Experimental Section/Methods are described in the Supporting Information (SI).
2.1. Material Composition
Monofunctional acrylate monomer 4-(6-Acryloxy-hex-1-yl-oxy) phenyl-4-(hexyloxy) benzoate (M1) and bifunctional monomer 1,4-Bis-[4-(6-acryloyloxyhexyloxy) benzoyloxy]-2-methylbenzene (M2) were purchased from Synthon chemicals. Ionic liquid 1-Hexyl-3-methylimidazolium bis(trifluormethyl sulfonyl)imide (HMIM-TFSI), photoinitiator 2,2-Dimethoxy-2-phenylacetophenone (Irgacure® 651), ionic salt lithium bis(trifluoromethyl sulfonyl)imide (LITFSI), plasticizer succinonitrile (SCN) were acquired from Sigma-Aldrich, Milwaukee, US. Lithium metal disks (about 600 µm thick) and lithium iron phosphate (LFP) were purchased from MSE supplies (Error! Reference source not found.).
Figure 1.
The molecular structures of the components of the studied PEM: M1 and M2 are mesogenic units. HMIM-TFSI is the ionic liquid, Irgacure 651 is the photoinitiator, SCN is the plasticizer, LiTFSI is the ionic salt.
Figure 1.
The molecular structures of the components of the studied PEM: M1 and M2 are mesogenic units. HMIM-TFSI is the ionic liquid, Irgacure 651 is the photoinitiator, SCN is the plasticizer, LiTFSI is the ionic salt.
2.1.1. Synthesis of Ionic Liquid Crystal Elastomer
M1, M2, and the photoinitiator were mixed in 87:12:1 weight ratio to form the LCE precursors (see
Figure S6(a)), because it gave the optimum LCE properties [
66,
68,
69,
70,
71,
72]. Subsequently, an ionic liquid (HMIM-TFSI) was added to the LCE precursor solution and mechanically stirred for 24
h using a magnetic stirrer in an Ar-filled glovebox after heating to 80 ºC to achieve complete mixing. The mixture was then stored in a glass amber vial to prevent photopolymerization and kept in the glovebox at room temperature for future use.
Figure S6 shows a synthesis route of UV crosslinking reaction between M1 and M2 functional groups using a 1 wt% Irgacure 651 photoinitiator to obtain LCE co-networks. The same reaction was undertaken in the presence of LiTFSI salt and SCN plasticizer that afforded a transparent, homogeneous film before and after UV curing, suggestive of a miscible character among the PEM constituents.
2.1.2. Fabrication of iLCE Electrolyte Membrane
The binary eutectic mixture of LiTFSI and SCN at weight ratios of 25:50 was melt-mixed at 60 ºC under Ar-gas environment. Subsequently, the homogeneously dissolved LiTFSI and SCN binary mixture was further mixed with the pre-mixed iLCE precursors at various wt% for an additional 24 h. The SPE membrane based on iLCE was prepared through UV photopolymerization of iLCE, SCN, and LiTFSI. The homogeneous mixture was poured into a polytetrafluoroethylene mold with a thickness of 300 µm and a diameter of 15 mm approximately, where the solution was kept at various preset temperatures for 0.5 h. Subsequently, the electrolyte membrane was exposed to 365 nm UV light at 280 mW cm-2 intensity for 30 s at the pre-set temperatures to activate the photoinitiator to polymerize the mono- and bi-functional LCE monomers. These functional groups reacted in the presence of SCN plasticizer and LiTFSI salt in various ratios. Upon curing, a binary iLCE co-network-based electrolyte membrane (PEM) was formed with a film thickness of approximately 300 µm. Finally, the flexible electrolyte films were dried at RT under vacuum for 24 h. Fourier Transform Infrared Spectroscopy (FTIR) was performed on the electrolyte film and on the individual components to check the formation of fully cross-linked LCE co-networks, as shown in Figure S1.
Error! Reference source not found. shows SEM images of the electrolyte film. As SEM is carried out in vacuum, the low molecular weight components (ionic liquids) are evaporated, and their former places appear dark. This way we get information about the distribution of the ionic channels that is crucial for the operation of the battery. The images in Error! Reference source not found.(a, c) show that the ionic liquid and the polymers form bi-continuous structure, thus allowing ionic pathway from one end to the other. The diameter of the ion channels varies from 40 µm (see Error! Reference source not found.(a, b)) down to 100 nm range as seen in Error! Reference source not found.(d). Error! Reference source not found.(b) illustrates phase separated polymer (LCE and SCN) domains. The minority component (LCE) phase separates into spherical shape droplets from the continuous majority component (SCN) – more evidence on this is shown by POM (Error! Reference source not found.(a-d)), as will be discussed in later sections. Importantly, Error! Reference source not found.(d), where we see a part of the minority droplet, shows that ion channels interpenetrate even in the discontinuous minority domain.
