A versatile single-ion electrolyte with a Grotthuss-like Li conduction mechanism for dendrite-free Li metal batteries

Shouyi Yuana, Junwei Lucas Baob, Jishi Weia, Yongyao Xiaa, Donald G. Truhlar*b and Yonggang Wang*a
aDepartment of Chemistry, Shanghai Key Laboratory of Catalysis and Innovative Materials, Center of Chemistry for Energy Materials, Fudan University, Shanghai, 200433, China. E-mail: ygwang@fudan.edu.cn
bDepartment of Chemistry, Chemical Theory Center, and Minnesota Supercomputing Institute, University of Minnesota, 207 Pleasant Street SE, Minneapolis, Minnesota 55455-0431, USA. E-mail: truhlar@umn.edu

Received 7th May 2019 , Accepted 17th July 2019

First published on 18th July 2019


Batteries with Li metal anodes have the desirable feature of high energy density; however, the notorious problem of Li dendrite formation has impeded their practical applications. Herein, we present a versatile single-ion electrolyte, which is achieved by a different strategy of coordinating the anions in the electrolyte on the open metal sites of a metal organic framework. Further investigations of the activation energy and theoretical quantum mechanical calculations suggest that Li ion transport inside the pores of Cu-MOF-74 is via a Grotthuss-like mechanism where the charge is transported by coordinated hopping of Li ions between the perchlorate groups. This single-ion electrolyte is versatile and has wide applications. When the single-ion electrolyte is used for Li‖Li symmetric cells and Li‖LiFePO4 full cells, Li dendrites are suppressed. As a result, an ultralong cycle life is achieved for both cells. In addition, when the single-ion electrolyte is assembled into Li‖LiMn2O4 batteries, the dissolution of Mn2+ into the electrolyte is suppressed even at elevated temperatures, and a long cycle life with improved capacity retention is achieved for Li‖LiMn2O4 batteries. Finally, when the single-ion electrolyte is applied to Li–O2 batteries, an improved cycle life with reduced overpotential is also achieved.



Broader context

Given the limitations inherent in current battery technologies, much of the research on next generation energy storage technologies has been focused on potential successors to lithium ion batteries such as Li–S batteries or Li–O2 batteries. However, the use of alkali metal anodes is needed to realize the potential of these batteries, but these anodes are likely to suffer from dendrite formation, which results in excessive electrolyte decomposition. In this work, we present a versatile single-ion electrolyte, which is achieved by a different strategy of coordinating the anions in the electrolyte on the open metal sites of a metal organic framework. Further theoretical calculations suggest that the Li ion transport in this electrolyte is via a Grotthuss-like mechanism involving hopping of Li ions between ClO4. The single-ion electrolyte is versatile and has wide applications including suppressing Li dendrites, suppressing the dissolution of Mn2+ in Li‖LiMn2O4 full cells and reducing the polarization of Li–O2 batteries. As a result, improved performances are achieved for all the batteries.

Introduction

With the rapid popularization of electric vehicles, there is an urgent demand for batteries with high energy density and high specific capacity such as Li–sulfur batteries and Li–air batteries.1 Li metal anodes would be especially suitable for this kind of application because of their high specific capacity (3860 mA h g−1), which is far beyond that of commercial graphite anodes (372 mA h g−1).2–4 However, although there has been steady progress in improving the cathode side of Li–air batteries and Li sulfur batteries,5–10 the use of Li metal anodes still suffers from irregular deposition of Li ions on the anode and the resulting formation of Li dendrites, which causes internal short circuits, which will lead to hazardous explosions. The Li dendrites also lead to capacity loss due to depletion of Li from the anode.2–4 Therefore, it is critical to address the issue of Li dendrite formation.

Various strategies have been put forward to address the issue of Li dendrites, including optimizing the electrolyte,11–18 engineering artificial solid electrolyte interfaces,19–23 and designing structured anodes.24–32 However, although those strategies effectively prolong the cycle life of Li-metal batteries, they do not solve the intrinsic instability of Li metal in liquid-electrolyte batteries. Hence, the Li dendrites still form in the long term.

It is widely acknowledged that the conventional liquid electrolytes for Li-metal batteries contain lithium salts and organic solvents, such as 1 M LiPF6 in ethylene carbonate/diethyl carbonate (EC/DEC). During charge/discharge, the cations (e.g., Li+) and anions (e.g., PF6) in the electrolyte move towards the opposite direction along with electron transfer between the cathode and anode. The mobility of the anions actively contributes to the ionic conduction process in batteries, but they do not participate in the electrode reactions. Thus, during the Li stripping process, the electrical field drives the anions in the electrolyte towards the anodic side, where they gradually accumulate near the anode. This causes concentration polarization, which may lead to the formation of an ion depletion layer.33,34 Previous reports33,34 have shown that tiny and random lithium deposition takes place at the ion depletion layer on the surface of lithium metal, triggering the growth of Li dendrites. Therefore, immobilizing the anions in the electrolyte would be an efficient solution for Li dendrites in the electrolyte.

Inorganic solid-state electrolytes are the ultimate solution for Li dendrites, because they fundamentally change the behavior of Li deposition.35–37 Inorganic solid-state electrolytes have the advantage of being single-ion conductors that can have a high Li-ion transference number close to 1 and negligible electronic conductivity. Theoretical studies have also demonstrated that the utilization of active materials may remain close to 100% even at relatively high charge/discharge currents with single-ion electrolytes.33,34,38 In addition, Li dendrite growth in the electrolyte has been investigated, where it is shown that Li dendrite growth is suppressed when the anions in the electrolyte are partially immobilized.39–41 As a result, single-ion electrolytes have demonstrated an excellent capability to suppress Li dendrites.34,42 Countering these advantages though are new challenges arising from the interfacial contact between the electrode (especially the cathode side) and the inorganic solid-state electrolyte.36 Another traditional approach to realize single-ion transport in the electrolyte is to employ a polymer as the electrolyte backbone with some strategies to immobilize the anions including covalently linking the anions to the polymeric backbone, attaching anions to the inorganic backbone, and addition of trapping agents for anions to dual-ion conducting single-ion polymer electrolytes.33 Nevertheless, the ionic conductivity of single-ion polymer electrolytes is inferior especially at room temperature.33 Despite the rapid progress made in solid-state batteries, major obstacles still remain for the commercialization of all solid-state batteries. Hence, we are motivated to realize single-ion transport in a liquid electrolyte. Zhou et al.42 previously demonstrated that single-ion transport in batteries can be obtained with a MOF membrane infiltrated with electrolyte. However, since the anions in their paper have no interreaction with the MOF matrix, they can also cross the membrane through the interparticle voids between MOF particles.

