Novel gel polymer electrolyte for high-performance lithium–sulfur batteries

Novel gel polymer electrolyte for high-performance lithium–sulfur batteries

Nano Energy (2016) 22, 278–289 Available online at www.sciencedirect.com journal homepage: www.elsevier.com/locate/nanoenergy FULL PAPER Novel gel...

3MB Sizes 7 Downloads 119 Views

Nano Energy (2016) 22, 278–289

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/nanoenergy

FULL PAPER

Novel gel polymer electrolyte for highperformance lithium–sulfur batteries Ming Liua,b, Dong Zhoua,b, Yan-Bing Hea,n, Yongzhu Fuc, Xianying Qina,e, Cui Miaoa, Hongda Dua, Baohua Lia, Quan-Hong Yanga, Zhiqun Lind, T.S. Zhaoe, Feiyu Kanga,b,nn a

Engineering Laboratory for the Next Generation Power and Energy Storage Batteries, Graduate School at Shenzhen, Tsinghua University, Shenzhen 518055, PR China b Laboratory of Advanced Materials, Department of Materials Science and Engineering, Tsinghua University, Beijing 100084, PR China c Department of Mechanical Engineering, Indiana University-Purdue University Indianapolis, Indianapolis, IN 46202, USA d School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA e Department of Mechanical and Aerospace Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong Received 17 October 2015; received in revised form 9 January 2016; accepted 1 February 2016 Available online 19 February 2016

KEYWORDS

Abstract

Lithium–sulfur battery; Gel polymer electrolyte; Pentaerythritol tetraacrylate; In-situ synthesis; Polysulfides immobilization

The ability to suppress the dissolution of lithium polysulfides in liquid electrolyte (LE) is emerging and scientifically challenging, representing an important endeavor toward successful commercialization of lithium–sulfur (Li–S) batteries. In this context, a common and effective strategy to address this challenge is to replace the LE with a gel polymer electrolyte (GPE). However, the limited ionic conductivity of state-of-the-art GPEs and poor electrode/GPE interfaces greatly restrict their implementation. Herein, we report, for the first time, a facile in-situ synthesis of pentaerythritol tetraacrylate (PETEA)-based GPE with an extremely high ionic conductivity (1.13  10  2 S cm  1). Quite intriguingly, even interfaced with a bare sulfur cathode, this GPE rendered the resulting polymer Li–S battery with a low electrode/GPE interfacial resistance, high rate capacity (601.2 mA h g  1 at 1 C) and improved capacity retention (81.9% after 400 cycles at 0.5 C). These remarkable performances can be ascribed to the immobilization of soluble polysulfides imparted by PETEA-based GPE and the construction of a robust integrated GPE/electrode interface. Notably, due to the tight adhesion between the PETEA-based GPE and electrodes, a high-performance flexible polymer Li–S

n

Corresponding author. Corresponding author at: Engineering Laboratory for the Next Generation Power and Energy Storage Batteries, Graduate School at Shenzhen, Tsinghua University, Shenzhen 518055, PR China. E-mail addresses: [email protected] (Y.-B. He), [email protected] (F. Kang). nn

http://dx.doi.org/10.1016/j.nanoen.2016.02.008 2211-2855/& 2016 Elsevier Ltd. All rights reserved.

Novel gel polymer electrolyte

279 battery was successfully crafted. This work therefore opens up a convenient, low-cost and effective way to substantially enhance the capability of Li–S batteries, a key step toward capitalizing on GPE for high-performance Li–S batteries. & 2016 Elsevier Ltd. All rights reserved.

Introduction Recent research has witnessed rapid progress in lithium-ion batteries (LIBs) over the past two decades. However, due to the insufficient specific energy (o200 W h kg  1), LIBs still cannot meet the requirements of electric vehicles (EV) and energy storage systems (EES) [1,2]. In sharp contrast, lithium–sulfur (Li–S) battery has a theoretical specific energy of 2500 W h kg  1, more than ten times higher than that of conventional lithium-ion batteries [3,4]. Owing to the advantageous attributes, that is, abundance, non-toxicity, and environmental benignity of elemental sulfur [5], Li–S battery is widely recognized as one of most promising systems for next-generation energy storage devices. However, commercial success of Li–S battery is limited because of a set of shortcomings, including low electronic conductivity of sulfur particles and their huge structural/volumetric changes (79%) during the cathodic reaction from S to Li2S, thereby resulting in tremendous capacity fading and rate capability loss [6]. More importantly, lithium polysulfides formed as intermediates during the charge/discharge processes are highly soluble in liquid electrolyte (LE) and easily shuttle to the lithium metal anode. Subsequently, these polysulfides are reduced to Li2S2 and Li2S [7], depositing on the lithium anode and leading to the loss of active materials and the interfacial deterioration that greatly decrease the cycling stability of Li–S battery [4,8]. To overcome the dissolution of lithium polysulfides in Li–S battery as noted above, great effort has been paid to innovate the current LE system. As a rational substitute for the conventional LE, the use of gel polymer electrolyte (GPE) has been recognized as one of the most promising routes to addressing the solubility issue of polysulfides in Li– S batteries [8,9]. The implementation of GPE also offers additional benefits such as suppressing the formation of dendrite on the surface of lithium metal and improving safety [10]. To this end, a diversity of GPEs prepared by hotpress [11], phase inversion [12,13], or electrospinning [14] approaches have been employed in Li–S batteries. However, these polymer Li–S batteries generally suffer from rapid capacity decay and large polarization, which are due primarily to the limited ionic conductivity of GPEs (o3  10  3 S cm  1) and the large electrode/GPE interfacial resistance. In addition, it should be noted that the complexity in the synthesis of GPE and the assembly of GPEbased battery also profoundly restrict the application of GPEs in Li–S batteries. The key to further developing polymer Li–S batteries is to identify a novel GPE material which can be synthesized in a facile manner and, at the same time, possesses high ionic conductivity and low interfacial resistance. Recently, we developed a versatile synthetic route to a series of polymer electrolytes,

