Printer Friendly

Lithium ion transport through nonaqueous perfluoroionomeric membranes.

INTRODUCTION

The development of advanced secondary (rechargeable lithium batteries has generated a tremendous amount of research in recent years. Most of the work involves constructing improved electrodes. A smaller effort has been devoted to developing polymeric rigid or gelled separators doped with lithium salts. These lithium polymer batteries have obtained good performance in the lab, but improvements, especially in developing more efficient separators, are needed to achieve the high theoretical energy output capability of these systems.

Research on separators for lithium cells initially focused on inert polymer matrices imbibed with a solvent containing a dissociable lithium salt. The polymer acts to physically separate the two electrodes and itself has negligible ionic conductivity: only when the solvent and salt are added to the polymer does the separator system become conductive. The Department of Energy, in conjunction with the United States Advanced Battery Consortium, has established a set of goals for such multiple ion-conducting separators (1). One of the goals, an ionic conductivity of [10.sup.-3] S/cm, has been addressed through various combinations of polymer, solvent, salt, and preparation methods. Some examples include poly(vinylidene fluoride) (PVdF] homopolymer cast by evaporation of methyl ethyl ketone from a heated solution and then imbibed with [LiBF.sub.4] in PC (2), poly(vinyl chloride-co-vinyl acetate) cast from a heated solution of NMP and EC with [LiCIO.sub.4] (3), and thermal extrusion of PVdF in a EC/PC solution with LiN([SO.sub.2][CF.sub.3]).sub.2] (4). All chemical abbreviations are listed in Table 1. Although one goal may be attained, meeting all of the goals simultaneously is still a difficult and elusive challenge.

One important property not being addressed by the present multiple ion separator systems is a high transference number ([t.sub.+]), defined as the fraction of current carried by the active species. A polymer system may be an excellent overall conductor, but it must specifically conduct the species of interest ([Li.sup.+]). A significant disadvantage of the present binary salt separator systems is that [t.sub.+] is typically between 0.1 and 0.4: i.e., only 10 to 40% of the current is carried by the Li cations; the anions (e.g., [PF.sub.6.sup.-] carry the remainder of the current. The anionic current is wasted since the anions do not react at the electrodes and, therefore, do not aid in charging the battery. The lower the transference number, the lower the efficiency and capacity of the battery.

A fundamental change in the current lithium polymer battery configuration is presented herein. The idea is to use lithiated ion-conducting fluoropolymers formed into membranes possessing dissociable lithium ions to replace the current doped polymer separators. Ion-exchange polymers consist of a charged species covalently bound to the backbone or sidegroup of the polymer. An oppositely charged ion forms the salt with the bound charge. In the presence of an electric field and an appropriate solvent, these free charges are forced to migrate in the direction of the electric field. Since the anions are bound to the polymer, the only mobile species are the [Li.sup.+] ions resulting in a transference number of unity. The membrane, therefore, controls the mass transport of lithium ions through the polymer microstructure. It has been proposed that an ionomer membrane can actually have an ionic conductivity an order of magnitude smaller than one using a binary salt separator and still achieve similar performance since the transference numbers are dramatically different (5).

Solid polymer electrolytes or ion-exchange membranes serve as both a separator and an electrolyte in a battery. Research on ionomers in the past has been almost exclusively done in aqueous systems. Theories to date suggest that water swells the polymer and promotes conduction within and between ion clusters. In nonaqueous systems, the solvent must be able to swell the polymer in a similar manner, initiate dissociation of the cation, and promote ion conduction.

Several properties of ionomer membranes are desired (5, 6): ionic conduction above [10.sup.-4] S/cm: no electronic conduction: mechanical and thermal stability over a wide temperature range; ability to withstand chemical attack by lithium and electrochemical potentials up to 4.5 V; [t.sub.+] of unity; no degradation over thousands of cycles; be simple and inexpensive to make; and safe to use. These properties are not independent of one another and are related to polymer properties, such as molecular weight (MW), MW distribution, equivalent weight (EM), EW distribution, chemistry, morphology, crystallinity, cluster size, and thermal, mechanical, and chemical properties. The goal of the current research is to present experimental data aimed at understanding these relationships in order to develop the next generation ion-exchange membrane suitable for use in advanced energy storage devices.

There have been different approaches to developing an ionomeric membrane for use in lithium cells (7-9). Some even surpass the goal of [10.sup.-4] S/cm when imbibed with an organic plasticizing solvent. Examples include the free-radical copolymerization of N,N-dimethylacryl amide and lithium 2-acrylamido-2-methyl-1-propane sulfonate with tetraethyleneglycol diacrylate as a crosslinking agent (10) or alloying a synthesized lithiated salt of poly(2-oxo-1-difuluoroethylene sulfonylimide) in a poly(ethylene oxide-co-propylene oxide) triacrylate and poly(ethylene oxide-co-propylene oxide) monoacrylate polyether matrix with 2,2-dimethoxy-2-phenyl acetophenone as a photo-initiator (11). However, in addition to meeting all of the transport, stability, and strength requirements, another issue to consider is the ease of transition into a potential full-scale production product (12).