Herein, we show that another possible kind of single-ion conductor in a liquid electrolyte can be prepared by coordinating the anions (in particular ClO4) in the liquid electrolyte on the open metal sites of a metal organic framework (MOF) matrix as a way to bring about homogeneous single Li-ion transport in the electrolyte. Further investigations of the activation energy and theoretical calculations suggest that the solvated Li+ outside the MOF pores is preferentially adsorbed into the pores of the MOF due to abundant lithiophilic and electronegative ClO4 inside the MOF pores, and the Li+ ions migrate within the pores of the MOF by a Grotthuss-like mechanism involving hopping of solvated Li+ between ClO4 groups. Thus, the Li ion transport in the single-ion electrolyte is dominated by this transport within the pores of the MOF.

The single-ion electrolyte presented here is versatile and has a wide range of applications. When it is used in Li‖Li symmetric batteries, Li dendrites are effectively suppressed, and an ultralong cycle life of over 2000 h with low voltage hysteresis is obtained. When the single-ion electrolyte is used in Li‖LiFePO4 full cells, we achieve a long cycle life of over 2000 cycles. Furthermore, when the single-ion electrolyte is assembled into Li‖LiMn2O4 batteries, it can effectively suppress the dissolution of Mn2+, and thus the cycling stability of the Li‖LiMn2O4 full cells is improved even at elevated temperatures. Finally, when the single-ion electrolyte is used for a Li–O2 battery, we can achieve improved cycling stability with reduced overpotential.

Experimental section

Preparation of the single-ion electrolyte

Cu-MOF-74 was synthesized via a hydrothermal method. In particular, Cu(NO3)2 (0.966 g, 4 mmol) and 2,5-dihydroxyterephthalic acid (0.396 g, 2 mmol) were dissolved into a mixture solvent of N,N-dimethyformamide and ethanol in a ratio of 20[thin space (1/6-em)]:[thin space (1/6-em)]1 (40 ml). Then, the solution was transferred into an autoclave and hydrothermally-treated at 80 °C for 20 h. After that, the precipitates were washed with ethanol several times and then collected by centrifugation. Subsequently, the precipitates were immersed into hot ethanol at 70 °C for 7 days. Finally, the precipitates were dried at 120 °C under a vacuum overnight and further activated at 200 °C under a vacuum overnight.

To prepare the single-ion electrolyte pellets, the as-synthesized Cu-MOF-74 powder was mixed with PTFE in a ratio of 9[thin space (1/6-em)]:[thin space (1/6-em)]1 and pressed into a pellet. The pellets were further compacted under 20 tons of pressure to remove the interparticle pores. Eventually, the pellets were immersed into 1 M LiClO4 DOL/DME electrolyte for Li‖Li symmetric batteries, 1 M LiClO4 PC electrolyte for LiFePO4–Li full cells and 1 M LiClO4 TEGDME for Li–O2 batteries at 80 °C overnight. Then, the pellets were taken out, wiped with tissues and dried under a vacuum for 2 hours at 40 °C.

Material characterization

The morphologies of all the materials were investigated by Scanning Electron Microscopy (SEM) (FE-SEM S-4800) and Transmission Electron Microscopy (TEM) (JEOL JEM-2100 F microscope (Japan) operated at 200 kV). Energy Dispersive Spectroscopy (EDS) mapping was carried out in conjunction with SEM; scanning transmission electron microscopy (STEM) mapping was carried out in conjunction with TEM. X-ray photoelectron spectroscopy (XPS) was carried out on an XSAM800 Ultra spectrometer. Powder X-ray diffraction was carried out on an X-ray diffractometer (Bruker D8 Advance, Germany) with Cu Kα radiation (λ[thin space (1/6-em)] = [thin space (1/6-em)]0.15406 nm). The specific surface area was investigated by the Brunauer–Emmett–Teller (BET) method. The samples for BET measurements were degassed at 150 °C under a vacuum overnight.

Electrochemical measurements

The ionic conductivity was measured by electrochemical impedance spectroscopy (EIS) by sandwiching the electrolyte between two stainless steel plates. The EIS spectrum was measured on an AUTOLAB electrochemical workstation (PGSTAT 302N) from 1 MHz to 1 Hz. The electrochemical window of the single-ion electrolyte was measured by cyclic voltammetry with a scan rate of 1 mV S−1.

The LiFePO4 cathodes were prepared by grinding commercialized LiFePO4 powder with Super P and PVDF powder in a ratio of 7[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]1. Subsequently, the mixture was stirred in NMP solvent and slurried on aluminum foil. Finally, the cathode was dried under a vacuum at 80 °C overnight. The LiMn2O4 cathode was also prepared in the same way. The average cathode loading is about 3.0 mg cm−2 based on the weight of the active materials.

To assemble the LiFePO4 full cells, Li metal was employed as the counterpart electrode. Then, the cells were assembled by sandwiching the single-ion electrolyte between the LiFePO4 cathode and the Li metal anode. For comparison, Li‖LiFePO4 full cells were also prepared with Celgard 2300 dipped in 1 M LiClO4 PC electrolyte. The Li‖LiMn2O4 full cells were also assembled in the same way.

To prepare the Li–O2 cells, the cathode was prepared by mixing Kejenblack with PVdF in a ratio of 9[thin space (1/6-em)]:[thin space (1/6-em)]1. Then, the mixture was slurried on carbon paper, which was dried at 120 °C overnight. The average cathode loading is 0.5 mg cm−2 (based on the weight of Kejenblack).

The Li–O2 batteries were assembled with CR2036 Li–O2 coin cells in a glovebox filled with Ar. The O2 cathode and Li metal were separated by the single-ion electrolyte. For comparison, the O2 cathode and Li metal were also separated by Celgard dipped in 1 M LiTFSI TEGDME electrolyte.

Computational details

Computational details are given in the ESI.

Results and discussion

Synthesis and characterization of the single-ion electrolyte

MOFs have been previously investigated for ion conduction of various cations43 such as proton conductors44,45 or Li+ conductors.46–50 As shown in Fig. 1a, Cu-MOF-74 is a nanoporous material comprised of Cu2+ metal ion nodes and 2,5-dihydroxyterephthalic acid organic ligands; it has hexagonal, one-dimensional pores lined with Cu2+ ions whose open metal sites point directly into the channel. Cu MOF-74 is carefully chosen from thousands of MOFs because this kind of MOF has one of the highest open metal site densities known51 along with suitable electrochemical stability against Li metal. Thus, when ClO4 is coordinated with the open metal center, it will be inside the nanopores of Cu MOF-74.
image file: c9ee01473j-f1.tif
Fig. 1 Synthesis and characterization of the single-ion electrolyte: (a) schematic illustration of the structure of Cu-MOF-74; (b) TEM image and corresponding elemental mapping of Cu-MOF-74; (c) photograph of the single-ion electrolyte; (d) XRD pattern of Cu-MOF-74 and the single-ion electrolyte; (e) FT-IR spectrum of Cu-MOF-74 and the single-ion electrolyte; (f) XPS spectrum of Cl 2p 1/2.

