Electrochimica Acta 50-200*-****-****

Single-ion conducting polymer silicate nanocomposite electrolytes

for lithium battery applications

Mary Kuriana,1, Mary E. Galvina,, Patrick E. Trapab,

Donald R. Sadowayb, Anne M. Mayesb,2

a Department of Materials Science and Engineering, University of Delaware, Newark, DE 19716, USA

b Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, USA

Received 28 April 2004; received in revised form 12 September 2004; accepted 13 September 2004

Abstract

Solid-state polymer silicate nanocomposite electrolytes based on an amorphous polymer poly[(oxyethylene)8 methacrylate], POEM, and

lithium montmorillonite clay were fabricated and characterized to investigate the feasibility of their use as salt-free electrolytes in lithium

polymer batteries. X-ray scattering and transmission electron microscopy studies indicate the formation of an intercalated morphology in the

nanocomposites due to favorable interactions between the polymer matrix and the clay. The morphology of the nanocomposite is intricately

linked to the amount of silicate in the system. At low clay contents, dynamic rheological testing veri es that silicate incorporation enhances the

mechanical properties of POEM, while impedance spectroscopy shows an improvement in electrical properties. With clay content 15 wt.%,

mechanical properties are further improved but the formation of an apparent superlattice structure correlates with a loss in the electrical

properties of the nanocomposite. The use of suitably modi ed clays in nanocomposites with high clay contents eliminates this superstructure

formation, yielding materials with enhanced performance.

2004 Elsevier Ltd. All rights reserved.

Keywords: Polymer silicate nanocomposites; Salt-free nanocomposite electrolytes; Lithium polymer battery; Nanocomposite morphology; Organic silicate

modi ers

and exible [1]. However, to realize these performance ad-

1. Introduction

vantages, several challenges related to materials design must

be overcome, including the ability to make LPBs mechan-

The increasing energy needs of modern society have

ically robust and operable at high current rates while still

spurred extensive research and development in the areas of

retaining their lightweight construction. This is particularly

energy production, storage and distribution. Devices incor-

important in applications where batteries are designed to

porating solvent-free polymer electrolytes, in particular the

serve a secondary function as a structural or insulating el-

lithium polymer battery (LPB), are highly desirable due

ement that might commonly encounter signi cant stress or

to characteristics such as inherent low safety risks and their

deformation.

ability to be formed into thin lm structures of large surface

Poly(ethylene oxide) (PEO) doped with alkali metal salts

area, yielding high energy density cells that are lightweight

has long been favored a candidate for use as an electrolyte in

solid-state rechargeable lithium batteries [2]. However, since

Corresponding author. Tel.: +1-302-***-****; fax: +1-302-***-****.

the local relaxations and segmental motions of the polymer

E-mail addresses: abqq6c@r.postjobfree.com (M.E. Galvin),

host that are typically needed [3] to allow ef cient Li+ trans-

abqq6c@r.postjobfree.com (A.M. Mayes).

1 Current address: Air Products and Chemicals, Inc., Corporate Science

port in the electrolyte are best achieved in an amorphous poly-

& Technology Center, 7201 Hamilton Blvd., Allentown, PA 18195, USA.

mer, it is only above the melting temperature of crystalline

2 Tel.: +1-617-***-****; fax: +1-617-***-****.

0013-4686/$ see front matter 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.electacta.2004.09.020

2126 M. Kurian et al. / Electrochimica Acta 50-200*-****-****

PEO LiX complexes ( 60 C) that appreciable conductivi-

ties ( > 10 4 S/cm) are seen [4]. Most efforts aimed at low-

ering the operation temperatures of PEO LiX systems to the

ambient region have focused on the development of copoly-

merization [4 12] or cross-linking [13 17] strategies and the

use of suitable plasticizers [1,18,19] to create completely

amorphous systems with enhanced conductivity. Incorpora-

tion of inorganic particles in the polymer matrix to impart Fig. 1. Schematic representation of structure of poly[(oxyethylene)8

mechanical stability [20], and enhance interfacial properties methacrylate] (POEM) polymer.

[21] and conductivity by suppressing crystallization of the

be very strongly dependent on clay and co-solvent/plasticizer

PEO host has also been investigated [22 27].

content.

An alternative strategy for creating polymer electrolyte

Our work in this paper reports on the development of

systems with improved electrical and mechanical properties

dry polymer silicate nanocomposite electrolytes based

is through fabrication of polymer silicate nanocomposites

on a lithium cation exchanged montmorillonite clay (Li-

(PSNs). PSNs are a class of materials in which nanoscale clay

mmt) and an amorphous PEO variant, poly[(oxyethylene)8

particles are molecularly dispersed within a polymer matrix

methacrylate], POEM, without added plasticizers or dopant

[28 34]. Recent commercial interest in these nanocompos-

salts (Fig. 1 is a schematic representation of the structure of

ites arises from the fact that they exhibit dramatic increases

POEM). Here the cations serve as the charge carriers in the

in tensile strength [30], heat resistance [29] and solvent re-

PSN electrolyte, while the clay sheets themselves provide im-

sistance [35] as well as decreases in gas permeability when

mobilized counterions, leading to the formation of a nearly

compared with the bulk polymer [29] properties also desir-

single-ion conductor where the Li+ transport number should

able in electrolytes for LPBs.

