Macrophase-Separated Organic Ionic Plastic Crystals/PAMPS-Based Ionomer Electrolyte: A New Design Perspective for Flexible and Highly Conductive Solid-State Electrolytes.

A material design approach was taken for the preparation of an organic ionic plastic crystal (OIPC)-polymer electrolyte material that exhibited both good mechanical and transport properties. Previous attempts to form this type of electrolyte material resulted in the solvation of the OIPC by the ionomer and loss of the plastic crystal component. Here, we prepared, in situ, a macrophase-separated OIPC-polymer electrolyte system by adding lithium bis(fluorosulfonyl)imide (LiFSI) to a (PAMPS-N1222) ionomer. It was found that an optimal compositional window of 40-50 mol % LiFSI exists whereby the electrolyte conductivity suddenly increased 4 orders of magnitude while exhibiting elastic and flexible mechanical properties. The phase behavior and transport properties were studied using differential scanning calorimetry and 7Li and 19F solid-state nuclear magnetic resonance spectroscopy. This is the first example of a fabrication principle that lends itself to a wide range of promising OIPC and ionomeric materials. Subsequent studies are required to characterize and understand the morphology and conductive nature of these systems and their application as electrolyte materials.

Introduction

The development of new solid polymer electrolytes (SPEs) for application in lithium ion batteries and for the next generation of high-capacity lithium metal-based devices is currently an intense area of research. The ability to utilize a lithium metal anode would increase the storage capacity owing to the high theoretical capacity of lithium, 3860 mAh·g–1. However, there are many issues associated with the reactive lithium metal surface, including dendrite formation and propagation that can result in a fire hazard.[1,2] Additionally, the low conductivities of SPEs at practical operating temperatures have, to date, impeded widespread SPE uptake in commercial Li ion devices, while the mechanical properties of the SPE also need to be considered for applications in lithium metal-based devices to impede dendrite propagation.[3,4]

One avenue of material research has focused on tethering the anion to the polymer backbone, essentially creating an ionomer (also known as single-ion conductor when associated with an alkali metal counterion). With such materials, the issues with concentration polarization are overcome, as the only mobile ion is the ion of interest. Despite having an alkali metal transport number of unity by design (tLi = 1.0), ionomers typically still suffer from poor alkali metal ion dissociation from the anionic polymer backbone, resulting in lower ionic conductivity and poor cycling performance when used as a single-component polymer system. To overcome this, an ionomer block is generally coupled with an polyethylene oxide (PEO)-based block as either a random or block copolymer to improve ion dissociation.[5−10] However, the lithium ion mobility in a PEO-based single-ion conductor is still bound to the local motion of the PEO chain, resulting in poor ionic conductivity at temperatures below the melting point of PEO (65 °C).[5,6,9,10]

Alternatively, Colby and co-workers have recently demonstrated that the ion motion in ionomers can be significantly enhanced by substituting the alkali metal counterion with a bulky organic cation.[11] The increase in ion motion in these systems is mainly due to the plasticizing effect from the bulky organic cation, effectively lowering the glass transition temperature (Tg) of the ionomer. Additionally, we have previously studied the use of random copolymer ionomers based on 2-acrylamido-2-methyl-1-propane-sulfonate triethylmethylammonium (N1222AMPS) and sodium vinyl sulfonate (NaVS) monomers as solid-state Na ion conductors.[12] Significant enhancement of the ionic conductivity, by 3–5 orders of magnitude, was observed when the molar fraction of N1222AMPS monomer was increased from 0 to 80–90 mol %.[12] However, the Tg in these systems remained relatively unaffected when varying the N1222AMPS mole fraction, suggesting a progressive decoupling of the conductivity from the Tg of the polymer as the N1222AMPS content is increased. To further understand how the decoupling of the ion dynamics occurs in ionomers, a series of random PAMPS-based copolymer ionomers with a lithium counterion and different quaternary ammonium-based cations were also studied in various molar ratios.[13,14] In these studies, clear decoupling behaviors were observed by solid-state NMR spectroscopy for the Li-based copolymer ionomer containing a triethylammonium cation ((N1222)1–-Li-PAMPS).[13,14] It was hypothesized that the introduction of a bulky ammonium cation favors a Li ion hopping transport mechanism, thereby decoupling the Li ion motion from the polymer segmental mobility.[13,14] Interestingly, the introduction of an ether containing moiety in these Li-based copolymer ionomers by either ammonium cation modification or the addition of tetraglyme plasticizer induced a recoupling of the ionic conductivity back to the polymer segmental motion.[14]