Cu MOF-74 was prepared by mixing 4 mmol Cu(NO3)2 with 2 mmol 2,5-dihydroxyterephthalic acid in a mixed solvent of N,N-dimethyformamide (DMF) and ethanol; then, after hydrothermal treatment at 80 °C for 12 h, Cu-MOF-74 is formed. After activating Cu-MOF-74 at 200 °C under a vacuum, the as-obtained activated Cu-MOF-74 is further thermally-treated with 1 M LiClO4 in various solvents (propylene carbonate (PC) for LiFePO4 and LiMn2O4 full cells, 1,3-dioxolane/ethylene glycol dimethyl ether (DOL/DME (1[thin space (1/6-em)]:[thin space (1/6-em)]1)) for Li‖Li symmetric cells, and tetraglyme (TEGDME) for Li–O2 batteries) at 80 °C for 12 h in a sealed vessel. Eventually, the single-ion electrolyte is formed. The as-obtained powders were then compacted under 20 tons of pressure to remove the interparticle pores, and finally the pellet was dried for 30 min at 40 °C under a vacuum to remove the excess solvent.

To investigate the morphology and the composition of Cu-MOF-74, transmission electron microscopy (TEM) was carried out. Fig. 1b shows the TEM image of Cu-MOF-74, which shows a rodlike morphology, and the elemental mapping of Cu-MOF-74 confirms that Cu, C, and O elements are uniformly distributed in the MOF particles, which is in accordance with EDS mapping in SEM (Fig. S1, ESI). To further investigate the morphology of Cu-MOF-74, scanning electron microscopy (SEM) was also carried out. It can be identified in the SEM images that the cross section of Cu-MOF-74 shows a hexagonal morphology (Fig. S1, ESI).

Fig. 1c shows photographs of the single-ion electrolyte. When the liquid electrolyte (LiClO4 in solvent) is dripped on the surface of the compacted single-ion electrolyte, the liquid electrolyte does not permeate into the single-ion electrolyte, indicating that most interparticle voids of the electrolyte have been squeezed out after the compaction process. Further evidence of SEM images of the electrolyte sheet (Fig. S2, ESI) also confirms that no obvious piled pores are observed on the surface of the electrolyte, suggesting that the MOF particles of the electrolyte are densely stacked. Further investigation by EDS mapping in conjunction with SEM suggests that C, O, Cu, Cl and F are uniformly distributed in the electrolyte sheet (the F element observed in the electrolyte sheet arises from the PTFE binder). The SEM image of the single-ion electrolyte after cycling 50 times is also given in Fig. S3 (ESI); the result suggests that the morphology of the single-ion electrolyte remains unchanged.

To further investigate the structure of the single-ion electrolyte, X-ray Diffraction (XRD) is also carried out on Cu-MOF-74 before and after treating with LiClO4 solution (Fig. 1d). The XRD pattern of pristine Cu-MOF-74 is consistent with the previously published XRD pattern of Cu-MOF-74,52 confirming that we successfully synthesized Cu-MOF-74. After treating with LiClO4 solution, the XRD pattern of Cu-MOF-74 remains unchanged, indicating the structural stability of Cu-MOF-74 after reaction with LiClO4. The XRD pattern of the single-ion electrolyte after compaction (Fig. S4, ESI) and after cycling 50 times (Fig. S5, ESI) suggests that the structure of the single-ion electrolyte is stable during the compaction process and cycling.

The reaction between LiClO4 and Cu-MOF-74 was investigated through Fourier transform infrared spectroscopy (FT-IR) and X-ray photoelectron spectroscopy (XPS). As shown in Fig. 1e, after treating with LiClO4, some new peaks (630 cm−1, 870 cm−1 and 950 cm−1) appear in Cu-MOF-74; these can be assigned to Cu(ClO4)2.48 We also performed X-ray photoelectron spectroscopy on Cu-MOF-74 after treating with LiClO4. Cl 2p 1/2 (Fig. 1f) is broad, extending from 204 to 210 cm−1, and is fitted by two peaks, one assigned to LiClO4 (∼208.4 eV) and one assigned to Cu(ClO4)2 (∼207.2 eV).48 Notably, the peak area assigned to Cu(ClO4)2 is significantly larger than that of LiClO4, indicating that most of the LiClO4 salt has reacted with the open metal sites of Cu2+ in the MOF and eventually converted to Cu(ClO4)2 after thermal treatment at 80 °C. Thus, we conclude that after treating with the LiClO4 electrolyte, the ClO4 anions in the electrolyte react with Cu-MOF-74 and are anchored on the open metal site of Cu-MOF-74, leaving free solvated Li+ ions in the pores of Cu-MOF-74. The specific surface area and pore distribution before and after treating with LiClO4 were also measured by the Brunauer–Emmett–Teller method (see Fig. S6, ESI). The result shows that after treating with LiClO4, the surface area decreases from 630 m2 g−1 to 35 m2 g−1 and the pore volume sharply reduces from 0.35 cc g−1 to 0.065 cc g−1, indicating that LiClO4 is incorporated into the pores of Cu MOF-74. We also perform N2 adsorption and desorption measurements on the Cu MOF-74-PTFE pellet to study the porosity of the single-ion electrolyte (see Fig. S7, ESI). The results suggest that even after densification under pressure, some stacked pores still exist.