approach a value of unity. Such systems would be expected

Aranda and Ruizhitzky [36] have shown that the incor-

to hold performance advantages over salt-doped systems at

poration of lithium/sodium montmorillonite clay in a PEO

high currents if suf cient conductivities could be achieved

matrix using solution processing results in composite elec-

[44].

trolytes with enhanced ionic conductivity in the tempera-

ture range 400 700 K. Vaia et al. reported that composite

electrolytes with a high inorganic content (60 wt.% clay)

2. Experimental

fabricated by melt intercalation had higher conductivity

( 10 6 S/cm) than LiBF4 /PEO electrolytes ( 10 8 S/cm)

2.1. POEM synthesis

at 30 C [37]. In most studies, intercalation of the poly-

mer chains in the silicate galleries appears to suppress their Poly[(oxyethylene)8 methacrylate] was synthesized by

tendency to crystallize, resulting in enhanced ionic con- solution-free radical polymerization of the monomer us-

ductivity. There are also reports of annealing leading to ing ethyl acetate as the solvent. Both the OEM monomer

enhanced conductivity due to an increase in PEO intercalation (polyethylene glycol methyl ether methacrylate) and the

into the clay galleries [38]. Smectic clay/polymer nanocom- initiator 2,2 -azobisisobutyronitrile (AIBN) were purchased

posites [39] based on [polystyrene/clay] and [poly(methyl from Sigma-Aldrich and used as received. A clean, dry

methacrylate)-co-acrylamide/clay] prepared by in situ in- 500 mL round bottom ask was used as the reactor. Ethyl

tercalative polymerization of the host polymer in the pres- acetate was introduced into the reactor and monomer was

ence of clay mineral were also studied, but showed very added in the amount that provided a 10% monomer solu-

low ionic conductivities. Additionally, gel composite elec- tion. AIBN was added in concentration to yield the desired

trolytes [40] of lithium hectorite dispersed in high dielectric molecular weight. The reactor was degassed to remove traces

solvents (ethylene carbonate and propylene carbonate) exhib- of oxygen from the reaction mixture by bubbling argon gas

ited a room temperature conductivity above 10 4 S/cm and through the mixture for about 15 min, after which the reactor

a transference number of 0.8, while PEO intercalated into was clamped into an oil bath held at 65 C. Both the reac-

lithiated taeniolite showed conductivities of 3 10 7 S/cm tants and the oil were stirred individually to achieve good

at room temperature [41]. mixing and promote even heating, respectively. The reaction

There have also been studies on PSN electrolytes doped was allowed to proceed for approximately 48 h. The result-

with lithium metal salts. [PEO/poly(oxypropylene) diamine ing polymer solution was concentrated on a rotary evaporator,

modi ed sodium montmorillonite/LiCF3 SO3 ] nanocompos- precipitated in a hexane methanol mixture (40:1) and nally

ites have shown higher ionic conductivities than conventional dried under vacuum for 24 h to isolate the colorless polymer.

PEO/LiX electrolytes [42]. Gel electrolytes based on [poly-

acrylonitrile/organophilic clay/propylene or ethylene car- 2.2. Clay puri cation and handling

bonate co-solvent/LiClO4 ] [43] showed an enhancement in

dimensional and electrochemical stability due to clay incor- Sodium montmorillonite (SWy-2, Source Clay Minerals

poration. However, in these cases, properties were found to Repository, University of Missouri) was used as the starting

M. Kurian et al. / Electrochimica Acta 50-200*-****-**** 2127

Table 1

Sample nomenclature and estimated [Li+ :EO] ratios for POEM Li+ mont-

morillonite nanocomposite electrolytes

Estimated [Li+ :EO]

Sample Composition

[Li-clay/POEM] (w/w) ratio

7 10 4

2% PSN 2/98

1.8 10 3

5% PSN 5/95

3.9 10 3

10% PSN 10/90

6.1 10 3

15% PSN 15/85

8.8 10 3

20% PSN 20/80

11.6 10 3

25% PSN 25/75

2.3. Nanocomposite fabrication

A solution technique using dry, distilled acetonitrile as

solvent was employed in all of the nanocomposite prepara-

tions. All nanocomposite handling was done inside an argon-

Fig. 2. Schematic representation of the structure of the unit cell in Na+

lled glove box (moisture and oxygen levels were measured

montmorillonite clay mineral (adapted from Ref. [45]).

to be less than 1 ppm) to maintain dry conditions and avoid

contamination of the electrolytes. POEM/lithium montmo-

material. Montmorillonite has a net negative charge that is

rillonite nanocomposites with 2, 5, 10, 15, 20 and 25 wt.% of

compensated with sodium cations. In the presence of water,

clay were prepared (Table 1 lists sample nomenclature used

the compensating sodium cations on the clay layer surfaces

in this report). Measured amounts of POEM and Li-mmt were

can be easily exchanged for other cations when available in

each dissolved separately in 250 mL acetonitrile and stirred

excess amount in solution. (Fig. 2 shows a schematic repre-

for 24 h. The polymer solution was then added to the clay

sentation of the unit cell of montmorillonite clay [45] used in

suspension with rapid stirring to ensure maximum contact

this work.)

