*2.2. Electrolyte Preparation*

1-Ethyl-3-methyl-imidazolium bis(trifluoromethanesulfonyl)imide (EMI-TFSI), was synthesized by reacting HPLC water with a 1:1 LiTFSI:EMI-Br molar ratio mixture. This solution was stirred for 24 h in an oil bath at 70 ◦C. After the reaction took place, an aqueous layer and an ionic liquid (EMI-TFSI) rich layer were formed, and the solution was extracted from the oil bath. Once the solution was cooled to room temperature, the EMI-TFSI was separated from its aqueous counterpart and decanted into a separator funnel. HPLC water was poured into the separator funnel and mixed with the EMI-TFSI. The mixture was left to rest until the two layers were formed again. The EMI-TFSI layer was once again removed. This process was repeated two more times. Then, the EMI-TFSI was placed in a 500 mL round bottom flask to be dissolved with a sufficient amount of DCM. The dissolved EMI-TSFI was decanted into a chromatography column in order to filter any remaining impurities. The chromatography column contained one inch of sand, followed by silica oxide fully covering the remaining of the column up to the beginning of reservoir. The collected solution was then placed in a rotavap to remove the solvent (DCM) from the EMI-TSFI. Finally, the obtained EMI-TSFI was placed in a vacuum oven at 100 ◦C for 48 h to remove any water and excess DCM. The purity of the synthesized IL electrolyte was confirmed by 1H NMR spectroscopy.

The organic liquid electrolyte was prepared in a glove box (MBRAUN, Garching, Germany) with a controlled argon atmosphere. A 20 mL solution was prepared by combining a 1:1 *v*/*v* ratio of EC and DMC followed by 2 h of magnetic stirring. This OLE solution was stored and used to make 5 mL batches of 1 M LiTFSI in 60% EMI-TFSI and 40% EC/DMC. First, 1.435 g of LiTFSI and 2.564 g of EC/DMC solution were stirred until the LiTFSI was fully dissolved. Next, 4.590 g of EMI-TFSI were added and stirred for 24 h. The final weight of the solution was 8.590 g. Using this weight, an additional 0.429 g of SN was added to compare the ionic conductivity of the MOILE with one containing 5 wt.% of SN.

The ionic conductivity of the MOILEs was measured by assembling coin-type cells (CR2032) composed of two stainless-steel spacers as the positive and negative terminals, and a Teflon washer filled with MOILE. The LiCoO2 cathode was assembled with a common half-cell configuration to investigate the electrochemical performance of the cell. The as-prepared MOILE was used with the commercial LiCoO2 cathode. The active material loading in the electrode was 6.2–8.0 mg/cm2. The coin cells were assembled in a glove box using the cathode as the working electrode, with a Li counter electrode and microfiber glass mat separator (Whatman).

#### *2.3. Preparation of SnO2*/*C Composite Fiber Membranes*

The SnO2/C composite fibers were prepared by forcespinning of PAN/SnO2 precursor fibers followed by a thermal treatment. The PAN/SnO2 solution was prepared by dissolving 12 wt% PAN in DMF. A tin (II) 2-ethylhexanoate solution to 2:1 weight ratio of PAN solution was added and stirred for 24 h. The forcespinning process relies on applying centrifugal forces at high rotational speeds to a polymer solution or melt to produce microfibers with different structure and morphology. A description of the forcespinning process was given previously [19,32,33]. The PAN/SnO2 precursor solution was spun using the FiberRio L-1000 cyclone at a rotational speed of 8000 rpm for 1 min. The PAN/SnO2 fibrous mat was collected, stabilized in air at 280 ◦C for 5 h, and subsequently carbonized under argon atmosphere at 700 ◦C for 3 h (heating rate: 3 ◦C/min). The SnO2/C composite fibers were removed from the tube furnace, punched into 0.5 in (0.0127 m) diameter anodes, then weighed and used directly as working electrodes in lithium anode half-cells.

#### *2.4. Fiber Membrane Characterization*

The morphology and structure of composite fiber membranes were investigated using a scanning electron microscope (SEM) from Sigma VP Carl Zeiss, Oberkochen, Germany while energy-dispersive X-ray spectroscopy (EDS) from EDAX Inc., Mahwah, NJ, USA was used to confirm the elemental composition of the fibers. The crystal structure of the composite fiber membranes was evaluated by X-ray diffraction (XRD) using a Bruker D8 Advanced X-ray Diffractometer at a scan rate of 1 ◦C/min over a range of 2θ angle from 10◦ to 70◦.

