivities of
network single ion conductors having same cross-linker

2224 Sun et al.
Macromolecules, Vol. 37, No. 6, 2004
Ta ble 2. Gla ss Tr a n sition Tem p er a tu r e (Tg) a n d Am bien t Con d u ctivity (log σ) of Sin gle Ion Con d u ctor s
polymers
R
m
n
x
EO/Li
polymer code
Tg/°C
log σ/S cm-1
H
H
H
H
H
H
H
H
3
5
5
5
5
8
8
8
8
8
1
1
1
1
1
1
1
3
5
1
6.33
7.80
3.80
1.80
0.80
2.27
1.08
0.84
0.60
1.06
40/1
80/1
40/1
20/1
10/1
40/1
20/1
20/1
20/1
20/1
PAE3XE1-40
PAE5XE1-80
PAE5XE1-40
PAE5XE1-20
PAE5XE1-10
PAE8XE1-40
PAE8XE1-20
PAE8XE3-20
PAE8XE5-20
PMAE8XE1-20
-52.2
-55.4
-54.7
-52.5
-37.2
-56.4
-50.4
-51.5
-53.0
-48.5
-7.75
-7.96
-7.57
-7.94
N/A
-6.73
-7.04
-6.76
-6.46
-7.34
polyacrylates
H
CH3
polymethacrylates
F igu r e 3. Glass transition temperatures Tg and ionic con-
ductivities for PAE5XE1-based network single ion conductors
at varying temperatures.
F igu r e 2. Arrhenius plots of ionic conductivities of single ion
conductors with general formaula of PAE8XEn (eight EO units
in the side chain and different EO units in cross-linker chain)
at the same concentration of EO/Li ) 20.
Effect of Salt Concentration-PAE5XE1. Figure 3 shows
the glass transition temperatures Tg and ionic conduc-
tivities for PAE5XE1-based network single ion conduc-
tors at several temperatures. As observed in conven-
tionalsolid polymer electrolytes,22-24 the ionicconductivity
initially increases with increasing salt concentration,
passes a maximum, and then decreases. The best
conductivity was found at EO/Li ) 40 with a value of
1.2 × 10-8 S cm-1 at 25 °C. In contrast to the ionic
conductivity, the glass transition temperature continu-
ally increases with increasing salt concentration. The
increase of glass transition temperature with salt
concentration is not only due to the increase of chemical
cross-linking density but also due to the intermolecular
and intramolecular coordination of lithium cations,
which act as transient physical cross-linking points in
the polymer electrolytes. The increase in Tg results in
a decrease of segmental motion of the matrix polymer
and thus reduces the mobility of free cations; however,
the number of free cations increases with increasing salt
concentration. The combination of these two opposite
effects results in the ionic conductivity maximum ob-
served in Figure 3.22-24
Effect of Plasticizer. The conductivities of these new
single ion conductors as seen in Figures 1-3 are low,
suggesting that the severe ion pairing occurred in these
systems. To improve the ionic conductivity, a solvent
mixture of EC/DMC (1/1, wt/wt) was added to form gel
polyelectrolytes. This particular combination was se-
lected due to its best combination of dielectric constant
and viscosity. The gels were only formed from the series
of PAE5XE1 with different salt concentrations to eluci-
date the effect of plasticizer. The isothermal ionic
Effect of Cross-Linker Chain Length-PAE8XEn-20.
Figure 2 shows the ionic conductivity of single ion
conductors having same side chain and different anion
cross-linker chain lengths at the same salt concentration
of EO/Li ) 20. In this case, the number of EO units in
the side chain was kept at eight while the number of
EO units in the cross-linker changes. It should be noted
that the number of EO units in the cross-linker of the
prepolymer is different from that in the cross-linking
bridge of the single ion conductor, since each lithium
salt, LiBAMB, was connected to two cross-linkers of the
prepolymer. Thus, the number of EO units in each
bridge of the single ion conductor is double the number
of EO units in the prepolymer. The conductivity in
Figure 2 showed similar behavior as observed in Figure
1 for the effect of side chain length; i.e., conductivity
increases with increasing the cross-linker length. As
discussed in the previous paragraph, the main contribu-
tion to the ionic conductivity difference is from the ion
mobility difference. In these polymers, it is observed
that a longer bridge between the polymer main chains
will result in a more flexible network and thus a higher
ionic mobility. The higher mobility of networks contain-
ing longer cross-linkers is confirmed from the lower
glass transition temperatures, as shown in Table 2.
