Optimizing conductivity and cationic transport in crosslinked solid polymer electrolytes

Solid State Ionics 345 (2020) 115161

Contents lists available at ScienceDirect

Solid State Ionics journal homepage: www.elsevier.com/locate/ssi

Optimizing conductivity and cationic transport in crosslinked solid polymer electrolytes

T

⁎

Elyse A. Baroncini, Dominique M. Rousseau, Christopher A. Strekis IV, Joseph F. Stanzione III Henry M. Rowan College of Engineering, Rowan University, 201 Mullica Hill Road, Glassboro, NJ 08028, United States

A R T I C LE I N FO

A B S T R A C T

Keywords: Solid polymer electrolyte Ion transport Lithium ion Bio-based Vanillyl alcohol

Solid polymer electrolytes (SPEs) are prepared through thiol-ene polymerization with functionalized, potentially bio-based, aromatic monomers. Differing functionality and aromatic content of the monomers vary the glass transition temperatures (Tgs) and crosslink densities of the resulting polymers, allowing for analysis of the structure-property relationships. Though the SPEs contain repeating PEO segments, the formation of crystalline regions is avoided through the crosslinked nature of the networks. The solid polymer electrolytes exhibit high conductivity values at room temperature, the highest reaching 7.65 × 10−4 S cm−1 for the DAVA-containing SPE with 50 mol% LiPF6, and moderate lithium ion transference numbers, the highest reaching 0.39 for the DAGd-containing SPE with 25 mol% LiPF6. Lower polymer Tg is associated with higher overall conductivity, but, the SPEs with higher crosslink densities display higher cationic transport, though the Tgs are higher. Therefore, it is determined that there exists an optimal range of polymer Tgs, crosslink densities and neat polymer dielectric constants for high cationic transport through the SPEs; extremes of these parameters are not necessarily beneficial. The promising conductivity and ion transference results, in addition to high initial decomposition temperatures in N2 (> 300 °C) and great electrochemical stability, reveal potential for these crosslinked, aromatic, thiol-ene polymers in electrolyte applications in lithium-ion batteries.

1. Introduction The desire for robust secondary battery technologies has pushed the development of advanced materials for lithium-ion batteries (LIBs) to the forefront of research across the globe. Liquid electrolytes that are commonly used in LIBs exhibit safety, stability and toxicity concerns, thus, emphasis has been placed on research into solid electrolyte materials [1–3]. Promising alternatives are solid polymer electrolytes (SPEs) that possess no liquid electrolyte, thereby eliminating leakage concerns, and exhibit better thermal and electrochemical stabilities and higher mechanical strengths [1–3]. However, mediocre conductivity at room temperature and low mobility of cations compared to liquid electrolytes are a couple of areas where SPEs require improvement [1,4–5]. SPEs can be easily created by dissolving lithium salt in a resin and subsequently curing the system into a polymer network. Both conductivity and cationic transport can be optimized in SPEs through judicious selection of the base polymer matrix. Electrolytes based on poly (ethylene oxide) s (PEOs) have been extensively studied because of the ability of lithium salts to dissociate in the presence of PEOs and the propensity for movement of the Li+ cation through the polymer

⁎

network and/or chains [3,6–7]. The cations have been shown to coordinate with Lewis base polar atoms in polymers, such as eCeOe, eSe and eC]O groups [3,8]. Ionic transport through the PEOs is primarily a result of motion of the cations along or with the polymer chains containing repeating ethylene oxide (EO) units [7]. When creating SPEs, it is therefore advantageous to design a polymer network containing flexible, long chains with repeating ether oxygen or carbonyl units. Much SPE research has focused on linear PEO-based polymers and reducing crystallinity to assist with ionic movement [1]. Crystalline and semi-crystalline regions are typically found in thermoplastic materials. Crosslinking a polymer, thereby making it a thermoset, is a way to reduce the possibility of crystallization [9]. Therefore, this study chose to explore a crosslinked thermoset network assembled via thiol-ene chemistry as flexible but stable polymer electrolytes. Our previous work [10] identified multi-functional thiols that could incorporate structurally advantageous long, flexible, polyether chains into the resulting polymer network. Incorporation of long PEO-segments into thermosets can sometimes result in formation of crystalline regions within those sections of the polymer network. However, such crystallization will alter the thermal characterization of the polymers, presented in the

Corresponding author. E-mail address: [email protected] (J.F. Stanzione).

https://doi.org/10.1016/j.ssi.2019.115161 Received 7 August 2019; Received in revised form 10 October 2019; Accepted 18 November 2019 0167-2738/ © 2019 Published by Elsevier B.V.

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

form of a melting point [9] that can be easily identified through DSC analysis. Additionally, such crystallization usually occurs with very long, repeating PEO-segments – previous studies reporting crystallization use PEO with average Mw = 100,000 [11]. Fortunately, there is a wide selection of thiols of various sizes and shapes commercially available. The thiol utilized in this study has a total molecular weight of 1300 g mol−1 and is tri-functional. Each polyether-containing branch is calculated to have a molecular weight of approximately 300 g mol−1. The relatively short polyether chains are not long enough to promote crystalline formation but are thought to be long enough to assist with ionic transport – a point explored in this work through the evaluation of the conductivity and ionic transport properties of the thiol-ene SPEs. Through the nature of thiol-ene polymerizations [12], the selected thiol could be photopolymerized with a variety of functionalized monomers, allowing for incorporation of structural benefits through strategic monomer selection. Though a flexible polymer network is desirable for this application, so are thermal stability and mechanical integrity [13]. To achieve this balance of traits, aromatic compounds are selected for functionalization and subsequent polymerization with the desired thiol compound. Aromatic content is known to increase structural and thermal stability of polymer networks [14]. The increased stability can allow for creation of a very thin film polymer with high thermal resistance (especially compared to liquid electrolytes) and potentially high decomposition voltages [15–16]. Compounds with aromaticity are often found in nature; thus, monomers explored in this work were variations of potentially bio-derived compounds vanillyl alcohol [17], gastrodigenin [17], 4-methylcatechol [18] and pyrogallol [19]. The aromatic compounds are functionalized with allyl groups (Fig. 1) and combined with a multifunctional thiol (Fig. 2) in a radically-induced, step-growth addition polymerization known as a thiolene polymerization [12]. Thiol-ene polymerizations are largely insensitive to oxygen and will occur in a step-growth mechanism if the “ene” (or allylated monomer) and thiol each have a functionality of at least two and a total combined functionality of at least five [12]. Thiolene polymers typically have a high conversion of reactive groups and are less susceptible to shrinkage than comparable systems making them coveted polymer systems [12,20]. The thiol-ene polymerizations are radically-induced in a fast, solvent-free ultraviolet (UV) curing process at ambient temperature, see previous publication for general photopolymerization schematic [10]. UV curing is deemed an efficient, cost-

Fig. 2. Structures of thiol-containing ETTMP (top), which, in this study, has a molecular weight of 1300 g mol−1 and lithium salt LiPF6 (bottom).