between the polymer chains and clay platelets. This mixed

Prior to use in the nanocomposites, the clay was puri ed

solution was stirred for 24 h, after which it was evaporated

and treated to obtain clean lithium montmorillonite of parti-

to dryness on a rotary evaporator. The composite was further

cle size

dried under vacuum at 60 C for 24 h.

sodium montmorillonite were dispersed in 1 L of de-ionized

water and the dispersion was allowed to stand for 24 h to

allow impurities and heavy mineral fractions to settle. The 2.4. Preparation of organically modi ed Li-mmt and

clean clay suspension was poured into a graduated beaker and fabrication of the corresponding nanocomposites

the residual heavy sediment was discarded. To ensure com-

plete saturation of the cation exchange capacity positions of A simple solution technique was used to modify Li-

the clay with lithium cations, the suspension was saturated mmt using diamine-terminated PEO, [NH2 ]2 -PEO (Scien-

ti c Polymer Products, Inc., molecular weight 2000 g/mol,

with lithium by adding suf cient lithium chloride salt (Fisher

amine content 0.88 meq/g). A measured amount of Li-mmt

Chemicals) to achieve a 1 M solution. (This provides Li+ in

over 100 times excess of the amount required to completely was dispersed in de-ionized (DI) water and the solution was

exchange all available cation exchange sites, based on the allowed to stir for 12 h. In a similar manner, the modi er solu-

cation exchange capacity determined for the clay.) The solu- tion was made by dissolving a calculated amount of [NH2 ]2 -

tion was stirred for 24 h to allow for complete replacement PEO separately in DI water and allowing it to stir for 12 h.

of sodium cations by lithium cations and then centrifuged The modi er solution was then added to the clay dispersion

to separate the lithium montmorillonite clay from the super- and the mixed solution stirred for a further 12 h. Two dif-

natant containing the excess ions. The clay was re-suspended ferent modi ed clay solutions were made based on the ex-

in de-ionized water and stirred for 12 h to remove excess ions. tent of modi cation of the clay by the organic modi er, the

This suspension was then re-centrifuged to recover the clay. rst such that only 10% of the cation exchange sites were

modi ed ( 2 wt.% [NH2 ]2 -PEO) and the second with 50%

The conductivity of the supernatant was monitored and the

of the sites modi ed ( 11 wt.% [NH2 ]2 -PEO). Solutions of

clay was repeatedly washed, centrifuged and separated till

the supernatant showed no further decrease in conductivity, measured amounts of the matrix polymer POEM were also

indicating the removal of all excess cations from the clay made in DI water. The polymer solution was then added to a

[45]. The puri ed and exchanged clay was freeze-dried for a measured amount of the clay suspension with rapid stirring

week to remove all traces of water which could be adsorbed to ensure maximum contact between the polymer chains and

on the clay layers and possibly interfere with the battery per- clay platelets. This mixture was stirred for 24 h, after which it

formance in the applications we seek to investigate. The dried was evaporated to dryness on a rotary evaporator and further

dried under vacuum at 60 C for 24 h. Two sets of composites,

clay was then powdered using a mortar and pestle.

2128 M. Kurian et al. / Electrochimica Acta 50-200*-****-****

each incorporating PSNs with 10, 15 and 20 wt.% clay, were with valves, served as inlet and outlet for argon or nitrogen

made using the two different organically modi ed clays. ow. The last port accommodated an O-ring sealed glass tube

through which a type-K thermocouple was fed and positioned

directly next to one of the stainless steel electrodes. The use of

3. Characterization compression ttings rendered the cell cap vacuum tight. The

sealed cell was removed from the glove box and annealed at

70 C for 12 h, blanketed by a continuous ow of dry, grade

The cation exchange capacity for the native sodium mont-

morillonite clay and the exchanged lithium montmorillonite 5.0 argon. Conductivity measurements were carried out over

a temperature interval spanning 20 70 C. The specimen was

were determined using standard soil chemical analysis tech-

niques [46 48]. protected at all times during the experiment by owing dry

As described in the experimental procedures, all nanocom- nitrogen gas.

posite fabrication and handling was done in the controlled at- Lithium symmetric cells tted with the PSN electrolytes

mosphere of an argon- lled glove-box (moisture and oxygen were used to obtain a measure of lithium transference num-

levels were measured to be less than 1 ppm), since resid- ber. The electrodes were active to lithium ions but blocked the

ual water in the electrolytes would greatly affect the ionic anions; upon application of a stepped potential, the initial cur-

conductivity of the electrolytes. Thermogravimetric analysis rent (I0 ) re ected both cation and counter ion contributions

(TGA) was used to verify that the nanocomposites were com- whereas the long-term, steady-state value (Iss ) arose solely

pletely dry and moisture-free. A sample 5 10 mg in weight from lithium ion motion. The transference number was then

was analyzed using a Perkin Elmer TGA 7 instrument by taken to be Iss /I0 .

recording the weight loss of the sample as a function of tem- To determine the effect of clay content on the mechani-

perature in the range 30 900 C. The presence of water in cal performance of the PSN electrolytes, rheological charac-

the sample can be determined by the weight loss at 100 C, terization was performed using an ARES rheometer (Rheo-

while the amount of residue that remains at the end of the metrics, Inc.) with a parallel plate xture. Temperature was

maintained at 25 C during the measurements using a Peltier

experimental run provides a measure of the inorganic or ash

content in the material. system. Samples were pressed to a gap width of 1 mm under

Small angle X-ray scattering (SAXS) was used to monitor a stable normal force of approximately 100 g. The complex

the gallery height of lithium montmorillonite in the different shear modulus, G = G + iG, was measured as a function

nanocomposites. Room temperature scans were collected in of frequency by dynamically shearing the sample at a xed

transmission mode using a Rigaku Roto ex diffractometer strain rate over the low frequency regime 0.1 10 rad/s.