Figure 2.
SEM images of the iLCE SPE. (a) Electrolyte cross-section tilted at 45º, (b) Surface image, (c) Bulk image at electrolyte cross-section, (d) Zoomed in texture on the surface of liquid crystal droplet.
Figure 2.
SEM images of the iLCE SPE. (a) Electrolyte cross-section tilted at 45º, (b) Surface image, (c) Bulk image at electrolyte cross-section, (d) Zoomed in texture on the surface of liquid crystal droplet.
2.2. Electrochemical Performance
An impedance test for various PEM compositions was performed from 20 ºC to 100 ºC using a homemade rectangular cell having an area of 1mm × 1mm and a 0.1 mm gap. The ionic conductivity ) for each PEM composition was found to decrease with increasing temperature in reasonably good approximation following Arrhenius behavior, , where kB = 1.38 × 10-23 J K-1, is the temperature in Kelvin scale, and is the activation energy. Figure 1 shows the electric conductivity as a function of 103/T for various LCE (Figure 1(a)), SCN (Figure 1(b)), ionic liquid (IL) and Li+ ratios (Figure 1(c)), and at various crosslink temperatures (Figure 1(d)). The straight lines are the best fits corresponding to the Arrhenius behavior.
In comparison of 30, 40 and 50 % of LCE contents (Figure 1(a)), one finds that the highest ionic conductivity values (9.0 × 10-4 S cm-1 at 20 ºC, and 5.39 × 10-3 S cm-1 at 90 – 100 ºC) were found for 30 % LCE. The activation energies obtained from the slopes of the best fits are 0.12, 0.27 and 0.44 eV for the 30, 40 and 50 % of LCE contents, respectively. This means the material is becoming increasingly solid at increasing LCE concentration.
The effect of a plasticizer on lithium-ion conduction was examined by comparing 10, 20, 30, 40, 50 wt% SCN, keeping the liquid crystal elastomer composition to be the same (
Figure 1(b)). It is found that the ionic conductivity is increasing with increasing SCN concentration below 30 wt%. This is because an increasing plasticizer amount enhances the polymer chain dynamics, promoting ion conductions from an anode to a cathode and vice versa [
31,
32,
33,
42]. At higher SCN concentrations the conductivity saturates providing about the same conductivity for the 40 and 50 wt% SCN. In accordance with the enhanced polymer chain dynamics, the activation energies decrease from 0.22 eV to 0.16 eV between 10% to 30 % plasticizer concentrations. However, at excessive plasticizer amount of SCN (> 40 wt%) the activation energies are higher (0.22 and 0.19 eV at 40 and 50 wt%, respectively) indicating phase separation between the iLCE and SCN and the small-molecule weight SCN tends to ooze out to the PEM surface. This phase separation interrupts the ion channels that leads to saturation (an even slight decrease) of the conductivity and an increase of the activation energy.
Figure 1.
Arrhenius plots of ionic conductivity (logarithmic conductivity versus 1000/T). (a) For different ratios of LCE in electrolyte mixture; (b) For different ratios of SCN in electrolyte mixture; (c) For different ratios of HMIM-TFSI and LiTFSI in electrolyte mixture; and (d) For different crosslinking temperature of PEM.
Figure 1.
Arrhenius plots of ionic conductivity (logarithmic conductivity versus 1000/T). (a) For different ratios of LCE in electrolyte mixture; (b) For different ratios of SCN in electrolyte mixture; (c) For different ratios of HMIM-TFSI and LiTFSI in electrolyte mixture; and (d) For different crosslinking temperature of PEM.
The effect of the ionic liquid (IL) to Li salt ratio on the ionic conductivity was investigated by comparing the 5:35, 10:30, 15:25, 20:20, 25:15, 30:10 and 35:5 IL:Salt ratios so that the total weight percentage added to 40 wt% of the total mixture, keeping the same amounts of SCN and LCE (Figure 1(c)). At room temperature the ionic conductivity of the 20:20 IL-LiTFSI PEM was the largest, 1.76 × 10-3 S cm-1. The activation energies are found to be 0.20, 0.18, 0.24, 0.21, 0.26, 0.22 and 0.20 eV for the 5:35, 10:30, 15:25, 20:20, 25:15, 30:10 and 35:5 IL:Salt ratios, respectively.