exhibiting satisfactory ionic conductivity and good affinity with electrodes [15,16]. Notably, such synthesis process allows for an integrative formation of a polymer electrolyte and an electrolyte/electrode assembly. Herein, we report an in-situ synthetic strategy for novel pentaerythritol tetraacrylate (PETEA)-based GPE with an extremely high ionic conductivity (1.13  10  2 S cm  1; to the best of our knowledge, it is the highest ionic conductivity among GPEs) crafted via an in-situ synthesis approach (details about the monomer selection and concentration optimization for GPEs please see Table S1, Figures S1–S4 in Supporting information). It is noteworthy that an integrated structure was, for the first time, developed between the GPE and a bare sulfur cathode, carrying excellent rate performance and cycling stability due primarily to the marked immobilization of polysulfides by the PETEA-based GPE. Importantly, the GPE promotes the formation of a flexible and stable passivation layer on the sulfur electrode, which can effectively inhibit the polysulfide diffusion and retain a strongly integrated electrolyte/electrode structure. Clearly, the replacement of conventional LEs with the novel PETEA-based GPE developed in this study overcomes the solubility and shuttle issues of polysulfides in Li–S batteries.

Experimental section Preparation and characterization of gel polymer electrolytes PETEA-based GPE was prepared by in-situ gelation of a precursor solution in a sealed container. The precursor solution was composed of 1.5 wt% PETEA (C17H20O8, Tokyo Chemical Industry Co., Ltd.) monomer and 0.1 wt.% azodiisobutyronitrile (AIBN, C8H12N4, Aldrich) initiator dissolved in a LE consisting of 1 M bis(trifluoromethane) sulfonamide lithium (LiTFSI) salt in a non-aqueous mixture of 1,2dioxolane (DOL)/dimethoxymethane (DME) (1:1 by volume) with 1 wt% LiNO3 additive. The precursor solution was polymerized at 70 1C for half an hour to obtain translucent GPE films. All procedures were carried out in an Ar-filled glove box (Mbraum) with the concentrations of moisture and oxygen below 1 ppm. The polymer matrix was separated and purified from GPEs and the detailed procedure is provided in Supporting information. The morphology of PETEA-based GPE was examined by field emission scanning electron microscopy (FE-SEM, HITACH S4800) at 5 kV. The Fourier transform infrared (FTIR) spectra of the PETEA monomer and the resulting polymer matrix were recorded with a Bruker Vertex70 instrument at ambient temperature. The ionic conductivity, lithium ion

280 transference number and electrochemical stability of PETEA-based GPE were measured (see Supporting information). To characterize the thermal stability of PETEA-based GPE, thermogravimetric analysis (TGA) measurements were performed using a Mettler Toledo TGA thermoanalyzer under a flow of air at a rate of 10 1C min  1.