(Rigaku/USA Inc.), equipped with a rotating anode X-ray

generator (operated at a 30 kV 100 mA setting) with a Cu tar-

get and a Bruker 2D area detector (Bruker Analytical X-ray 4. Results and discussion

Systems). Samples (5 mm diameter, 2 mm thick), sandwiched

using Kapton* polyimide lm tape (Precision PCB Services, The molecular weight and polydispersity of the synthe-

Inc.) were mounted on a custom-made vertical sample posi- sized POEM polymer were measured to be approximately

tioning stage situated in an evacuable beam path to reduce 100 kg/mol and 2.1, respectively, by gel permeation chro-

background scattering due to air. Background subtractions matography using polystyrene calibration standards. The

were made for all sample spectra. glass transition temperature (Tg ) was determined by differ-

ential scanning calorimetry (DSC) to be 65 5 C.

Transmission electron microscopy (TEM) was used to di-

rectly observe the nanocomposite morphology. Since POEM The cation exchange capacity (CEC) of the clay was de-

has a low Tg 65 C, sections were microtomed from bulk termined to be 67 meq/100 g clay using standard soil chemi-

samples under cryogenic conditions using a Leica Ultracut cal analysis techniques [46 48]. Since these values are fairly

ultramicrotome (Leica Inc.). Sections 50 nm thick were dependent on the exact technique and soil handling method

mounted on 1000 mesh carbon coated grids (Ted Pella Inc.) used, we can consider it to be in good agreement with the

for observation. Microscopy was performed on a JEOL 2010F manufacturer s reported value of 77 meq/100 g clay. We use

TEM, operated at an accelerating voltage of 200 kV. the measured value of the CEC in all relevant calculations.

Electrical conductivities of the PSNs were measured by TGA results for the nanocomposite electrolytes showed

impedance spectroscopy using a Solatron 1260 impedance no detectable weight loss that could be associated with the

gain/phase analyzer (Solatron Instruments, Allentown, PA). presence of either residual water or organic solvents from

The nanocomposite electrolytes were placed between a pair the fabrication techniques used, thus con rming that the

of blocking electrodes made of 316 stainless steel and pressed materials are dry and measured materials properties are

to a thickness of 250 m. The electrode assembly was then not in uenced by such contaminations. A nanocomposite

placed inside a brass cell equipped with seven hermetically with 2 wt.% clay exhibited a decomposition temperature of

400 C, roughly 50 C above that of neat POEM.

sealed ports. To maintain dry conditions, the sample prepa-

ration was carried out in an argon- lled glove box (moisture X-ray diffraction is an ideal tool to study the structure of

level measured to be less than 2 ppm). Two ports, equipped clay minerals [49]. In order to monitor the interactions of

M. Kurian et al. / Electrochimica Acta 50-200*-****-**** 2129

the silicate with the polymer, the gallery height of the clay

was measured as a function of POEM content. SAXS plots

of intensity versus angle of scattering, 2, for the neat clay

and the different nanocomposites, along with corresponding

d-spacings, are shown in Fig. 3.

These results give direct evidence of the formation of in-

tercalated morphologies in the PSNs. There is an increase

in the spacing of (0 0 1) planes from about 10 A in the na-

tive lithium montmorillonite clay to between 17 and 18 A in

the 2, 5, and 10 wt.% clay nanocomposites due to favorable

interactions between the polar polymer molecules and the

hydrophilic silicate sheets. This primary spacing is similar to

values seen in other reports, obtained for the intercalation of

PEO into layered silicates [36,37,50 53]. With 15 wt.% or

more clay incorporation, a second peak appears at 2 2.8

corresponding to a d-spacing of about 31 A. These results

suggest the formation of a superlattice structure of tactoids

containing two silicate layers at higher clay loadings. Ogata et

al. [54] observed a similar low angle peak in studies of PEO-

intercalated montmorillonite containing 5 15 wt.% organo-

modi ed clay. In their case, the corresponding long period

was 70 A, suggesting a possible superlattice arrangement

Fig. 3. Room temperature X-ray scattering data for POEM Li+ montmoril-

of 4-layer tactoids [54]. Hence at higher clay loadings, bun- lonite nanocomposites.

dles of sheets, which are approximately 100 200 nm in lateral

dimension, apparently interact to form a superstructure.

Fig. 4. TEM micrographs for POEM Li+ montmorillonite nanocomposites with different clay loadings: (a) low (5 wt.%) clay content, (b) and (c) high (15 wt.%)

clay content.

2130 M. Kurian et al. / Electrochimica Acta 50-200*-****-****

to 28 wt.% PEO in montmorillonite. For higher polymer con-

tents, a PEO-rich phase was also observed.