Instead of developing an entirely new polymeric structure, perhaps existing polymers hold the key. Perfluorinated sulfonate ionomers consist of a polymer with a fluorocarbon mainchain and pendant ionomeric sidegroups. These sidegroups are normally terminated with either sulfonate or carboxylate structures which can be ion-exchanged to a variety of ionic groups. The polymeric chain is known to be very stable, both chemically and mechanically. Some examples of the use of ionomers in lithium electrochemical cells are given below.

Sata (13) swelled a cation exchange membrane with PC and coupled the membrane with a piece of lithium foil; the cell exchanged [Li.sup.+] ions at 3.0 to 3.2 V but displayed a high self discharge. Li et al. (14) combined Nafion[R] and polyaniline to make a composite film. A [LiClO.sub.4] salt was used in PC. The charge/discharge performance was related to the cation size.

Mori et al. (15) coated a Pt sheet with a 200 [micro]m Nafion[R] layer. The electrode was electrolyzed in a solution containing 0.1 M [P.sub.h][NH.sub.2] and 0.2M HCl to deposit 100 [micro]m polyaniline in the Nafion[R] layer. A lithium battery was made using this cathode and a [LiBF.sub.4]/PC electrolyte, and was subjected to 100 charge/discharge cycles at high energy density. Mori (16) also used a polypyrrole/Nafion[R]/Li-Pt system to make a lithium battery.

The use of Nafion[R]-1100 [Li.sup.+] membranes imbibed with PC was investigated independently by Armand (17) and Aldebert et al. (18), which resulted in ionic conductivities of 2.05 X [10.sup.-5] S/cm and 6 X [10.sup.-5] S/cm, respectively. Noda and Yoshihisa (19) used a Nafion[R] membrane containing -[SO.sub.3]Li groups with a [LiCIO.sub.4] containing hexamethylene diisocyanatecrosslinked polyether triol salt. The cathode was an amorphous [V.sub.2][O.sub.5] sheet containing the same polymer electrolyte.

Genies (20) produced a battery using a Nafion[R] membrane, an electrically conductive polymer anode (e.g., polypyrrole and ferrocene in [LiClO.sub.4] containing methyl cyanide PC electrolyte), and a Li-plated stainless steel grid cathode. The battery produced 1 mA/[cm.sup.2] at 3.3 V.

Perusich et al. (21-24) enhanced the conductivity of lithiated Nafion[R] based membranes in nonaqueous solvents without the use of a lithium salt. Through treatments of the polymer membrane, the ionic conductivity was increased by two orders of magnitude over previous studies.

The general patent literature on salt containing separators and ionomers for battery applications is summarized by references (25-35).

EXPERIMENTAL APPARATUS AND PROCEDURES

The four-point probe impedance technique was used for determining the ionic conductivity of the perfluorinated ionomer membranes. The electrical measurements discussed herein involve the application of an alternating voltage to a cell containing a solid polymer electrolyte. Measurement of the resulting alternating current passing through the electrolyte allows computation of the ionic conductivity. Note that the names and properties of all chemicals are given in Tables 1 and 2.

Polymers

Nafion[R] (DuPont) is a perfluorosulfonic membrane that has a polytetrafluoroethylene backbone and long perfluorovinyl ether pendant sidechains terminated by a sulfonate ionic group. Specifically, Nafion[R] in the sulfonate form is a copolymer of tetrafluoroethylene (TFE) and 2,2,3,3-tetrafluoro-3-[1',2',2'-trifluoro-1'-trifluoromethyl-2'-(l",2 ",2"-trifluoro-ethyloxy) ethoxy] sulfonic acid, the polymeric structure of which is given below

~ [CF.sub.2] - CF - [([CF.sub.2] - [CF.sub.2]).sub.n] ~

O - [CF.sub.2] - [CF([CF.sub.3]) - O - [CF.sub.2] - [CF.sub.2] - [SO.sup.-.sub.3] [Li.sup.+]

where n represents the moles of tetrafluoroethylene monomer units per mole of the vinyl ether sidechain. The polymer composition is expressed in terms of the equivalent weight defined as the weight in grams of polymer containing one equivalent of acid or, alternatively, the comonomer molecular weight plus 100n. Thus, the lower the EW, the lower the TFE concentration, and the higher the vinyl ether concentration. The polymer is semicrystalline since the backbone TFE segments tend to crystallize the structure while the sidegroups hinder crystallization.

The Dow polymer (Dow Chemical), illustrated below, has a much shorter side chain than the Nafion[R] polymer, which alters the crystallinity and ion-exchange capacity of the polymer.

~ [CF.sub.2] - CF - [([CF.sub.2] - [CF.sub.2]).sub.n] ~

O - [CF.sub.2] - [CF.sub.2] - [SO.sup.-.sub.3][Li.sup.+]

Conductivity Measurements

Alternating voltage and current methods represent a reliable approach to the determination of the electrical properties of polymer electrolytes. Inert blocking electrodes are used to determine bulk electrolyte properties. Compared to DC techniques, the equipment required and the theory necessary to interpret the AC techniques are more complex; however, AC measurements yield information about long-range migration of ions and polarization phenomena occurring within the cell.