Ionic conductivity investigation of the single-ion electrolyte

To investigate the ionic conductivity of the single-ion electrolyte, Electrochemical Impedance Spectroscopy (EIS) was carried out at various temperatures ranging from 298 K to 333 K. The single-ion electrolyte was sandwiched between two stainless steel pellets in a CR2032. Since the ionic conductivity of the MOF based ion conductor is highly dependent on the amount of solvent in the electrolyte, we also investigated the influence of the solvent on the ionic conductivity. To investigate the influence of the amount of solvent on the ionic conductivity, an electrolyte without vacuum drying was also prepared. Fig. 2a shows Nyquist plots at 298 K, 313 K, 323 K and 333 K of the electrolyte without vacuum drying. The ionic conductivity (S cm−1) was calculated as L/(RS), where L (cm) is the thickness of the electrolyte, R (Ohm) is the impedance, and S (cm2) is the area of the electrolyte. The electrolyte without drying presents a high ionic conductivity of 10−3 S cm−1. In addition, the electrolyte shows temperature-dependent conductivity with typical Arrhenius-like behavior with a low activation energy of 0.13 eV. The low activation energy of the single-ion electrolyte suggests that the mechanism of Li+ transport in the MOF matrix is via a Grotthuss-like mechanism.46 Fig. 2c shows Nyquist plots at 298 K, 313 K, 323 K and 333 K of the electrolyte after drying under a vacuum at 40 °C for 30 min. The ionic conductivity of the electrolyte after drying reduces to 10−5 S cm−1 with an activiation energy of 0.29 eV (Fig. 2d). Although the activation energy slightly increases after drying under a vacuum, it is still below 0.4 eV, suggesting that the mechanism of Li+ transport in the MOF matrix is also via a Grotthuss-like mechanism even with less solvent in the electrolyte. Above all, the results suggest that a higher amount of solvent in the single-ion pellet not only facilitates the ionic conductivity of the single-ion pellet but also reduces the activiation energy of ion transport in the single-ion pellet. For comparison, the ionic conductivity of the conventional single-ion polymer electrolyte is usually 10−7 S cm−1 at room temperature (see Table S1 for comparison, ESI).
image file: c9ee01473j-f2.tif
Fig. 2 Ionic conductivity investigation of the single-ion electrolyte at various temperature: (a) EIS of the electrolyte without vacuum drying at various temperature; (b) Arrhenius plot of the single-ion electrolyte without vacuum drying; (c) EIS of the single-ion electrolyte after vacuum drying for 30 min at various temperature; (d) Arrhenius plot of the single-ion electrolyte after vacuum drying for 30 min; (e) schematic illustration of the Li diffusion path in the pore of the MOF (O atoms: red, Cl atoms: yellow; Cu atoms: blue; Li atoms: grey.)

Fig. 2e schematically illustrates the Li ion transport path in the pores of the MOF. Since the O containing groups in the MOF are electronegative and highly lithiophilic,30 the charge is transported by the coordinated hopping of solvated Li+ between the oxygen groups in the pores of Cu-MOF-74. Hence, we propose two possible pathways for Li ion transport: (a) Li transitions between the oxygen sites on the Cu-MOF-74 backbone; and (b) Li transitions between the perchlorate fragments which are added into the pores of Cu-MOF-74. When the solvated Li ions diffuse out of the pores of Cu-MOF-74, they will become free-moving solvated Li ions. However, the free-moving Li-ion prefers to be adsorbed into the pores of a neighboring Cu-MOF-74 particle, because of the presence of highly lithiophilic and electronegative ClO4 inside the pores of Cu-MOF-74.

To confirm the single-ion transport mechanism, the Li transference number was also calculated by the potentiostatic polarization method on the basis of the Bruce–Vincent–Evans equation53 (see Fig. S8 for details, ESI). The calculated Li ion transference number is as high as 0.82, indicating that most of the anions are immobilized on the open metal sites. However, the typical Li transference number in conventional liquid batteries with a Celgard separator is only around 0.2–0.442.34,48

To further probe the electrochemical window of the single-ion electrolyte, Cyclic Voltammetry (CV) was also carried out with a scan rate of 1 mV s−1 from −0.2 to 5.0 V. The electrochemical window of the single-ion electrolyte is from 2.1–4.9 V, which makes it appropriate for use with most cathode materials (see Fig. S9 and the associated discussion, ESI). To probe whether the MOF backbones in the single-ion electrolyte paticipate in the formation of a solid electrolyte interface (SEI), the XPS spectrum was also measured for Li metal taken from the Li‖Li symmetric cell (see Fig. S10 and the corresponding discussion, ESI). As shown in Fig. S10 (ESI), some peaks correponding to carboxyl groups and hydroxyl groups are observed on the surface of the Li metal. Since the solvent in the Li‖Li symmetric cells is DOL/DME, which is stable against Li metal, those carboxyl groups and hydroxyl groups based on the SEI layer come from the MOF backbone. In conclusion, the MOF backbones in the single-ion electrolyte participate in the formation of the SEI layer on the surface of the Li metal.

Theoretical investigation of the Li+ transport mechanism in the single-ion electrolyte

Due to their unique structure, MOFs have been widely investigated for proton conduction,44,45 and two key factors are found to be indispensable for proton conduction in MOFs. The first factor is a certain degree of humidity, while the second factor is the presence of highly hydrophilic and electronegative functional groups (most commonly –SO32−) in the nanopores of MOFs. For proton transport inside the pores of a MOF, the most common mechanism is the Grotthuss mechanism, where solvated protons hop between two neighboring sulfonate groups inside the pores of the MOF matrix. When the protons diffuse out of the nano-pores of the MOF into the void between MOF particles, they will become free-moving hydrated protons between two adjacent MOF particles. However, it is favorable for them to be adsorbed into the pores of the neighboring MOF particles because there are abundant highly hydrophilic and electronegative sulfonate groups in the pores of the MOF. Hence, the proton conduction in MOFs is dominated by proton transport in the pores of the MOF matrix. This mechanism has also been proposed for Li ion conduction in MOFs;46–50 when the two indispensable components (i.e., the highly lithiophilic groups inside the pores of the MOFs and solvent) are present, the MOF can support Li+ transport in the nanopores. For example, a sulfonate-functionalized MOF separator was previously reported as an ion sieve for separating cations (e.g., H+, Li+, Na+, and Mg2+) in water.46

To investigate the Li+ transport mechanism in the single-ion electrolyte proposed here, quantum chemistry calculations were performed to compute the free energies of activation ΔGact for the two possible pathways mentioned above. For this purpose, we first optimized the bulk structure of Cu-MOF-74 by spin-polarized Kohn–Sham density functional theory (DFT) with the PBE-D3(BJ) exchange–correlation functional and periodic boundary conditions. We then carved a cluster model with the formula C181H87Cu24O94 from the periodic-DFT optimized bulk structure, and free energies of activation were computed by finite-cluster calculations in the presence of a solvent reaction field with the solvent modeled as diethyl ether. The DFT-optimized crystalline structure of Cu-MOF-74 is shown in Fig. S11 (ESI), and the cluster model is shown in Fig. S12 (ESI).