Measured values of the electrical conductivities for the

PSN electrolytes in the temperature range 20 70 C, are

shown in Fig. 5. The results show that there is up to an

order of magnitude increase in conductivity of neat POEM

with the addition of 2 and 5 wt.% clay, the highest value of

3.75 10 7 S/cm being obtained in the former case at a tem-

perature of 70 C. At 10 wt.% clay incorporation, the compos-

ite performs better than neat POEM at higher temperatures

(>40 C) but its performance decays to below that of POEM

at lower temperatures. With 15 wt.% and higher clay content

(conductivity data for the 20 and 25 wt.% PSNs were identi-

cal), the conductivities fall to lower values, distinctly below

that of the neat polymer.

As discussed in the scattering analysis, there seems to

be a transition in the morphology associated with systems

incorporating 15 wt.% clay, which may account for the de-

Fig. 5. Temperature dependence of electrical conductivity for POEM Li+

cay in electrical properties. Discrepancies between the shape

montmorillonite nanocomposites of varying clay content.

of the impedance spectra of the low and high clay-content

PSNs (Fig. 6) suggest that different conduction mechanisms

TEM micrographs yield more information about the mor-

are present in the two electrolytes. Data obtained from the

phologies formed in the nanocomposite electrolytes. Repre-

2 wt.% clay sample appear semicircular (Fig. 6(a)), and can

sentative micrographs for the nanocomposites incorporating

be t well using a common equivalent circuit consisting of a

low (5 wt.%) and high (15 wt.%) clay contents are shown in

resistor and capacitor in parallel. In contrast, the spectrum for

Fig. 4. At lower clay loadings (Fig. 4(a)), in addition to inter-

the 25 wt.% clay electrolyte exhibits a depressed semicircle

calated regions, the presence of a signi cant number of indi-

(Fig. 6(b)), characteristic of a system where more than one

vidual delaminated clay platelets dispersed homogeneously

conduction process is present simultaneously. TEM micro-

in the polymer matrix indicates that some exfoliation has

graphs show that, at higher clay loading, uniform dispersion

occurred in the nanocomposites. In contrast, at higher load-

of the clay in the POEM matrix is more dif cult to achieve,

ings, a limited number of exfoliated clay sheets are seen, but

leading to the formation of phase-separated morphologies.

more prominently, bundles of intercalated tactoids (Fig. 4(b))

This is expected to affect the conductivity of the system, since

are non-homogeneously dispersed in the polymer matrix

the Li+ carriers are no longer uniformly distributed through-

(Fig. 4(c)), indicating a tendency towards possible phase sep-

out the material. Moreover, the clay platelets could be acting

aration. The formation of such phase-separated materials at

as physical barriers to the effective motion of carriers, thus

intermediate polymer contents is consistent with recent stud-

leading to a decrease in conductivity of the electrolyte below

ies by Shen et al. [55] on the saturation limit of PEO in mont-

that of POEM.

morillonite clay. These authors observed intercalation of up

Fig. 6. Impedance plots at 70 C for PSNs with: (a) 2 wt.% clay and (b) 25 wt.% clay content.

M. Kurian et al. / Electrochimica Acta 50-200*-****-**** 2131

Evidence for single-ion conduction in these PSN elec- (Table 1). Further work is needed to assess the promise of

trolytes was obtained from transference number measure- these systems as battery electrolytes, focusing on the de-

velopment of oriented morphologies with a higher [Li+ :EO]

ments. For example, in measurements of the 2 wt.% clay

nanocomposite electrolyte using a symmetric lithium cell, the ratio.

Given the low glass transition temperature (Tg 65 C)

application of a stepped potential led to a simultaneous jump

in the current that was subsequently maintained. Less than of the POEM matrix and the resulting gel like nature of the

1% current decay was observed, indicating that anions were nanocomposites at room temperature [56], one of the best

effectively immobile over the course of the measurement, methods of characterizing the mechanical behavior of this

yielding Iss /I0 = tLi+ 1. To con rm that this current arose system is through rheological measurements. Fig. 7 shows

from ionic rather than electronic processes, a second stepped plots of the complex moduli for POEM and the nanocom-

potential measurement using stainless steel blocking elec- posites as a function of frequency in the low frequency

trodes was performed, which resulted in a 92% current decay regime. Rheological studies indicate an enhancement in the

arising from the blocking of Li+ motion at the electrodes,

demonstrating that the majority of the current is carried by

these ions. Hence these materials can be viewed as single-ion

conductors.

Conductivities between 10 4 and 10 5 S/cm have been

reported for [clay/PEO] [36] nanocomposites (70 75 wt.%

clay, 250 C). [PEO/NH4 + smectic] [50] composites with

similarly high inorganic contents had a conductivity

10 7 S/cm in the temperature range 125 280 C, while

other [clay/polymer] systems [39] exhibited conductivities in

the range of 10 8 to 10 10 S/cm (10 wt.% clay, 85 110 C).