In an AC measurement, a sinusoidal voltage is applied to a cell, and the sinusoidal current passing through the cell as a result of this perturbation is determined. In a perturbation, two parameters are required to relate the current flowing to the applied potential. One represents the opposition to the flow of charge and is equal to the ratio of voltage and current maxima, [V.sub.max]/[I.sub.max], and is analogous to the resistance in DC measurements. The other parameter is the phase angle ([delta]) between the applied voltage and the measured current. The combination of these parameters represents the impedance (z in [ohm]) of the cell, as defined in Eq 1:

Z = [V.sub.max]/[I.sub.max] = [V.sub.max]/[I.sub.in] + [iI.sub.out] = [V.sub.max]/[I.sub.o]sin[delta] + [iI.sub.o]cos[delta] (1)

where [I.sub.max] is the total current (Amps), [V.sub.max] is the voltage (Volts), [I.sub.in] is the in-phase current with the voltage (Amps), [I.sub.out] is the out-of-phase current with the voltage (Amps), [I.sub.0] is the amplitude of the current (Amps), and [delta] is the phase angle (degrees).

The experimental apparatus consisted of a four-electrode cell for the measurement of the conductivity of perfluorinated membranes, as shown in Fig. 1. Cahan and Wainright (36) reported the use of a four-electrode system for impedance measurements and successfully measured the membrane impedance, which was separated from interfacial capacitance over a wide range of frequency from DC to [10.sup.5] Hz. The four electrode measurement in the present system consists of four gold-plated electrodes mounted on the same side of a Teflon[R] (DuPont) block. The membrane is sandwiched between this block and another block of Teflon[R]. The Teflon[R] blocks are held together using spring clamps. The outer two electrodes serve to apply a sinusoidal potential (at each frequency), and the resulting current passing through the inner two electrodes is measured. This technique is employed to avoid the complications arising from a nonuniform potential field near the outer two electrodes. The conductivity ([[sigma] in 1/[ohm]-cm or S/cm) of the membrane was obtained from the measured impedance by using Eq 2.

[sigma] = {d/zlt' for l<l'

d/zl't' for l>lt} (2)

where z is the measured impedance of the membrane ([ohm]), d is the distance between the inner sensing electrodes (0.2 cm), l is the length of each electrode (1.0 cm), l' is the length of the polymer film (cm), and t' is the film thickness (cm).

A potentiostat (EG&G, Model 273A) was used for the electrochemical analysis. An option board (EG&G, Model 92) allowed the potentiostat to be interfaced with the EG&G PARC electrochemical impedance systems. The electrochemical interface option (EG&G, Model 273A/92) was designed to allow superposition of an externally generated AC excitation signal on the DC signal generated by the potentiostat.

A two-phase lock-in amplifier (EG&G, Model 5210) was used for the impedance studies to generate the excitation signal and to analyze the response. The lock-in amplifier and the potentiostat were interfaced by connecting the multiplexed output of the potentiostat to the lock-in amplifier. The oscillator output of the lock-in amplifier was connected to the AC input of the potentiostat.

EG&G's electrochemical impedance spectroscopic software, Powersine, was used to perform the experiments. Powersine applies an AC excitation to an electrochemical system, measures the response of the system to the excitation, digitizes and stores the resulting data, and allows data display and analysis. Powersine was operated in the single-sine fixed frequency mode. A small amplitude, fixed frequency sinusoidal voltage was applied to the test cell, and the current response was measured. The response signal was then used to determine the in-phase (real) and out-of-phase (imaginary) components of the total impedance as well as the phase shift between the output and the input signals as a function of frequency.

Membrane Pretreatment

The Nafion[R] and Dow membranes were first converted from the -[SO.sub.2]F form to the -[SO.sub.3]H form by immersing the membranes in 1.0 M [HNO.sub.3] prepared from deionized [H.sub.2]O at 25[degrees]C for 30 minutes. Two methods for conversion to the lithium salt form were used and compared. The first method involved boiling the membranes in 1.0 M LiOH in a 1:1 volume ratio solution of DMSO and deionized [H.sub.2]O at 75[degrees]C for 2 hours. The membranes were then rinsed in deionized [H.sub.2]O for 2 hours at 90[degrees]C to 100[degrees]C. The other method involved boiling the membranes in 2.7 M LiOH in a 1:2 volume ratio solution of ethanol and deionized [H.sub.2]O at 82[degrees]C for 2 hours; a similar procedure was performed using methanol. The membranes were then repeatedly rinsed in room temperature deionized [H.sub.2]O. The conversion of the membranes to the salt form was done in a batch reactor with a total condenser. The temperature was controlled using a fitted heating mantle. The membranes wer e then placed in a vacuum oven (VWR) at 120[degrees]C under 20 inches Hg vacuum for over 24 hours. FP-IR studies were performed and compared to standard results to verify the chemical conversion of the membranes (37). The membranes were then transferred to the glove box.

Membrane Measurements

All the measurements were performed in a glove box (PlasLabs, Model 855-AC/SP). The glove box was kept dry and purged with Grade V Argon. The atmosphere in the glove box was circulated through a molecular sieve (4 A) dry train. The moisture content in the glove box was checked using [P.sub.2][O.sub.5], which remained dry during the entire experimental analysis indicating the water content was less than 10 ppm.