Fig. 3 depicts the two possible migration pathways, with the MOF backbone drawn in wireframe style and the atoms that are allowed to relax during the cluster geometry optimization shown explicitly. To simulate the Li transport process in batteries, the Gibbs free energy barriers are corrected with self-consistent reaction field (SCRF) calculations in diethyl ether (DME) solvent. The first path is Li+ hopping between the O hollow sites on the Cu-MOF-74 matrix (Path I), while the second path is Li+ hopping between the ClO4 anchored on the open metal sites of Cu-MOF-74 (Path II). The Gibbs free energies of activation were computed at 298 K for these two possible pathways. As shown in Fig. 3, ΔGact for Li+ hopping between oxygen hollow sites (Path I) on Cu-MOF-74 is about 1.2 eV, and ΔGact for Li+ hopping between perchlorate fragments (Path II) in Cu-MOF-74 is only about 0.4 eV. Considering the lower migration energy barrier of Path II compared with Path I, we conclude that Li+ migration in the single-ion electrolyte is dominated by the hopping of Li ions between the ClO4 anchored on the open metal sites of Cu-MOF-74. As a result, a homogeneous Li ion flux is achieved by the single-ion electrolyte.


image file: c9ee01473j-f3.tif
Fig. 3 Quantum chemistry investigation of Li migration in the single-ion electrolyte. The MOF backbone is drawn in the wireframe style, and the atoms that are allowed to be relaxed during geometry optimization are explicitly shown (colour code: Cu orange, O red, C grey, H white, Li violet, and Cl green).

Morphology of Li metal with the single-ion electrolyte

The surface morphology of Li metal in different electrolytes was investigated by ex situ SEM. Li‖Li symmetric batteries were assembled with different electrolytes and then cycled for a certain period of time. Fig. 4 shows the morphology of the Li metal in different electrolytes and obvious differences are evident in the morphology of the Li metal cycling in different electrolytes. The Li‖Li symmetric batteries with LiClO4 DOL/DME can only cycle for 10 h, and the SEM images of the Li metal after cycling for 10 h with LiClO4 DOL/DME electrolyte present a rough surface with loosely stacked deposits. This loosely stacked Li metal is especially vulnerable to attack from the electrolyte and can be easily peeled off from the Li metal. When employing LiTFSI DOL/DME, the cycle life was prolonged to over 100 h. The SEM images of the Li metal also show a loose and rough surface with obvious needle-like Li dendrites coating the surface of the Li metal, while a compact and smooth surface is observed on Li cycling with the single-ion electrolyte even after 100 h. Above all, a smooth surface with compact stacking of Li is achieved with the single-ion electrolyte.
image file: c9ee01473j-f4.tif
Fig. 4 Morphology of Li metal with different electrolytes.

Electrochemical performance of Li‖Li symmetric batteries and Li‖LiFePO4 full cells with the single-ion electrolyte

To investigate the electrochemical performance of the single-ion electrolyte, Li‖Li symmetric cells were assembled by separating two pieces of Li metal with the single-ion electrolyte. As shown in Fig. 5a, the Li‖Li symmetric cell exhibits an ultralong cycle life of over 2000 h without short circuits at a current density of 0.2 mA cm−2, indicating the superior ability of the single-ion electrolyte to suppress Li dendrites. In addition, the Li‖Li symmetric cell with the single-ion electrolyte shows a low voltage hysteresis of 40 mV for the first several cycles, which reduces to 20 mV after 400 h and then remains at 20 mV over 2000 h. Even after 2000 h, no sign of short circuits is observed for the Li‖Li symmetric cells with the single-ion electrolyte. When the Li‖Li symmetric cells are tested at higher current density with higher deposition capacity, excellent performance is also achieved (Fig. S13, ESI). For comparison, the Li‖Li symmetric cell was also assembled with Celgard infiltrated with LiTFSI DOL/DME electrolyte, which was cycled at a low current rate of 0.2 mA cm−2; the polarization suddenly increased after only 400 h (Fig. S14, ESI). To our knowledge, the cycle life of Li‖Li symmetric batteries with our single-ion electrolyte is the best ever reported (see Table S2 for comparison, ESI).
image file: c9ee01473j-f5.tif
Fig. 5 Electrochemical performance of Li-metal batteries with the single-ion electrolyte: (a) Li‖Li symmetric cell with the single-ion electrolyte at a current density of 0.2 mA cm−2; (b) rate performance of a Li‖LiFePO4 full cell with the single-ion electrolyte; (c) corresponding voltage profile of Li‖LiFePO4; (d) cycling performance of Li‖LiFePO4 full cells with the single-ion electrolyte at a rate of 1C; (e) corresponding voltage profiles of the single-ion electrolyte at 1C during the cycles.

To further investigate the properties of the single-ion electrolyte, Li‖LiFePO4 full cells were also assembled in a glovebox. As shown in Fig. 5b, the Li‖LiFePO4 full cell shows a specific capacity of 163 mA h g−1 at a rate of 0.2C (1C = 170 mA h g−1), which slightly reduces to 160 m A h g−1 at a rate of 0.5C. When the current density increases to 1C, a specific capacity of 150 mA h g−1 is also achieved. In addition, when the current density further increases to 2C and 5C, a high capacity of 140 mA h g−1 and 110 mA h g−1 is also achieved with the single-ion electrolyte. The corresponding voltage profiles at different current densities are presented in Fig. 5c. It can be identified that even at a high current rate of 5C, the polarization of the full cell is only 0.4 V.

Fig. 5d shows the cycling performance of Li‖LiFePO4 full cells with the single-ion electrolyte at a rate of 1C. It is identified that the cell with the single-ion electrolyte presents an initial discharge capacity of 142 mA h g−1, which slightly increases to 151 mA h g−1 after 10 times. The specific capacity of the Li‖LiFePO4 full cell remains over 150 mA h g−1 without any decay after 200 times. Then, the specific capacity slightly decreases to 142 mA h g−1 after 300 cycles, and 138 mA h g−1 after 500 cycles, which corresponds to a capacity retention of 100% and 97.2% of the initial capacity respectively. Even after 2000 cycles, a specific capacity of 106.5 mA h g−1 is also achieved for the single-ion electrolyte, corresponding to a capacity retention of over 75%. To our knowledge, this is the longest cycle life ever reported in Li‖LiFePO4 full cells (see Table S3 for comparison, ESI). For comparison, the cycling stability of the Li‖LiFePO4 full cell was also investigated in the conventional electrolyte, which exhibits an inferior cycle life. Fig. 5e presents the corresponding voltage curves of Li‖LiFePO4 with the single-ion electrolyte at a rate of 1C. It can be identified that the overpotential of Li‖LiFePO4 remains almost unchanged for the first 500 cycles. Even after 2000 cycles, the overpotential is only slightly increased. However, the polarization of Li‖LiFePO4 full cells with the conventional liquid electrolyte became severer after only 100 cycles (Fig. S15, ESI). The cycling performance of the Li‖LiFePO4 full cell and the corresponding voltage profiles of the Li‖LiFePO4 full cells at 0.2C and 0.5C are also presented in the ESI as Fig. S16 and S17 (ESI), which also shows excellent cycling performance.