Electrochemical performance measurements CR2032-type Li–S coin cells were assembled in an Ar-filled glove box. The PETEA-based polymer coin cells were fabricated in-situ by the direct polymerization. The cells comprised bare sulfur as the cathode, a polyolefin separator (Shenzhen Senior Technology Material Co., LTD., SD216), and lithium foil as the anode. The sulfur cathode slurry was prepared by mixing 60 wt% nano sulfur powder (Dk Nano technology, Beijing), 30 wt% Super P and 10 wt% polyvinylidene fluoride (PVDF) in N-methyl-pyrrolidone (NMP) as dispersant. The positive electrodes were fabricated by coating the slurry on the carbon-coated aluminum foil and drying at 60 1C for 24 h. The mass loading of the sulfur active material on the electrode is 1.2–1.5 mg cm  2. The precursor solution containing 1.5 wt% PETEA and 0.1 wt% AIBN dissolved in 1 M LiTFSI/DOL: DME (1:1 by volume) with 1 wt% LiNO3 electrolyte was injected into the separator and filled into the cells. To ensure the accuracy of experiment, the sulfur/electrolyte ratio in each cell is uniformly set as 30 g L  1. Subsequently, the assembled cells were aged for 2 h to ensure the precursor solution well-wetted into the electrodes. The cells were then heated at 70 1C for 2 h in a vacuum oven to ensure the complete polymerization of monomers. The [email protected] nanocomposite was prepared following a melt-diffusion strategy [17]. CMK-3 (0.3 g) and sulfur (0.7 g) were ground together, and heated to 155 1C for 10 h. The weight ratio of sulfur/carbon was adjusted to be equal to 7:3, to allow for expansion of the pore content on full lithiation to Li2S. The [email protected] electrode was prepared by mixing 80 wt% [email protected] powder, 10 wt% Super P and 10 wt% PVDF in a NMP solvent dispersant and then coating the slurry on aluminum foil. [email protected]/GPE/Li batteries were assembled with the same procedure as S/GPE/Li batteries. The assembled Li–S cells were cycled at various charge/discharge rates between 1.7–2.8 V on a Land 2001A battery testing system at 25 1C. Cyclic voltammograms (CVs) of the assembled polymer Li–S cells were tested using a VMP3 electrochemical working station (Bio Logic Science Instruments, France) at a scanning rate of 0.05 mV s  1 with sulfur electrode as the working electrode and lithium coil as both reference and counterelectrodes. Electrochemical impedance spectrum (EIS) of Li–S cells at half discharge state was examined using the VMP3 multichannel electrochemical station in the frequency range of 10  2–105 Hz by applying a 5 mV ac oscillation. The Li–S coin cells after designated cyclic tests were transferred into the glove box and dissembled for further examination. Their sulfur electrodes and separators were repeatedly rinsed with DME and vacuum dried at 50 1C for 6 h to remove the residual solvent. The morphology and microstructure of electrodes and separators were characterized by FE-SEM at 5 kV and transmission electron microscopy (TEM, JEOL-2100F) at 20 kV. X-ray photoelectron spectroscopy (XPS) measurements were conducted on a Physical Electronics PHI5802 instrument using an X-rays

M. Liu et al. magnesium anode (monochromatic Ka X-rays at 1253.6 eV) as the source. C 1 s region were used as references and set at 284.8 eV.

Assembly and characterization of flexible polymer Li–S batteries A flexible (soft packing) Li–S battery was assembled in a glove box. The compositions of electrodes and precursor solution were the same as the above-mentioned coin cells. Aluminum and nickel strips were joined anchored to the side of cathode and anode as the electrode tabs, respectively. The electrodes and separator were laminated together to form the battery core and assembled into aluminum-plastic film packages, followed by injecting the precursor solution into the packages and sealing batteries under vacuum. Subsequently, the assembled cells were aged at room temperature for 6 h to ensure the precursor solution wellwetted into the electrodes, and then subjected to a formation process (galvanostatically charge/discharge at 0.05 C for 2 cycles) at 70 1C under 0.25 MPa. Finally, batteries were aged at 25 1C for 12 h and degassed. This assembly route is illustrated in Figure S19. The assembled polymer Li–S batteries were cycled at between 1.7–2.8 V under a flat, bent and clustered conditions, respectively.

Results and discussion Characterization of PETEA-based GPE The radical polymerization of PETEA thermally initiated by azobisisobutyronitrile (AIBN) is illustrated in Figure 1a. The polymerization reaction occurs in LE. The primary radicals derived via the thermal decomposition of AIBN attack the C = C double bond of the PETEA monomer to create four free radicals on the monomer as PETEA possesses four C = C double bonds to be initiated, followed by the chain growth reaction by sequentially adding PETEA monomers to the active sites (i.e., four free radical ends) of initiated monomer. Finally, a three-dimensional network-like polymerized PETEA is formed in LE due to the presence of four active sites for the growth of polymer chains, and a translucent GPE is thus obtained (inset of Figure 1b) [18]. A representative field emission scanning electron microscopy (FESEM) image is shown in Figure 1b, displaying that the surface of the PETEA-based GPE has a porous structure, in which the polymerized PETEA acts as the scaffold, with which the LE is filled. The polymerized PETEA provides a mechanical support while the LE offers the pathway for ion transport. To verify the polymerization mechanism described above, LE was removed from the PETEA-based GPE, yielding the polymer matrix. Figure 1c compares the Fourier transform infrared spectroscopy (FTIR) spectra of the PETEA monomer and the polymer matrix of PETEA-based GPE. It is clear that the appearance of peaks at 1156 cm  1 (C–O, symmetrical stretching), 1260 cm  1 (C–O antisymmetric stretching), 1453 cm  1 and 1407 cm  1 (CH2 bending), and 1738 cm  1 (C = O stretching) are in good accordance with previously reported results [19]. After polymerization, the absorption peak at approximately 1633 cm  1 assigned to the stretching

Novel gel polymer electrolyte

281

o

t/ C

δ CH

2

υ C=O

-1.7

After polymerization δ CH2

υ C-O

υ C-O

Before polymerization

100 80 60

40

20

0

-20

-1.8 log σ / S cm-1

Transmittance / a.u.