The effect of polymerization temperature on the ionic conduction was examined by polymerizing in the nematic phase of the liquid crystal at 30 ºC and in the isotropic phase at 40, 60, 80 °C, during UV irradiation (Figure 1(d)). Comparing samples polymerized in the isotropic phase, we find that at room temperature the ionic conductivity increases with increasing polymerization temperatures, most likely due to decreased cross-link density toward higher polymerization temperatures. Interestingly, the conductivity of the sample polymerized in the nematic phase is even higher than that of the PEM polymerized at 80 °C. This suggests that the elastomer network in the nematic phase forms a directional network which in turn acts as ion conduction pathways for improved ion conduction. The slopes of the best fits gave 0.19, 0.22, 0.22, 0.13 eV for the samples polymerized in 30, 40, 60, 80 °C, respectively.
Based on these findings, subsequent experiments were focused on the LCE/HMIM-TFSI/LiTFSI/SCN: 30/17.5/17.5/35 PEM, as it showed the highest (1.76 × 10
-3 S cm
-1) ionic conductivity at ambient temperature. Such conductivity value is comparable to those of lower end organic liquid electrolyte batteries [
34,
35,
36]. We conservatively chose this composition containing lower LCE content due to its mechanical integrity during assembling. Moreover, at room temperature this chosen 30/17.5/17.5/35 PEM exhibits an ion transference number (the ratio of the electric current derived from the cation to the total electric current) of t
+ ~ 0.61, which is considerably higher than those (0.22 - 0.35) of organic electrolyte systems, suggesting the domination of ion transport by lithium cations over the TFSI anion [
40] (
Figure 2).
Figure 2.
Chronoamperometry of a symmetric (Li−PEM−Li) cell at ambient temperature in response to a UDC=0.01 V bias for the LCE/HMIM-TFSI/LiTFSI/SCN:30/17.5/17.5/35 PEM. The inset shows the Nyquist plot of the PEM impedance before DC polarization and after steady-state current conditions.
Figure 2.
Chronoamperometry of a symmetric (Li−PEM−Li) cell at ambient temperature in response to a UDC=0.01 V bias for the LCE/HMIM-TFSI/LiTFSI/SCN:30/17.5/17.5/35 PEM. The inset shows the Nyquist plot of the PEM impedance before DC polarization and after steady-state current conditions.
To determine the transference number of lithium ions through the PEMs, Li/PEM/Li symmetric coin cells were assembled and measured by using Autolab PGSTAT302N galvanostat. A constant DC bias of 10 mV was applied to determine the initial (I0) and steady-state (IS) currents, and the response of current was monitored for 3600 s until a steady-state current was reached. The before (R0) and after (RS) resistances of PEMs were determined by means of an impedance analyzer in the frequency range from 1 Hz to 100 kHz. The transference number (t+) of the PEMs was calculated in accordance with the following equation:, where ΔV is the DC polarization voltage applied through the PEMs.
2.3. Structure of iLCE electrolytes
Polarized Optical Microscopy (POM) studies before and after cross-linking, and inspection of mechanical properties of the LCE/HMIM-TFSI/LiTFSI/SCN: 30/17.5/17.5/35 PEM are shown in
Error! Reference source not found.. POM studies showed that the mixture before crosslinking is in the isotropic phase above 34.2 °C; it has a nematic phase between 34 °C and 24 °C and it is crystalline below 24 °C.