effective and environmentally-friendly polymerization method with implications of additive manufacturing for these polymeric electrolyte systems. The SPEs examined in this work are synthesized with the multifunctional, relatively high molecular weight thiol and four potentially bio-derived monomers, each with varying functionality and aromatic content. Contrary to previous studies on polymer electrolytes that varied monomer or PEO concentration to vary polymer structures [2,11,21], this work chose to employ variations of monomer functionality and aromatic content to vary polymer structure. Thus, the relationships between Tg, crosslink density and cationic transport in crosslinked, thiol-ene polymers are determined. Ion transport through

Fig. 1. Structures of the four allylated monomers synthesized in this study. Clockwise from top left: di-allylated vanillyl alcohol (DAVA), di-allylated gastrodigenin (DAGd), tri-allylated vanillyl alcohol-4-methyl catechol (TAVC) and tri-allylated pyrogallol (TAPg). 2

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

5.27–5.39 (m, 3H); 5.91–6.07 (m, 3H); 6.52 (d, 1H); 6.71–6.85 (m, 4H). NMR spectra with peak assignments are available in Supporting Information.

the SPEs is characterized by spectroscopic, thermal and electrochemical techniques to determine structure-property relationships of the aromatic containing SPEs. By doing so, this work aims to incorporate research from both niche studies and well-explored fields, such as, gel/ solid polymer electrolytes [13,21–24], polyether-based electrolytes [2,4,25–27], thiol-based polymers/electrolytes [19,28–29] and biobased electrolyte materials [30] while also bridging gaps in knowledge between such areas of study.

2.1.4. Synthesis of allylated pyrogallol The procedure was adapted from Uemura et al. [19] as follows: pyrogallol (1 equivalent) and potassium carbonate (3 equivalents) were added to a round-bottomed flask with acetone. The flask was purged with argon in an ice bath and allowed to stir for 10 min. Allyl bromide (3.3 equivalents) was added dropwise after which the solution was removed from ice and allowed to stir for 24 h. Potassium carbonate was filtered and the product was purified on a flash chromatography system with n-hexanes and ethyl acetate. The resulting product was tri-allylated pyrogallol (TAPg), a light yellow liquid with 60% yield. 1,2,3-Tris(allyloxy)benzene (TAPg) (C15H18O3): Light yellow liquid. 1 H NMR (CDCl3): δ 4.53–4.61 (m, 6H); 5.16–5.46 (m, 6H); 6.02–6.20 (m, 3H); 6.58 (d, 2H); 6.91–6.96 (m, 1H). NMR spectra with peak assignments are available in Supporting Information.

2. Experimental method 2.1. Materials and synthesis 2.1.1. Materials Vanillyl alcohol (4-hydroxy-3-methoxy benzyl alcohol, 99%), Gastrodigenin (4-hydroxybenzyl alcohol, 99%), 4-methyl catechol (4methylbenzene-1,2-diol, 98%), Pyrogallol (benzene-1,2,3,-triol, 99%), ethylene carbonate (EC) (99%), diethyl carbonate (DEC) (99%) and allyl bromide (99%, stabilized) were purchased from Acros Organics. LiPF6 (lithium hexafluorophosphate, 98%) and Amberlyst 15(H) were purchased from Alfa Aesar and TEBAC (benzyltriethylammonium chloride, 99%), photoinitiator 1-hydroxycyclohexyl phenyl ketone (99%) and deuterated dimethyl sulfoxide d6 (99.8%) were purchased from Sigma Aldrich. From VWR was purchased potassium carbonate, sodium chloride, acetone, dichloromethane, n-hexanes and ethyl acetate, while sodium hydroxide and sodium sulfate were obtained from Fisher Scientific. Thiol terminated ETTMP (ethoxylated-trimethylolpropan tri(3 mercaptopropionate)) with molecular weight of 1300 g mol−1 was purchased from Bruno Bock Thiochemicals.

2.1.5. Synthesis of polymer electrolytes Each monomer (DAVA, DAGd, TAVC or TAPg) was mixed with thiol ETTMP 1300 in a 1:1 allyl:thiol ratio. The photoinitiator (2 wt% relative to the total weight) was added and the mixture stirred until dissolved. The resin was UV cured in a UVP Ultra-Violet Crosslinker model CL-1000 L at 365 nm and 9000 μJ cm−2 for multiple sessions depending on sample thickness. Sample bars for DMA analysis were made according to previous work in a UV-transmissible Plexiglas and silicone mold [10]. Thin films of 0.4 mm thickness were drawn with an applicator rod, cut to appropriate dimensions and measured with a digital caliper. To create solid polymer electrolytes, LiPF6 was added to the resin in varying mole percentages relative to the molar amount of monomer and thiol. Complete dissolution was ensured with aid of a Thinky Mixer. The polymer electrolytes were synthesized and assembled for further testing in a dry room with < 1% humidity.

2.1.2. Synthesis of allylated vanillyl alcohol and gastrodigenin Vanillyl alcohol (VA) and gastrodigenin (Gd) were allylated in a procedure adapted from Fache et al. [31] as described in our previous work [10]. The resulting products were di-allylated vanillyl alcohol (DAVA), a light yellow liquid with a 79% yield, and di-allylated gastrodigenin (DAGd), a clear liquid with an 81% yield. See previous work for NMR assignments and spectra [10].

2.2. Polymer characterization 2.1.3. Synthesis and allylation of vanillyl alcohol- 4 methylcatechol A trifunctional bisphenol was synthesized from the reaction of vanillyl alcohol and 4-methylcatechol according to a procedure adapted from Hernandez et al. [17]: vanillyl alcohol (1 equivalent), 4-methylcatechol (1 equivalent) and Amberlyst (10 wt%) were added to a roundbottomed flask with dichloromethane. The flask was heated to 50 °C under reflux for while stirring. After 24 h, the reaction was cooled and the Amberlyst was filtered. The solution was reduced by half under pressure and stored in the refrigerator to allow the product to crystallize. The resulting product was vanillyl alcohol-4-methyl catechol (VAMC), a white powder with a 95% yield. 4-(4-Hydroxy-3-methoxybenzyl)-5-methylbenzene-1,2-diol (VAMC) (C15H16O4): White solid. 1H NMR (DMSO-d6): δ 2.01 (s, 3H); 3.61 (s, 2H); 3.67 (s, 3H); 6.43 (m, 2H); 6.49 (m, 1H); 6.64 (m, 2H); 8.45 (d, 2H); 8.65 (s, 1H). To allylate the trifunctional bisphenol, a procedure was adapted from Paduraru et al. [32] as follows: VAMC (1 equivalent) was placed in a round-bottomed flask with potassium carbonate (4 equivalents) and acetone. The flask was plugged with a rubber stopper and needle. After stirring for 10 min while heating to 50 °C, allyl bromide (4 equivalents) was added. The reaction was stirred on heat overnight. The potassium carbonate was filtered with a Buchner funnel and the product purified on a flash chromatography system with n-hexanes and ethyl acetate. The resulting product was tri-allylated vanillyl alcohol-4-methyl catechol (TAVC), a white powder with 80% yield. 1,2-Bis(allyloxy)-4-(4-(allyloxy)-3-methoxybenzyl)-5-methylbenzene (TAVC) (C24H28O4): White solid. 1H NMR (DMSO-d6): δ 2.08 (s, 3H); 3.67 (s, 3H); 3.74 (s, 2H); 4.47 (m, 6H); 5.14–5.23 (m, 3H);