In parallel work, our group has previously shown that maintaining a very high mole fraction of alkali metal salts (Li or Na) in an ionic liquid solvent can also result in a Li ion hopping transport mechanism, which promotes excellent lithium transport properties (tLi ≥ 0.4 as measured by the Bruce–Vincent method) and good cycling performance.[2,15−19]

In the present work, we have studied the behavior of the N1222-PAMPS ionomer mixed with lithium bis(fluorosulfonyl)imide (LiFSI) in various molar ratios, whereby the ionomer essentially acts as the solvent for the LiFSI salt. Figure shows the chemical structure of the N1222-PAMPS ionomer and LiFSI salt. While these systems are not considered as single-ion conductors, the aim of this study was to investigate if a decoupling behavior could also be achieved by the use of high lithium content in the lithium salt/ionomer system. Two distinct regimes were observed in N1222-PAMPS:LiFSI mixtures. At low LiFSI contents, a typical salt-in-polymer behavior was observed, where the LiFSI salt was fully solvated by the polymer chains. At high LiFSI contents, an unexpected macrophase-separated system was observed. The thermal behavior of these mixtures changes drastically with the formation of a plastic crystal phase, which results in a significant enhancement in both ionic conductivity and mechanical properties. This behavior represents a simultaneous improvement in both mechanical and transport properties without the use of any additional plasticizing additives.

Figure 1

Chemical structure of the N1222-PAMPS ionomer (a) and the LIFSI salt (b).

Results and Discussion

A series of N1222-PAMPS:LiFSI mixtures were prepared with various LiFSI molar ratios, ranging from 0 to 75 mol % LiFSI. The SPE membranes made of the pure N1222-PAMPS ionomer are quite brittle, limiting their implementation from a device perspective. Increasing the LiFSI content to 40 mol % results in a drastic change of the physical properties of the SPE membrane, going from glassy PAMPS-like behavior to a film that is flexible, strong, and elastic. These physical properties, which are favorable characteristics for application in a device, were observed across an optimal compositional window of 40–50 LiFSI mol %.

DSC measurements were performed to characterize the phase behavior of N1222-PAMPS:LiFSI systems and better understand the changes observed in the physical properties of the resulting SPEs. The DSC profiles of N1222-PAMPS:LiFSI mixtures shown in Figure a reveal some remarkable behaviors as the composition is varied. Pure N1222-PAMPS undergoes a glass transition at 89.9 °C, while for the 80/20 and 70/30 compositions, this Tg is increased to 102.0 and 116.2 °C, respectively. Recently, our group reported the synthesis and characterization of a series of random copolymer PAMPS ionomers, varying the molar ratio of quaternary ammonium ion to lithium ion (as the PAMPS counterion). In this study, random copolymer ionomers based on N1222-Li-PAMPS were synthesized.[14] Increasing the molar ratio of the lithium counterion from 10 to 50 mol % resulted in an increase of Tg from 105 to 171 °C. This difference in Tg was attributed to the cation size difference, with the larger N1222 cation suggested to disrupt to a greater extent the polymer chain interactions. Therefore, the increase of the glass transition temperature as LiFSI is first introduced into the N1222-PAMPS ionomer (up to 30 mol %) can be likely attributed to the strong association of the Li ion with the sulfonate anions of the polymer backbone, resulting in the formation of dynamic ionic interchain cross-links.[11,14] Although the N1222-PAMPS:LiFSI system studied here has some similarities to the previously reported N1222-Li-PAMPS ionomer,[14] it is important to note that in the N1222-PAMPS:LiFSI system, LiFSI also contributes a large FSI anion, which is likely to further disrupt the polymer chain interaction and thereby enhance ion dynamics when compared to the previously reported N1222-Li-PAMPS copolymer ionomer.[14]