Error! Reference source not found.(a, b) show POM images of a 20 µm film between two bare glass plates (no rubbing) before polymerization under cross polarizers in the isotropic (60 °C) and nematic (31 °C) phase, respectively. The isotropic structure is evidenced by the uniformly dark image. In the nematic phase, the iLC prepolymer and the SCN plasticizer phase separate and form a structure like polymer dispersed nematic liquid crystals (PDLCs) [
73]. The droplets exhibit four dark brushes (so-called Maltese crosses) along the cross polarizers indicating that the LC director orients either tangentially or radially inside the droplets [
74]. Because of the radial symmetry, the optical texture remains unchanged when the sample is rotated between the crossed polarizers. After UV induced crosslinking, the PDLC-type structure remains unchanged although the size of the iLCE micro droplets depends on the UV intensity, polymerization time and temperature. At 280 mW cm
-2 UV intensity applied at 60 °C for 30
s cross linking time the packing of iLCE droplets is quite dense with droplets sizes approximately 3 - 10 µm as shown in
Error! Reference source not found.(c). After polymerization, the iLCE electrolyte shows a nematic phase between 34 °C and 56 °C. At higher temperatures, the droplets become isotropic, but the overall structure does not change, and on cooling the nematic droplets appear again in their original locations. Heating above 110 ºC will start to melt the electrolyte membrane to liquid.
Error! Reference source not found.(d) shows the surface structure without polarizers after low intensity slow (10 mW cm-2 for 10 min) polymerization. One can see that slow polymerization causes uneven distribution of droplet sizes and occasionally forms large domains of iLCEs in SCN matrix. Thus, we opted for fast polymerization at high intensities. Error! Reference source not found.(e, f) show 300 µm thick free-standing films of the 30/17.5/17.5/35 PEM at room temperature. In Error! Reference source not found.(e) one can see that the 300 µm thick film is self-standing and solid, while Error! Reference source not found.(f) demonstrates that the iLCE PEM is flexible. The toughness of a 2 mm × 1 cm × 1cm (m ~ 0.262 g) thick film was tested by hanging weights of up to 60 g masses when it started tearing up, as shown in Figure S2.
Figure 5.
Polarized Optical Microscopy (POM) studies before (a, b) and after (c) cross-linking, Microscopy image of the surface after (d) cross-linking and inspection of mechanical properties (e, f) of the 30/17.5/17.5/35 PEM after cooling at 0.1 ºC min-1 rate. (a) POM image in the isotropic phase at 60 ºC, (b) Nematic droplets at 31 °C. (c) Nematic elastomer droplets after crosslinking by 280 mW cm-2 UV intensity at 80 ºC applied for 30 s; (d) Surface images of nematic elastomer droplets after crosslinking by 10 mW cm-2 for . between parallel polarizers (e) Self-standing 300 µm thick film; (f) Demonstration of the flexibility of the iLCE PEM film.
Figure 5.
Polarized Optical Microscopy (POM) studies before (a, b) and after (c) cross-linking, Microscopy image of the surface after (d) cross-linking and inspection of mechanical properties (e, f) of the 30/17.5/17.5/35 PEM after cooling at 0.1 ºC min-1 rate. (a) POM image in the isotropic phase at 60 ºC, (b) Nematic droplets at 31 °C. (c) Nematic elastomer droplets after crosslinking by 280 mW cm-2 UV intensity at 80 ºC applied for 30 s; (d) Surface images of nematic elastomer droplets after crosslinking by 10 mW cm-2 for . between parallel polarizers (e) Self-standing 300 µm thick film; (f) Demonstration of the flexibility of the iLCE PEM film.
2.4. Cell Performance
To understand the Li-ion storage mechanism in the PEM during prelithiation, cyclic voltammetry (CV) and galvanometric charge/discharge cyclic tests were carried out in Li metal/PEM/LiFePO
4 (LFP) configuration using CR2032 coin cells at various potential ranges 2.5 - 4.0 V. In the linear sweep voltammetry (LSV) scans, PEMs appear stable against the stainless-steel (SS) electrode up to 3.7 V (see
Error! Reference source not found.(a)), as indicated by the arrows at the onsets of the CV curves (see
Error! Reference source not found.(b)). The PEM's stability up to 3.75 V in LSV data (
Error! Reference source not found.(a)) likely indicates a stability limitation, which is typical for organic and polymer electrolytes [
75,
76].
Figure 6.
Summary of voltammetry results with potential scan rate of 5 mV s-1. (a) Linear sweep voltammogram of PEM at 24 °C, (b) Cyclic voltammetry test of the Li/SPE/LFP configuration in cathode range (2.5 - 4.0 V) at 24 °C.
Figure 6.
Summary of voltammetry results with potential scan rate of 5 mV s-1. (a) Linear sweep voltammogram of PEM at 24 °C, (b) Cyclic voltammetry test of the Li/SPE/LFP configuration in cathode range (2.5 - 4.0 V) at 24 °C.