υ C=C

-1.9 -2.0 -2.1 -2.2

LE GPE * Solid lines represent VTF fitting results.

1800

1600 1400 -1 Wavenumber / cm

1200

-2.3 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 1000 T-1 / k-1

Figure 1 Synthesis and characterization of the PETEA-based GPE: polymerization mechanism of the PETEA monomers (a); the FESEM image of the PETEA-based GPE (the optical images of PETEA-based GPE and its precursor solution are shown in inset) (b); FTIR spectra of the PETEA monomer and the polymer matrix of the PETEA-based GPE (c); ionic conductivities for the PETEA-based GPE and blank LE (1 M LiTFSI/DOL: DME (1:1 by volume) with 1 wt% LiNO3) at various temperatures (d). The symbols are the experimental data and the solid lines represent the VTF fitting.

vibration of C = C bonds almost disappears, indicating a high conversion of PETEA monomers [20]. These results confirm that the PETEA monomers have been successfully polymerized in LE, leading to the formation of the mechanically robust, self-standing GPE. Ionic conductivity is a key property for GPEs to be used in energy storage devices [21]. The ionic conductivity of GPE is known to be influenced by the polymer matrix as well as the incorporated LE [22]. Unfortunately, GPEs usually suffer from dramatically reduced conductivities compared with the blank LE, due mainly to the poor ionic conductivity of polymers in GPEs and the immobilization of LE in the polymer matrix [23]. The ionic conductivities for the PETEA-based GPE film and the blank LE (1 M LiTFSI/DOL: DME (1:1 by volume) with 1 wt% LiNO3) as a function of temperature from  20 to 90 1C are presented in Figure 1d. Remarkably, the conductivity of the PETEA-based GPE reaches an extremely high value of 1.13  10  2 S cm  1 at 25 1C. To the best of our knowledge, this is the highest value for GPEs and quite close to the conductivity of the blank LE (1.19  10  2 S cm  1 at 25 1C). The high ionic conductivity of PETEA-based GPE can be attributed to the unique molecular structure of PETEA monomer. Firstly, the structural similarity between the ether-bond-rich polymerized PETEA and ether-based LE ensures a good compatibility

between polymer matrix and LE. Secondly, PETEA monomer has a symmetrical star structure with four C =C bonds in each molecule, which provides the polymerized PETEA with higher crosslink density compared with other monomers. Such high-crosslinked polymerized PETEA matrix negligibly inhibits the ion mobility in GPE. Furthermore, it can be seen that for either GPE or LE, the plots of log s versus T  1 exhibit a non-linear relationship, which can be well described by Vogel–Tamman–Fulcher (VTF) empirical equation below [22]:   Ea s ¼ so T  1=2 exp  ð1Þ RðT T o Þ where Ea is the activation energy, so is the pre-exponential factor, To is a parameter correlated to the glass transition temperature, and R is the ideal gas constant. It is worth noting that the value of Ea for PETEA-based GPE obtained by fitting using the VTF equation (2.94  10  2 eV) is quite close to that for the blank LE (2.60  10  2 eV). Ea is considered to be the barrier for ionic conduction [24]. The similar Ea between GPE and blank LE signifies that, in contrast to copious past work, the inhibition effect on the ion mobility is negligible for the polymerized PETEA matrix, which may be attributed to the structural similarity between the etherbond-rich polymerized PETEA and the ether-based LE.