A dry membrane sample was weighed (Denver Instrument Co., Model TR 203 balance) and its thickness was determined prior to performing a dry impedance measurement. The polymer equivalent weight was measured with FT-IR (37). The film was then soaked in a nonaqueous solvent or solvent mixture for 10 minutes. All the solvents were used as obtained and had moisture contents of less than 20 ppm. The membrane was removed from the solvent and patted dry to remove excess solvent before weight, thickness, and impedance measurements were taken. The procedure was repeated in 10 minute intervals until an apparent maximum had been obtained. The weight uptake was calculated based on a percentage of the dry membrane weight. The impedance measurements were performed in potentiostatic mode over a frequency range of 10 to 35,000 Hz. The conductivities were calculated according to Eq 2.

DISCUSSION OF RESULTS

Fourier-Transform Infrared

Spectroscopy Studies

The goal of the FT-IR studies of the perfluorinated ionomers was to verify the conversion of the ionomers from either the [SO.sub.2]F or [H.sup.+] form to the [Li.sup.+] salt form. A thorough analysis of the IR behavior of Nafion[R] polymers has been done by Perusich (37). FT-IR scans of the [SO.sub.2]F, [H.sup.+], and [Li.sup.+] forms of Nafion[R] polymers are shown in Fig. 2. Note that in Fig. 2, the absorbance scale is common and relative, and each spectrum is simply offset from one another to show the representative peaks for the various polymeric forms. The conversion from the [H.sup.+] form to the [Li.sup.+] form was verified by the disappearance of the 924 and 1413 [cm.sup.-1] bands and the appearance of the 1630 [cm.sup.-1] band. The conversion from the [SO.sub.2]F form to any of the other forms was verified by the disappearance of the 2704 (not shown in the figure) and 1469 [cm.sup.-1] bands.

The common treatment uses DMSO to swell the polymer. DMSO is an excellent swelling agent, and the polymer can be hydrolyzed completely in under 2 hours. The problem is that residual DMSO remains in the polymer after hydrolysis, as is detected with FT-IR by the doublet around 2950 [cm.sup.-1] and the broad 1050 [cm.sup.-1] peak, and even numerous washings fail to completely remove all the DMSO. In precise conductivity measurements, this residual DMSO will cause a mixed solvent effect. Thus, the measured conductivity will be that of the mixed solvent rather than the solvent of interest.

To alleviate this problem, new procedures using ethanol or methanol as the swelling agents were developed. Complete hydrolysis was achieved in roughly the same time as using the DMSO treatment. Figure 2 shows that the ethanol and methanol treated polymers display the requisite peak at 1630 [cm.sup.-1] and no peaks at 924 or 1413 [cm.sup.-1] indicating complete hydrolysis. Both ethanol and methanol are readily extracted from the polymers leaving no residual trace amounts, thus yielding an extremely solvent-free polymer network to enable accurate conductivity measurements.

Solvent Uptake Studies

Solvent uptake studies were performed on all the ionomer membranes. The time-dependent swelling behavior for the ethanol-lithiated Nafion[R]-1100 (1100 g/equiv is the nominal equivalent weight of the ionomer) membranes imbibed with several nonaqueous solvent mixtures is shown in Fig. 3. The imbibing time in the figure indicates the total time the polymer film was exposed to the solvents.

The PC and PC/EC mixed solvents reach a plateau saturation state after about an hour of immersion. All four of these polymer films attained roughly the same saturation value (~90 wt%). The three other solvent mixtures took a longer time and reached much lower saturation levels. Data on each of the polymer films were also taken at 24 hours, and the plateau values did not change from those shown in Fig. 3. The solvent uptake and swelling of the polymer are dependent on the solvent properties and solvent-polymer interactions, as discussed in later sections.

The saturation values of the Dow membranes in various aqueous and nonaqueous solvents were also measured. To compare the two polymers, their equivalent weights were normalized such that the solvent uptake was expressed as moles of solvent per equivalent (discussed in a later section). Other methods of comparison can also be used, but the conclusion will be unaltered. In general, the solvent uptake capability in the Dow membranes is three to five times the solvent uptake in Nafion[R]. For example, when PC was used as the imbibed solvent, 9.198 mol solvent/equiv were taken up by the Nafion[R] polymer, whereas 31.073 mol solvent/equiv were taken up by the Dow polymer. For methanol, the Naflon[R] uptake was 27.933 mol solvent/equiv, whereas the Dow polymer imbibed 100.625 mol solvent/equiv.

Ionic Conductivity Studies

The lithium ion ([Li+) conductivity through Nafion[R] membranes imbibed with various solvents is shown in Fig. 4. The trends in Fig. 4 are very similar to those in Fig. 3, indicating a direct relationship between the amount of solvent imbibed in the polymer and the ionic conductivity. The ionic conductivity becomes nearly constant after a certain amount of solvation of the ionomer. After this point, the increase in ionic conductivity is not appreciable. The solvent helps to dissociate the lithium ions from the polymer sidechain and creates liquid pathways through the cluster network to allow ion transport.