To identify the origin of oxidation and reduction peaks, a full battery cell was assembled in the Li-metal anode, /SPE/LFP cathode configuration based on 30/17.5/17.5/35 LCE/HMIM-TFSI/LiTFSI/SCN PEM composition. When the CV test was carried out in 2.5 V to 4.0 V range using the LFP cathode, the oxidation and reduction peaks of the Li-metal anode and those of PEM can be observed in the corresponding potential ranges, which continue to increase from the 2nd to the 4th cycles. As can be expected, the oxidation and reduction peaks of the LFP cathode are observable in the vicinity of 3.6 V ~ 4.0 V and 3.0 V, respectively, which fluctuates slightly with increasing number of cycles (Error! Reference source not found.(b)). The continued increase in the peak strength may be attributed to the continued lithiation occurring in situ during repeated cycles, hereafter termed “in situ lithiation”. More importantly, these CV results are indeed reproducible, although the resulting oxidation and reduction peaks could increase or decrease depending on the competition between the in situ lithiation versus the capacity fading.
To further confirm the electrochemical stability of a PEM-based rechargeable battery, the galvanometric charge/discharge cycling test was performed at 0.1C rate from 3.0 V to 3.7 V (Error! Reference source not found.(a, b)). Note, XY C-Rate means X Current × Y Hours the battery can provide X current. The initial specific capacity was ∼124 mAh g-1 which dropped to ~100 mAh g-1 with the capacity retention of about 80% and the Coulombic efficiency of ~99 %. To avoid any potential overcharging issues in the LFP cathode, the test was operated up to 3.7 V instead of 4.0 V.
Figure 7.
Charge and discharge cycling characteristics of the iLCE based PEM in coin cell battery. (a) Specific capacity and coulombic efficiency over 160 charge and discharge cycles; (b) Charge and discharge profiles for Li/SPE/LFP cathode cells with a cathode range of 3.0-3.7V at a C-rate of 0.1 at 25 °C; (c) Charge/discharge curves of Li/SPE/LFP cell at different C-rates after 100 cycles; (d) The cycle performance of the Li/SPE/LFP battery during galvanostatic cycling at 0.1, 0.2, 0.5 and 1 C-rate.
Figure 7.
Charge and discharge cycling characteristics of the iLCE based PEM in coin cell battery. (a) Specific capacity and coulombic efficiency over 160 charge and discharge cycles; (b) Charge and discharge profiles for Li/SPE/LFP cathode cells with a cathode range of 3.0-3.7V at a C-rate of 0.1 at 25 °C; (c) Charge/discharge curves of Li/SPE/LFP cell at different C-rates after 100 cycles; (d) The cycle performance of the Li/SPE/LFP battery during galvanostatic cycling at 0.1, 0.2, 0.5 and 1 C-rate.
Error! Reference source not found.(c, d) show the charge/discharge voltage profiles and capability of the PEM at different C-rates. The cell clearly delivers outstanding average discharge capacities of 98.2, 83.4, 65.7, and 48.8 mAh g
-1 at C-rates of 0.1, 0.2, 0.5, and 1 at room temperature, respectively. It can be observed that the cell exhibits excellent reversibility and the charge–discharge capacities are almost the same at a given C-rate. Unfortunately, the cell exhibits poor performance at higher C-rates. According to the manufacturer, the specific capacity of the LFP sheet used as the cathode is 130 mAh g⁻¹, which is only slightly larger than the initial 124 mAh g⁻
1 we obtained. A challenge arises from the interaction between the nitrile groups in the SCN plasticizer and the lithium metal anode. Li⁰ can catalyze nitrile polymerization, leading to side reactions at the anode, potentially affecting cathodic stability [
77]. Additionally, our composite electrolyte functions as a dual-ion conductor, with a non-unity transference number. This implies that lithium ions at the interface may be consumed or generated faster than they can migrate, creating a salt concentration gradient. Such a gradient can deplete the salt at the electrode, increasing ionic resistance, or cause salt precipitation due to enrichment. Despite this, cyclic voltammetry results and post charge-discharge analysis show minimal evidence of a salt gradient forming during operation. At 0.1 C-rate from cycles 80 to 160, the iLCE/LiTFSI/SCN cell maintained a discharge capacity of around 98.8 mAh g⁻¹ with nearly perfect coulombic efficiency (~99%). The stable interfacial properties and excellent capacity retention are largely attributed to the flexibility of the ILC-based electrolyte film, which ensures long-term cycling stability at room temperature.