Electrochemical performance of Li‖LiMn2O4 full cells with the single-ion electrolyte

LiMn2O4 with a spinel structure is a very promising cathode for Li batteries owing to the high specific capacity, high operating voltage of 4 V, low cost, superior safety and long cycle life within a wide working temperature range.54 However, the poor cycling performance of this material, in particular at elevated temperatures, limits its wide application as a cathode material for Li metal batteries.54 A previous report54 has shown that the capacity fade of 4 V LiMn2O4 cells is directly associated with the dissolution of Mn2+ from the cathode/electrolyte interface because of the disproportionation reaction of Mn(III), and the subsequent deposition of Mn2+ on the anode. Herein, we show that our single-ion electrolyte can effectively suppress the dissolution of Mn2+ into the electrolyte and thus improve the cycling stability of the LiMn2O4‖Li full cell.

Fig. 6a shows the cycling performance of the LiMn2O4‖Li full cell at room temperature. It is identified in Fig. 6a that the LiMn2O4‖Li full cell shows an initial discharge capacity of 116 mA h g−1. After 250 cycles a discharge capacity of 106 mA h g−1 is still achieved, corresponding to a capacity retention of 92.2%. For comparison, the cycling performance of the Li‖LiMn2O4 full cell in the traditional liquid electrolyte only shows a capacity retention of 83%. The corresponding voltage profiles over the cycles are given in Fig. 6b. For comparison the corresponding voltage curve of the Li‖LiMn2O4 full cell in the conventional electrolyte is given in Fig. S18a (ESI). When the working temperature increases to 55 °C, the comparison of the capacity retention becomes much more obvious. Fig. 6c shows the cycling performance of LiMn2O4‖Li full cells at 55 °C. It can be noted that the LiMn2O4‖Li full cell exhibits an initial discharge capacity of 114 mA h g−1, which reduces to 81.3 mA h g−1 after 250 cycles with the single-ion electrolyte, corresponding to a capacity retention of 72% after 250 cycles at 55 °C. However, the capacity retention is only 52% with the conventional liquid electrolyte after 250 cycles. Fig. 6d shows the corresponding voltage curves of the LiMn2O4‖Li full cell with the single-ion electrolyte. The corresponding voltage curves with the conventional electrolyte at 55 °C are also given in Fig. S18b (ESI).


image file: c9ee01473j-f6.tif
Fig. 6 Li‖LiMn2O4 full cells at a rate of 1C: (a) cycling performance at a rate of 1C at room temperature; (b) corresponding voltage profiles of the cell with the single-ion electrolyte at room temperature; (c) cycling performance at a rate of 1C at 55 °C; (d) corresponding voltage profiles of the single-ion electrolyte at 55 °C.

A previous report55 has shown that the dissolution of Mn2+ in LiMn2O4 batteries involves in a series of processes including the dissolution of Mn2+ into the electrolyte, and then the migration of Mn2+ from the cathode region to the anode region. Eventually, Mn2+ will deposit on the surface of the anode. This process will result in a structural change of the LiMn2O4 cathode, which is responsible for the capacity decay of LiMn2O4‖Li batteries. Since the ion diameter of Mn2+ is significantly larger than that of Li ions, the migration rate of Mn2+ inside the pores of Cu MOF-74 is much slower than that of Li+. As a result, this single-ion electrolyte can suppress the dissolution of Mn2+. Thus, improved cycling stability is achieved for LiMn2O4‖Li full cells with the single-ion electrolyte even at elevated temperatures.

Electrochemical performance of Li–O2 batteries with the single-ion electrolyte

To further demonstrate the application of the single-ion electrolyte, Li–O2 batteries were also assembled with the single-ion electrolyte. Fig. 7a schematically illustrates the structure of the Li–O2 cells. The cathode was prepared by mixing 90 wt% Kejenblack and 10 wt% PVDF. Then, CR2036 Li–O2 coin cells were assembled by separating the cathode and Li metal with the single-ion electrolyte. Fig. 7b shows the midpoint voltage of the Li–O2 batteries with the single-ion electrolyte and the traditional G4 electrolyte. It can be identified that the polarization of the Li–O2 cell with the single-ion electrolyte is much lower than that of the Li–O2 batteries with the G4 (1 M LiTFSI in TEGDME) electrolyte. The decrease in the polarization may be attributed to the catalytic activity of MOF-74 for Li peroxide, which has been demonstrated by previous reports.56,57 Fig. 7c shows the voltage profiles of the Li–O2 cell with the G4 electrolyte. After only 40 cycles the cut-off discharge voltage drops below 2.0 V, while the cycle life of the Li–O2 cells with the single-ion electrolyte (Fig. 7d) is further prolonged 50 times with lower polarization. Even after 50 cycles, the cut-off discharge voltage of the Li–O2 cells is still above 2.0 V. It should be noted that the cycle life of Li–O2 batteries may be limited by the KB cathode, as the lithium peroxides formed during discharge may clog the pores of KB, leading to the decay of the batteries. Above all, improved performance of Li–O2 batteries is achieved by the single-ion electrolyte.
image file: c9ee01473j-f7.tif
Fig. 7 Performance of Li–O2 cells with the single-ion electrolyte: (a) Schematic illustration of the Li–O2 cell structure with the single-ion electrolyte; (b) comparison of the mid-point voltage of Li–O2 cells with the single-ion electrolyte (red) and traditional G4 electrolyte (blue); (c) voltage profiles of Li–O2 batteries with the G4 electrolyte; (d) voltage profiles of Li–O2 batteries with the single-ion electrolyte.

Conclusions

We report a versatile single-ion electrolyte, which is achieved by coordinating the anion on the open metal sites of Cu-MOF-74. The single-ion electrolyte prepared in this way radically changes the behavior of Li+ migration in the electrolyte, which opens a new pathway departing from the traditional solid electrolyte or polymer electrolyte to realize homogeneous single-ion transport in a liquid electrolyte. When the electrolyte is assembled into LiFePO4–Li metal batteries, Li dendrites are suppressed. As a result, an ultralong cycle life of 2000 times with a capacity retention of nearly 75% is achieved for Li‖LiFePO4 full cells with this single-ion electrolyte. When the single-ion electrolyte is applied for LiMn2O4‖Li batteries, the dissolution of Mn2+ is suppressed and improved performance is achieved even at high temperatures. Finally, when the single-ion electrolyte is employed as the electrolyte for Li–O2 batteries, an improved cycle life with reduced overpotential is also achieved.