## *2.5. Electrochemical Measurements*

Lithium anode (or cathode) half-cells were assembled in an argon-filled glove box with SnO2/C composite fibers as a binder-free anode and Li-metal as the counter electrode, using the MOILE. The electrothermal performance of the SnO2/C composite-fiber anode was evaluated by conducting galvanostatic charge/discharge experiments on CR2032 coin cells at 100 mA g<sup>−</sup>1. The active material loading in the anode was 2.4–4.5 mg/cm2. The ionic conductivity experiments on half cells with MOILEs were performed at different temperatures using a home-built heating block chamber. The design was based on a home-built sealed conducting cell in use at the University of Minnesota [23]. The impedance of the MOILEs at different temperatures was measured using a Metrohm Autolab (PGSTAT128N) connected to the heating chamber, over a frequency range from 0.1 Hz to 1 kHz. The ionic conductivity of the electrolyte was determined using coin cells with two stainless steel blocking electrodes filled

with the electrolyte. For accurate measurements of the ionic conductivity, a Teflon spacer was placed between the stainless-steel electrodes to hold the electrolyte inside the cell. The sample (electrolyte) preparation was conducted in an argon-filled glove box. The cell was then taken outside the glove box and inserted in the heating chamber. The ionic conductivity, σ, was calculated as L/(RA), where L and A are the sample thickness and superficial area of the sample and R is the bulk resistance [23]. The bulk resistance was determined from the frequency-independent plateau of the real part of the impedance (Z). The temperature was controlled and monitored using thermocouples and heating cartridges connected to a temperature process control CN 7500 purchased from Omega. The experimental setup was connected to a personnel computer using a RS485 USB converter to monitor the time and temperature during the impedance measurements.

The electrochemical performance of the LiCoO2 half-cells was evaluated at 60 ◦C. The LiCoO2 half cells were placed in a controlled temperature oven (ESPEC BTZ – 133). LiCoO2 cathode with electrolytes 1 M LiTFSI in 60% EMI-TFSI and 40% EC/DMC with and without SN were tested at a current density of 100 mAh g<sup>−</sup><sup>1</sup> for 50 cycles. Arbin's MTIS Pro was employed to conduct the galvanostatic charge/discharge experiments over a voltage range of 2.5–4.2 V. A port extension was connected between the Arbin instrument and the ESPEC oven to conduct the electrochemical experiments at di fferent temperatures.

#### **3. Results and Discussion**

## *3.1. Materials Characterization*

Figure 1 shows SEM images of SnO2/C composite fibers. It can be seen in Figure 1 that the SnO2 nanoparticles tend to aggregate, forming large clusters on the fibers. Some of these nanoparticles are embedded in the fibers and some are deposited on the fiber strands [20]. The average fiber diameter of the SnO2/C composite fibers was 1.86 m.

**Figure 1.** SEM images of a SnO2/C composite-fiber membrane [20], with copyright permission from the IOP Publishing.

Figure 2 shows an SEM image of SnO2/C composite fibers and the corresponding EDS mapping. Figure 2 shows that the composite fibers consist of C, O, and Sn that are distributed over the fibers. The EDS results confirm that the aggregated nanoparticles on the fibers contain Sn and O (i.e., SnO2 nanoparticles), which are attached to the surface of the carbon-fiber matrix

**Figure 2.** SEM image of SnO2/C composite fibers (left) and corresponding EDS mapping of the SnO2/C composite fibers (right).

Figure 3 shows an XRD pattern for the carbon fibers, where a broad diffraction peak is observed at 2θ = 27.8◦ corresponding to the (002) lattice plane of graphite [34–36]. It is observed in Figure 3 that this peak is weak and broad, which is the result of the formation of an amorphous carbon fiber structure after carbonization of the precursor PAN fibrous membrane.

**Figure 3.** XRD pattern of the carbon-fiber membrane prepared after the carbonization of polyacrylonitrile (PAN) fibers at 700 ◦C.

Figure 4 shows XRD analysis of the SnO2/C composite-fiber membrane. The observed pattern has predominantly crystalline peaks corresponding to (110), (101), (200), (211), and (310) planes. The observed peaks overlap with five of the seven peaks of the SnO2 crystal structure published by the (JCPDS 41-1445), further confirming the formation of SnO2 nanoparticles in the carbon matrix.

**Figure 4.** XRD pattern of the SnO2/C composite-fiber membrane prepared after calcination of PAN/SnO2 precursor fibers at 700 ◦C.

#### *3.2. Ionic Conductivity Measurement of Electrolytes at Di*ff*erent Temperature*

Figure 5 shows the ionic conductivity of the ILE and MOILEs as a function of temperature. The ILE was prepared from LiTFSI salt dissolved in 100% EMI-TFSI while the MOILE was prepared by dissolving LiTFSI salt in 60% EMI-TFSI and 40% EC/DMC, with and without the addition of 5% SN. The results show that the ILE (100% EMI-TFSI) delivered an ionic conductivity ~5 mS·cm<sup>−</sup><sup>1</sup> at room temperature, which is lower than that for the MOILE (60% EMI-TFSI and 40% EC/DMC) (~14 mS·cm<sup>−</sup>1). It is also clear in Figure 5 that the ionic conductivity of the three electrolytes increases with increasing temperature. The MOILE with 5% SN shows the highest ionic conductivity at 100 ◦C (70 mS·cm<sup>−</sup>1) among these three electrolytes. Despite its lower conductivity at room temperature, the ILE ionic conductivity increased significantly as the temperature was increased. At 150 ◦C, the ILE conductivity was ~30 mS·cm<sup>−</sup>1. This behavior is expected since the viscosity of ILEs tends to decrease with increasing temperature. The addition of 40% organic liquid, EC/DMC (1:1 *v*/*v* ratio), to 1 M LiTFSI in 60% EMI-TFSI resulted in an increased ionic conductivity of ~14 mS·cm<sup>−</sup><sup>1</sup> at room temperature while the MOILE with the addition of 5 wt% SN exhibited the highest room temperature ionic conductivity of ~26 mS·cm<sup>−</sup>1. Note here that the ionic conductivity of the OLE (EC/DMC/LiTFSI) is not shown in Figure 5 since there are data available in the literature on LiTFSI in binary EC/DMC mixtures. In fact, LiTFSI salt in EC/DMC binary system shows a higher ionic conductivity than that for ILE and MOILE. For example, results reported by Dahbi et al. showed that the LiTFSI in EC/DMC (1:1 *v*/*v* ratio), which is the same OLE used in the present work, exhibited an ionic conductivity of 8.6 mS·cm<sup>−</sup><sup>1</sup> at 25 ◦C. This value was increased to 11.5 and 14.9 mS·cm<sup>−</sup><sup>1</sup> when the temperature was increased to 40 and 60 ◦C, respectively [37]. The results also showed the ionic conductivity of LiPF6 in EC/DMC mixture was higher than that with EC/DMC/LiTFSI electrolyte over the entire temperature range [37].

**Figure 5.** Ionic conductivity vs temperature for the ionic liquid electrolyte (ILE), mixed organic/ionic liquid electrolyte (MOILE), and MOILE with 5 wt.% SN.

#### *3.3. Electrochemical Performance of A LiCoO2 Electrode in Lithium Cathode Half-Cells*

The commercial LiCoO2 electrode was employed in lithium cathode half-cells with a single-coated lithium foil to investigate its electrochemical performance. The MOILE with and without 5 wt% SN was used with the commercial LiCoO2. The purpose was to evaluate the behavior of the MOILEs in high voltage cathode materials such as LiCoO2, which has a larger voltage range than LiFePO4. LiCoO2 still dominates the portable electronics market due to its high voltage plateau and easy synthesis compared to LiFePO4 [38]. Galvanostatic charge/discharge experiments were performed for 50 cycles at different temperatures and at a current density of 100 mA g<sup>−</sup>1.

Figure 6a,b shows the charge/discharge profiles at 60 ◦C and at 100 mA g<sup>−</sup><sup>1</sup> of the commercial LiCoO2 cathode in MOILEs without SN and with 5 wt% SN, respectively. As can be observed in Figure 6a, the LiCoO2 cathode in the MOILE without SN maintained a consistent specific capacity of 148 mAh g<sup>−</sup><sup>1</sup> up to the 10th cycle. However, significant irreversibilities were observed at the 25th and 50th cycles. After 50 cycles, the cathode delivered a discharge capacity of 91 mAh g<sup>−</sup>1, indicating a capacity retention of 61.5% at a current density of 100 mA g<sup>−</sup>1. The discharge capacity retention is equal to the capacity after the 50th cycle divided by the capacity at the 1s<sup>t</sup> cycle (i.e., 61.5% = (100–38.5)%). On the other hand, the LiCoO2 cathode in the MOILE with SN (Figure 6b) exhibited an initial discharge capacity of 150 mAh g<sup>−</sup><sup>1</sup> at 100 mA g<sup>−</sup>1, and after the 50th cycle, the discharge capacity reached a value of 129 mA g<sup>−</sup><sup>1</sup> indicative of acceptable capacity retention of 86%. The improvement in the electrochemical performance of the LiCoO2 cathode is attributed to the effect of the SN additive on the ILE, and to the high conductivity of MOILEs at high temperature (60 ◦C). The high volatility and evaporation (high vapor pressure) of DMC at high temperature might influence the ionic conductivity of electrolytes containing a high percentage of DMC, thus affecting the electrochemical performance of the electrode. The effect of DMC on the ionic conductivity of MOILEs was not investigated since the amount of DMC in MOILEs is only 20% (1 M LITFSI in 1:1 *v*/*v* EC/DMC) and this should affect the performance of the electrode only slightly. However, results reported in the literature show that the ionic conductivity of 1 M LiPF6 in DMC remains significant (i.e., 9 mS·cm<sup>−</sup>1) at 55 ◦C [39]. Results reported by Aurbach et al. on a LNMO cathode at 60 ◦C in a liquid electrolyte (DMC–EC (2:1)/LiPF6 1.5 M), over a 3.5–4.9 V potential range showed that the cycling behavior of the cathode was explored without any observed degradation of the electrolyte solution [40].

**Figure 6.** Charge/discharge profiles of a commercial LiCoO2 cathode at 60 ◦C with MOILEs (**a**)1M LiTFSI 60% EMI-TFSI 40% EC/DMC EC/DMC (1:1 *v*/*v*), and (**b**) 1 M LiTFSI 60% EMI-TFSI 40% EC/DMC EC/DMC (1:1 *v*/*v*) containing 5 wt% SN. Current density = 100 mA g<sup>−</sup>1.

It is worth noting here that the LiCoO2 cathode in the MOILEs shows moderate capacity fading and voltage change in the plateau of Figure 6a,b. This might be caused by a decrease in active material (lithium) on the current collector after the 25th cycle. Another important factor that could affect this loss in capacity of the LiCoO2 cathode is that the corrosion of the Al current collector on the cathode side by the TFSI, thereby contributing to the loss of active material from the Al current collector [41]. More work will be conducted to investigate these effects on the LiCoO2 cathode in LiTFSi/MOILEs systems.

Figure 7a,b shows the cycling performance corresponding to the charge/discharge curves shown in Figure 6a,b. Although the capacity is stable within the first twenty cycles, the LiCoO2 cathode in the MOILE without SN suffered from a steady loss in specific capacity after 20 cycles. In contrast, the same cathode in the MOILE with 5 wt% SN maintained a stable specific capacity for the first 20 cycles; there was a slight decrease in capacity between 20th and 30th cycles, while thereafter the cathode maintained a constant capacity of ~129 mAh g–1. The LiCoO2 cathode in both electrolytes maintained a similar high coulombic efficiency of 98% for 50 cycles.

**Figure 7.** Cycling performance and coulombic efficiency of LiCoO2 commercial cathode at 60 ◦C in (**a**) 1 M LiTFSI in 60% EMI-TFSI 40% EC/DMC (1:1 *v*/*v*) electrolyte, and (**b**) 1 M LiTFSI in 60% EMI-TFSI 40% EC/DMC (1:1 *v*/*v*) with 5 wt% SN. Current density = 100 mA g<sup>−</sup>1.

Figure 8a,b shows the charge/discharge curves at 100 mA g<sup>−</sup><sup>1</sup> for the SnO2/C composite-fiber anode in two different electrolytes, OLE and MOILE with SN. The cycle performance of the SnO2/C composite electrode was evaluated by conducing galvanostatic charge/discharge experiments at room

temperature and at a current density of 100 mA g<sup>−</sup>1. The voltage range for lithium anode half-cells tested with the SnO2/C composite-fiber anode was 0.05–3 V (versus Li+/Li). The SnO2/C composite-fiber anode with the 1 M LiPF6 in EC/DMC (1:1 *v*/*v*) electrolyte showed an initial discharge capacity of 785 mAh g<sup>−</sup>1. The reversible specific capacity after 100 cycles was 319 mAh g–1. Nevertheless, the SnO2/C composite anode showed a stable specific capacity after the 25th cycle, with a capacity retention of 98% after the 2nd cycle. Improved cycling stability of the SnO2/C composite fibers in 1 M LiTFSI in 60% EMI-TFSI 40% EC/DMC with 5% SN electrolyte (Figure 8b) was observed after the 2nd cycle, with a specific capacity of 382 mAh g<sup>−</sup>1, having a ~20% increase compared to the SnO2/C composite-fiber anode cycled with the OLE.

**Figure 8.** Charge/discharge profiles for a SnO2/C composite-fiber anode at 25 ◦C in two different electrolytes: (**a**) 1M LiPF6 in EC/DMC (1:1 *v*/*v*) electrolyte, and (**b**) 1 M LiTFSI in 60% EMI-TFSI/ 40% EC/DMC (1:1 *v*/*v*) with 5 wt% SN.

Figure 9 shows the cycling performance (charge/discharge capacity vs. cycle number) of the SnO2/C composite-membrane anode in the OLE and the MOILE with 5 wt% SN at a current density of 100 mA g<sup>−</sup>1. It is observed in Figure 9a,b that the discharge and charge capacities of the SnO2/C composite-fiber anode in the MOILE with 5 wt% SN are higher than in the OLE. The improvement in the specific capacity of the composite-membrane anode was attributed to the addition of the high polarity SN to the MOILE and its ability to dissolve the LiTFSI salt, which resulted in enhanced ionic conductivity and improved cycling stability of the electrode in the MOILE. The SnO2/C composite-membrane anode in MOILE with 5 wt.% SN shows (Figure 9b) improved cycling stability and capacity retention after the 2nd cycle.

**Figure 9.** Cycling performance of a SnO2/C composite-fiber anode at 25 ◦C in two different electrolytes: (**a**) 1M LiPF6 in EC/DMC (1:1 *v*/*v*) electrolyte, and (**b**) 1 M LiTFSI in 60% EMI-TFSI/ 40% EC/DMC (1:1 *v*/*v*) with 5 wt% SN.

The rate performance of the SnO2/C composite fibers was further evaluated by conducting current rate (or rate capability) experiments on the lithium anode half-cells at di fferent current densities between 50 and 500 mA g<sup>−</sup>1. The SnO2/C composite fibers were cycled ten times at current densities of 50, 100, 200, 400, 500, and then again at the initial value of 50 mA g<sup>−</sup><sup>1</sup> (Figure 10). The results exemplify the SnO2/C composite anode's ability to perform at higher current densities, as well as demonstrating the capacity recovery after being cycled from high to low current density. Figure 10 shows the rate performance (charge capacity vs cycle number at di fferent current densities) of the SnO2/C composite-fiber anode in the OLE and in the MOILE with 5 wt.% SN. As expected, the composite-fiber anode delivered a higher specific charge capacity at lower current density, and vice versa. At 50 mAh g–1, the specific capacity decreased after 10 cycles to 418 mAh g–1 for the Li-ion cell cycled with the 1 M LiPF6 in EC/DMC (1:1 *v*/*v*) electrolyte, but only to 579 mAh g–1 for the 1 M LiTFSI in 60% EMI-TFSI 40% EC/DMC (1:1 *v*/*v*) with 5 wt% SN electrolytes. This can be attributed to the stresses and strains caused by the high-volume change of the SnO2/C composite fibers after repeated charge/discharge cycles. At a current density of 100 mA g<sup>−</sup>1, the charge capacity was stable at ~315 mAh g<sup>−</sup><sup>1</sup> for 1 M LiPF6 in EC/DMC (1:1 *v*/*v*) and at ~441 mAh g<sup>−</sup><sup>1</sup> for 1 M LiTFSI in 60% EMI-TFSI 40% EC/DMC (1:1 *v*/*v*) 5 wt% SN. The SnO2/C composite anode in the MOILE with 5 wt% SN had a higher percentage increase in specific capacity with 25% at 100 mAh g<sup>−</sup>1, 23% at 200 mAh g<sup>−</sup>1, 30% at 400 mAh g<sup>−</sup>1, and 1% at 500 mAh g<sup>−</sup>1, compared to 1 M LiPF6 in EC/DMC (1:1 *v*/*v*). However, after cycling back to 50 mA g<sup>−</sup>1, the SnO2/C composite fibers with 1 M LiPF6 in EC/DMC (1:1 *v*/*v*) had less specific charge capacity than with the MOILE with 5 wt% SN. However, the SnO2 electrode in both electrolytes (OLE and MOILE with SN) shows relatively low capacity at higher current density. Thus, the improvement in the charge capacity of the SnO2/C composite anode with MOILE and SN can be attributed to the high Li-ion conductivity and di ffusion caused by the addition of SN to the ionic liquid electrolyte.

**Figure 10.** Rate performance (charge capacity vs cycle number at different current densities) of SnO2/C composite fibers at 25 ◦C with two di fferent electrolytes: OLE (1M LiPF6 in EC/DMC (1:1 *v*/*v*)), and MOILE (1 M LiTFSI in 60% EMI-TFSI/ 40% EC/DMC (1:1 *v*/*v*) with 5 wt% SN).