282

M. Liu et al.

3.3

LE

0.1 C 0.5 C

3.3

0.3 C

2.7 2.4 ~0.434 V

2.1 1.8 1.5

0

200

400

600

800

~ 0.285V

2.1

1000

1.5

1200

0

200

400

600

800

1000

1200

Specific capacity / mAhg -1 1000

1200

LE charge GPE charge

1000 800 600

0.1 C charge and discharge

200

800

10

20

30

40

50

60

70

80

400 0.5 C charge and discharge 200

90 100

1600 1200

0

50

100

0.1 C 0.2 C

1000

0.1C

0.3 C

800

0.5 C 1C

600 400 200

150 200 250 Cycling number

300

350

400

1600 -1

LE charge GPE charge

Specific capaicity / mAhg

-1

LE discharge GPE discharge

1400

LE charge GPE charge

600

0 0

LE discharge GPE discharge

-1

LE discharge GPE discharge

1400

Specific capaicity / mAhg

-1

Specific capaicity / mAhg

2.4

Cycling number

Specific capaicity / mAhg

0.3 C

2.7

-1

1600

0

0.2 C 1C

1.8

Specific capacity / mAhg

400

0.1 C 0.5 C

GPE

3.0

Voltage / V

Voltage / V

3.0

0.2 C 1C

1400

LE discharge GPE discharge

0.1 C

1200

LE charge GPE charge 0.1C

0.5 C

1000

1C 3C

800

5C

600 400 200 0

0 0

10

20

30

40

50

60

Cycling number

0

10

20

30

40

50

60

Cycling number

Figure 2 Electrochemical performances of Li–S batteries by capitalizing on PETEA-based GPE as electrolyte: typical charge/ discharge voltage curves of S/LE/Li cell (a) and S/GPE/Li cell (b) at various C-rates; cycling performances of S/LE/Li cell and S/GPE/ Li cell at 0.1 C (c); cycling performances of S/LE/Li cell and S/GPE/Li cell at 0.5 C (d); rate performances of the S/LE/Li cell and S/GPE/Li cell from 0.1 to 1 C (the current density of 1 C is 1675 mA g  1) (e); rate performances of the [email protected]/LE/Li cell and [email protected]/GPE/Li cell from 0.1 to 5 C (f). Specific capacity values were calculated based on the mass of sulfur.

Lithium ion transference number (tLi+ ) is significant for electrolytes, since a low tLi+ releases abundant mobile anions which could increase the electrode polarization [25]. The tLi+ of PETEA-based GPE obtained using a Li/GPE/Li symmetrical cell is presented in Figure S5. It can be seen that the PETEA-based GPE exhibits a significant high tLi+ of 0.47. By contrast, the tLi+ of LE is only 0.35. The obvious improvement of tLi+ for PETEA-based GPE may be attributed to the fact that the polymerized PETEA framework can greatly limit the movement of anions [21]. This high tLi+ of PETEA-based GPE is expected to reduce the polarization and raise the rate performance of Li–S batteries. Notably, electrochemical and thermal stabilities are also important parameters for the application of GPEs. The electrochemical stability of the PETEA-based GPE was examined by linear sweep voltammetry (LSV). As shown in Figure S6, no peak or noticeable oxidation current is observed in the voltammogram of the PETEA-based GPE up to 4.73 V vs. Li/Li + , implying that the GPE is stable up to 4.73 V (more details please see Figure S7), higher than that of LE (4.59 V). This enhanced electrochemical stability may

be ascribed to the strong interactions between the polymer chain possessing high electrochemical oxidation resistance and the LE, [26] and will be sufficient to meet the requirements for Li-ion and Li–S batteries. Furthermore, the 300 h galvanostatic cycling curves indicate an excellent compatibility between PETEA-based GPE with Li electrode, which assures a long operation of a Li metal-based battery without undergoing safety hazards caused by the Li dendrite growth (Figure S8). The thermal stability of the PETEAbased GPE polymerized in a commercial separator and LE absorbed in a commercial separator was examined by thermogravimetric analysis (TGA) (Figure S9). The LE rapidly evaporates even at room temperature, leading to a significant weight loss due to the low boiling temperature of the ether-based solvent. In contrast, the PETEA-based GPE has a negligible volatility below 45 1C (more details please see Figure S10) and presents a more moderate weight loss than LE at higher temperature. Such outstanding electrochemical and thermal stabilities of the PETEA-based GPE are of key importance for the safety and electrochemical improvement of Li batteries. Li4Ti5O12/GPE/Li and LiFePO4/GPE/Li cells were then assembled to evaluate the

Novel gel polymer electrolyte rate performance (from 0.1 to 10 C) of the PETEA-based GPE in Li-ion batteries. As shown in Figure S11a and b, the Li4Ti5O12/GPE/Li and LiFePO4/GPE/Li cells exhibit nearly the same rate performances to the cells assembled using LE. Even at a high rate of 10 C, the Li4Ti5O12 and LiFePO4 electrodes still deliver reversible capacities of 125.4 mA h g  1 and 96.3 mA h g  1, respectively, which are at the same level as cells using the LE (127.8 mA h g  1 and 93.5 mA h g  1). We note that, to date, this is the best high ratecapability for GPE-based lithium-ion batteries, demonstrating that the extremely high ionic conductivity of the PETEAbased GPE successfully overcomes the low ionic conductivity issue that is widely existed in traditional GPEs [27]. Clearly, this novel GPE carries promising potential for use in highperformance power batteries.