Authors contributions

Shouyi Yuan conceived this idea and designed the experiments. Junwei Lucas Bao carried out the theoretical calculations. Jishi Wei helped to plot the 3D matrix scheme in Fig. 2e. Yonggang Wang, Yongyao Xia and Donald G. Truhlar directed the project. Shouyi Yuan performed the material synthesis, characterization, electrochemical measurements and data analysis. Shouyi Yuan, Junwei Lucas Bao and Donald G. Truhlar co-wrote the paper. All authors discussed the results and commented on the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We acknowledge funding from the National Natural Science Foundation of China (21622303), National Key R&D Programme of China (2018YFE0201702), the State Key Basic Research Program of China (2016YFA0203302), and the U.S. Department of Energy, Office of Basic Energy Sciences (DE-FG02-17ER16362).

Notes and references

  1. M. Armand and J. M. Tarascon, Nature, 2008, 451, 652–657 CrossRef CAS PubMed.
  2. X. B. Cheng, R. Zhang, Z. C. Zhao and Q. Zhang, Chem. Rev., 2017, 117, 10403–10473 CrossRef CAS PubMed.
  3. W. Xu, J. L. Wang, F. Ding, X. L. Chen, E. Nasybulin, Y. H. Zhang and J. G. Zhang, Energy Environ. Sci., 2014, 7, 512–537 Search PubMed.
  4. R. Zhang, N. W. Li, X. B. Cheng, Y. X. Yin, Q. Zhang and Y. G. Guo, Adv. Sci., 2017, 4, 1600445 CrossRef PubMed.
  5. S. Y. Yuan, J. W. L. Bao, L. N. Wang, Y. Y. Xia, D. G. Truhlar and Y. G. Wang, Adv. Energy Mater., 2016, 6, 1501733 CrossRef.
  6. J. Zhang, C. P. Yang, Y. X. Yin, L. J. Wan and Y. G. Guo, Adv. Mater., 2016, 28, 9539–9544 CrossRef CAS PubMed.
  7. G. Y. Xu, B. Ding, P. Nie, L. F. Shen, H. Dou and X. G. Zhang, ACS Appl. Mater. Interfaces, 2014, 6, 194–199 CrossRef CAS PubMed.
  8. F. Li, Y. Chen, D. M. Tang, Z. Jian, C. Liu, D. Golberg, A. Yamadad and H. Zhou, Energy Environ. Sci., 2014, 7, 1648–1652 RSC.
  9. X. G. Yang, J. J. Xu, Z. W. Chang, D. Bao, Y. B. Yin, T. Liu, J. M. Yan, D. P. Liu, Y. Zhang and X. B. Zhang, Adv. Energy Mater., 2018, 8, 1702242 CrossRef.
  10. Y. N. Chen, Q. Zhang, Z. Zhang, X. L. Zhou, X. L. Y. R. Zhong, M. Yang, Z. J. Xie, J. P. Wei and Z. Zhou, J. Mater. Chem. A, 2015, 3, 17874–17879 RSC.
  11. N. Chen, Y. J. Dai, Y. Xing, L. Wang, C. Guo, R. J. Chen, S. J. Guo and F. Wu, Energy Environ. Sci., 2018, 9, 1803372 Search PubMed.
  12. S. H. Jiao, X. D. Ren, R. G. Cao, M. H. Engelhard, Y. Z. Liu, D. H. Hu, D. H. Mei, J. M. Zheng, W. G. Zhao, Q. Y. Li, N. Liu, B. D. Adams, C. Ma, J. Liu, J. G. Zhang and W. Xu, Nat. Energy, 2018, 3, 739–746 CrossRef CAS.
  13. X. L. Fan, L. Chen, O. Borodin, X. Ji, J. Chen, S. Hou, T. Deng, J. Zheng, C. Y. Yang, S. C. Liou, K. Amine, K. Xu and C. S. Wang, Nat. Nanotechnol., 2018, 13, 715–722 CrossRef CAS PubMed.
  14. L. M. Suo, Y. S. Hu, H. Li, M. Armand and L. Q. Chen, Nat. Commun., 2013, 4, 1481 CrossRef PubMed.
  15. C. Yan, X. Y. Yao, X. Chen, X. B. Chen, X. Q. Zhang, J. Q. Huang and Q. Zhang, Angew. Chem., Int. Ed., 2018, 57, 1807034 Search PubMed.
  16. Q. Pang, X. Liang, A. Shyamsunder and L. F. Nazar, Joule, 2017, 1, 871–886 CrossRef CAS.
  17. J. M. Zheng, P. F. Yan, D. H. Mei, M. H. Engelhard, S. S. Cartmell, B. J. Polzin, C. M. Wang, J. G. Zhang and W. Xu, Adv. Energy Mater., 2016, 6, 1502151 CrossRef.
  18. F. Chu, J. Hu, C. Wu, Z. Yao, J. Tian, Z. Li and C. Li, ACS Appl. Mater. Interfaces, 2019, 11, 3869–3879 CrossRef CAS PubMed.
  19. B. Zhou, L. Guo, Y. Zhang, J. Wang, L. Ma, W. Zhang, Z. Fu and Z. Peng, Adv. Mater., 2017, 29, 1701568 CrossRef PubMed.
  20. X. Liang, Q. Pang, I. R. Kochetkov, M. S. Sempere, H. Huang, X. Sun and L. F. Nazar, Nat. Energy, 2017, 2, 17119 CrossRef CAS.
  21. J. Hu, K. Y. Chen and C. Li, ACS Appl. Mater. Interfaces, 2018, 10, 34322–34331 CrossRef CAS PubMed.
  22. B. Zheng, J. Zhu, H. Wang, M. Feng, E. Umeshbabu, Y. Li, Q. Wu and Y. Yang, ACS Appl. Mater. Interfaces, 2018, 10, 25473–25482 CrossRef CAS PubMed.
  23. J. Yan, J. Yu and B. Ding, Adv. Mater., 2018, 30, 1705105 CrossRef PubMed.
  24. H. Zhao, D. Lei, Y. He, Y. Yuan, Q. Yun, B. Ni, W. Lv, B. Li, Q. Yang, F. Kang and J. Lu, Adv. Energy Mater., 2018, 8, 1800226 Search PubMed.
  25. L. Lu, J. Ge, J. Yang, S. Chen, H. B. Yao, F. Zhou and S. H. Yu, Nano Lett., 2016, 16, 4431 CrossRef CAS PubMed.