Others report conductivities of 1.6 10 6 S/cm (60 wt.%

clay, 30 C) [37] and 10 5 S/cm (75 wt.% clay, RT) [38]

for [PEO/clay] nanocomposites. As can be seen from the

preceding discussion, other studies dealing with similar

dry electrolyte systems incorporate signi cantly higher

amounts of clay (60 70 wt.%) and hence a larger concen-

tration of carriers in comparison with the current system

under consideration. Most of these studies also measure con-

ductivities at much higher temperatures. Though the highest

values of conductivity reported herein are low in compari-

son with many PEO-based electrolytes reported in the lit-

erature, it must be noted that the systems described in

this work are salt-free and much lower in [Li+ :EO] ratios

Fig. 8. Room temperature X-ray scattering data for POEM/Li-

montmorillonite nanocomposites fabricated using clays modi ed to

Fig. 7. Room temperature complex moduli for POEM and nanocomposites. different extents with diamine-terminated PEO.

2132 M. Kurian et al. / Electrochimica Acta 50-200*-****-****

mechanical properties of POEM with the incorporation of ing. Fig. 9 compares the scattering data for a representa-

clay. We observe an increase in the complex modulus with tive nanocomposite incorporating unmodi ed clay (15 wt.%

increasing clay content by up to six orders of magnitude at clay) with PSNs having similar inorganic contents but in-

a loading of 20 wt.%. The loss modulus G is greater than corporating clay modi ed to different extents with [NH2 ]2 -

the storage modulus G at lower clay loadings, with a re- PEO. The use of [NH2 ]2 -PEO modi ed clay appears to dis-

versal in the trend at higher ( 15 wt.% clay) loadings. This rupt the superlattice structure seen for the unmodi ed sys-

reversal, indicative of a switch from liquid-like to solid-like tem, as evidenced by the disappearance of the low angle

re ection at 2.8 . The results indicate that incorporation of

behavior, is expected with increasing clay content. It is in-

teresting to note that in these measurements too, the tran- [NH2 ]2 -PEO in the PSNs improves the dispersion of the clay

sition in the mechanical behavior of the PSNs is strongly platelets in the POEM matrix, yielding morphologies simi-

linked to the clay content and occurs at 15 wt.% clay loading, lar to those observed at lower clay loadings in unmodi ed

similar to the morphological and electrical characteristics. PSNs.

The rheology resembles that exhibited by a strongly gelled The change in morphology due to clay modi cation has a

pseudo-solid network [57]. The moduli are fairly constant corresponding effect on the electrical properties of the mate-

in the frequency regime employed and the small slope of rial. Electrical conductivities for the [NH2 ]2 -PEO modi ed

the curves is indicative of the high strength of the compos-

ites.

From the above ndings, the electrolyte properties ap-

pear closely linked with clay content. At higher loadings, the

clay is non-uniformly dispersed in the matrix and may form

physical barriers to the motion of the charge carriers. In an

attempt to promote more uniform dispersion of the clay at

higher loadings, nanocomposites incorporating organically

modi ed Li-mmt were fabricated, where the modi cation of

the clay is expected to aid its dispersion in the POEM matrix.

Two sets of nanocomposites incorporating clay modi ed with

diamine-terminated PEO to different extents (2 and 11 wt.%)

were made. For each set, PSNs with clay contents of 10, 15,

20 wt.% were fabricated.

SAXS data for the two modi ed clays and the corre-

sponding sets of PSNs are shown in Fig. 8(a) and (b). The

data exhibit a single re ection corresponding to a spacing of

about 17.7 A, slightly larger than the spacing for nanocom-

posites prepared with unmodi ed clay at the same load-

Fig. 10. Temperature dependence of electrical conductivity for POEM/Li-

montmorillonite PSNs fabricated using clays modi ed by diamine-

Fig. 9. X-ray scattering data for PSNs (15 wt.% clay content) use of un-

terminated PEO: (a) clay modi ed with 2 wt.% [NH2 ]2 -PEO; (b) clay mod-

modi ed clay vs. clay modi ed to different extents with diamine-terminated

i ed with 11 wt.% [NH2 ]2 -PEO.

PEO.

M. Kurian et al. / Electrochimica Acta 50-200*-****-**** 2133

employed for such measurements, we did not undertake them

here due to concerns over thermal degradation.

5. Conclusions

Solid-state electrolytes based on polymer silicate

nanocomposites have been fabricated using an amorphous

PEO-based polymer, POEM, and lithium montmorillonite

clay. The incorporation of the silicate in the POEM matrix

leads to a considerable enhancement in mechanical prop-

erties. Scattering studies indicate that the nanocomposites

primarily exhibit an intercalated morphology with the

formation of an apparent superlattice structure at higher

clay contents. At lower loadings, TEM results suggest the

clay is more homogeneously distributed in the matrix in

comparison with higher clay loadings. The presence of

completely exfoliated sheets at lower loadings indicates that

Fig. 11. Temperature dependence of electrical conductivity for PSNs

partial exfoliation has occurred in these materials, whereas

(15 wt.% inorganic content) use of unmodi ed clay vs. clay modi ed to

at higher loadings, non-uniformly distributed intercalated

different extents with diamine-terminated PEO.

tactoids are more predominant. Partial modi cation of the

clay by an organic modi er appeared to aid the dispersion

clay PSNs, measured by impedance spectroscopy are shown of the clay at higher loadings, consequently enhancing elec-

in Fig. 10(a) and (b). Compared with the PSNs incorpo- trical performance. The effect of the level of modi cation

rating unmodi ed clay, an increase in conductivity is seen of the clay on electrical properties, though not explicitly

with the use of modi ed clays. Representative data for investigated in the current report, could be an important

nanocomposites incorporating 15 wt.% clay are shown in parameter, which could be controlled to create electrolytes