Electrochemical performance of Li–S battery using PETEA-based GPE S/GPE/Li cells containing lithium metal anodes, bare sulfur cathodes, and PETEA-based GPEs were fabricated via in-situ formation of GPEs on the sulfur electrodes. The sulfur particles used have a hexagonal structure with a particle size of 30 nm (Figure S12). Figure 2e compares the rate performance of the S/LE/Li cell and S/GPE/Li cell at various rates from 0.1 to 1 C (the current density of 1 C is 1675 mA g  1). The corresponding initial charge/discharge curves are shown in Figure 2a and b. It is clear that the S/GPE/Li cell exhibits nearly the same initial specific capacity (1219.8 mA h g  1) as the S/LE/Li cell (1212.2 mA h g  1) at 0.1 C. Interestingly, the S/GPE/Li cell shows a much better high rate performance than the S/LE/Li cell. The specific discharge capacities of the S/GPE/Li at 0.2, 0.3, 0.5 and 1 C are 920.2, 814.4, 687.2 and 601.2 mA h g  1, respectively, while the corresponding capacities of the S/LE/Li are only 606.7, 501.7, 215.5 and 168.4 mA h g  1. When the Crate was switched abruptly from 1 to 0.1 C again (Figure 2e), the original capacity of S/GPE/Li was largely recovered, reflecting that the S/GPE/Li cell is robust and highly stable [28]. Moreover, it is worth noting that the charge/discharge potential gap of the S/GPE/Li cell at 0.2 C is only 0.285 V, which is much smaller than that of the S/LE/Li cell (0.434 V) (Figure 2a and b). The reduced potential gap suggests that the PETEA-based GPE significantly decreases the polarization of Li– S battery. The cycling performances of the S/LE/Li cell and S/GPE/ Li cell are presented in Figure 2c and d. Figure 2c shows that the S/GPE/Li cell can deliver a discharge capacity of 744.1 mA h g  1 after 100 cycles at 0.1 C with a capacity retention of 63.4%. In sharp contrast, the corresponding retention for the S/LE/Li cell is only 31.2%. More surprisingly, the S/GPE/Li cell exhibits a dramatically enhanced cycling performance at high current density. It is seen from Figure 2d that the S/GPE/Li cell achieves a discharge capacity of 529.7 mA h g  1 (a capacity retention of 81.9%) after 400 cycles at 0.5 C. In comparison, the S/LE/Li cell delivers a low discharge capacity of only 70.2 mA h g  1 after 200 cycles. These results indicate that the sulfur active material in the S/GPE/Li cell is largely retained in the cathode. In other words, the shuttle effect owing to polysulfide dissolution has thus been successfully inhibited in the polymer Li–S battery. Such electrochemical

283 performance is a significant breakthrough in current electrolyte systems for Li–S battery as shown in Table S2. Impressively, when a widely employed mesoporous carbon material, CMK-3 [17], was introduced to the sulfur electrode to enhance the electrical conductivity network, the [email protected]/GPE/Li cell can yield a further improved discharge capacity of about 1100 mA h g  1 at 0.1 C, which obtains an energy density of about 2365 W h kg  1 considering an average operational voltage value of 2.15 V (Figure 2f). Besides, the [email protected]/GPE/Li cell can successfully deliver a high reversible capacity of over 600 mA h g  1 even at a high current of 5 C, while the corresponding capacity of the [email protected]/LE/Li cell is only about 152.4 mA h g  1. This outstanding electrochemical performance is of great significance to the development of energy storage system with high energy density.

The immobilization mechanism for polysulfides by PETEA-based GPE Optically transparent cells were assembled to preliminarily illustrate the polysulfides immobilization of PETEA-based GPE (Figure S13). Moreover, Figure 3a and b shows the cyclic voltammograms (CVs) of the S/LE/Li cell and S/GPE/Li cell obtained at a scanning rate of 0.05 mV s  1. For the S/LE/Li cell, the cathodic peaks appear at approximately 2.28 and 2.03 V and the corresponding overlapped anodic peaks locate at approximately 2.35 and 2.45 V, which are the characteristic peaks of Li–S batteries [29]. It is found that the main cathodic peaks for the S/GPE/Li cell shift to slightly higher potentials (2.30 and 2.07 V) while the main anodic peaks shift to lower potentials (2.32 and 2.37 V). In addition, the peak current densities for the S/GPE/Li cell are obviously higher than those of the S/LE/Li cell. The higher peak current densities and lower potential gap between main anodic and cathodic peaks for the S/GPE/Li cell (0.301 V, compared with 0.432 V for the S/LE/Li cell) signifies a significantly less polarization and smaller impedance for the polymer Li–S battery [30]. This observation explains the excellent rate performance of the S/GPE/Li cell. Furthermore, the peak current density of the S/GPE/Li cell exhibits a more stable tendency with the increase of CV cycles than that of the S/LE/Li cell, reflecting that a stable electrode/electrolyte interface is yielded in the S/GPE/Li cell which dramatically improves the cycling stability of the bare sulfur cathode. The CV results are consistent with the electrochemical performances shown in Figure 2. Electrochemical impedance spectroscopy (EIS) is used to evaluate the interfacial resistance and reversibility of Li–S batteries using PETEA-based GPE. The EISs of the S/LE/Li cell and S/GPE/Li cell after different cycles are simulated (Figure 3c and d) using an equivalent circuit shown in Figure S14 and the simulation results are summarized in Table 1. It is seen that the interfacial resistance (Rf) in the S/LE/Li cell increases from 16.19 to 94.32 Ω as the cycle number increases from 2 to 100 cycles, and the corresponding charge transfer resistance (Rct) rises from 45.15 to 187.4 Ω. Figure 4 illustrates the immobilization mechanism for polysulfides by capitalizing on PETEA-based GPE as electrolyte. It is widely accepted that the surfaces of both cathode and anode are passivated by a layer of the solid