  26. D. C. Lin, J. Zhao, J. Sun, H. B. Yao, Y. Y. Liu, K. Yan and Y. Cui, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 4613–4618 CrossRef CAS PubMed.
  27. S. Liu, A. X. Wang, Q. Li, J. S. Wu, K. V. Chiou, J. X. Huang and J. Y. Luo, Joule, 2018, 2, 184–193 CrossRef CAS.
  28. W. D. Zhang, H. L. L. Zhuang, L. Fan, L. N. Gao and Y. Y. Lu, Sci. Adv., 2018, 4, eaar4410 CrossRef PubMed.
  29. D. Zhou, R. L. Liu, Y. He, F. Li, M. Liu, B. Li, Q. Yang, Q. Cai and F. Kang, Adv. Energy Mater., 2016, 6, 1502214 CrossRef.
  30. S. Yuan, J. L. Bao, C. Li, Y. Xia, D. G. Truhlar and Y. Wang, ACS Appl. Mater. Interfaces, 2019, 11, 10616–10623 CrossRef CAS PubMed.
  31. W. Liu, Y. Mi, Z. Weng, Y. Zhong, Z. Wu and H. Wang, Chem. Sci., 2017, 8, 4285–4291 RSC.
  32. B. Q. Li, X. R. Chen, X. Chen, C. X. Zhao, R. Zhang, X. B. Cheng and Q. Zhang, Research, 2019, 2, 4608940 Search PubMed.
  33. H. Zhang, C. M. Li, M. Piszcz, E. Coya, T. Rojo, L. M. Rodriguez-Martinez, M. Armand and Z. B. Zhou, Chem. Soc. Rev., 2017, 46, 797–815 RSC.
  34. C. C. Li, B. S. Qin, Y. F. Zhang, A. Varzi, S. Passerini, J. Y. Wang, J. M. Dong, D. L. Zeng, Z. H. Liu and H. S. Cheng, Adv. Energy Mater., 2019, 9, 1803422 CrossRef.
  35. C. Yang, K. Fu, Y. Zhang, E. Hitz and L. B. Hu, Adv. Mater., 2017, 29, 1701169 CrossRef PubMed.
  36. Z. Wang, Z. Wang, L. Yang, H. Wang, Y. Song, L. Han, K. Yang, J. Hu, H. Chen and F. Pan, Nano Energy, 2018, 49, 580–587 CrossRef CAS.
  37. X. Yan, Z. Li, Z. Wen and W. Han, J. Phys. Chem. C, 2017, 121, 1431–1435 CrossRef CAS.
  38. K. E. Thomas, S. E. Sloop, J. B. Kerr and J. Newman, J. Power Sources, 2000, 89, 132–138 CrossRef CAS.
  39. M. D. Tikekar, Lynden A. Archer and D. L. Koch, Sci. Adv., 2016, 2, e1600320 CrossRef PubMed.
  40. Y. Lu, M. D. Tikekar, R. Mohanty, K. Hendrickson, L. Ma and L. A. Archer, Adv. Energy Mater., 2015, 5, 1402073 CrossRef.
  41. M. D. Tikekar, L. A. Archer and D. L. Koch, J. Electrochem. Soc., 2014, 161, A847–A855 CrossRef CAS.
  42. S. Bai, Y. Sun, J. Yi, Y. He, Y. Qiao and H. Zhou, Joule, 2018, 2, 2117–2132 CrossRef CAS.
  43. S. Horike, D. Umeyama and S. I. Kitagawa, Acc. Chem. Res., 2013, 46, 2376–2384 CrossRef CAS PubMed.
  44. F. Yang, G. Xu, Y. Dou, B. Wang, H. Zhang, H. Wu, W. Zhou, J. Li and B. Chen, Nat. Energy, 2017, 2, 877–883 CrossRef CAS.
  45. Y. Guo, Z. Q. Jiang, W. Ying, L. P. Chen, Y. Z. Liu, X. B. Wang, Z. J. Jiang, B. L. Chen and X. B. Peng, Adv. Mater., 2017, 30, 1705155 CrossRef PubMed.
  46. Y. Guo, Y. L. Ying, Y. Y. Mao, X. S. Peng and B. L. Chen, Angew. Chem., Int. Ed., 2016, 55, 15120–15124 CrossRef CAS PubMed.
  47. Z. Wang, R. Tan, H. Wang, L. Yang, J. Hu, H. Chen and F. Pan, Adv. Mater., 2018, 30, 1704436 CrossRef PubMed.
  48. L. Shen, H. B. Wu, F. Liu, J. L. Brosmer, G. R. Shen, X. F. Wang, J. I. Zink, Q. F. Xiao, M. Cai, G. Wang, Y. F. Lu and B. Dunn, Adv. Mater., 2018, 30, 1707476 CrossRef PubMed.
  49. S. S. Park, Y. Tulchinsky and M. Dinca, J. Am. Chem. Soc., 2017, 139, 13260–13263 CrossRef CAS PubMed.
  50. B. M. Wiers, M. L. Foo, N. P. Balsara and J. F. Long, J. Am. Chem. Soc., 2011, 133, 14522–14525 CrossRef CAS PubMed.
  51. E. Haldoupis, J. Borycz, H. L. Shi, K. D. Vogiatzis, P. Bai, W. L. Queen, W. L. L. Gagliardi and J. I. Siepmann, J. Phys. Chem. C, 2015, 119, 16058–16071 CrossRef CAS.
  52. Y. J. Liao, L. Zhang, M. H. Weston, W. Morris, J. T. Hupp and O. K. Farha, Chem. Commun., 2017, 53, 9376–9379 RSC.
  53. J. Evans, C. A. Vincent and P. G. Bruce, Polymer, 1987, 28, 2324–2328 CrossRef CAS.
  54. J. Lu, C. Zhan, T. P. Wu, J. G. Wen, Y. Lei, A. J. Kropf, H. M. Wu, D. J. Miller, J. W. Elam, Y. K. Sun, X. P. Qiu and K. Amine, Nat. Commun., 2014, 5, 5693 CrossRef CAS PubMed.
  55. C. Zhan, T. Wu, J. Lu and K. Amine, Energy Environ. Sci., 2018, 11, 243–257 RSC.
  56. W. J. Yan, Z. Y. Guo, H. S. Xu, Y. B. Lou, J. X. Chen and Q. W. Li, Mater. Chem. Front., 2017, 1, 1324–1330 RSC.
  57. D. F. Wu, Z. Y. Guo, X. B. Yin, Q. Q. Pang, B. B. Tu, L. J. Zhang, Y. G. Wang and Q. W. Li, Adv. Mater., 2014, 26, 3258–3262 CrossRef CAS PubMed.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ee01473j

This journal is © The Royal Society of Chemistry 2019