Figure 2

(a) DSC traces (vertical bars are a scale for the Y-axis) and (b) room temperature XRD patterns of N1222-PAMPS:LiFSI mixtures as a function of LiFSI molar fraction.

Interestingly, the 60/40 and 50/50 N1222-PAMPS:LiFSI systems show unexpected phase behavior with the appearance of a solid–solid transition at −9.3 and −7.8 °C (Ts-s) and a melting transition at 101.0 and 121.6 °C (Tm). As the LIFSI molar fraction is increased from 40 to 50 mol %, both the solid–solid and melting transitions become sharper and more pronounced. The appearance of the solid–solid transition for the 60/40 and 50/50 compositions suggests the formation of a macrophase-separated system with the presence of a plastic crystal phase in the temperature range between Ts-s (e.g., −7.8 °C) and Tm (121.6 °C), which agrees with the room temperature physical property observations. The enthalpy associated with these transitions, as well as the temperature at which these transitions occurs, does not match with that of the pure LiFSI salt (solid–solid transition: 331 vs 953 J·mol–1 for LiFSI and 50/50 N1222-PAMPS:LiFSI system, respectively),[20] suggesting that this new plastic crystal phase is not from the LiFSI salt precipitating out. On the other hand, the temperatures at which the solid–solid and melting transitions are observed for the 50 mol % sample are very similar to those reported for the pure N1222FSI organic ionic plastic crystal (OIPC) (see Figure S1). OIPCs are analogous to ionic liquids but are in the solid state and have pliable (or plastic) mechanical properties.[21] Previous attempts to combine both OIPC and polymer electrolyte properties into SPE have failed primarily due to the OIPC being solvated by the polymer chains.[22] To our knowledge, this is the first time that the properties of an OIPC and polymer electrolyte have been combined successfully to give a true OIPC–polymer electrolyte. However, it is worth mentioning that the enthalpy associated with the melting transition for the 50/50 N1222-PAMPS:LiFSI system suggests that only ∼21% of the total mole amount of N1222FSI is macrophase separated (solid–solid transition: 5402 vs 953 J·mol–1 for N1222FSI and 50/50 N1222-PAMPS:LiFSI system, respectively), indicating that the remaining ∼79% of N1222FSI in mole is still likely solvated within the N1222-PAMPS:LiFSI ionomeric matrix. The glass transition of the remaining N1222-PAMPS:LiFSI ionomeric matrix is not observable in the temperature range studied. Indeed, we previously reported a glass transition temperature of 171 °C for a (N1222)0.5-Li0.5-PAMPS random copolymer ionomer.[14] An attempt to extend the temperature range studied fails in detecting the glass transition of the remaining N1222-PAMPS:LiFSI ionomeric matrix of the 50/50 N1222-PAMPS:LiFSI system, due to the presence of a broad exothermic peak (∼179–223 °C), related to early decomposition processes of FSI anions[20] (see Figure S2). Interestingly, further increase of the LiFSI content to 75 mol % results in the disappearance of the melting transition associated with the OIPC phase, while the solid–solid transition is consequently shifted to a lower temperature and becomes broader. Additionally, a glass transition at 119.0 °C is now observed, suggesting that most N1222FSI is now solvated by the N1222-PAMPS:LiFSI ionomeric matrix. All of these suggest that the macrophase separation observed in the N1222-PAMPS:LiFSI system is only favorable in a compositional range of 40–50 LiFSI mol %. Table summarizes both the physical and the thermal properties of N1222-PAMPS:LiFSI mixtures.