284

M. Liu et al.

LE Current density / A g

0.0

1st 2nd 3rd 4th 5th

-0.2

~ 0.432 V

-0.3 -0.4 1.6

1.8

2.0

2.2 2.4 2.6 Voltage / V

2.8

0.1 0.0

-0.2

2.0

2.2 2.4 2.6 Voltage / V

GPE after 2 cycles GPE after 10 cycles GPE after 100 cycles

50

30 20 10 0 0

200

10

20

30

Z' / Ω

40

50

-Z'' / Ω

300

400

40

-Z'' / Ω

LE after 2 cycles LE after 10 cycles LE after 100 cycles

400 -Z'' / Ω

1.8

2.8

3.0

500

500

100 0

~ 0.301 V

-0.3 -0.4 1.6

3.0

1st 2nd 3rd 4th 5th

-0.1

50 40 30

-Z'' / Ω

Current density / A g

0.1

-0.1

GPE

0.2 -1

-1

0.2

20

300

10 0 0

200

10

20

30

Z' / Ω

40

50

100 0

100

200 300 Z' / Ω

400

500

0

0

100

200 300 Z' / Ω

400

500

Figure 3 Electrochemical measurements of Li–S batteries by employing LE and PETEA-based GPE as electrolyte: cyclic voltammograms (CVs) of S/LE/Li cell (a) and S/GPE/Li cell (b) at a scan rate of 0.05 mV s  1; electrochemical impedance spectroscopy (EIS) plots of S/LE/Li cell (c) and S/GPE/Li cell (d) after different cycles at 0.1 C. Impedances are measured at half state of the charge (2.1 V).

reaction products during cycling [31–33]. Since the bare sulfur particles in the cathode of Li–S batteries suffer from a 79% volumetric change during the cathodic reaction from S to Li2S [6], some cracks on the passivation layer would be generated during the repeated huge volumetric change in charge/discharge cycles, leading to the exposure of fresh sulfur surface in LE [34]. The exposed active sulfur surface would accelerate the dissolution of polysulfides and further reacts with the solution molecules of LE to reconstruct the passivation layer (upper panels in Figure 4) [35]. This repeated breakdown-reconstruction process results in a gradually thickened passivation layer on the sulfur electrode upon prolonged cycling (third and fourth upper panels in Figure 4), which ultimately results in continuous increase of Rf and Rct. However, the changes of Rf and Rct for the S/ GPE/Li cell during charge/discharge cycles are apparently lower than those of the S/LE/Li cell. Notably, comparing with that of the S/LE/Li cell, the Rf for the S/GPE/Li cell

Table 1

exhibits an obviously less increase (from 3.76 Ω after 2 cycles to 21.22 Ω after 100 cycles), demonstrating a stable passivation layer is formed on the surface of the sulfur electrode [36]. This may be due primarily to the highstrength PETEA-based gel matrix that pre-covers the cathode surface, inducing the formation of a flexible passivation layer (first lower panel in Figure 4), which makes the sulfur electrode surface maintain the integrity and stability against the volumetric change of sulfur particles during the charge/discharge process (third and fourth lower panels in Figure 4). The flexible passivation layer can also effectively separate the sulfur electrode from organic electrolyte, and thus suppress the continuous interfacial reaction and polysulfide dissolution[37], as schematically illustrated in lower panels in Figure 4. Moreover, the strong interaction between lithium sulfide and the oxygen donor atoms in ester (C = O) groups of polymerized PETEA also benefits the immobilization of polysulfides [38]. The Rct for the S/GPE/

Summary of simulation results from Figure 3c and d.