Figure 5 is a comparison of the long term (saturated polymer) ionic conductivity in lithiated Nafion [R] films in different pure solvents. The polymer ionic conductivity is highest for aqueous solvents and lower for the nonaqueous solvents. The ionic conductivity is independent of frequency for high conductivity solvents, such as [H.sub.2]O, methanol, ethanol, DMSO, and PC. However, for less-conductive solvents, such as THF and acetonitrile, the ionic conductivity varied with frequency at the higher frequencies due to instrumental limitations; only the frequency independent portions of the curves are shown in Fig. 5. In addition, an experiment was performed on extremely dry Nafion [R] (dried over [P.sub.2][O.sub.5]); the resulting conductivity was extremely small and resulted in a linear frequency dependence again indicating measurement limitations.

Ionic conductivity was also measured in perfluorinated Dow membranes in a host of aqueous and non-aqueous so[vents, as shown in Fig. 6. Figure 6 is a plot of the ionic conductivity vs. frequency in the lithiated Dow-805 (805 g/equiv is the nominal equivalent weight of the ionomer) ionomer membrane imbibed with these solvents. The conductivities for the PC and PC/EC imbibed polymers are around [10.sup.-3] S/cm, some of the highest conductivities measured for lithiated ion-exchange polymers in nonaqueous solvents.

In all the solvents, the Dow ionomer membrane has greater swelling and higher conductivity than the Nafion [R] membrane. One of the reasons for this result could be the shorter sidechain length in Dow membranes in comparison to the Nafion [R] membrane. The sidechain length affects the chain packing, crystallinity, and cluster size, all of which affect the swelling and the ionic conductivity.

Ionic Conductivity in Solvent Mixtures

Ionic conductivity was also measured in mixtures of solvents. Figure 7 is a plot of the ionic conductivity in Nafion [R] and Dow membranes imbibed with different nonaqueous solvent mixtures including 2:1, 1:1, and 1:2 volume ratio mixtures of PC/EC, PC/DMC, and PC/DEC. The solvent concentration in the membrane was constant throughout the course of an experiment. Since pure EC is a solid at room temperature, the ionic conductivity could not be measured with the current apparatus; even at PC volume fractions less than 0.3 (> 70% EC), the EC would begin to crystallize in the polymer making the conductivity measurement invalid. Pure DMC and DEC solvents in the Nation [R] polymers had conductivities much too low to be measured with the current 4-point probe technique.

For the PC/DMC and PC/DEC mixtures, the conductivity increased at higher PC concentrations. For the PC/EC mixture, though, the conductivity increased at higher EC concentrations as long as the EC was prevented from crystallizing. The lithium ionic conductivity of any of the mixtures lies in between the conductivities measured in the pure solvents.

Ionomer-Solvent Interactions

The ability of a solvent to imbibe and swell a polymer is thermodynamically related to the change in enthalpy upon mixing ([DELTA][H.sub.mix] in cal/g) of the solvent and the polymer. The enthalpy is in turn related to the solvent and polymer solubility parameters by Eq 3 (38)

[DELTA][H.sub.mix] = [[phi].sub.p][[phi].sub.s]V[([[delta].sub.p] - [[delta].sub.s]).sup.2] (3)

where [[phi].sub.p] is the polymer volume fraction, [[phi].sub.s] is the solvent volume fraction in the polymer, V is the specific volume ([cm.sup.3]/g), [[delta].sub.p] is the polymer solubility parameter ([cal.sup.1/2]/[cm.sup.3/2]), and [[delta].sub.s] is the solvent solubility parameter ([cal.sup.1/2]/[cm.sup.3/2]). For the solvent and polymer to be compatible, thereby maximizing the swelling, the difference in the solubility parameters should be minimized. Since it was shown in previous sections that the swelling and the ionic conductivity are directly related, the difference in solubility parameters must also relate to the conductivity.

The solubility parameters of the ionomers were calculated using group contribution methods (39) to be roughly 16 [cal.sup.1/2]/[cm.sup.3/2]. The solubility parameters of the solvents are given in Table 2. Figure 8 shows that the ionic conductivity is indeed a function of the solubility parameter difference squared or the [DELTA][H.sub.mix] for both the Dow and Nafion [R] polymers. As the difference approaches zero, the ionic conductivity increases.

The data trend, however, is certainly not without scatter. Solubility parameters are based purely on thermodynamic considerations, but other phenomena, such as mass transport, ion dissociation, solvent percolation, viscosity, dielectric constant, among others, occur which couple to the thermodynamic processes. The data for water in Fig. 8 is a dramatic example of how strong hydrogen bonding forces can overwhelm the solubility behavior. In addition, although the effect of dielectric constant on the conductivity is not fully understood, Table 2 shows that a polymer solvated with a solvent of higher dielectric constant (like water and propylene carbonate) is more likely to have a higher conductivity than a polymer solvated with a solvent of low dielectric constant (methanol is an exception to this statement). Research is ongoing to uncover the fundamental relationships involved in nonaqueous ion transport.