Fig. 11, where electrical conductivities for 15% PSN (un- with optimal properties.

modi ed clay) and POEM, are contrasted with the conduc- Although electrical conductivities of the PSNs investi-

tivities for 15% PSN (2 wt.% [NH2 ]2 -PEO) and 15% PSN gated in this work are low in comparison with conventional

(11 wt.% [NH2 ]2 -PEO). Similar behavior is seen for the 10 polymer electrolytes, these materials are salt-free and incor-

and 20 wt.% clay systems. Inspection of these results indi- porate signi cantly lower lithium ion concentrations. We ex-

cates no systematic trend in the electrical properties with pect that the formation of a fully exfoliated system, perhaps

clay content. It is expected that properties would depend on through the use of shear, would lead to improved electrical

both the extent of modi cation of the silicate by the organic properties, as would the use of clays having higher cation

modi er, as well as the subsequent level of clay loading in exchange capacities and hence Li+ concentration.

the nanocomposite. It should be noted that the two levels of

modi cation of the clay by the diamine-terminated PEO mod-

i er (2 wt.% [NH2 ]2 -PEO, i.e. exchange of 10% of CEC, and Acknowledgments

11 wt.% [NH2 ]2 -PEO, i.e. exchange of 50% of CEC), were

chosen arbitrarily and no attempt was made to optimize the The authors thank Dr. Chaoying Ni and Mr. Frank Kriss

level of modi cation. The results seen here could indicate from the W.M. Keck Electron Microscopy Facility at the Uni-

that there exists an optimal level of organic modi cation of versity of Delaware for their help with transmission electron

the clay with regard to electrical properties. microscopy studies and Dr. Sudhir Shenoy for assistance with

Transference number measurements on the modi ed clay the rheological experiments. M.K. gratefully acknowledges

PSNs gave similar results to those obtained for the un- partial funding by the University of Delaware Competitive

modi ed clay/POEM nanocomposites, with Li+ carriers ac- Fellowships. This work was sponsored by the Of ce of Naval

counting for 90% of the current, and the rest arising from Research through Grants N00014-00-1-0356 and N00014-

electronic processes. 02-1-0226.

The use of modi ed clays also enhanced the mechani-

cal properties of the PSNs beyond those of the unmodi ed

clay systems. Attempts to quantify the mechanical properties

References

of modi ed clay nanocomposites with dynamic rheological

measurements in the frequency range used for the unmodi- [1] F.M. Gray, Solid Polymer Electrolytes: Fundamentals and Techno-

ed clay PSNs were unsuccessful due to the more solid-like logical Applications, VCH Publishers, Inc., New York, 1991.

nature of these systems. While higher temperatures could be [2] P. Lightfoot, M.A. Mehta, P.G. Bruce, Science 262 (1993) 883.

2134 M. Kurian et al. / Electrochimica Acta 50-200*-****-****

[3] F. Chia, Y. Zheng, J. Liu, N. Reeves, G. Ungar, P.V. Wright, Elec- [28] K. Yano, A. Usuki, A. Okada, T. Kurauchi, O. Kamigaito, J. Polym.

trochim. Acta 48 (2003) 1939. Sci. Polym. Chem. 31 (1993) 2493.

[4] D. Fauteux, A. Massucco, M. McLin, M. Vanburen, J. Shi, Elec- [29] P.B. Messersmith, S.I. Stupp, J. Mater. Res. 7 (1992) 2599.

trochim. Acta 40 (1995) 2185. [30] Y. Kojima, A. Usuki, M. Kawasumi, A. Okada, T. Kurauchi, O.

[5] P.P. Soo, B.Y. Huang, Y.I. Jang, Y.M. Chiang, D.R. Sadoway, A.M. Kamigaito, J. Polym. Sci. Polym. Chem. 31 (1993) 983.

Mayes, J. Electrochem. Soc. 146 (1999) 32. [31] R. Krishnamoorti, R.A. Vaia, E.P. Giannelis, Chem. Mater. 8 (1996)

[6] H.R. Allcock, P.E. Austin, T.X. Neenan, J.T. Sisko, P.M. Blonsky, 1728.

D.F. Shriver, Macromolecules 19 (1986) 1508. [32] H.Z. Shi, T. Lan, T.J. Pinnavaia, Chem. Mater. 8 (1996) 1584.

[7] K.M. Abraham, M. Alamgir, S.J. Perrotti, J. Electrochem. Soc. 135 [33] Z. Wang, T.J. Pinnavaia, Chem. Mater. 10 (1998) 3769.

(1988) 535. [34] Z. Wang, T.J. Pinnavaia, Chem. Mater. 10 (1998) 1820.

[8] J.S. Tonge, D.F. Shriver, J. Electrochem. Soc. 134 (1987) 269. [35] S.D. Burnside, E.P. Giannelis, Chem. Mater. 7 (1995) 1597.

[9] D.W. Xia, J. Smid, J. Polym. Sci. Polym. Lett. 22 (1984) 617. [36] P. Aranda, E. Ruizhitzky, Chem. Mater. 4 (1992) 1395.