2.2.1. FTIR analysis Fourier transform infrared (FT-IR) spectroscopy was performed on a Nicolet iS50 in the near IR range from 4000 to 8000 cm−1 in transmission mode at room temperature with a resolution of 8 cm−1. Additional attenuated total reflectance (ATR) FT-IR in the mid IR range was performed on a Nicolet 6700 from 650 to 4000 cm−1 with an ATR crystal accessory. Extent of cures of the polymer samples were calculated from the decrease in the absorption band associated with the allylated monomer at 6165 cm−1 as described in previous work [10]. 2.2.2. Density measurements Archimedes' principle [33] was employed to measure the density of each neat polymer sample in distilled water. Measurements were triplicated and the average density was used in the theory of rubber of elasticity to calculate the molecular weight between crosslinks (equation shown in next section). 2.2.3. Dynamic mechanical analysis (DMA) A TA Instruments DMA Q800 was used at 1 Hz frequency, 7.5 μm amplitude, 0.35 Poisson's ratio and heating rate of 2 °C min−1 over a range of −70 °C to 70 °C to obtain the glass transition temperature (Tg) and storage modulus at 25 °C of each neat polymer. Samples of dimensions 35 × 12 × 5.0 mm3 were used and samples were run in triplicate. The molecular weight between crosslinks (Mc), in g mol−1, was measured as the point at which the storage modulus began to increase in the 5 mm sample in accordance with the theory of rubber 3

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

reached steady state. EIS with 20 mV AC and 0 V DC from 1 MHz to 0.1 Hz was taken before and after DC polarization. The cation transference number was then calculated as [37–38]:

elasticity [34] using the following equation:

3ρRT E′ = Mc

(1) −3

where E′ is the storage modulus in MPa, ρ is the density in g cm , R is the universal gas constant, 8.314 MPa cm3 K−1 mol−1, and T is the temperature in Kelvin. The crosslink density (ν) was calculated as the following:

ν=

ρ Mc

(2)

2.3.3. Dielectric constants The dielectric constants of the neat polymers were determined through EIS. The neat polymers were cut into disks of 8 mm diameter and assembled in Swagelok cells with stainless steel electrodes. EIS with 20 mV AC and 0 V DC was run from 1 MHz to 100 Hz. The complex capacitances were calculated from impedances according to the following [39–41]:

|Z (ω)| =

2.2.5. Thermogravimetric analysis (TGA) A TA Instruments Discovery TGA 550 was used to monitor the decomposition of the neat polymers and determine the initial decomposition temperature at 95% weight loss (IDT95%) and the temperature at maximum weight loss rate (Tmax). Samples of approximately 10 mg were placed on platinum pans and heated from 40 °C to 700 °C at 10 °C min−1 in N2 (25 mL min−1 sample gas flow, 40 mL min−1 purge gas flow.) Samples were run in triplicate to ensure truthful averages and standard deviations.

Z ′2 + Z ′′2

(5)

C′ =

−Z′ ′ (ω) ω |Z (ω)|2

(6)

C′ ′ =

Z′ (ω) ω |Z (ω)|2

(7)

where Z′ and Z″ are the real and imaginary parts of complex impedance noted as Z(ω), C′ and C″ are the real and imaginary parts of complex capacitance and ω is angular frequency equal to 2πf. The dielectric constants were calculated from [42–43]:

ε′ =

Cd ε0 A

(8)

where C is capacitance, d is sample thickness, ε0 is the permittivity of free space, 8.854E−12 F m−1 and A is the cross-sectional area of the electrode. Samples were run in triplicate to ensure truthful averages and standard deviations.

2.3. Electrochemical characterization 2.3.1. High frequency electrochemical impedance spectroscopy Solid polymer electrolytes were evaluated for conductivity using high frequency electrochemical impedance spectroscopy (EIS) on a Solartron SI 1260 Impedance/Gain-phase Analyzer with 1296 Dielectric Interface and 12962 sample holder with 19 mm brass electrodes, courtesy of the US Army CCDC C5ISR Center in Aberdeen Proving Ground, Maryland. Samples of approximately 0.4 mm thickness were cut into circles of 19 mm diameter. A 100 mV AC voltage and 0 V DC voltage were applied over a frequency range from 30 MHz to 100 kHz. The high frequency intercepts of the resulting Nyquist plots yielded the resistances of the polymer electrolytes and allowed for calculation of the conductivities through the following equation [35–36]:

l RA

(4)

where Iss is the current at steady state, ΔV is the applied potential, I0 is the initial current calculated by Ohm's law, R0 is the initial resistance and Rss is the steady state resistance. Samples were run in triplicate to ensure truthful averages and standard deviations.

2.2.4. Differential scanning calorimetry (DSC) A TA Instruments DSC 2500 was used to obtain the glass transition temperatures (Tg) of the solid polymer electrolytes. Samples of 5–10 mg were sealed in aluminum pans and run from −80 °C to 50 °C at a rate of 10 °C min−1 in a N2 atmosphere. The temperature ramp was completed twice to erase thermal history and the Tg was obtained from the second ramp through. Samples were run in triplicate to ensure truthful averages and standard deviations. Additional samples were run from −80 °C to 100 °C at a rate of 10 °C min−1 in a N2 atmosphere to verify absence of crystallinity.

σ=

Iss (∆V − I0 R 0 ) I0 (∆V − Iss Rss )

t+ =

2.3.4. Linear sweep voltammetry The onset decomposition voltages and electrochemical stabilities of the solid polymer electrolytes were determined through linear sweep voltammetry as follows: the polymer electrolytes were assembled in Swagelok cells between two stainless steel electrodes. To evaluate the upper voltage stability limit of the individual polymer electrolytes, a constant rate voltage was swept through and a current-voltage curve was generated. From the curve, the voltage at intersection of the rising current and the baseline current tangential lines was the decomposition voltage of the electrolyte [44–45]. Samples were 0.4 mm thick and 8 mm in diameter and run from 0 to 10 V at 100 mV s−1 on the Arbin MSTAT4 Cycler.

(3)

where l is the polymer electrolyte thickness, R is the bulk resistance and A is the electrode area in contact with the polymer. Samples were run in triplicate to ensure truthful averages and standard deviations.