Comparison of Ionomer Systems

A comparison of the ionic conductivities of the Dow and Nafion[R] membranes was conducted. Since these polymers have different chemical compositions, a basis of comparison needed to be established. The basis chosen involved selecting polymers with the identical number of TFE segments per vinyl ether chains (comparable concentration of ion-exchange groups). The EW of Nafion[R] ([EW.sub.N]) is computed from adding the molecular weights of n number of TFE segments and one vinyl ether segment per repeat unit.

[EW.sub.N] = l00n + 530 (4)

For an 1100 g/equiv Nafion[R] polymer, n = 5.7. For the Dow polymer, the EW is given by

[EW.sub.D] = l00m + 348 (5)

To compare the polymer based on identical ion-pair concentrations, m=n=5.7. Therefore, a comparable Dow polymer would be one with an equivalent weight ([EW.sub.D]) of 918 g/equiv. Since the Dow polymer is not commercially produced. the choice of [EW.sub.D] polymers is limited. The closest match available was the Dow-805 polymer, which was used as a basis for comparison. Note that other basis of comparisons could also have been used, such as crystallinity, cluster size, and others.

Dow-805 and Nafion[R]-1100 ionomers in the presence of the same solvent are compared in Fig. 9. In all solvents, the ionic conductivity of the Dow polymer is equal to or greater than the ionic conductivity for the Nafion[R] polymer. The same differences shown by some solvents could be attributed to the lower EW used for the Dow polymer as explained above. The large differences for the PC and PC/EC solvents, though, must be due to other factors, such as differences in crystallinity, solubility parameter, EW and MW distributions, among others.

Effect of Equivalent Weight on Ionic Conductivity

The effect of the ionomer equivalent weight on the ionic conductivity was studied in the Nafion[R] and Dow membranes. The ionic conductivity in Nafion[R] 1100, Nafion[R]-1050 Dow-805, and Dow-1030 polymers imbibed with PC or methanol were measured. In all cases, the conductivity increased as the EW decreased.

As the EW is decreased, more ion-pairs are present in the polymer. For the same dissociation level, greater concentrations of [Li.sup.+] ions are present in the low EW polymers. These higher [Li.sup.+] concentrations lead to larger migration and diffusion fluxes through the membrane yielding a higher ionic conductivity.

CONCLUSIONS

The ionic conductivities of perfluoroionomeric polymer membranes were measured in an array of non-aqueous solvents and solvent mixtures. The conductivity increased with solvent content, solvent solubility parameter, and solvent dielectric constant and decreased with the ionomer EW. The conductivity was highly dependent on the solvent or solvent mixture imbibed in the polymer as was the time to achieve saturation of the solvent. Solvent mixtures had conductivities between the two pure solvents. Under similar conditions, the Dow membranes had higher ionic conductivities than the Nafion[R] membranes.

To further the present analysis, other polymer, solvent, and processing factors which could influence the ion transport need to be studied. The most important feature of the present research is that certain solvent/ionomer combinations allow the ionic conductivity to exceed [10.sup.-4] S/cm. Conductivities this large, especially with a transference number of unity with no doped salt, allow these polymers to be viable candidates for practical lithium ionomer battery systems.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]
Table 1

Chemical Nomenclature



Chemical Symbol

Acetone [CH.sub.3][COCH.sub.3]
Acetonitrile [CH.sub.3]CN
Cyclohexane [C.sub.6][H.sub.12]
Diethyl Carbonate DEC
Dimethyl Carbonate DMC
N-N-Dimethylformamide DMF
Dimethyl Sulfoxide DMSO
Ethanol EtOH
Ethylene Carbonate EC
Lithium Hydroxide LiOH
Methanol MeOH
Nitric Acid [HNO.sub.3]
N-methylpyrrolidinone NMP
Phosphorous Pentoxide [P.sub.2][O.sub.5]
Propylene Carbonate PC
Sulfuric Acid [H.sub.2][SO.sub.4]
Tetrahydrofuran THF
Water [H.sub.2]O

 Molecular
 Weight
Chemical Structure (g/mol)

Acetone [CH.sub.3][COCH.sub.3] 58.08
Acetonitrile [CH.sub.3]CN 41.05
Cyclohexane [C.sub.6][H.sub.12] 84.16
Diethyl Carbonate [C.sub.5][H.sub.10][O.sub.3] 118.13
Dimethyl Carbonate [H.sub.3][COCO.sub.2][CH.sub.3] 90.08
N-N-Dimethylformamide HCON[([CH.sub.3]).sub.2] 73.10
Dimethyl Sulfoxide [([CH.sub.3]).sub.2]SO 78.13
Ethanol [CH.sub.3][CH.sub.2]OH 46.07
Ethylene Carbonate [([CH.sub.2]).sub.2][OCO.sub.2] 88.06
Lithium Hydroxide LiOH 41.96
Methanol [CH.sub.3]OH 32.04
Nitric Acid [HNO.sub.3] 63.01
N-methylpyrrolidinone [C.sub.5][H.sub.9]NO 99.13
Phosphorous Pentoxide [P.sub.2][O.sub.5] 141.94
Propylene Carbonate [C.sub.4][H.sub.6][O.sub.3] 102.09
Sulfuric Acid [H.sub.2][SO.sub.4] 98.07
Tetrahydrofuran [C.sub.4][H.sub.B]O 72.11
Water [H.sub.2]O 18.02