Table 1

Physical and Thermal Properties of N1222-PAMPS:LiFSI Mixtures

	physical properties	thermal properties
N1222-PAMPS/LiFSI molar fraction	visual observations	Tg (°C)b	Tsolid-solid (°C)c	ΔHs-s (J/g)	Tm (°C)c	ΔHm (J/g)
0/100	crystalline		–51.3	1.77	140.2	91.24
25/75	rubbery	119.0	–26.7	2.48
50/50	rubbery	d	–7.8	7.49	121.6	5.14
60/40	rubbery	d	–9.3	5.34	101.0	3.23
70/30	rubbery	116.2
80/20	glassy	102.0
100/0	glassy	89.9

DSC data.

Midpoint.

Onset.

Not detectable due to early decomposition of FSI anions.

To better understand the macrophase separation that occurs in the N1222-PAMPS:LiFSI system, XRD analysis was also performed, and the results are presented in Figure b. As can be seen in Figure b, three diffraction peaks are observed for the pristine N1222-PAMPS ionomer at a diffraction angle 2θ of ∼7.5, 12.8, and 17.9°, strangely characteristic of amorphous PAMPS.[23] The addition of 30 mol % of LiFSI to the N1222-PAMPS ionomer results in a significant broadening of the diffraction peaks associated with N1222-PAMPS, likely due to the presence of LiFSI salt disrupting the original polymer chain packing and thus promoting disorder. Interestingly, the opposite trend is observed at 50 mol % of LiFSI, where the diffraction peaks from the PAMPS become sharper. This coincides with the formation of a plastic crystal phase, as suggested by the DSC results, supporting the presence of a macrophase separation between the N1222-PAMPS:LiFSI ionomeric matrix and the OIPC phase. The macrophase separation likely results in a reduction of the total amount of N1222FSI solvated by the N1222-PAMPS:LiFSI ionomeric matrix (e.g., 79% according to DSC results), thus promoting the packing of the polymer chains in a more ordered manner. Further increase of the LiFSI content to 75 mol % results, this time, in a broadening of the diffraction peaks of N1222-PAMPS, indicating again that the macrophase separation observed in N1222-PAMPS:LiFSI system is only favorable in a compositional range of 40–50 LiFSI mol %. Both DSC and XRD data indicate that the formation of a macrophase-separated system is likely due to the strong interactions between the sulfonate anions of the polymer backbone and the lithium ions from the LiFSI salt, as suggested by the progressive increase of the glass transition for the N1222-PAMPS:LiFSI compositions when LiFSI concentration is increased from 0 to 30 mol %. Additionally, these strong interactions also likely trigger a rearrangement of the polymer chains in a more ordered manner (i.e., close to the pristine polymer), as observed by the XRD data when the molar ratio between the lithium ions and sulfonate groups of the polymer backbone is close to unity. Further increase of the LiFSI concentration results in the disordering of the polymer chain packing, which is not favorable for the formation of a macrophase-separated system.

The ionic conductivity of N1222-PAMPS:LiFSI mixtures was measured by electrochemical impedance spectroscopy (EIS) to determine how the ion dynamics are affected by LiFSI content. Figure shows the ionic conductivity of N1222-PAMPS:LiFSI mixtures, from 30 to 110 °C. These data, like the DSC traces, show that there is a distinct difference in the ion dynamics of the 60/40 and 50/50 compositions compared to the other compositions. While the conductivity of the other compositions varies from 10–10 to 10–6 S·cm–1, these two compositions possess conductivities that are several orders of magnitude higher, ranging from 10–6 to 10–5 S·cm–1. The higher ionic conductivity observed for the 60/40 and 50/50 compositions coincides with the formation of a plastic crystal phase, as suggested by DSC data.

Figure 3

Temperature-dependent ionic conductivity of N1222-PAMPS:LiFSI mixtures as a function of LiFSI molar fraction. Solid lines are fitted functions based on Arrhenius equation.