3. Results and discussion SPEs are created with the four allylated, aromatic compounds: diallylated vanillyl alcohol (DAVA), di-allylated gastrodigenin (DAGd), tri-allylated pyrogallol (TAPg) and tri-allylated vanillyl alcohol-4-methyl catechol (TAVC), shown in Fig. 1. All polymers are cured with appropriate amounts of ETTMP 1300 and the photo-initiator using the same curing procedure, see experimental procedure in Section 2.1.5. Extent of cures of the thiol-ene SPEs are verified by FT-IR in the near-IR range. All systems exhibit near complete disappearance of the reactive allyl bands demonstrating consistent near complete conversions, all at least 88% extent of cure and most at least 90% extent of cure, see supplemental information for tabulated values. Additionally, the disappearance of the allyl bands at 6165 cm−1 in the near IR range and the disappearance of the free thiol band at 2550 cm−1 in all the samples are

2.3.2. DC polarization The lithium ion transference numbers (t+) of the solid polymer electrolytes were determined by AC impedance and DC polarization analysis. The polymer electrolytes were sandwiched between lithium disks, 16 mm in diameter, in Teflon cells. The cells were analyzed using EIS on a Solatron SI 1260 Impedance/Gain-phase Analyzer with 1287 Electrochemical Interface and fittings provided by ZView2 software from Princeton Applied Research. The cells were also monitored on an Arbin MSTAT4 Cycler (4 channels, 110 V, 1 phase, 50/60 Hz, 1 kVA), both instruments courtesy of the US Army CCDC C5ISR Center in Aberdeen Proving Ground, Maryland. Polarization of the cells was accomplished by applying 0.01 V DC over 24 h or until the current 4

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

Table 1 Thermomechanical, network and thermogravimetric properties of the neat polymers. Neat polymera

Tg (°C) (via DSC)b

Tg (°C) (via DMA)c

ρ (g cm−3)d

DAVA DAGd TAPg TAVC

−48.3 −52.0 −48.5 −41.3

−37.1 −42.1 −40.0 −33.4

1.15 1.15 1.15 1.17

a b c d e f g h i

± ± ± ±

0.4h 0.2h 1.3h 0.5h

± ± ± ±

0.4i 0.8i 0.9h 0.5h

± ± ± ±

0.01i 0.00i 0.02h 0.0h 1

ν (mmol cm−3)e 0.26 0.21 0.58 0.80

± ± ± ±

0.03i 0.02i 0.09h 0.02h

IDT95% (°C)f 308 308 332 329

± ± ± ±

2h 3h 16h 2h

Tmax (°C)g 373 374 372 374

± ± ± ±

2h 1h 1h 2h

Polymers cured with ETTMP 1300 and 2 wt% photo-initiator. Tg measured through Differential Scanning Calorimetry (DSC). Tg measured through Dynamic Mechanical Analysis (DMA) as temperature at peak of loss modulus. Density measured according to Archimedes' principle. Crosslink density, ν = ρ/Mc where molecular weight between crosslinks, Mc, is measured at point where storage modulus begins to increase. Initial Decomposition Temperature (IDT) at 5% decomposition as determined by TGA. Temperature at maximum weight loss rate as determined by TGA. Reference [47]. Reference [10].

used to confirm no disruption of the cures by interaction of the monomers with LiPF6 salt, see supplemental information for spectra. The four allylated monomers vary in functionality, from di-functional DAVA and DAGd to tri-functional TAVC and TAPg. DAVA and DAGd differ by a methoxy substituent group, whereas TAVC and TAPg possess differing aromatic content and position of the allyl functional moieties relative to each other. Thermomechanical analysis of the neat polymers reveals that aromatic content, not allylated monomer functionality, most affected the formation of the crosslinked polymer network, see Table 1. The neat polymer containing TAVC possesses both the highest Tg and crosslink density. Though TAPg is also tri-functional, the resulting polymer network possesses a similar Tg and crosslink density as polymers with di-functional DAVA and DAGd. As shown in previous work, a polymer with a low crosslink density is desirable for this application [10]. However, polymers with great stability are also desired, making exploration of the TAVC-containing polymers worthwhile due to the known structural and thermal stability imparted by increased aromatic content [14], see IDT95% and Tmax values in Table 1. The thermal stability of these polymers is a great advantage compared to commonly used LIB systems in which the polymer separator and liquid electrolyte are susceptible to deterioration at 130 °C–150 °C [46]. Thiol ETTMP 1300 is a relatively high molecular weight comonomer, compared to other thiol-containing molecules explored in previous work [10], that contains many repeating ethylene oxide units. The presence of repeating ethylene oxide units in polymer electrolytes have been shown to increase ionic transport and conducting ability of the resulting polymer networks through coordination with Li+ ions [2,22,28]. Interactions of the LiPF6 salt with the cured polymer networks are monitored by FT-IR. Dissolution of LiPF6 is verified in the mid-IR range with the appearance of a peak at 839 cm−1 corresponding to the PF6- free anion, see Fig. 3 [48]. Distortion of the C]O band at 1730 cm−1 is indicative of Li+ ion coordination with the few C]O sites in ETTMP 1300, see Fig. 4 [2]. Similarly, formation of a side peak on the CeO stretching at 2870 cm−1 is indicative of Li+ ion coordination with the many CeO sites in the repeating ethylene oxide units in the thiol, also see Fig. 4. The wide ranges of LiPF6 concentrations utilized results in exploration of SPEs containing a wide range of [EO]/[Li] ratios: from 240/1 of the TAVC-containing SPE with 5 mol% LiPF6 to 11/1 of the TAPg-containing SPE with 110 mol% LiPF6. The tabulated values of all [EO]/[Li] ratios are included in the supporting information. Even in thermoset networks, the presence of long PEO segments can result in formation of crystalline regions within the polymer network [9]. It has been shown that in PEO-containing polymer electrolytes with low concentrations of LiPF6, there are not many Li+ ions to coordinate with the PEO segments, thus, increasing the likelihood of crystalline phase formation [11]. Lithium transport is known to be hindered by polymer crystallinity [5]. Therefore, DSC traces of the SPEs are analyzed for

Fig. 3. FT-IR spectra of appearance of peak at 839 cm−1 corresponding to dissociated PF6− anion. Exemplary spectra of DAVA + 1300 thiol-ene polymer at four different LiPF6 concentrations. For spectra of the other SPEs, see Supplementary Information and Fig. 6. Figure adapted from Reference [47].

presence of a melting point as an indication of crystallinity within the SPEs [2]. As can be seen in Fig. 5, the thermoset SPEs demonstrate defined glass transitions and no melting points, confirming the absence of crystalline regions in the thermoset SPEs of all LiPF6 concentrations. The interaction of the dissociated lithium salt with the polymer network is known to affect the Tg of the resulting conductive polymer electrolyte [2,25]. For each polymer formulation, a range of LiPF6 concentrations, expressed in mol%, are analyzed for Tg via DSC and for room temperature conductivity. As seen in Fig. 6, the addition of LiPF6 in the resin increases the Tg of the SPEs. This is due to coordination between the Li+ ions and the ether oxygens that has a stabilizing effect inhibiting motion of the crosslinked polymer chains [2]. SPEs containing TAVC sees the smallest change in Tg indicating that the more stable crosslinked polymer network formed by the TAVC, as seen by the higher neat Tg, is not as affected by disruptive physical crosslinking by the dissociated lithium salt. Polymers containing difunctional DAVA and DAGd have the largest change in Tg (−48.3 °C for the neat DAVA containing polymer to −29.4 °C for the 109 mol% LiPF6 DAVA containing SPE and −52.0 °C for the neat DAGD containing polymer to −31.53 °C for the 106 mol% LiPF6 DAGd containing SPE), whereas the Tg difference for the polymer containing trifunctional TAPg is less (−48.5 °C for the neat polymer to −35.27 °C for the 110 mol% LiPF6 5