Chemical Description

Acetone 99.6% (0.3% [H.sub.2]0)
Acetonitrile Nonaqueous
Cyclohexane HPLC Grade
Diethyl Carbonate Anhydrous 99.98% (15 ppm [H.sub.2]0)
Dimethyl Carbonate Anhydrous 99.98% (15 ppm [H.sub.2]0)
N-N-Dimethylformamide HPLC Grade 99.9+% (<0.03% [H.sub.2]0)
Dimethyl Sulfoxide 99.9% (0.03% [H.sub.2]0)
Ethanol Anhydrous, Denatured
Ethylene Carbonate Anhydrous 99.98% (20 ppm [H.sub.2]0)
Lithium Hydroxide Purified Crystal
Methanol 99.9% (0.01% [H.sub.2]0)
Nitric Acid 69.6%
N-methylpyrrolidinone Peptide Synthesis Grade
Phosphorous Pentoxide 99.6%
Propylene Carbonate Anhydrous 99.98% (15 ppm [H.sub.2]0)
Sulfuric Acid 96.2%
Tetrahydrofuran Anhydrous
Water Deionized, ultrafiltered



Chemical Source

Acetone Fisher
Acetonitrile EM Industries
Cyclohexane Fisher
Diethyl Carbonate EM Industries
Dimethyl Carbonate EM Industries
N-N-Dimethylformamide Aldrich
Dimethyl Sulfoxide Fisher
Ethanol Aldrich
Ethylene Carbonate EM Industries
Lithium Hydroxide Fisher
Methanol Fisher
Nitric Acid Fisher
N-methylpyrrolidinone Fisher
Phosphorous Pentoxide Fisher
Propylene Carbonate EM Industries
Sulfuric Acid Fisher
Tetrahydrofuran Acros
Water Fisher
Table 2

Solvent Properties (40-42). *

 Solubility Vapor
 Parameter Dielectric Pressure
Solvent ([ca.sup.1/2]/[cm.sup.3/2]) Constant (kPa)

Water 23.53 80.16 17.536
Methanol 14.5 32.66 16.937
Ethanol 12.78 24.55 7.87
DMSO 12 46.45 0.08
PC 13.3 64.92 0.16
EC 14.7 89.78 0.449
THF 9.9 7.58 21.6

Acetonitrile 12.11 35.94 11.84
DMC 9.97 3.1
DEC 8.8 2.82 1.3

 Nafion[R] Ionic Dow Ionic
 Viscosity Conductivity Conductivity
Solvent (cP) (S/cm) (S/cm)

Water 1 2.6 x [10.sup.-2] 2.9 x [10.sup.-2]
Methanol 0.5929 4 x [10.sup.-3] 8.5 x [10.sup.-3]
Ethanol 1.0826 2.8 x [10.sup.-3]
DMSO 2.2159 1.8 x [10.sup.-3]
PC 2.76 1.6 x [10.sup.-4] 8 x [10.sup.-4]
EC 1.93
THF 0.575 8 x [10.sup.-6] 7.5 x [10.sup.-6]
Acetonitrile 0.341 9 x [10.sup.-7] 1.6 x [10.sup.-6]
DMC 0.63 9.5 x [10.sup.-7]
DEC 0.748 6 x [10.sup.-7]

* All values are given at 20 to 25[degrees]C except the dielectric
constant and viscosity of EC, which are given at 40[degrees]C.


ACKNOWLEDGMENTS

Our most sincere thanks to Dr. Mark Doyle (DuPont) for fruitful conversations concerning the data interpretation and Dr. Mark Lewittes (DuPont) for designing and building the 4-point probe. Part of the research was sponsored by a DOD STTR Grant Number F33615-00-C-2070.

REFERENCES

(1.) K. L. Heitner, J. Power Sources, 89, 128 (2000).

(2.) T. Michot, A. Nishimoto, and M. Watanabe, Electrochim. Acta, 45, 1347 (2000).

(3.) H. Sung, Y. Wang. and C. Wan, J. Electrochem. Soc., 145, 1207 (1998).

(4.) Z. Jiang, B. Carroll, and K. M. Abraham, Electrochim. Acta, 42, 2667 (1997).

(5.) K. E. Thomas, S. E. Sloop, J. B. Kerr. and J. Newman, J. Power Sources, 89, 132 (2000).

(6.) J. Y. Song, Y. Y. Wang, and C. C. Wan, J. Power Sources, 77, 183 (1999).

(7.) Z. Florjanczyk, W. Bzducha, W. Wieczorek, E. Zygadlo-Monikowska, W. Krawiec, and S. H. Chung, J. Phys. Chem. B, 102, 8409 (1998).

(8.) T. Fujinami, A. Tokimune. M. A. Mehta, D. F. Shriver, and G. C. Rawsky, Chem. Mater., 9, 2236 (1997).

(9.) K. Onishi, M. Matsumoto, Y. Nakacho, and K. Shigebara, Chem. Mater., 8, 469 (1996).

(10.) J. Travas-Sejdic, R. Steiner, J. Desilvestro, and P. Pickering, Electrochim. Acta, 46, 1461 (2001).