However, it should also be noted that the effect of cross-
linker on conductivity is not as pronounced as that of
the side chain. The highest ambient temperature con-
ductivity is 9.1 × 10-8 S cm-1 for n ) 1, 1.7 × 10-7
S
cm-1 for n ) 3, and 3.5 × 10-7 S cm-1 for n ) 5.

Macromolecules, Vol. 37, No. 6, 2004
Network Type Single Ion Conductors 2225
F igu r e 4. Dependence of ionic conductivity and glass transi-
tion temperature on salt concentration for PAE5XE1-based gel
single ion conductors containing 50 wt % EC/DMC (1/1, wt/
wt).
F igu r e 6. Cycling profile the Li/Li cell of a dry single ion
conductor with a salt concentration of EO/Li ) 80/1 under the
sequence of 1 h relaxation, 2 h charge, 1 h relaxation, and 2 h
discharge at a current density of 50 µA cm-2
.
based single ion conductor is slightly higher than that
of polymethyacrylate-based single ion conductor; how-
ever, at higher temperatures their conductivities tend
to merge. The conductivity difference at lower temper-
atures is primarily due to the differences in their glass
transition temperatures which are listed in Table 2.
3.4. Li/Li Cell Cyclin g P er for m a n ce. The dry
single ion conductors were used to construct Li/Li cells,
which were cycled under the sequence of 1 h relaxation,
2 h charge, 1 h relaxation, and 2 h discharge at different
current densities. One typical cycling performance of dry
single ion conductor PAE5XE1-80 is shown in Figure 6.
The cell was cycled at a current density of 50 µA cm-2
.
Since there are no published reports describing the
behavior of single ion conductors between nonblocking
lithium electrodes under charge/discharge conditions,
there are no direct comparisons to mention here.
However, the result shown in Figure 6 has two distinct
features, which, we believe, is unique to the behavior
of single ion conductors. First of all, for all cycles (except
the first cycle) shown in the figure, once the current was
turned on, the cell potential immediately increased to
a value determined by the ohmic and interfacial imped-
ances and did not increase further with time, which is
distinctly different from the behavior of binary salt
conductors in which cell potential increases further over
time due to the concentration polarization.25 Second, the
behavior of the cell during the very first charge cycle,
i.e., the cell potential falls with time, also appears to be
a special feature of this network type single ion conduc-
tor. The cations were complexed by either the side
chains or the cross-linker chains, and the anions were
fixed between the polymer chains and cannot move on
the time scale of the initial current pulse (due to the
low ionic conductivity), which resulted in an initial
potential higher than that determined by the combined
bulk and interfacial impedance (Figure 7). The polymer
chains themselves are not fully extended like the ideal
two-dimensional network structure drawn on paper.
However, once the current is turned on, the lithium
cations move through the membrane by means of the
polymer segmental motion, which results in some
relaxation of the polymer chains. The anions will also
move around due to segmental motion but the range of
F igu r e 5. Comparison of ionic conductivity of polyacrylate-
based single ion conductor with that of polymethacrylate-based
single ion conductor at the same salt concentration of EO/Li
) 20.
conductivities of the resulting gel single ion conductors
containing 50 wt % of EC/DMC at 25 °C is shown in
Figure 4. To facilitate a direct comparison, the conduc-
tivities of the dry single ion conductors are also included
in the figure as dashed lines. As can be seen in the
figure, the conductivities of the gel single ion conductors
are 2 orders of magnitude higher than those of the dry
single ion conductors across all the salt concentrations.
The addition of plasticizer not only increases the ion
mobility by providing a microscopically liquid environ-
ment but also apparently increases the number of
conducting ions through better dissociation in the high
dielectric constant media. The highest ambient temper-
ature conductivity of 7.9 × 10-6 S cm-1 is obtained for
the gel single ion conductor with a salt concentration
of EO/Li ) 40.
Effect of Polymer Backbone. Figure 5 compares the
ionic conductivity of polyacrylate-based single ion con-
ductor with that of polymethacrylate-based single ion
conductor at the same salt concentration of EO/Li ) 20.
At lower temperatures the conductivity of polyacrylate-

2226 Sun et al.
Macromolecules, Vol. 37, No. 6, 2004
F igu r e 8. Gas chromatograph (GC) of the extraction of the
aqueous solution of failed Li|dry single ion conductor with EO/
Li ) 80| Li cell with dichloromethane.
chain lengths of three and eight EO units. This GC
result is consistent with the observed increases in both
cell potential and interfacial impedance in Figures 6 and
7, respectively, which appear to be mainly due to the
decomposition of the comb branches at the ester
sites.
The result shown in Figure 8 suggests that the ester
group is not stable against the lithium metal under
charge/discharge conditions. In addition, the conductiv-
ity of these dry single ion conductors is too low for
practical applications. To improve ionic conductivity,
different polymers need to be synthesized. Usually
polyepoxides have lower glass transition temperatures
than those of polyacrylates, so higher mobility of ions
is expected. In addition, the ether bond of polyepoxide
is less easily reduced than the ester bond of polyacrylate
and should be more stable against the lithium metal
when cycled under charge/discharge condition. Through
the same variation scheme shown for polyacrylate-based
single ion conductors, the conductivity of polyepoxide-
based single ion conductors can also be optimized.26 The
results of the optimization of ionic conductivity and cell
performance of polyepoxide-based single ion conductors
will be reported in a separate paper.27
F igu r e 7. Impedance spectra of the cell from Figure 6 before
(A) and after (B) cycling.
motion will be limited by the length of the side chain
that tethers them to the polymer backbone. If observed
carefully, the final cell potential during the first charge
cycle is very close to the calculated cell potential from
the combined bulk and interfacial impedances (Figure
7). The decrease of the potential may be due to polymer
relaxation processes in the bulk or on the surface but
may equally be due to nucleation on the lithium metal
electrode. Further experimentation is required to elu-
cidate the sources of the potential variations. The
potential increase with continued cycling is apparently
related to reactions on the surface of lithium metal,
leading to a buildup of the SEI layer and thus the
increase of the interfacial impedance. The source of the
SEI formation, which was confirmed by the following
post mortem analysis result, is from the decomposed
small species (organic or inorganic lithium salts). These
mobile species were responsible for the potential tail up
in the final charge/discharge period as observed after
10 cycles, which is due to concentration polarization.
4. Con clu sion s
Polyacrylate/polymethacrylate-based network single
ion conductors containing the LiBAMB salt have been
successfully synthesized and tested as electrolytes for
rechargeable lithium batteries. An increase in the
number of ethylene oxide units in both the side chain
and the cross-linker chain leads to an increase in the
ionic conductivity of new single ion conductors. An
optimum in conductivity is observed for salt concentra-
tion that results from the increase in charge carriers
vs the increase in Tg with salt concentration. Plasticizer
increases the ionic conductivity of gel single ion conduc-
tor by more than 2 orders of magnitude relative to that
of the dry single ion conductor. The Li/Li cycling
performance of dry single ion conductor showed no
polarization, and nearly no relaxation was observed. An
increase of cell potential with cycling is caused by the
increase of the interfacial impedance arising from the
formation of a SEI layer. The main reason for cell failure
appears to be due to the decomposition of the ester
group in the side chains.
3.5. P ostm or tem An a lysis. The failed Li/Li cells
were disassembled inside a glovebox. The polymer
electrolyte was scraped off the electrode, dissolved in a
small amount of distilled water, and extracted with
dichloromethane. GC was used to analyze the extracted
compounds, and a typical chromatogram is shown in
Figure 8. The peak with a retention time of 20.355 min
is identical to that of the synthesized pure side chain
starting material, penta(ethylene glycol) monomethyl
ether, which means that the polymers were decomposed
from the ester sites of the side chains. Similar results
were observed for other single ion conductors with side
Ack n ow led gm en t. This work was financially sup-
ported by the NASA PERS program (NASA Glenn).

Macromolecules, Vol. 37, No. 6, 2004
Network Type Single Ion Conductors 2227
(14) Sun, X. G.; Lin, Y. Q.; J ing, X. B. Solid State Ionics 1996,
83, 79.
(15) Barthel, L.; Buestrich, R.; Carl, E.; Gores, H. J . J . Electro-
chem. Soc. 1996, 143, 3572.
(16) Legoff, E.; Ulrich, S. E.; Denney, D. B. J . Am. Chem. Soc.
1958, 80, 622.
(17) Hergott, H. H.; Simchen, G. Synthesis 1980, 626.
(18) Ma, Y. P.; Doyle, M.; Fuller, T. F.; Doeff, M. M.; De J onghe,
L. C.; Newman, J . J . Electrochem. Soc. 1995, 142, 1859.
(19) Doeff, M. M.; Edman, L.; Sloop, S. E.; Kerr, J . B.; De J onghe,
L. C. J . Power Sources 2000, 89, 227.
(20) Buriez, O.; Han, Y. B.; Hou, J .; Kerr, J . B.; Qiao, J .; Sloop, S.
E.; Tian, M.; Wang, S. J . Power Sources 2000, 89, 149.
(21) Nishimoto, A.; Agehara, K.; Furuya, N.; Watanabe, T.;
Watanabe, M. Macromolecules 1999, 32, 1541.
(22) Kobayashi, N.; Uchiyama, M.; Shigehara, K.; Tsuchida, E.
J . Phys. Chem. 1985, 89, 987.
(23) Tsuchida, E.; Kobayashi, N.; Ohno, H. Macromolecues 1988,
21, 96.
(24) Synder, J . E.; Ratner, M. A.; Shriver, D. F. J . Electrochem.
Soc. 2003, 150, A1090.
Refer en ces a n d Notes
(1) Gray, F. M., Ed.; Solid Polymer Electrolytes: Fundamentals
and Technological Applications; VCH Publishing: New York,
1991.
(2) (a) Mehta, M. A.; Fujinami, T.; Inoue, T. J . Power Sources
1999, 81-82, 724. (b) Mehta, M. A.; Fujinami, T.; Inoue, S.;
Matushita, K.; Inoue, T. Electrochim. Acta 1999, 45, 1175.
(3) (a) Sun, X. G.; Angell, C. A. Electrochim. Acta 2001, 46, 1467.
(b). Sun, X. G.; Xu, W.; Zhang, S. S.; Angell, C. A. J . Phys.:
Condens. Matter 2001, 13, 8235. (c) Xu, W.; Sun, X. G.; Angell,
C. A. Electrochim. Acta 2003, 48, 2255.
(4) Ito, Y.; Ohno, H. Solid State Ionics 1995, 79, 300.
(5) Zhou, G. B.; Khon, I. M.; Smid, J . Polym. Commun. 1989,
30, 52.
(6) Kobayashi, N.; Uchiyama, M.; Tsuchida, E. Solid State Ionics
1985, 17, 307.
(7) Tsuchida, E.; Shigehara, K. Mol. Cryst. Liq. Cryst. 1984, 106,
361.
(8) Benrabah, D.; Sylla, S.; Alloin, F.; Sanchez, J . Y.; Armand,
M. Electrochim. Acta 1995, 40, 2259.
(9) Onishi, K.; Matsumoto, M.; Nakacho, M.; Shigehara, K.
Chem. Mater. 1996, 8, 469.
(25) Kerr, J . B.; Sloop, S. E.; Liu, G.; Han, Y. B.; Hou, J .; Wang,
S. J . Power Sources 2002, 110, 389.
(10) Fujinami, T.; Tokimune, A.; Mehta, M. A.; Shriver, D. F.;
Rawsky, G. C. Chem. Mater. 1997, 9, 2236.
(11) Xu, W.; Angell, C. A. Solid State Ionics 2002, 147, 295.
(12) Sun, X. G.; Angell, C. A. Solid State Ionics, in press.
(13) Xu, W.; Angell, C. A. Electrochem. Solid-State Lett. 2001, 4,
E1.
(26) Marchese, L.; Andrei, M.; Roggero, A.; Passerini, S.; Prosperi,
P.; Scrosati, B. Electrochim. Acta 1992, 37, 1559.
(27) Sun, X. G.; Craig, R.; Liu, G.; Han, Y. B.; Kerr, J . B. J . Am.
Chem. Soc., in preparation.
MA035690G