[10] J.M.G. Cowie, R. Ferguson, J. Polym. Sci. Polym. Phys. 23 (1985) [37] R.A. Vaia, S. Vasudevan, W. Krawiec, L.G. Scanlon, E.P. Giannelis,

2181. Adv. Mater. 7 (1995) 154.

[11] J.M.G. Cowie, A.C.S. Martin, Polym. Commun. 26 (1985) 298. [38] B. Liao, M. Song, H.J. Liang, Y.X. Pang, Polymer 42 (2001) 10007.

[12] F.M. Gray, J.R. Maccallum, C.A. Vincent, J.R.M. Giles, Macro- [39] M. Okamoto, S. Morita, T. Kotaka, Polymer 42 (2001) 2685.

molecules 21 (1988) 392. [40] M. Riley, P.S. Fedkiw, S.A. Khan, J. Electrochem. Soc. 149 (2002)

[13] J.R. Maccallum, M.J. Smith, C.A. Vincent, Solid State Ionics 11 A667.

(1984) 307. [41] G.J. Moore, M.S. Whittingham, Mater. Res. Soc. Symp. Proc. 548

[14] M. Watanabe, S. Nagano, K. Sanui, N. Ogata, Polym. J. 18 (1986) (1999) 173.

809. [42] H.W. Chen, F.C. Chang, Polymer 42 (2001) 9763.

[15] M. Andrei, L. Marchese, A. Roggero, P. Prosperi, Solid State Ionics [43] J.J. Hwang, H.J. Liu, Macromolecules 35 (2002) 7314.

72 (1994) 140. [44] D.R. Sadoway, B.Y. Huang, P.E. Trapa, P.P. Soo, P. Bannerjee, A.M.

[16] A. Killis, J.F. Lenest, H. Cheradame, A. Gandini, Makromol. Chem. Mayes, J. Power Sources 97 (8) (2001) 621.

183 (1982) 2835. [45] H. Van Olphen, An Introduction to Clay Colloid Chemistry, Inter-

[17] H. Cheradame, J.F. LeNest, in: J.R. Maccallum, C.A. Vincent science Publishers, New York, 1963.

(Eds.), Polymer Electrolyte Reviews, vol. 1, Elsevier, London, 1987, [46] J.D. Rhoades, Cation Exchange Capacity, vol. 2, American Society

p. 103. of Agronomy, Madison, WI, 1982.

[18] K.M. Abraham, in: B. Scrosati (Ed.), Applications of Electroactive [47] D.S. Ross, Recommended Methods for Determining Soil Cation Ex-

Polymers, Chapman & Hall, London, 1993, p. 75. change Capacity, vol. 2, Northeastern Regional Publication No. 493

[19] I.E. Kelly, J.R. Owen, B.C.H. Steele, J. Power Sources 14 (1985) Revised, University of Delaware Experimental Station, Newark, DE,

13. 1995.

[20] J.E. Weston, B.C.H. Steele, Solid State Ionics 7 (1982) 75. [48] G.W. Thomas, Exchangeable Cations, vol. 2, American Society of

[21] F. Capuano, F. Croce, B. Scrosati, J. Electrochem. Soc. 138 (1991) Agronomy, Madison, WI, 1982.

1918. [49] D.M. Moore, R.C. Reynolds Jr., X-ray Diffraction and the Identi -

[22] F. Croce, G.B. Appetecchi, L. Persi, B. Scrosati, Nature 394 (1998) cation and Analysis of Clay Minerals, Oxford University Press, New

456. York, 1997.

[23] F. Croce, R. Curini, A. Martinelli, L. Persi, F. Ronci, B. Scrosati, [50] P. Aranda, E. Ruiz-Hitzky, Appl. Clay Sci. 15 (1999) 119.

R. Caminiti, J. Phys. Chem. B 103 (1999) 10632. [51] E. Ruizhitzky, P. Aranda, Adv. Mater. 2 (1990) 545.

[24] W. Wieczorek, Z. Florjanczyk, J.R. Stevens, Electrochim. Acta 40 [52] P. Aranda, E. Ruizhitzky, Acta Polym. 45 (1994) 59.

(1995) 2251. [53] J.H. Wu, M.M. Lerner, Chem. Mater. 5 (1993) 835.

[25] W. Krawiec, L.G. Scanlon, J.P. Fellner, R.A. Vaia, E.P. Giannelis, J. [54] N. Ogata, S. Kawakage, T. Ogihara, Polymer 38 (1997) 5115.

Power Sources 54 (1995) 310. [55] Z.Q. Shen, G.P. Simon, Y.B. Cheng, Eur. Polym. J. 39 (2003) 1917.

[26] A.S. Best, A. Ferry, D.R. MacFarlane, M. Forsyth, Solid State Ionics [56] F. Yan, P. Dejardin, Y. Frere, P. Gramain, Makromol. Chem. 191

126 (1999) 269. (1990) 1197.

[27] C. Capiglia, P. Mustarelli, E. Quartarone, C. Tomasi, A. Magistris, [57] R.G. Larson, The Structure and Rheology of Complex Fluids, Oxford

Solid State Ionics 118 (1999) 73. University Press, New York, 1999.

materials with enhanced performance.

© 2004 Elsevier Ltd. All rights reserved.

Contact this candidate