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

containing polymer). However, all three systems see similar jumps in Tg from the neat polymer to the 10 mol% LiPF6 concentration. Room temperature conductivities are lowest in the SPEs containing bisphenolic TAVC, 1.04E−4 S cm−1 at 50 mol% LiPF6, and highest in those with DAVA and TAPg, 7.65E−4 S cm−1 at 50 mol% LiPF6 and 6.21E−4 S cm−1 at 90 mol% LiPF6, respectively, illustrating the importance of a low Tg, flexible network (see Fig. 6). Overall, with roomtemperature conductivities on the order of 10−4 S cm−1, all of the SPEs exhibit excellent conducting ability compared to similar solid or gel polymer electrolytes (the majority exhibiting 10−6–10−4 S cm−1) though conductivities are still lower than liquid electrolyte systems [1,13]. The SPEs containing TAVC show an increase and leveling off in conductivity as the mol% of LiPF6 is increased. The DAVA, DAGd and TAPg containing SPEs, however, display a more irregular pattern of conductivity as the LiPF6 content in the resin is increased. The DAVA and DAGd containing SPEs both drop in conductivity after 10 mol% LiPF6 but exhibit a resurgence at the 50 mol% LiPF6 concentration. After 50 mol% LiPF6, conductivity decreases again. The TAPg containing SPEs have a similar spike in conductivity at the low 25 mol% LiPF6 concentration, followed by a slight drop until another spike at 75 mol% LiPF6 and 90 mol% LiPF6 after which the conductivity drops again significantly. As the Tg of the SPEs increases due to interaction with the ethylene oxide units and Li+ cations, the polymer segmental motion and ability to assist in ionic transport is inhibited causing a drop in conductivity at the 100 and 110 mol% LiPF6 concentrations which have the highest Tgs [25]. However, the fluctuation in conductivity across the low and mid concentration ranges of lithium salt in the cured polymer networks is an interesting result further explored through lithium ion transference numbers. The lithium ion transference number, t+, is a measure of the amount of current carried through the polymer electrolyte by the cation Li+ species. A high t+ is desirable as it is related to prevention of salt concentration gradients and salt precipitation that can occur when the ionic conductivity is carried primarily by anionic motion, i.e. at low t+ [49]. Therefore, it is important to characterize conductivity as well as t+ values when evaluating the feasibility of SPEs. The t+ numbers of the most conductive SPEs of each formulation are evaluated through use of AC impedance and DC polarization techniques [37–38]. Table 2 shows Tg, conductivity and lithium ion transference values of the SPEs to discern the impact of lithium salt concentration in the cured resins on cationic flow. The SPEs containing difunctional DAVA and DAGd exhibit lower t+ numbers with 10 mol% LiPF6 than with 50 mol% LiPF6 even though conductivities at these concentrations are similarly high. The low t+ number indicates that the conductivity at the 10 mol% LiPF6 concentration is mostly a result of anionic motion through the polymer. It is hypothesized that with the small addition of LiPF6 salt, the few Li+ ions in the polymer are strongly coordinated with the carbonyl and ethylene oxide groups and are not free to move throughout network. This is confirmed by the slight increase in Tg from the neat polymer to the 10 mol% LiPF6 containing polymer. Instead, the anions are free to move. When the concentration of Li salt is increased to 50 mol% LiPF6, though the Tg increases and segmental motion is slightly inhibited, the t+ number also increases, indicating that some Li+ ions coordinate with the polymer chains but there are enough free Li+ ions in the system that are free to move. After 50 mol% addition of LiPF6 salt, the Tg increases and restriction of segmental motion is so great that conductivity drops, see the 65 mol% LiPF6 DAVA and DAGD containing polymers conductivities shown in Fig. 6. This pattern is observed in both DAVA and DAGd containing SPEs, however, those with DAVA have higher overall conductivity values, indicating that the methoxy substituent on the aromatic ring of DAVA aids slightly in the ionic transport. Looking at the trifunctional TAPg containing SPEs, a similar pattern to that of DAVA- and DAGd- containing SPEs is seen where the increase in LiPF6 concentration leads to an increase in conductivity up to a certain point after which the polymer segmental motion is so restricted

Fig. 4. FT-IR spectra of shift in shape, highlighted by arrows, of C]O band at 2870 cm−1 and CeO band at 1730 cm−1 due to Li+ ion coordination. Exemplary spectra of DAVA + 1300 thiol-ene polymer at four different LiPF6 concentrations. For spectra of other SPEs, see Supplementary Information. Figure adapted from Reference [47].

Fig. 5. DSC traces of SPEs with varying concentrations of LiPF6. As can be seen by the distinct Tg and absence of melting point, all the SPEs lacked crystallinity. Exemplary traces of TAPg-containing SPEs; all synthesized SPEs displayed similar traits. Figure adapted from Reference [47].

6

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

Fig. 6. Conductivity and Tg data for SPEs with the four monomers. The pink bars correspond to conductivity (×103 S cm−1) on the left y axis. The blue dots and line correspond to the Tg (°C) on the right y axis. Samples were run in triplicate to ensure accurate average and standard deviation analysis. Error bars are displayed for each sample. Figure adapted from Reference [47]. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

triplet is seen as an additional peak at 876 cm−1 next to the PF6- anion at 839 cm−1 in the mid-IR range, see Fig. 7 for FT-IR spectra and tables with normalized areas under the 876 cm−1 peaks for quantitative comparison of the triplet peaks [48]. The larger area under the additional peak at 876 cm−1 confirms that anion triplets are most abundant in the 90 mol% LiPF6 TAPg containing SPEs; however, the peak does exist in the 25 mol% LiPF6 and 75 mol% LiPF6 TAPg containing SPEs as well. The same can be said for the DAVA- and DAGd- containing SPEs (see Fig. 3 and supporting information.) Transference numbers for similar polymer electrolyte systems have been shown to be in the range of 0.3–0.4 [49] which indeed matches the results of the best-performing SPEs obtained here. Ideally, t+ values

that the conductivity drops, see Fig. 6. The t+ values seem to follow the same pattern where the increase in concentration from 75 mol% LiPF6 to 90 mol% LiPF6 lead to a drop in t+ though the conductivities are comparable. Whereas the low t+ values at the low concentrations in the DAVA- and DAGd-containing SPEs are a result of lack of free cations and surplus of free anions, it is hypothesized that the low t+ at the high concentration of LiPF6 in TAPg-containing SPEs are a result of ion cluster formation of the surplus of free anions. It has been shown that lithium salt species can form negatively charged triple ion clusters in PEO-based polymer electrolytes [37] and in lithium salt/ionic liquid mixtures [50] resulting in a reduction in t+ values. The presence of these negatively charged triplets are confirmed via FT-IR where the

Table 2 Tg, conductivity and t+ for the SPEs with the four monomers. Solid polymer electrolyte

LiPF6 in resin (mol%)

Tg (°C) (via DSC)

Conductivity (×103 S cm−1)