(11.) M. Watanabe, Y. Suzuki and A. Nishimoto, Electrochim. Acta, 45, 1187 (2000).

(12.) R. Koksbang, I. I. Olsen, and D. Shackle, Solid State Ionics, 69, 320 (1994).

(13.) T. Sata, J. Membr. Sci., 70, 267 (1992).

(14.) N. Li, J. Y. Lee, and L. H. Ong, J. Appl. Electrochem., 22, 512 (1992).

(15.) T. Mori, Y. Ohashi, and M. Okada, Japanese Patent 02-174076 (1990).

(16.) T. Mori, Japanese Patent 02-168578 (1990).

(17.) M. B. Armand, "Ionically Conductive Polymers," in Solid State Batteries, C. A. C. Sequeira and A. Hooper, eds., Martinus Nijhoff Pub., Boston (1985).

(18.) P. Aldebert, M. Guglielmi, and M. Pineri, Polym. J., 23, 399 (1991).

(19.) T. Noda and H. Yoshihisa, Japanese Patent 01-189872 (1989).

(20.) E. Genies, U.S. Patent 4,547,439 (1985).

(21.) S. Sachan and S. Perusich, Electrochemical Society Meeting, Seattle (1999); S. Sachan, MS Thesis, Auburn University (1999).

(22.) C. M. Doyle, M. E. Lewittes, S. A. Perusich, G. Rajendran, and M. G. Roelofs, U.S. Patent 6,033,804 (2000).

(23.) C. A. Ray and S. A. Perusich, Electrochemical Society Meeting, Washington, D.C. (2001).

(24.) M. Doyle, M. E. Lewittes, M. G. Roelofs, S. A. Perusich, and R. E. Lowrey, J. Mem. Sci., 184, 257 (2001).

(25.) S. C. Narang and D. D. MacDonald, U.S. Patent 5,061,581 (1991).

(26.) 5. Narang and S. Ganapathia, U.S. Patent 5,102,751 (1992).

(27.) D. G. Fauteux, A. A. Massucco, J. R. Powell, and M. F. van Buren, U.S. Patents 5,433,876 and 5,453,335 (1995).

(28.) W.-T. Whang, U.S. Patent 5,407,593 (1995).

(29.) E. J. Plichta and W. K. Behl, U. S. Patent 5,300,376 (1994).

(30.) K. M. Abraham, M. Alamgir, and H. S. Choe, U.S. Patent 5,474,860 (1995).

(31.) A. Agrawal, F. R. Grimes, P.-M. Allemand, and M. Trosky, Jr., U.S. Patent 5.300,374 (1994).

(32.) Y. Yokomichi, S. Tada, H. Nishino, and K. Seki, U.S. Patent 5,204,196 (1993).

(33.) D. D. MacDonald and S. C. Narang, U.S. Patent 5,061,581 (1991).

(34.) D. B. Cotts, U.S. Patent 4,832,869 (1989).

(35.) S. C. Narang, U.S. Patent 4,888,257 (1989).

(36.) B. D. Cahan and J. S. Wainright, J. Electrochem. Soc., 140, L185 (1993).

(37.) S. A. Perusich, Macromolecules, 33, 3437 (2000).

(38.) J. R. Fried, Polymer Science and Technology, Prentice-Hall, Englewood Cliffs, N.J. (1995).

(39.) D. W. van Krevelen, Properties of Polymers: Their Correlation with Chemical Structure: Their Numerical Estimation and Prediction From Additive Group Contributions, 3rd Ed., Elsevier (1990).

(40.) J. A. Riddik, W. B. Bunger, and T. K. Sakano, "Organic Solvents-Physical Properties and Methods of Purification," in Techniques of Chemistry, 4th Ed.. Vol. II, A. Weissberger, ed., John Wiley & Sons (1986).

(41.) H. Uramoto, ElectroCommunications, 1(3), 1 (1996).

(42.) W. V. Steele, R. D. Chirico, S. E. Knipmeyer, and A. Nguyen, J. Chem. Eng. Data, 42, 1008 (1997).

STEPHEN A. PERUSICH *

* To whom correspondence should be addressed.
COPYRIGHT 2002 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2002 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Sachan, Sunil; Ray, Cameron A.; Perusich, Stephen A.
Publication:Polymer Engineering and Science
Date:Jul 1, 2002
Words:6439
Previous Article:Techniques for reducing residual monomer content in polymers: a review.
Next Article:Strategies for achieving ultra low-density polypropylene foams.
Topics:


Related Articles
New lithium battery design charges up.
The First Fluorescent Fiber-Optic Lithium Ion Sensors.
New material charges up lithium-ion battery work. (Bigger, cheaper, safer batteries).
Images of tiny ion may help battery designers. (Lithium Sees the Light).
Hydrogen fuel cell powers latest GM E-Flex.
Why A123 may put the Chevy Volt on the road in 2010.
Solvent solution could delay 'peak lithium'.
EnerDel to invest $3.3 million in research for battery charging performance.
PHET to standardize LifePO4 li-ion batteries for vehicle applications.
Porous power technologies LLC.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters