Stable Cycling of Sodium Metal All-Solid-State Batteries with

Article pubs.acs.org/acsapm

Cite This: ACS Appl. Polym. Mater. XXXX, XXX, XXX−XXX

Stable Cycling of Sodium Metal All-Solid-State Batteries with Polycarbonate-Based Polymer Electrolytes Christofer Sångeland, Ronnie Mogensen, Daniel Brandell, and Jonas Mindemark* Department of ChemistryÅngström Laboratory, Uppsala University, Box 538, SE-751 21 Uppsala, Sweden

ACS Appl. Polym. Mater. Downloaded from pubs.acs.org by UNIV PARIS-SUD on 04/02/19. For personal use only.

S Supporting Information *

ABSTRACT: Solid polymer electrolytes based on high-molecular-weight poly(trimethylene carbonate) (PTMC) in combination with NaFSI salt were investigated for application in sodium batteries. The polycarbonate host material proved to be able to dissolve large amounts of salt, at least up to a carbonate:Na+ ratio of 1:1. Combined DSC, conductivity, and FTIR data indicated the formation of a percolating network of salt clusters along with the transition to a percolation-type ion transport mechanism at the highest salt concentrations. While the highest total ionic conductivities were seen at the highest salt concentrations (up to a remarkable 5 × 10−5 S cm−1 at 25 °C at a 1:1 carbonate:Na+ ratio), the most stable battery performance was seen at a more moderate salt loading of 5:1 carbonate:Na+, reaching >80 cycles at a stable capacity of ∼90 mAh g−1 at 60 °C in a sodium metal/Prussian blue cell. The results highlight the importance of the choice of salt and salt concentration on electrolyte performance as well as demonstrate the potential of utilizing polycarbonate-based electrolytes in sodium-based energy storage systems. KEYWORDS: polymer electrolytes, polycarbonates, sodium, batteries, ionic conductivity

■

INTRODUCTION

conductivities in fully amorphous matrices that can support battery cycling at down to ambient temperature.8−10 These developments, however, have not yet fully translated into similar progress in alternative host materials for Na+conducting SPEs. In an effort to rectify this, we recently communicated on the use of poly(trimethylene carbonate) (PTMC) as a host material for sodium battery electrolytes, including demonstrating short-term cycling of sodium/ Prussian blue solid-state battery cells.11 Building on these developments, we here present how, by a salt substitution from NaTFSI to NaFSI, it is possible to attain significant improvements in battery stability for solid-state Na-ion cells with PTMC-based electrolytes. A high capacity of >90 mAhg−1 of Prussian blue for an excess of 80 cycles is demonstrated at 60 °C, which is superior to other SPE-based Na-batteries demonstrated in scientific literature. We further show the effects of salt concentration on both conductivity and battery performance to reach the conclusion that a high conductivity is far from the only factor contributing to long-term battery cycling performance.

Energy storage is an essential ingredient in our quest for an energy system not dependent on fossil fuels. Today, lithiumion batteries (LIBs) are on their way to become the dominant battery chemistry, but in light of the expected surge in LIBs for electric vehicles and grid storage, the future sustainability of Libased battery chemistries is debated.1 Consequently, to secure future large-scale energy storage needs, there is a need to move beyond lithium-based battery chemistries. This need for energy storage based on more abundant elements has fuelled research interest in alternative chemistries that have the potential to complement or replace LIBs.2 Of immediate interest in this context is sodium;3,4 because of the similarities between lithium and sodium in terms of reduction potentials, ionic radii, and energy densities,5 sodium in many ways constitutes a natural progression of the plentiful research on LIBs. For batteries based on both lithium and sodium chemistries, solid polymer electrolytes (SPEs) have been highlighted as a means to improve safety and stability, in particular when employing metallic alkali metal anodes.6 While, by far, most of the research in this area has been directed toward polyether host materials, with a particular focus on poly(ethylene oxide) (PEO), there has lately been an upsurge in research activity on “alternative” host materials stimulated by potential gains in, e.g., ionic conductivity, cation transport numbers, and functional diversity.7 Notable among these are in particular materials where the ion coordination is based on carbonyl groups, i.e., polycarbonates and polyesters, where tremendous progress has been made to reach materials with high Li+ © XXXX American Chemical Society

■

EXPERIMENTAL SECTION

Materials. Sodium fluorosulfonylimide (NaFSI; Solvionic) was dried under vacuum at 80 °C for 24 h and kept in a glovebox under Ar before use. All other chemicals were obtained from commercial sources and used as received. Received: January 24, 2019 Accepted: March 22, 2019

A

DOI: 10.1021/acsapm.9b00068 ACS Appl. Polym. Mater. XXXX, XXX, XXX−XXX

Article

ACS Applied Polymer Materials Synthesis of PTMC. High-molecular-weight PTMC was prepared using a method described previously.12 In brief, trimethylene carbonate was polymerized via bulk ring-opening polymerization catalyzed by Sn(Oct)2 at 130 °C for 72 h. The polymer was obtained as a transparent, rubbery mass with a typical Mn in the range of 185 000−334 000 g mol−1 and a DPI of 1.8−1.9 according to GPC. The structure was confirmed through 1H NMR spectroscopy. The polymerization reagents as well as the obtained polymer were strictly handled in a glovebox under Ar to avoid moisture contamination. Synthesis and Characterization of Prussian Blue Cathode Material. Nanosized cubic crystals of Prussian blue, with a low Fe(CN)6 vacancy and H2O content, were precipitated using the procedure described in detail by You et al.13 In short, 2 mmol of Na4Fe(CN)6·10H2O was dissolved in 100 mL of deionized water and heated to 60 °C. One milliliter of 10.15 M HCl was added while the solution was stirred. The crystals were allowed to precipitate for 4 h, and the final product was centrifuged and rinsed with excess deionized water and then ethanol. On a precautionary note, ferrocyanides generate hydrogen cyanide gas when in contact with acid or undergo thermal decomposition.14 In this work, the excess hydrogen cyanide was trapped in a gas scrubber containing a solution of sodium hydroxide and hypochlorite. The structure of the synthesized material was confirmed with X-ray diffraction, and the morphology was documented using scanning electron microscopy (Figure S1). Preparation of Polymer Electrolyte Films. Polymer electrolyte films were prepared by first dissolving the prepared PTMC together with NaFSI salt in anhydrous acetonitrile under stirring for 12 h. The solution was cast in PTFE molds and dried by gradually lowering the pressure to vacuum during 20 h at ambient temperature, followed by thorough drying in vacuum at 60 °C for 40 h. Samples are referred to as PTMCnNaFSI, where n denotes the carbonate:NaFSI ratio. Polymer Electrolyte Characterization. The thermal properties of the polymer electrolyte films were studied using differential scanning calorimetry (DSC) on a TA Instruments DSC Q2000. The samples were hermetically sealed in aluminum pans followed by cooling to −70 °C at a rate of 10 °C min−1 and then heating to 80 °C at 10 °C min−1. FTIR measurements were performed using a PerkinElmer Spectrum One using a ZnSe crystal ATR setup. Electrochemical Characterization. The total ionic conductivity of the electrolytes was determined by electrochemical impedance spectroscopy (EIS). The thickness of the electrolyte films was measured using a Mitutoyo digital indicator, and typical values ranged from 30 to 150 μm. The films (⌀14 mm) were sandwiched between stainless steel blocking electrodes and mounted in Swagelok-type cells; these were then annealed at 100 °C to ensure good electrolyte− electrode contact. The impedance−frequency response was measured using a Schlumberger impedance/Gain-Phase Analyzer SI 1260 from 100 to 25 °C between 107 to 10 Hz at 10 mV amplitude. The bulk resistance was determined by fitting a Debye equivalent circuit to the obtained Nyquist plots in the software ZView 3.2b (Scribner Associates).15 The cation transference number was determined using the Bruce− Vincent method.16 Symmetrical cells consisting of ⌀16 mm sodium electrodes, prepared by pressing sodium cubes at 25 MPa between aluminum foils and polyethylene sheets, and ⌀20 mm PTMC5NaFSI polymer films were assembled and hermetically sealed in pouch cells. Cells were allowed to rest for 24 h at room temperature followed by temperature equilibration to 80 °C for 3 h prior to measurement. Impedance spectroscopy and potentiostatic polarization were done using a BioLogic SP-240 Potentiostat. The chronoamperometric response was monitored with an applied 10 mV polarization bias. The impedance−frequency response was measured between 200 kHz and 100 mHz with an amplitude of 10 mV and a bias of 0 and 10 mV before and after polarization, respectively. The bulk and interface resistances were determined using equivalent circuit fitting to the obtained Nyquist plots in the software ZView 3.2b (Scribner Associates). To obtain an accurate initial current I0, this was calculated using the approach developed by Hiller et al.17 as the ratio between polarization potential and total resistance across the cell.

The electrochemical stability window of PTMC5NaFSI and PTMC1NaFSI was determined by linear sweep voltammetry at 60 °C with a scan rate of 0.1 mV s−1 using a Bio-Logic VMP2. The cell configuration was Na|PTMCnNaFSI|carbon-coated aluminum, where sodium was a combined counter and reference electrode, and the carbon-coated aluminum foil was the working electrode. Separate measurements were done to determine the anodic and cathodic stability limits. The ionic resistance of each film was determined using EIS prior to the LSV measurement in order to apply iR compensation. Cell Assembly and Galvanostatic Cycling. Composite cathodes were prepared by ball milling the synthesized Prussian Blue together with Ketjen black (EC-600, HPL), PVdF (Kynar, Arkema), and N-methyl-2-pyrrolidone solvent at 150 rpm for 3 h. The slurry was coated on aluminum foil and then dried overnight in ambient atmosphere, followed by vacuum drying for 24 h at 100 °C. In order to ensure intimate contact, dissolved polymer electrolyte was cast directly on top of the cathode composite (13 mm diameter) and then dried using the same procedure as for the electrolyte film preparation. Sodium metal electrodes were prepared by first stripping sodium cubes of oxide and then rolling the sodium between two polyethylene sheets. Half-cells comprising these components were assembled and hermetically sealed in pouch cells. The cells were galvanostatically cycled using a Digatron MBT 0.1-05-32. C-rates are given according to the mass of the cathode active material which had a theoretical capacity of 170 mAh g−1. When cycled in a conventional liquid electrolyte 0.5 M NaFSI in 1:1 EC:DMC (volume ratio) using a Solupor separator, a stable capacity of 100 mAh g−1 was attained at C/5.5.

■

RESULTS AND DISCUSSION PTMC:NaFSI Electrolyte System. According to the widely accepted description of cation conduction in polymer electrolytes as being dependent on polymer segmental motion for ion transport,18 a host material capable of facilitating fast cation transport should ideally fulfill a series of criteria that includes (1) high segmental mobility, including a subambient Tg; (2) insignificant crystallinity; and (3) cation-solvating properties with sufficiently weak ion coordination to allow for facile ion exchange between coordination sites.7 These considerations apply equally for both Li+ and Na+ transport. In this context, PEO, the most commonly applied host material for both Li+ and Na+ systems,19 typically falls short in terms of both points 2 and 3 outlined above; i.e., the high crystallinity at ambient temperature in combination with strong ion complexation severely restricts cation motion.7,20,21 As an alternative to this, we synthesized a high-molecular-weight PTMC (for the structure, see Figure 1a) through ring-opening polymerization of trimethylene carbonate to apply as a polymer electrolyte host material. In contrast with PEO, this polymer is fully amorphous, provided that the molecular weight is sufficiently high.22 The subambient Tg furthermore facilitates ion trans-

Figure 1. (a) Structures of high-molecular-weight PTMC and NaFSI salt. (b) PTMC1NaFSI electrolyte film, illustrating its amorphous character and excellent mechanical properties even at extreme salt concentrations. B


Article

ACS Applied Polymer Materials

Figure 2. (a) Arrhenius plot of PTMCnNaFSI. The dashed lines represent VFT fits (the fitting parameters can be found in Table S1). (b) Glass transition temperature vs salt concentration. The annotated numbers correspond to the [carbonate]:Na+ ratio (n). (c) Conductivity isotherms for different molar concentrations of NaFSI at different temperatures.

Perhaps counterintuitively, as the Tg increases, the ionic conductivity was also found to increase, as shown in Figure 2c. Again, this mimics what was previously seen in the PTMC:NaTFSI system and we attribute this effect to an increase in charge carriers as more salt is added to the electrolytes.11 At salt concentrations exceeding n = 5, no further stiffening of the polymer matrix is seen, and the Tg instead starts to decrease. This behavior is indicative of plasticization by ionic aggregates, as previously observed in, e.g., poly(ethylene carbonate) electrolytes with LiFSI.24,25 In PEO:NaFSI and PEO:NaTFSI systems, a clear maximum in conductivity is observed at intermediate concentrations.26 In the PTMCnNaFSI, on the other hand, there is only a slight dip in conductivity at n = 3 followed by a steep increase, which also coincides with the rapid drop in Tg. This indicates the emergence of a secondary transport mechanism at high salt concentrations, involving ion transport along percolating aggregates, as proposed by, for example, Forsyth et al. for PAN:LiCF3SO3 at high salt concentrations.27,28 Once a sufficient amount of salt has been added, the dispersed clusters of salt aggregates come into contact, forming a diffusion and migration “highway” for the cations. This secondary conduction mechanism explains the massive increase in ionic conductivity despite only modest changes in Tg. For the PTMC:NaFSI electrolyte system, at salt concentrations higher than the percolation threshold at n = 3, the emerging salt clusters push apart the polymer chains, thereby increasing the free volume. This lowers the Tg and results in a system that is more akin to being composed of a salt plasticized by the polymer, rather than a polymer plasticized by the salt, essentially forming an ionic liquid-like salt structure confined in a polymer matrix.27,28 In the polymer-in-salt electrolyte (PISE) domain,7 the percolation transport mechanism enables a conductivity as high as 5 × 10−5 S cm−1 at 25 °C for PTMC1NaFSI (Figure 2c). This level of conductivity is remarkable, particularly considering that the Tg is still as high as −4.7 °C. At 60 °C, the

port, and despite the lack of crystallinity or cross-links, at high molecular weights (exceeding ∼100 000−300 000 g mol−1), molecular entanglements and strain crystallization induce a sufficient level of mechanical stability for practical application.23 This was confirmed for the synthesized PTMC, which was obtained as a rubbery, transparent mass with typical molecular weights in the range of 185 000−334 000 g mol−1. These properties were largely retained when NaFSI salt was added to the system, i.e., the samples were fully amorphous, as indicated by DSC measurements, and had rubber-like mechanical properties even at extreme salt concentrations (Figure 1b). Polymer electrolytes were prepared with carbonate:Na+ ratios (n) in the range of 1−21; at n < 2, the salt content (by weight) exceeds that of the polymer. Thermal Properties and Ionic Conductivity. As expected, the ionic conductivity of the electrolytes was found to be linked to the segmental motions of the polymer host, as illustrated by the Vogel−Fulcher−Tammann (VFT)-type behavior seen in the Arrhenius plot in Figure 2a (details about the fitting can be found in the Supporting Information). The dependence of ion transport on polymer chain mobility indicates that the Tg should be a key performance indicator for these electrolytes. Adding NaFSI salt to the PTMC matrix results in the formation of transient physical cross-links as the polymer chains coordinate to the Na+ ions. As seen in Figure 2b, this causes the Tg to increase with salt concentration from n = 21 to 5. The close to linear increase mimics what has already been reported for PTMC with NaClO4 and NaTFSI salt in the same concentration range.11 Interestingly, Na+ appears to have a much larger effect on Tg in comparison to Li+ in PTMCbased systems.11,12 We attribute this to the larger ionic radius of Na+ relative to Li+ (116 and 90 pm, respectively), which allows for more carbonyl groups to coordinate to the Na+ ion and thereby causing a more substantial stiffening of the polymer host. We also note that the increase in Tg is higher for NaFSI than NaTFSI addition to PTMC,11 implying a certain plasticizing ability of the larger TFSI anion in comparison to the FSI anion. C


Article

ACS Applied Polymer Materials conductivity reaches 5.7 × 10−4 S cm−1, which is comparable to what was reported by Tominaga et al. for PEC:LiFSI in a similar extreme-salt concentration range but for a T g approaching −40 °C.24 These values also compare favorably to PEO:NaFSI and PEO:NaTFSI, particularly at low temperatures where crystallization becomes an issue in the polyether systems.20,21,26 We finally reiterate that the PTMCnNaFSI electrolytes maintain their mechanical integrity regardless of the salt concentration, appearing as transparent and amorphous rubbery films even at extreme salt concentrations. Ion Coordination. In Li+-conducting polymer electrolytes based on PTMC and other polyesters and polycarbonates, the cations are primarily solvated through coordination to the carbonyl oxygens of the polymer host.25,29,30 The peak arising from the carbonyl group in PTMC is found at approximately 1740 cm−1 in FTIR spectroscopy. When the carbonyl group is coordinating to cations, this peak shifts to lower wavenumbers.29 As can be seen in the spectra in Figure 3, this

1. The low coordination number for n = 21 is an effect of the difficulties in obtaining an accurate fit at this low ion Table 1. Portion of Coordinating Carbonyl Groups and Coordination Number (CN) for n = 21 to 1a n

NaFSI concentration (mol %)

fraction coordinating carbonyl groups

CN

21 13 8 5 3 2 1

4.5 7.1 11 17 25 33 50

0.06 0.25 0.36 0.58 0.67 0.92 0.95

1.19 3.19 2.88 2.91 2.02 1.84 0.95

a The estimations assume equal extinction coefficients for coordinating and noncoordinating carbonyl groups.

concentration, where small uncertainties in both ion concentration (n) and peak area have a large impact for low concentrations. As expected, the coordination number is found to increase with salt concentration until it peaks at n = 5. At this point, the system becomes saturated with salt, and there is no longer a sufficient amount of carbonyl groups to fully coordinate any additional Na+. We note that, this point (between n = 3 and 5) also coincides with the previously identified percolation threshold. The coordination number of Na+ is typically higher than that of Li+, and values of 5−7 have been cited for the former in both liquids and systems mimicking polymeric solvents.33 With the coordination numbers listed in Table 1, this translates into a roughly equal contribution to the cation coordination from the carbonyl groups of the polymer host and from the FSI anions up until the saturation point, with coordination to anions becoming more prevalent at the higher concentrations. The direct coordination of anions to Na+ is a clear indication of ion pairing in these systems. Cation Transference Number. One distinguishing feature of polymer electrolytes based on traditional polyethers, such as PEO, is low cation transference numbers (T+), often in the range of 0.1−0.2 for Na+, meaning that most of the conductivity is due to (unwanted) anion transport.20,21 With its alternative coordination chemistry based on nonchelating carbonyls rather than the oxyethylene motif of PEO, the coordination strength of the PTMC:NaFSI electrolytes is expected to be reduced, resulting in a higher cation transference number.7 Here, the T+ was determined for PTMC5NaFSI through potentiostatic polarization.16 In order to give a valid estimate of the transference number of the system, this method requires some assumptions regarding the sample, one of which is negligible ion association. With a percolation threshold at n = 3, this is clearly not the case for the highest-conducting samples of this study. Hence, the transference number was instead determined for PTMC5NaFSI, yielding a T+ of 0.48 at 80 °C (Figure 4). While this is lower than the T+ reported for PTMC8LiTFSI (0.8 at 60 °C),29 it is clearly superior to what is typical for polyether-based electrolytes in accordance with a weaker coordination structure. It should also be mentioned that, although it is not evident in the Nyquist plots in Figure 4, the measurements hinted at instabilities in the polymer electrolyte−sodium metal interface, making it difficult to reach steady-state in the polarization. As noted by other authors, this

Figure 3. Deconvoluted FTIR peaks (Voigtian lineshapes) corresponding to coordinating (blue dashed line) and noncoordinating (red dash-dotted line) carbonyl groups in PTMCnNaFSI. Experimental data shown in black dotted lines.

behavior is also observed in the PTMCnNaFSI system, and the peak corresponding to the coordinating carbonyl groups becomes more pronounced as the NaFSI concentration increases from n = 21 to 1. As the salt concentration increases beyond the percolation threshold, both the peak corresponding to interacting and the peak corresponding to “free” carbonyl groups are found to move toward a higher wavenumber (Figure S2). This is indicative of the formation of salt agglomerates as previously reported by Tominaga et al. for high-concentration poly(ethylene carbonate):LiFSI electrolytes.31 This initially subtle shift in peak position undergoes a sudden jump between n = 3 and 2, i.e., at the previously suggested percolation threshold. An estimation of the cation coordination number, considering only coordination by carbonyl groups, can be obtained by multiplying the fraction of Na+-coordinating carbonyls, obtained from areal peak ratios in the FTIR spectra, and the carbonyl:Na+ molar ratio n.32 The calculated coordination numbers for n = 21 to 1 can be found in Table D


Article


Figure 4. Cation transference number determination through potentiostatic polarization of a Na|PTMC5NaFSI|Na symmetrical cell at 80 °C. (a) Chronoamperometric response. (b) Nyquist plot of impedance response before (initial) and after (steady-state) potentiostatic polarization.

is inherent to the use of sodium metal electrodes in situations where a stable potential is required.34 Still, the high value obtained from the potentiostatic polarization indicates that a comparatively large fraction of the ionic current is carried by the cation in the PTMC:NaFSI system. Solid-State Battery Cycling. Parameters such as total ionic conductivity, glass transition temperature, and transference number are often cited in literature as key indicators of performance; however, they give little or no insight into the actual performance in its intended application in sodium battery cells. The high ionic conductivity seen for the highconcentration PTMC:NaFSI electrolytes hints that good rate performance in solid-state sodium cells should be possible but needs to be appropriately tested. To this end, a cell comprising a submicron-sized Prussian blue (partially sodiated Fe(Fe(CN)6); Figure S1) cathode and a sodium metal anode was assembled separated by the PTMC1NaFSI electrolyte and cycled at 60 °C. Unfortunately, as seen in Figure 5a, after an

Figure 6. (a) Discharge capacity and Coulombic efficiency per cycle and (b) voltage profiles for a Na|PTMC1NaFSI|Prussian blue half-cell cycled at 40 °C and C/5 (relative to active mass the loading of the cathode; 0.411 mg cm−2).

attributed to gradually improving electrode−electrolyte contact as the polymer electrolyte slowly infiltrates the porous cathode over time.12,35 Although more stable cycling was seen at this temperature compared to at 60 °C, a capacity fading over time could be noted. This was accompanied by an increasing polarization and unstable voltages (Figure 6b), indicating a gradual loss of conductivity during cycling. We also note Coulombic efficiencies exceeding 100%, indicating irreversible side reactions during discharge. As has already been discussed, at this extreme salt concentration, the PTMC1NaFSI essentially behaves as an ionic liquid confined in a polymer matrix.27,28 Of course, this has implications on the stability of the electrolyte compared to a salt-in-polymer electrolyte, where the material properties are more clearly influenced by the polymer host matrix, both in terms of electrochemical and chemical stability. Tominaga et al. showed that high-salt concentration electrolytes are prone to interfacial resistance increases over time when in contact with Li metal at low discharge currents or even when zero current is passed through the system.36 In the context of sodium batteries, a greater instability of the solid electrolyte interphase (SEI) layer is also expected, in particular with respect to dissolution of SEI components.37 Considering the ionic liquid character of the PTMC1NaFSI electrolyte, this naturally becomes more problematic than in an electrolyte with more pronounced solid characteristics. With regards to the chemical stability of high-salt (PISE) electrolytes, oversaturation leading to salt precipitation is detrimental to cell performance and is furthermore exacerbated by cycling.38 Precipitation of the electrolyte salt can, in some

Figure 5. (a) Discharge capacity and Coulombic efficiency per cycle and (b) voltage profiles for a Na|PTMC1NaFSI|Prussian blue half-cell cycled at 60 °C. The C-rates are given relative to the active mass loading of the cathode (0.306 mg cm−2).

initial increase in capacity followed by stable cycling for some 20 cycles at up to 1C, the cell showed erratic Coulombic efficiency and unstable capacities, indicating severe and detrimental side-reactions. This rapid performance degradation was also reflected in the growing polarization in the charge− discharge curves (Figure 5b). In an attempt to alleviate these side-reactions, another nominally identical cell was cycled at 40 °C. Cycling at this lower temperature also meant limiting the cycling current to C/5. As seen in Figure 6, an initial increase in discharge capacity with cycle number can be noted, which can be E


Article


a better ability of the latter salt of forming a stable passivation layer on the surface of Na metal, as shown by Hosokawa et al. for imidazolium ionic liquids.43

cases, be observed as a passive aging effect for metastable highsalt systems, leading to diminished conductivity over time.39 In order to gain more insight into the causes of the instability of the cells comprising the PTMC1NaFSI electrolyte, the electrochemical stability window was determined using linear sweep voltammetry. As can be seen in Figure S3, the electrochemical stability of the high-concentration PTMC1NaFSI is clearly inferior to that of the more moderate-concentration PTMC5NaFSI. Alas, the high-salt PTMC1NaFSI electrolyte, although high in conductivity and able to sustain high-rate cycling for a limited number of cycles, does not show sufficient stability for long-term operation of sodium cells. Instead, a half-cell comprising the more electrochemically stable PTMC5NaFSI was assembled and cycled at C/5 and 60 °C. As seen in Figure 7, the discharge capacity and Coulombic efficiency showed

■

CONCLUSIONS The combination of high-molecular-weight PTMC with NaFSI salt resulted in amorphous and mechanically stable SPEs that ranged from salt-in-polymer electrolytes at the low end of the concentration spectrum to polymer-in-salt electrolytes when the amount of salt approached that of the polymer host. Clear indications of a percolation threshold was seen in both DSC and conductivity data. The formation of a percolating network of salt clusters that supported a percolating ion transport mechanism at the highest salt concentrations was associated both with remarkably high total ionic conductivities but instabilities during cycling in sodium half-cells. Better longterm cycling stability was instead attained using an electrolyte with a more moderate salt concentration, which delivered stable discharge capacities for >80 cycles during galvanostatic cycling. This constitutes a substantial improvement compared to previously reported electrolytes not only based on the same host material in combination with NaTFSI salt, but also stateof-the-art PEO-based SPEs. The ability to dissolve large amounts of salt is potentially useful, as demonstrated by the possibility of attaining exceptionally high room-temperature conductivity, although the performance of battery cells was ultimately limited for this electrolyte system. Naturally, battery cycling at these extreme salt concentrations becomes increasingly dependent on salt stability and degradation behavior. Instead, a reasonable compromise between stability and fast ion transport was found at a lower salt concentration. Here, the cation transference number (0.48 at 80 °C) − although lower than in comparable Li+-conducting PTMC-based systems − was notably higher than in typical polyether systems, thus confirming the benefit of introducing carbonyl groups as ioncoordinating moieties also in sodium systems.

Figure 7. (a) Discharge capacity and Coulombic efficiency per cycle and (b) voltage profiles for a Na|PTMC5NaFSI|Prussian blue half-cell cycled at 40 °C and C/5 relative to the active mass loading of the cathode (0.253 mg cm−2).

much more stable behaviors at this lower salt concentration, with >80 cycles at a relatively constant discharge capacity. Unlike the other two cells, the PTMC5NaFSI cell also exhibited less pronounced polarization between cycles. This remarkable stability can be favorably compared to both allsolid and hybrid/gel polymer electrolytes based on traditional polyethers.20,21,40,41 Among some recent examples can be mentioned Qi et al., showcasing Na|PEO:NaFSI| Na3V2(PO4)3C and Na|PEO:NaFSI|Na0.67Ni0.33Mn0.67O2 with an average discharge capacity of ∼65 and 85 mAh g−1 for 50 and 30 consecutive cycles, respectively, at a high temperature of 80 °C.20 Colò et al. cycled Na|PEO:NaClO4|NaFePO4 at 60 °C with an average discharge capacity of ∼87 mAh g−1, albeit only for 20 cycles.21 Going further back in time to the mid1990s, in the pioneering works by Doeff et al., Na| PEO:NaCF3SO3|NaxCoO2 and Na|PEO:NaCF3SO3|NaxMnO2 thin-film batteries achieved initial capacities of ∼160 and ∼155 mAh g−1, respectively, at 85 °C but experienced capacity fading over time, which eventually resulted in cell death after 60 cycles.38,42 In terms of nonpolyether host materials for sodium batteries, we recently presented solid-state sodium metal cells with a PTMC3NaTFSI electrolyte.11 While high and reasonably stable capacities were attained at the operating temperature of 60 °C, we were unable to attain more than ca. 8 stable cycles before cell failure. In this context, the improvement on going from NaTFSI to NaFSI salt is monumental and can be attributed to

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsapm.9b00068. Prussian blue cathode material characterization and VTF fitting parameters (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Daniel Brandell: 0000-0002-8019-2801 Jonas Mindemark: 0000-0002-9862-7375 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work has been financed through support from the ERC, grant agreement No. 771777 FUN POLYSTORE. Parts of this work were included in the master thesis “PTMC: A polycarbonate candidate for polymer electrolytes in sodium batteries? A characterisation of the PTMC−NaFSI system” by Ronnie Mogensen in 2015. The full text of the final thesis is F


Article


(21) Colò, F.; Bella, F.; Nair, J. R.; Destro, M.; Gerbaldi, C. Cellulose-Based Novel Hybrid Polymer Electrolytes for Green and Efficient Na-Ion Batteries. Electrochim. Acta 2015, 174, 185−190. (22) Zhu, K. J.; Hendren, R. W.; Jensen, K.; Pitt, C. G. Synthesis, Properties, and Biodegradation of Poly(1,3-Trimethylene Carbonate). Macromolecules 1991, 24 (8), 1736−1740. (23) Pêgo, A. P.; Poot, A. A.; Grijpma, D. W.; Feijen, J. Copolymers of Trimethylene Carbonate and Epsilon-Caprolactone for Porous Nerve Guides: Synthesis and Properties. J. Biomater. Sci., Polym. Ed. 2001, 12 (1), 35−53. (24) Tominaga, Y.; Yamazaki, K. Fast Li-Ion Conduction in Poly(Ethylene Carbonate)-Based Electrolytes and Composites Filled with TiO2 Nanoparticles. Chem. Commun. 2014, 50 (34), 4448−4450. (25) Kimura, K.; Motomatsu, J.; Tominaga, Y. Correlation between Solvation Structure and Ion-Conductive Behavior of Concentrated Poly(Ethylene Carbonate)-Based Electrolytes. J. Phys. Chem. C 2016, 120 (23), 12385−12391. (26) Boschin, A.; Johansson, P. Characterization of NaX (X: TFSI, FSI) − PEO Based Solid Polymer Electrolytes for Sodium Batteries. Electrochim. Acta 2015, 175, 124−133. (27) Forsyth, M.; Sun, J.; Macfarlane, D.; Hill, A. Compositional Dependence of Free Volume in PAN/LiCF3SO3 Polymer-in-Salt Electrolytes and the Effect on Ionic Conductivity. J. Polym. Sci., Part B: Polym. Phys. 2000, 38 (2), 341−350. (28) Forsyth, M.; MacFarlane, D.; Hill, A. Glass Transition and Free Volume Behaviour of Poly(Acrylonitrile)/LiCF3SO3 Polymer-in-Salt Electrolytes Compared to Poly(Ether Urethane)/LiClO4 Solid Polymer Electrolytes. Electrochim. Acta 2000, 45 (8−9), 1243−1247. (29) Sun, B.; Mindemark, J.; Morozov, E. V.; Costa, L. T.; Bergman, M.; Johansson, P.; Fang, Y.; Furó, I.; Brandell, D. Ion Transport in Polycarbonate Based Solid Polymer Electrolytes: Experimental and Computational Investigations. Phys. Chem. Chem. Phys. 2016, 18 (14), 9504−9513. (30) Mindemark, J.; Imholt, L.; Montero, J.; Brandell, D. Allyl Ethers as Combined Plasticizing and Crosslinkable Side Groups in Polycarbonate-Based Polymer Electrolytes for Solid-State Li Batteries. J. Polym. Sci., Part A: Polym. Chem. 2016, 54 (14), 2128−2135. (31) Kimura, K.; Tominaga, Y. Understanding Electrochemical Stability and Lithium Ion-Dominant Transport in Concentrated Poly(Ethylene Carbonate) Electrolyte. ChemElectroChem 2018, 5 (24), 4008−4014. (32) Mindemark, J.; Tang, S.; Li, H.; Edman, L. Ion Transport beyond the Polyether Paradigm: Introducing Oligocarbonate Ion Transporters for Efficient Light-Emitting Electrochemical Cells. Adv. Funct. Mater. 2018, 28 (32), 1801295. (33) Åvall, G.; Mindemark, J.; Brandell, D.; Johansson, P. SodiumIon Battery Electrolytes: Modeling and Simulations. Adv. Energy Mater. 2018, 8 (17), 1703036. (34) Tchitchekova, D. S.; Monti, D.; Johansson, P.; Bardé, F.; Randon-Vitanova, A.; Palacín, M. R.; Ponrouch, A. On the Reliability of Half-Cell Tests for Monovalent (Li+, Na+) and Divalent (Mg2+, Ca2+) Cation Based Batteries. J. Electrochem. Soc. 2017, 164 (7), A1384−A1392. (35) Sun, B.; Mindemark, J.; Edström, K.; Brandell, D. Realization of High Performance Polycarbonate-Based Li Polymer Batteries. Electrochem. Commun. 2015, 52, 71−74. (36) Kimura, K.; Matsumoto, H.; Hassoun, J.; Panero, S.; Scrosati, B.; Tominaga, Y. A Quaternary Poly(Ethylene Carbonate)-Lithium Bis(Trifluoromethanesulfonyl)Imide-Ionic Liquid-Silica Fiber Composite Polymer Electrolyte for Lithium Batteries. Electrochim. Acta 2015, 175, 134−140. (37) Mogensen, R.; Brandell, D.; Younesi, R. Solubility of the Solid Electrolyte Interphase (SEI) in Sodium Ion Batteries. ACS Energy Lett. 2016, 1 (6), 1173−1178. (38) Doeff, M. M.; Ferry, A.; Ma, Y.; Ding, L.; De Jonghe, L. C. Effect of Electrolyte Composition on the Performance of Sodium/ Polymer Cells. J. Electrochem. Soc. 1997, 144 (2), L20−L22. (39) Łasińska, A. K.; Marzantowicz, M.; Dygas, J. R.; Krok, F.; Florjańczyk, Z.; Tomaszewska, A.; Zygadło-Monikowska, E.;

available online from Uppsala University at http://uu.divaportal.org/smash/record.jsf?pid=diva2:855729.

■

REFERENCES

(1) Vikström, H.; Davidsson, S.; Höök, M. Lithium Availability and Future Production Outlooks. Appl. Energy 2013, 110, 252−266. (2) Muldoon, J.; Bucur, C. B.; Gregory, T. Quest for Nonaqueous Multivalent Secondary Batteries: Magnesium and Beyond. Chem. Rev. 2014, 114 (23), 11683−11720. (3) Yabuuchi, N.; Kubota, K.; Dahbi, M.; Komaba, S. Research Development on Sodium-Ion Batteries. Chem. Rev. 2014, 114 (23), 11636−11682. (4) Palomares, V.; Serras, P.; Villaluenga, I.; Hueso, K. B.; CarreteroGonzález, J.; Rojo, T. Na-Ion Batteries, Recent Advances and Present Challenges to Become Low Cost Energy Storage Systems. Energy Environ. Sci. 2012, 5 (3), 5884−5901. (5) Kundu, D.; Talaie, E.; Duffort, V.; Nazar, L. F. The Emerging Chemistry of Sodium Ion Batteries for Electrochemical Energy Storage. Angew. Chem., Int. Ed. 2015, 54 (11), 3431−3448. (6) Hueso, K. B.; Armand, M.; Rojo, T. High Temperature Sodium Batteries: Status, Challenges and Future Trends. Energy Environ. Sci. 2013, 6 (3), 734−749. (7) Mindemark, J.; Lacey, M.; Bowden, T.; Brandell, D. Beyond PEOAlternative Host Materials for Li+-Conducting Solid Polymer Electrolytes. Prog. Polym. Sci. 2018, 81, 114−143. (8) Mindemark, J.; Sun, B.; Törmä, E.; Brandell, D. HighPerformance Solid Polymer Electrolytes for Lithium Batteries Operational at Ambient Temperature. J. Power Sources 2015, 298, 166−170. (9) Chai, J.; Liu, Z.; Ma, J.; Wang, J.; Liu, X.; Liu, H.; Zhang, J.; Cui, G.; Chen, L. In Situ Generation of Poly (Vinylene Carbonate) Based Solid Electrolyte with Interfacial Stability for LiCoO2 Lithium Batteries. Adv. Sci. 2017, 4 (2), 1600377. (10) Meabe, L.; Lago, N.; Rubatat, L.; Li, C.; Müller, A. J.; Sardon, H.; Armand, M.; Mecerreyes, D. Polycondensation as a Versatile Synthetic Route to Aliphatic Polycarbonates for Solid Polymer Electrolytes. Electrochim. Acta 2017, 237, 259−266. (11) Mindemark, J.; Mogensen, R.; Smith, M. J.; Silva, M. M.; Brandell, D. Polycarbonates as Alternative Electrolyte Host Materials for Solid-State Sodium Batteries. Electrochem. Commun. 2017, 77, 58− 61. (12) Sun, B.; Mindemark, J.; Edströ m , K.; Brandell, D. Polycarbonate-Based Solid Polymer Electrolytes for Li-Ion Batteries. Solid State Ionics 2014, 262, 738−742. (13) You, Y.; Wu, X.-L.; Yin, Y.-X.; Guo, Y.-G. High-Quality Prussian Blue Crystals as Superior Cathode Materials for RoomTemperature Sodium-Ion Batteries. Energy Environ. Sci. 2014, 7 (5), 1643−1647. (14) Lewis, R. J. Sax’s dangerous properties of industrial materials; John Wiley & Sons, Inc: Hoboken, NJ, 2012; p 2149. (15) Mindemark, J.; Törmä, E.; Sun, B.; Brandell, D. Copolymers of Trimethylene Carbonate and ε-Caprolactone as Electrolytes for Lithium-Ion Batteries. Polymer 2015, 63, 91−98. (16) Evans, J.; Vincent, C. A.; Bruce, P. G. Electrochemical Measurement of Transference Numbers in Polymer Electrolytes. Polymer 1987, 28 (13), 2324−2328. (17) Hiller, M. M.; Joost, M.; Gores, H. J.; Passerini, S.; Wiemhöfer, H.-D. The Influence of Interface Polarization on the Determination of Lithium Transference Numbers of Salt in Polyethylene Oxide Electrolytes. Electrochim. Acta 2013, 114, 21−29. (18) Ratner, M. A.; Shriver, D. F. Ion Transport in Solvent-Free Polymers. Chem. Rev. 1988, 88 (1), 109−124. (19) Xue, Z.; He, D.; Xie, X. Poly(Ethylene Oxide)-Based Electrolytes for Lithium-Ion Batteries. J. Mater. Chem. A 2015, 3 (38), 19218−19253. (20) Qi, X.; Ma, Q.; Liu, L.; Hu, Y.-S.; Li, H.; Zhou, Z.; Huang, X.; Chen, L. Sodium Bis(Fluorosulfonyl)Imide/Poly(Ethylene Oxide) Polymer Electrolytes for Sodium-Ion Batteries. ChemElectroChem 2016, 3 (11), 1741−1745. G


Article

ACS Applied Polymer Materials Ż ukowska, Z.; Lafont, U. Study of Ageing Effects in Polymer-in-Salt Electrolytes Based on Poly(Acrylonitrile-Co-Butyl Acrylate) and Lithium Salts. Electrochim. Acta 2015, 169, 61−72. (40) Colò, F.; Bella, F.; Nair, J. R.; Gerbaldi, C. Light-Cured Polymer Electrolytes for Safe, Low-Cost and Sustainable Sodium-Ion Batteries. J. Power Sources 2017, 365, 293−302. (41) Zheng, Y.; Pan, Q.; Clites, M.; Byles, B. W.; Pomerantseva, E.; Li, C. Y. High-Capacity All-Solid-State Sodium Metal Battery with Hybrid Polymer Electrolytes. Adv. Energy Mater. 2018, 8 (27), 1801885. (42) Doeff, M. M.; Peng, M. Y.; Ma, Y.; De Jonghe, L. C. Orthorhombic NaxMnO2 as a Cathode Material for Secondary Sodium and Lithium Polymer Batteries. J. Electrochem. Soc. 1994, 141 (11), L145−L147. (43) Hosokawa, T.; Matsumoto, K.; Nohira, T.; Hagiwara, R.; Fukunaga, A.; Sakai, S.; Nitta, K. Stability of Ionic Liquids against Sodium Metal: A Comparative Study of 1-Ethyl-3-Methylimidazolium Ionic Liquids with Bis(Fluorosulfonyl)Amide and Bis(Trifluoromethylsulfonyl)Amide. J. Phys. Chem. C 2016, 120 (18), 9628−9636.

H


Stable Cycling of Sodium Metal All-Solid-State Batteries with

Recommend Documents