Lithium ion transference number

DAVA + ETTMP 1300

10 50 10 50 25 75 90 10 25 50 75 100

−47.4 −36.4 −48.1 −37.5 −42.7 −36.3 −37.7 −40.2 −40.4 −34.9 −35.6 −32.1

0.60 0.77 0.47 0.46 0.37 0.51 0.62 0.07 0.08 0.10 0.09 0.09

0.07 0.26 0.11 0.31 0.22 0.23 0.05 0.18 0.39 0.08 0.05 0.01

DAGd + ETTMP 1300 TAPg + ETTMP 1300

TAVC + ETTMP 1300

± ± ± ± ± ± ± ± ± ± ± ±

0.2 1.4 0.1 0.1 0.3 0.2 0.2 0.4 1.1 0.0 0.2 0.6

Adapted from Reference [47]. 7

± ± ± ± ± ± ± ± ± ± ± ±

0.10 0.06 0.06 0.01 0.01 0.11 0.14 0.00 0.00 0.01 0.00 0.01

± ± ± ± ± ± ± ± ± ± ± ±

0.03 0.04 0.04 0.06 0.03 0.00 0.01 0.02 0.14 0.02 0.02 0.00

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

Fig. 7. Mid IR FT-IR of TAPg (top) and TAVC (bottom) showing dissociated PF6− anion peak at 839 cm−1 and the formation of ion complex peak at 876 cm−1, peaks highlighted by arrows. Figure adapted from Reference [47].

in the range of 0.7 and above are most desirable [49]. Previous reasoning for the low transference numbers in PEO-based systems have suggested that the Li+ ion too strongly coordinates with the EO groups resulting in the current largely carried by the anions [49]. This is observed at the low LiPF6 concentrations in these SPEs; however, at the higher concentrations, the presence of the negatively charged ion clusters likely causes the drop in t+ value. Interestingly, the TAVC containing SPEs, which exhibit the lowest conductivity values due to the higher Tgs, have the highest average t+ values, reaching 0.39. It is hypothesized that this occurs for two related reasons: one involving free anionic motion and one involving ion aggregation. In SPE systems such as those explored here, anions do not move along the polymer chains, like Li+ ions, but instead through free volume within the polymer matrix [3]. The more crosslinked network of the TAVC containing polymers, evidenced by higher Tgs and higher crosslink densities as shown in Table 1, reduces anionic motion between chains, allowing for more of the current to be carried by Li+ cations than in the DAVA, DAGd and TAPg containing SPEs. Additionally, looking at the FT-IR spectra in the mid-IR range of the 876 cm−1 ion cluster and 839 cm−1 dissociated PF6− peaks, the TAVC containing SPEs also have the lowest concentrations of anion clusters, see Fig. 7. Ion aggregation is suspected to be related to the polarity of the polymer matrix with more polar polymers allowing for more ion dissociation and lower aggregation [26]. However, polymers with too much of an increase in polarity are then said to be limited by segmental motion of the network resulting in reduction in ion conductivity [49].

Fig. 8. Average dielectric constants with standard deviations across frequency range of 100 Hz to 1 MHz for the neat polymers synthesized with the four aromatic monomers. Figure adapted from Reference [47].

8

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

Table 3 Average dielectric constants of neat polymers.

Table 4 Linear sweep voltammetry results of SPEs.

Neat polymer

ε′ at 1 kHz

Solid polymer electrolyte

Onset decomposition voltage (V)

Current at 6 V (A)

DAVA DAGd TAPg TAVC

25 26 25 12

DAVA + ETTMP 1300 DAGd + ETTMP 1300 TAPg + ETTMP 1300 TAVC + ETTMP 1300

5.0 5.5 6.0 1.5

1.8E−6 7.6E−6 5.0E−6 8.1E−6

± ± ± ±

10 5 5 4

Adapted from Reference [47].

Adapted from Reference [47].

decomposition can cause inaccurate trends in reported decomposition values. Therefore, current values at a specific voltage of 6 V are also reported in Table 4 in order to provide a more complete picture of electrochemical stability. As seen in Fig. 9 and Table 4, the SPEs show great electrochemical stability. Though onset decomposition voltages are observed at 5–6 V, the current-voltage curves do not exhibit spikes in current up to 10 V, i.e., the current even at 6 V remains relatively low. Conventional organic carbonate electrolytes can decompose at around 4.5 V [52] and even other polymer systems reported in literature can display sharp decomposition spikes at 5 V [2,22]. Therefore, the SPEs fabricated in this study have the ability to withstand a large operating voltage window and can potentially provide safer alternatives to incumbent liquid electrolytes.

Therefore, there exists an optimal intermediate neat polymer polarity that will limit ion aggregation [51]. The dielectric constants over a frequency range of 100 Hz–1 MHz are determined as a proxy measure of polarity, see Fig. 8. As seen in Table 3, the average dielectric constants at 1 kHz of the single aromatic DAVA, DAGd and TAPg containing neat polymers are around 25 while the average dielectric constant of the TAVC containing neat polymer is 12. According to the literature, a dielectric constant of 9 is on the high end for similar polymer systems [26]. The DAVA, DAGd and TAPg containing polymers are therefore well past the optimal range of neat polymer polarity, lending themselves to higher ion aggregation and lower cationic transport. The TAVC containing polymers, with the lowest neat polymer dielectric constants of the four, are closer to the range of optimal neat polymer polarity. Previous research on dielectric constants of neat polymers for electrolytes has hypothesized that conductivity increases with an increase in dielectric constant [23], whereas other groups have used computer simulations to identify an upper limit of desirable neat polymer polarity for such applications [49]. The authors of this study consider these results to be confirmation of the latter, i.e., there exists an optimal range for neat polymer polarity of the SPE systems. It is concluded that even in very low Tg polymer electrolytes (low Tg being an indicator of segmental mobility), if the ideal neat polymer polarity range is not accessed, cationic transport will not be optimized. In addition to allowing for flow of ions, the SPEs must display electrochemical stability over a wide voltage range. LSV was performed on the SPEs to determine the onset decomposition voltages through assembly in Swagelok cells. The SPEs were then subjected to a voltage sweep at a constant rate from 0 V to 10 V as shown in Fig. 9. Onset decomposition voltages are determined as the voltage at the intersection of the rising current and the baseline current tangential lines, as presented in Table 4. Slight increases in current not indicative of large

4. Conclusion The thiol-ene SPEs were prepared by mixing LiPF6 salt in resins containing functionalized, potentially bio-based, aromatic monomers and thiol terminated ETTMP 1300. After a fast and easy UV cure, the polymer electrolytes were analyzed for thermomechanical and electrochemical properties. Even at very low LiPF6 concentrations, the thermoset networks did not display any formation of crystalline regions, as evidenced by DSC analysis. The SPEs containing monophenolic DAVA, DAGd and TAPg had the lowest Tgs and crosslink densities and the highest conductivities, with the highest reaching 7.65 × 10−4 S cm−1. The low Tg polymers with long, flexible polymer chains allowed for better total ionic transport, resulting in the higher conductivities. The SPEs containing bisphenolic TAVC had the highest Tgs and crosslink densities and therefore the lowest conductivities, with the highest reaching 1.04 × 10−4 S cm−1. However, the larger crosslink densities of the TAVC containing polymer networks favored specifically cationic transport which resulted in desirable higher average t+ values. Additionally, the optimal polymer polarity, as indicated by neat polymer dielectric constant, of the TAVC containing polymers contributed to reduced ion aggregation. This study reveals that extremely low Tg or extremely polar neat polymers do not necessarily make for the best SPE; instead, there exists an optimal median for both parameters. Furthermore, higher crosslink density can assist with cationic transport in non-crystalline polymer networks. Linear sweep voltammetry reveals that these aromatic SPEs display great electrochemical stability at or above the 5 V threshold, providing potentially safer alternatives to incumbent liquid electrolyte systems. The promising results of the aromatic thiol-ene SPEs demand further electrochemical analysis. Assembly and galvanostatic cycling in coin cells must be performed to evaluate compatibility with various cathode and anode materials. Further, cycling in Li/Li symmetric cells to evaluate dendrite growth inhibition should commence.