To better understand the transport mechanism present in these electrolytes, an analysis of the temperature-dependent ionic conductivity was performed using the Arrhenius model. All N1222-PAMPS:LiFSI systems obey an Arrhenius behavior, which is to be expected since the temperature range studied is mostly below the glass transition temperature of all N1222-PAMPS:LiFSI systems. The activation energy (Ea) and the pre-exponential factor (σ0) extracted from the Arrhenius fitting are summarized in Table . Increasing the LiFSI concentration, from 20 to 30 mol %, results in a decrease in the activation energy, going from 110 to 69 kJ·mol–1. For a LiFSI concentration, ranging from 40 to 50 mol %, the activation energy reaches a minimum of around ∼34 kJ·mol–1. This minimum in activation energy coincides with the formation of a macrophase-separated system. Further increase of the LiFSI concentration, from 50 to 75 mol %, results, on the contrary, in an increase of the activation energy, going from 33 to 110 kJ·mol–1. Overall, the analysis of the activation energy as a function of LiFSI concentration suggests that the presence of the N1222FSI plastic crystal phase (i.e., macrophase-separated system) plays a significant role in lowering the displacement potential barrier and thus promoting ion conduction. On the other hand, the trend of the pre-exponential factor (σ0) as a function of LiFSI concentration does not show any minima, but rather, follows an exponential decay trajectory, reaching a plateau of around ∼1 S·cm–1 for LiFSI concentration above 40 mol %. This suggests that the high ionic conductivity observed for the macrophase-separated N1222-PAMPS:LiFSI systems, containing 40–50 mol % LiFSI, is likely not the result of an increase in the number of charge carriers, but rather, in a change of the transport mechanism, likely induced by the formation of a plastic crystal phase.

Table 2

Activation Energy and Pre-Exponential Factor for N1222/LiFSI Mixtures As Calculated from Temperature-Dependent Ionic Conductivity Using the Arrhenius Equation

	Arrhenius fitting parameters of the temperature-dependent ionic conductivity
N1222-PAMPS/LiFSI molar fraction	EA (kJ·mol–1)	σ0 (S·cm–1)
25/75	109.8 ± 2.3	3.4 × 10–9 ± 2.2
50/50	32.4 ± 0.3	0.3 ± 1.1
60/40	36.1 ± 0.7	2.3 ± 1.3
70/30	69.0 ± 1.8	139.2 ± 1.9
80/20	110.4 ± 1.1	6.0 × 108 ± 1.5

Although EIS gives detailed insights into the overall ion dynamics of N1222-PAMPS:LiFSI mixtures, it does not differentiate as to which ions are contributing to the conductivity. Therefore, 7Li and 19F solid-state nuclear magnetic resonance (NMR) spectroscopy experiments were performed to probe ion-specific dynamics as a function of temperature. Figure shows the 7Li and 19F NMR line widths of the N1222-PAMPS:LiFSI mixtures as a function of temperature.

Figure 4

(a) 7Li and (b) 19F static solid-state NMR line widths at 7.05 T as a function of temperature for the N1222-PAMPS:LiFSI mixtures. The dashed lines are fitted functions based on BPP theory (see the Experimental Section for details).

At 20 °C, the 7Li NMR line width for the 70/30 composition is around 5 kHz, which indicates that the Li ions are most likely immobile due to strong interactions with the sulfonate anion of the polymer backbone. Thus, the broader line width below Tg suggests a coupling of the ion dynamics to the macroscopic SPE dynamics.[24,25] This is also consistent with the low ionic conductivity observed for this composition, in the order of 10–10 S·cm–1 at room temperature. At temperatures above Tg, the 7Li line width significantly decreases and reaches 2.5 kHz at 140 °C, suggesting an increase in the Li ion dynamics. The decrease in 7Li NMR line width starts just before the Tg of the system, suggesting that the Li ion dynamics are only somewhat decoupled from the polymer segmental mobility in this composition.[13,14,26]

Additionally, since the 7Li NMR line widths are in the order of kHz over the temperature range studied, it can be assumed that this narrowing is only due to translational dynamics (i.e., translational jumps). A similar trend is observed in the 19F NMR line width for this system, with a sharp decrease of the 19F NMR line width as the temperature is increased. The onset temperature of the 19F NMR line narrowing also occurs slightly below the Tg of the ionomer, suggesting that the FSI ion motions are only partially coupled to the polymer mobility. The narrowing of the 19F NMR line width is more pronounced than that of the 7Li NMR line width, likely due to additional relaxation processes inherent to the FSI anion, which increase the FSI ion mobility, such as rotational motion. Upon increasing the LiFSI content to 50 mol %, the evolution of both the 7Li and 19F NMR line widths shows clear differences, with no sharp decrease as the temperature is increased, but rather a relatively linear decrease with increasing temperature. In the temperature range studied for NMR line width measurements (i.e., from 0 to 140 °C), only one endothermic transition at 121.6 °C, associated with the melting of the plastic crystal phase, is observed in the DSC traces of the 50 mol % N1222-PAMPS:LiFSI system. Since no sharp decrease in both 7Li and 19F NMR line widths is observed around this temperature, it could be concluded that both Li and FSI ion mobilities are not closely coupled to the dynamic of the plastic crystal phase, likely suggesting that the Li ions are closely associated to the sulfonate anions of the polymer backbone. The glass transition temperature of the remaining N1222-PAMPS:LiFSI ionomeric matrix of the 50/50 N1222-PAMPS:LiFSI system could not be observed by DSC, but as previously mentioned, can be estimated from previous work to be around 170 °C,[14] which is outside the temperature range studied. Therefore, the coupling of Li ion mobility to the segmental motion of the remaining N1222-PAMPS:LiFSI ionomeric matrix cannot be fully probed. Further increase of the LiFSI concentration, from 50 to 75 mol %, results in a trend of the 7Li NMR line width similar to the 30 mol % LiFSI system, suggesting a coupling of the Li ion dynamics to the macroscopic SPE dynamics. 19F NMR line width of the 75 mol % N1222-PAMPS:LiFSI system could not be measured due to the slow dynamics of the system.

A more in-depth analysis of the 7Li and 19F NMR line widths was also performed using the Bloembergen, Purcell, and Pound (BPP) model (see the Experimental section for more details).[27] From this analysis, the activation energy, EA, and the temperature-dependent correlation time, τ, which corresponds to the minimum energy required to induce a jump (or hop) and the inverse of the ion jump frequency at a defined temperature, respectively, can be determined. Table also summarizes the measured EA, τ0, and some temperature-dependent τ for 70/30, 50/50 and 25/75 N1222-PAMPS:LiFSI mixtures. Some degree of uncertainty is expected for these fitting parameters as both 7Li and 19F NMR line widths do not cover the full sigmoidal decay. However, these results obtained can be discussed qualitatively in relative terms and compared to the activation energy determined from the temperature-dependent ionic conductivity.

Table 3

Activation Energies (EA) and Some Selected Correlation Times (τ) for N1222-PAMPS:LiFSI Mixtures As Calculated from the 7Li and 19F NMR Line Widths

	Li translational dynamics				FSI translational dynamics
N1222-PAMPS/LiFSI molar fraction	EA (kJ·mol–1)	τ0 (s)	τ(293 K) (s)	τ(373 K) (s)	EA (kJ·mol–1)	τ0 (s)	τ(293 K) (s)	τ(373 K) (s)
25/75	43.9	3.8 × 10–10	7.0 × 10–3	3.6 × 10–4
50/50	14.0	3.0 × 10–6	9.2 × 10–4	2.7 × 10–4	2.1	6.9 × 10–6	1.6 × 10–5	1.4 × 10–5
70/30	28.8	3.4 × 10–8	4.5 × 10–3	3.6 × 10–4	49.6	2.1 × 10–8	1.4 × 10–5	1.8 × 10–7