Funding sources Fig. 9. Linear sweep voltammetry results of SPEs subject to a constant rate voltage sweep from 0 V to 10 V. Though currents begin to increase at ~5 V, the values remain low (< 4E−5 A) even at 10 V.

This was work was supported by The Henry M. Rowan College of Engineering and the Department of Defense SMART Scholarship program. 9

Solid State Ionics 345 (2020) 115161

E.A. Baroncini, et al.

Declaration of competing interest

(2010) 1540–1573. [21] R. Muchakayala, S.H. Song, S. Gao, X.L. Wang, Y.H. Fan, Structure and ion transport in an ethylene carbonate-modified biodegradable gel polymer electrolyte, Polym. Test. 58 (2017) 116–125. [22] J.R. Nair, C. Gerbaldi, G. Meligrana, R. Bongiovanni, S. Bodoardo, N. Penazzi, P. Reale, V. Gentili, UV-cured methacrylic membranes as novel gel-polymer electrolyte for Li-ion batteries, J. Power Sources 178 (2008) 751–757. [23] S.B. Aziz, T.J. Woo, M.F.Z. Kadir, H.M. Ahmed, A conceptual review on polymer electrolytes and ion transport models, J Sci 3 (2018) 1–17. [24] W.F. Kuan, R. Remy, M.E. Mackay, T.H. Epps, Controlled ionic conductivity via tapered block polymer electrolytes, RSC Adv. 5 (2015) 12597–12604. [25] Y. Tominaga, K. Yamazaki, Fast Li-ion conduction in poly(ethylene carbonate)based electrolytes and composites filled with TiO2 nanoparticles, Chem. Commun. 50 (2014) 4448–4450. [26] B.K. Wheatle, J.R. Keith, S. Mogurampelly, N.A. Lynd, V. Ganesan, Influence of dielectric constant on ionic transport in polyether-based electrolytes, ACS Macro Lett. 6 (2017) 1362–1367. [27] H. Nithya, S. Selvasekarapandian, D.A. Kumar, A. Sakunthala, M. Hema, P. Christopherselvin, J. Kawamura, R. Baskaran, C. Sanjeeviraja, Thermal and dielectric studies of polymer electrolyte based on P(ECH-EO), Mater. Chem. Phys. 126 (2011) 404–408. [28] M. Willgert, M.H. Kjell, G. Lindbergh, M. Johansson, New structural lithium battery electrolytes using thiol-ene chemistry, Solid State Ionics 236 (2013) 22–29. [29] A. Chiappone, C. Gerbaldi, I. Roppolo, N. Garino, R. Bongiovanni, Degradable photopolymerized thiol-based solid polymer electrolytes towards greener Li-ion batteries, Polymer 75 (2015) 64–72. [30] S.D. Gong, Y. Huang, H.J. Cao, Y.H. Lin, Y. Li, S.H. Tang, M.S. Wang, X. Li, A green and environment-friendly gel polymer electrolyte with higher performances based on the natural matrix of lignin, J. Power Sources 307 (2016) 624–633. [31] M. Fache, E. Darroman, V. Besse, R. Auvergne, S. Caillol, B. Boutevin, Vanillin, a promising biobased building-block for monomer synthesis, Green Chem. 16 (2014) 1987–1998. [32] P.A. Paduraru, R.T.W. Popoff, R. Nair, R. Gries, G. Gries, E. Plettner, Synthesis of substituted alkoxy benzene minilibraries, for the discovery of new insect olfaction or gustation inhibitors, J. Comb. Chem. 10 (2008) 123–134. [33] D. Halliday, R. Resnick, J. Walker, Fundamentals of Physics, 6th ed., John Wiley & Sons, New York, 2001. [34] L.H. Sperling, Introduction to Physical Polymer Science, 4th ed., Wiley, Hoboken, N.J, 2006. [35] Basics of Electrochemical Impedance Spectroscopy; Gamry Instruments, (2018), pp. 1–33. [36] X.M. Qian, N.Y. Gu, Z.L. Cheng, X.R. Yang, E.K. Wang, S.J. Dong, Methods to study the ionic conductivity of polymeric electrolytes using a.c. impedance spectroscopy, J Solid State Electr 6 (2001) 8–15. [37] D.M. Pesko, K. Timachova, R. Bhattacharya, M.C. Smith, I. Villaluenga, J. Newman, N.P. Balsara, Negative transference numbers in poly(ethylene oxide)-based electrolytes, J. Electrochem. Soc. 164 (2017) E3569–E3575. [38] S. Zugmann, M. Fleischmann, M. Amereller, R.M. Gschwind, H.D. Wiemhofer, H.J. Gores, Measurement of transference numbers for lithium ion electrolytes via four different methods, a comparative study, Electrochim. Acta 56 (2011) 3926–3933. [39] G.A.M. Ali, M.M. Yusoff, E.R. Shaaban, K.F. Chong, High performance MnO2 nanoflower supercapacitor electrode by electrochemical recycling of spent batteries, Ceram. Int. 43 (2017) 8440–8448. [40] R. Abd Aziz, I.I. Misnon, K.F. Chong, M.M. Yusoff, R. Jose, Layered sodium titanate nanostructures as a new electrode for high energy density supercapacitors, Electrochim. Acta 113 (2013) 141–148. [41] H. Cesiulis, The study of thin films by electrochemical impedance spectroscopy, in: I. Tiginyanu, P. Topala, V. Ursaki (Eds.), Nanostructures and Thin Films for Multifunctional Applications, Springer International Publishing, Switzerland, 2016, pp. 3–42 (Chapter 1). [42] J. Irvine, D.C. Sinclair, A.R. West, Electroceramics: characterization by impedance spectroscopy, Adv. Mater. 2 (1990) 132–138. [43] S. Nasreen, G. Treich, M. Baczkowski, A. Mannodi-Kanakkithodi, Y. Cao, R. Ramprasad, G. Sotzing, Polymer dielectrics for capacitor application, KirkOthmer Encyclopedia of Chemical Technology, John Wiley & Sons, Inc, 2017. [44] R.Y. Miao, et al., PVDF-HFP-based porous polymer electrolyte membranes for lithium-ion batteries, J. Power Sources 184 (2008) 420–426. [45] S. Monisha, et al., Investigation of bio polymer electrolyte based on cellulose acetate-ammonium nitrate for potential use in electrochemical devices, Carbohydr. Polym. 157 (2017) 38–47. [46] L.X. Kong, C. Li, J.C. Jiang, M.G. Pecht, Li-ion battery fire hazards and safety strategies, Energies 11 (2018) 2191. [47] E.A. Baroncini, Bio-Based Thiol-ene Polymer Electrolytes (Ph.D. dissertation), Chemical Engineering, Rowan University, New Jersey, 2019. [48] R. Aroca, R. Nazri, G.A. Nazri, A.J. Camargo, M. Trsic, Vibrational spectra and ionpair properties of lithium hexafluorophosphate in ethylene carbonate based mixedsolvent systems for lithium batteries, J. Solut. Chem. 29 (2000) 1047–1060. [49] K.M. Diederichsen, E.J. McShane, B.D. McCloskey, Promising routes to a high Li+ transference number electrolyte for lithium ion batteries, Acs Energy Lett 2 (2017) 2563–2575. [50] M. Gouverneur, F. Schmidt, M. Schonhoff, Negative effective Li transference numbers in Li salt/ionic liquid mixtures: does Li drift in the “wrong” direction? Phys. Chem. Chem. Phys. 20 (2018) 7470–7478. [51] B.K. Wheatle, N.A. Lynd, V. Ganesan, Effect of polymer polarity on ion transport: a competition between ion aggregation and polymer segmental dynamics, ACS Macro Lett. 7 (2018) 1149–1154. [52] Z. Liu, et al., Advanced materials for lithium-ion batteries, Electrochemical Energy: Advanced Materials and Technologies, CRC Press - Taylor & Francis Group, Boca Raton, FL, 2016, pp. 79–142 (Chapter 4).