Activation energy of 28.8 kJ·mol–1 was measured for the Li ion dynamics in the 70/30 N1222-PAMPS:LiFSI mixture. This activation energy is slightly lower than the one reported for the random N1222-Li-PAMPS ionomer with a lithium molar ratio of 10 mol % but still higher than those reported for traditional Li salt/uncharged polymer systems, where EA of 34.5 and 20–23 kJ·mol–1 were measured, respectively.[13,28] The relatively high value of the EA (30–35 kJ·mol–1) previously observed in random copolymer ionomers was attributed to a strong association between the Li ion and the polymer sulfonate groups and the formation of ionic aggregates. In the case of the 70/30 N1222-PAMPS:LiFSI mixture, the slightly lower EA could suggest that the association between the Li ion and the polymer sulfonate groups is weaker, likely due to the presence of the FSI ions. Interestingly, an EA of 14.0 kJ·mol–1 was measured when the LiFSI content is increased to 50 mol %. This value is significantly lower than those reported from either random copolymer ionomer or traditional Li salt/uncharged polymer systems. Such low activation energy could be due to the formation of a plastic crystal phase, as suggested by DSC results. Organic ionic plastic crystal electrolytes are known to exhibit high ionic conductivity due to enhanced translational motion.[21,29] These translational motions are allowed due to rotational or reorientational short-range motions within the long-range ordered plastic phase.[29] Therefore, this could suggest that the enhanced dynamics observed for the Li ion in the 50/50 N1222-PAMPS:LiFSI mixture may be due to some short-range jumps occurring within the plastic crystal phase. At a LiFSI concentration of 75 mol %, activation energy of 43.9 kJ·mol–1 was measured. This increase of activation energy, going from 14 to 43.9 kJ·mol–1, when increasing LiFSI concentration from 50 to 75 mol %, coincides with the loss of the plastic crystal phase. The trend in activation energy from the 7Li line width measurements, with the observation of a minima at a LiFSI concentration ranging from 40 to 50 mol %, is in agreement with the activation energy extracted from the temperature-dependent ionic conductivity measurements.

Figure shows the 7Li and 19F correlation time as a function of temperature for the N1222-PAMPS:LiFSI mixtures. As can be seen in Figure , the temperature-dependent correlation times allow us to clearly see the enhancement in ion motions observed for the 50/50 N1222-PAMPS:LiFSI mixture. 7Li correlation times of 4.5 and 0.9 ms were measured at 20 °C for the 70/30 and 50/50 N1222-PAMPS:LiFSI mixtures, respectively. This corresponds to a 4-fold increase in Li ion jump frequency, when the LiFSI content is increased from 30 to 50 mol %. Increasing LiFSI concentration to 75 mol % results in a decrease of the 7Li correlation time to 7.0 ms. In contrast, the 19F correlation times measured at 20 °C for the 70/30 and 50/50 N1222-PAMPS:LiFSI mixtures are very similar, with a correlation time of 0.014 and 0.016 ms, respectively. Interestingly, the 19F correlation time at a higher temperature is significantly higher, being almost 2 orders of magnitude greater for the 50/50 N1222-PAMPS:LiFSI mixture when compared to the 30 mol % sample. Overall, the 7Li and 19F correlation time results clearly suggest that the formation of a plastic crystal phase observed at a LiFSI content of 50 mol % significantly enhances Li translational dynamics at room temperature, while decreasing the FSI ion translational motion.

Figure 5

(a) 7Li and (b)19F correlation time as a function of temperature for the N1222-PAMPS:LiFSI mixtures.

Conclusions

The phase behavior and ion dynamics of N1222-PAMPS:LiFSI systems were characterized by means of DSC, EIS, and solid-state NMR spectroscopy as a function of LiFSI concentration. Two distinct regimes are observed in N1222-PAMPS:LiFSI mixtures. At low LiFSI content, a characteristic salt-in-polymer regime is observed, in which the Li mobility is only partially decoupled from the polymer segmental motion as suggested by solid-state NMR spectroscopy. A macrophase-separated system is observed at 40–50 mol % LiFSI, with the formation of a plastic crystal phase, likely from the precipitation of some amount of the N1222FSI OIPC. The presence of a plastic crystal phase results in the enhancement of both ionic conductivity and mechanical properties. Such improvements are unusual, as generally both mechanical properties and ionic conductivity decrease with increasing lithium salt content. Additionally, 7Li and 19F solid-state NMR spectroscopy shows a clear improvement of the Li ion mobility across an optimal compositional window of 40–50 LiFSI mol %. This is the first time we have observed such a microstructure for an OIPC–ionomer system that exhibits both promising ionic conductivity and mechanical properties. This new type of electrolyte material offers great scope for further optimization and investigations using a wide range of promising candidate materials. Further studies will focus on characterizing the microstructure of these materials and understanding how the choice of materials can affect the relationship between the microstructure and transport properties. The application of these materials in battery or capacitor systems will be presented in subsequent studies.

Experimental Section

Materials and Sample Preparation

A detailed experimental procedure for the synthesis of N1222-PAMPS has been published previously.[26] Lithium bis(fluorosulfonyl)imide (LiFSI, ≥99.0%) was purchased from CoorsTek Fluorochemical. Films of N1222-PAMPS:LiFSI were solvent cast from methanol solution, which was evaporated for 1 h at 300 mbar and 40 °C using a rotavapor. The film was then dried under a high vacuum for 72 h at 80 °C.

Differential Scanning Calorimetry (DSC)

DSC measurements were performed on a Netzsch DSC 214 Polyma instrument, which was calibrated using cyclohexane. Approximately 8–10 mg of the prepared electrolyte sample was tested over a temperature range of 233–423 K at a scanning rate of 10 K·min–1. The glass transition temperature was determined from the midpoint of the heat-capacity change on heating.

Ionic Conductivity

Electrochemical impedance spectroscopy (EIS) was performed on a Solartron Modulab 1296 impedance analyzer, equipped with a Eurotherm 2204 temperature controller. The sample was placed between two stainless steel electrodes and then sealed inside a custom-built barrel cell (Advanced Industrial Services, Moorabbin, Australia). EIS measurements were performed over a temperature range from 30 to 110 °C, in 10 °C intervals. An isothermal step of 20 min for each temperature was used. A frequency range from 1 MHz to 1 Hz and an amplitude of 100 mV were used. Two heating scans were conducted and the data shown in this study were extracted from the second heating scan.

Solid-State Nuclear Magnetic Resonance (NMR) Spectroscopy

Solid-state NMR experiments were recorded on a Bruker Avance III wide-bore spectrometer at a Larmor frequency of 282.4 and 116.6 MHz for 19F and 7Li, respectively. All 19F and 7Li spectra were referenced to trichlorofluoromethane and a 1 M lithium chloride aqueous solution, respectively. A 5 mm static variable-temperature NMR probe was used to record the NMR spectra. Spectra were acquired with a single pulse experiment using a 90° pulse length of 8.5 μs and a recycle delay of 15 s. Temperature was calibrated using lead nitrate to an accuracy of ±2 °C.[30] Reported line widths correspond to the full width at half-maximum (FWHM) intensity of the spectral peak and were measured as a function of temperature from 0 to 140 °C in 20 °C intervals, allowing 15 min for the sample to equilibrate at each temperature. The correlation time, τ, was determined by fitting the NMR line widths as a function of temperature using the following equationwhere Δν and δω0 are the line width and the rigid lattice line width in Hz. The temperature-dependent correlation time (τ) was then fitted with the following Arrhenius equationwhere EA is the activation energy in J·mol–1, k is the Boltzmann constant in J·K–1, τ0 is the correlation constant (i.e., the inverse of the ion jump frequency at infinite temperature), and T is the temperature in K.