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements We would like to thank the Department of Defense SMART Scholarship program and Dr. Terrill B. Atwater, Dr. Ashley L. Ruth and the rest of the team at the US Army CCDC C5ISR Center in Aberdeen Proving Ground, Maryland for their invaluable assistance and support. We would also like to thank Rowan University's Chemical Engineering Department, past and present members of the Sustainable Materials Research Laboratory who have assisted in various ways (Tristan Bacha, John Chea, Julia Reilly, Eric Hernandez, Daniel Rogers, Joseph Mauck, Alexander Bassett, Silvio Curia, Ph.D. and Minxue Shi), and Jonathan Foglein of the Rowan Department of Chemistry and Biochemistry for support with the NMR. Appendix A. Supplementary data Proton NMR spectra of the three monomers, tabulated extent of cure values, exemplary near-IR FTIR spectra of uncured resin versus cured polymers, mid-IR FTIR spectra of cured SPEs and tabulated values of [EO]/[Li] ratios are included in the associated Supporting Information. Supplementary data to this article can be found online at doi:https:// doi.org/10.1016/j.ssi.2019.115161 References [1] L. Fan, S.Y. Wei, S.Y. Li, Q. Li, Y.Y. Lu, Recent progress of the solid-state electrolytes for high-energy metal-based batteries, Adv. Energy Mater. 8 (2018) 1702657. [2] Y.C. Jung, M.S. Park, D.H. Kim, M. Ue, A. Eftekhari, D.W. Kim, Room-temperature performance of poly(ethylene ether carbonate)-based solid polymer electrolytes for all-solid-state lithium batteries, Sci Rep-Uk 7 (2017) 17482. [3] E. Strauss, S. Menkin, D. Golodnitsky, On the way to high-conductivity single lithium-ion conductors, J Solid State Electr 21 (2017) 1879–1905. [4] O. Borodin, G.D. Smith, Li+ transport mechanism in oligo(ethylene oxide)s compared to carbonates, J. Solut. Chem. 36 (2007) 803–813. [5] J.J. Zhang, J.F. Yang, T.T. Dong, M. Zhang, J.C. Chai, S.M. Dong, T.Y. Wu, X.H. Zhou, G.L. Cui, Aliphatic polycarbonate-based solid-state polymer electrolytes for advanced lithium batteries: advances and perspective, Small 14 (2018) 1800821. [6] S. Mogurampelly, O. Borodin, V. Ganesan, Computer simulations of ion transport in polymer electrolyte membranes, Annu Rev Chem Biomol 7 (2016) 349–371. [7] O. Borodin, G.D. Smith, Mechanism of ion transport in amorphous poly(ethylene oxide)/LiTFSI from molecular dynamics simulations, Macromolecules 39 (2006) 1620–1629. [8] L.Z. Long, S.J. Wang, M. Xiao, Y.Z. Meng, Polymer electrolytes for lithium polymer batteries, J. Mater. Chem. A 4 (2016) 10038–10069. [9] M. Willgert, Solid Polymer Lithium-Ion Conducting Electrolytes for Structural Batteries (PhD dissertation), KTH Royal Institute of Technology, Stockholm, 2012. [10] E.A. Baroncini, J.F. Stanzione, Incorporating allylated lignin-derivatives in thiol-ene gel-polymer electrolytes, Int. J. Biol. Macromol. 113 (2018) 1041–1051. [11] L. Porcarelli, C. Gerbaldi, F. Bella, J.R. Nair, Super soft all-ethylene oxide polymer electrolyte for safe all-solid lithium batteries, Sci Rep-Uk (2016) 6. [12] C.E. Hoyle, T.Y. Lee, T. Roper, Thiol-enes: chemistry of the past with promise for the future, J Polym Sci Pol Chem 42 (2004) 5301–5338. [13] X.L. Cheng, J. Pan, Y. Zhao, M. Liao, H.S. Peng, Gel polymer electrolytes for electrochemical energy storage, Adv. Energy Mater. 8 (2018) 1702184. [14] J.F. Stanzione, J.M. Sadler, J.J. La Scala, K.H. Reno, R.P. Wool, Vanillin-based resin for use in composite applications, Green Chem. 14 (2012) 2346–2352. [15] A. Swiderska-Mocek, P. Jakobczyk, A. Lewandowski, Kinetic and galvanostatic studies of a polymer electrolyte for lithium-ion batteries, J Solid State Electr 21 (2017) 2825–2831. [16] S.M. Chen, K.H. Wen, J.T. Fan, Y.S. Bando, D. Golberg, Progress and future prospects of high-voltage and high-safety electrolytes in advanced lithium batteries: from liquid to solid electrolytes, J. Mater. Chem. A 6 (2018) 11631–11663. [17] E.D. Hernandez, A.W. Bassett, J.M. Sadler, J.J. La Scala, J.F. Stanzione, Synthesis and characterization of bio-based epoxy resins derived from vanillyl alcohol, ACS Sustain. Chem. Eng. 4 (2016) 4328–4339. [18] D.K. Shen, S. Gu, K.H. Luo, S.R. Wang, M.X. Fang, The pyrolytic degradation of wood-derived lignin from pulping process, Bioresour. Technol. 101 (2010) 6136–6146. [19] Y. Uemura, T. Shimasaki, N. Teramoto, M. Shibata, Thermal and mechanical properties of bio-based polymer networks by thiol-ene photopolymerizations of gallic acid and pyrogallol derivatives, J. Polym. Res. 23 (2016) 1–10. [20] C.E. Hoyle, C.N. Bowman, Thiol-ene click chemistry, Angew Chem Int Edit 49

10