Battery Sample Cycle number

After 2 cycles After 10 cycles After 100 cycles

LE

GPE

Rb (Ω)

Rf (Ω)

Rct (Ω)

Rb (Ω)

Rf (Ω)

Rct (Ω)

3.55 4.661 7.529

16.19 18.47 94.32

45.15 66.51 187.4

4.67 4.566 4.781

3.76 13.36 21.22

32.08 33.26 46.91

Novel gel polymer electrolyte

285

Figure 4 The immobilization mechanism for polysulfides by capitalizing on PETEA-based GPE as electrolyte. For the Li–S battery using LE, the breakdown-reconstruction of passivation layer and the continuous growth of dendrites on the sulfur electrode surface (caused by the repeated volumetric changes of sulfur particles during the charge/discharge cycles) accelerate the dissolution of polysulfides and increase the interfacial resistance. However, in the case of Li–S battery employing PETEA-based GPE, high-strength PETEA-based gel matrixes are pre-covered on the cathode surface, which strongly enhances the flexibility of passivating layer against the volumetric changes of sulfur particles. The PETEA-based GPE combined with flexible passivating layer can effectively suppress the continuous interfacial reaction and polysulfide dissolution.

Li cell slightly increases from 32.08 Ω after 2 cycles to 46.91 Ω after 100 cycles, further demonstrating that the bare sulfur cathode can effectively retain excellent structural stability under the protection of PETEA-based GPE and flexible passivation layer. These lower Rf and Rct for the S/ GPE/Li cell contributes to the markedly improved performance as shown in Figure 2. In order to further verify the proposed immobilization mechanism for polysulfides by PETEA-based GPE depicted in Figure 4, the surface morphology and composition of the sulfur electrodes (Figure 5) and separators (Figure 6) after charge/discharge cycles were scrutinized. As shown in Figure 5a and b, the sulfur electrodes of the S/LE/Li and S/GPE/Li cells maintain the nearly same surface morphology in the initial state. However, after 2 cycles, the surface morphology of the sulfur electrode in the S/GPE/Li cell is observed to be totally different from that in the S/LE/Li cell. A distinct dendrite structure incompletely covered by the passivation layer produced by the deposition of interfacial reaction products is presented on the sulfur electrode surface of the S/LE/Li cell (Figure 5c) [39,40]. In stark contrast, an integral and smooth passivation layer is clearly seen on the sulfur electrode surface of the S/GPE/Li cell (Figure 5d) [41]. Moreover, massive dendrites grow on the surface of sulfur electrode in S/LE/Li cell after 50 cycles (Figure 5e), due possibly to the uneven current density distribution caused by the cracks on the passivation layer (third upper panel in Figure 4) [42,43]. Consequently, sulfurspecies tend to be deposited on these surface defects, leading to a rapid growth of filaments and dendrites [34,41,44]. The dendrites would strongly increase the interfacial resistance (Figure 3c), and be responsible for the poor cycling performance of S/LE/Li as shown in Figure 2d and e. Conversely, the surface of sulfur electrode

in the S/GPE/Li cell successfully maintains a smooth morphology after 50 cycles (Figure 5f), indicating that the sulfur electrode surface with high interfacial reactivity is completely covered by a stable passivation layer and a robust integrated GPE/electrode interface is formed (third and fourth panels in lower panel in Figure 4). Such flexible passivation layer with low resistance dramatically suppresses the interfacial dissolution reactions (Figure 3d) and thus contributes to the excellent cycling performance of the S/GPE/Li cell shown in Figure 2d and e. The result of X-ray photoelectron spectroscopy (XPS) further indicates that as a main composition of cathodic passivation layer for the S/LE/Li cell [41,45], the content of electrochemically irreversible sulfoxylate/sulfite (-SO2/-SO3) is obviously reduced in the passivation layer for the S/GPE/Li cell, signifying that the reactivity of the sulfur/GPE interface is greatly decreased due to the suppression of the shuttle effect as a result of the implementation of PETEA-based GPE (Figure S15). The surface morphologies and compositions of the separators from the disassembled Li–S batteries after charge/discharge cycles are shown in Figure 6, Figure S16 and Figure S17. The SEM image in Figure 6a clearly shows that the surface of separator in the S/LE/Li cell without charge and discharge possesses a typical porous structure with an average pore size of 300 nm. However, for the S/GPE/Li cell without cycling, the pores of separator are fully filled by the PETEA-based GPE (Figure 6b), which is expected to block the diffusion of polysulfides through the separator and thus inhibit the deposition/growth of polysulfides on the anode. It is interesting to note that in the S/ LE/Li cell, some depositions with ribbon-like structure are obviously adhered on the surface of separator after 2 cycles (Figure 6c), and the corresponding sulfur content

286

M. Liu et al.

Figure 5 Surface morphologies of the sulfur electrodes in the S/LE/Li cell (a, c, e) and the S/GPE/Li cell (b, d, f) after charge/ discharge cycles: in the initial state without charge and discharge (a, b), after 2 cycles (c, d), and after 50 cycles (e, f) at 0.1 C.

determined by energy dispersive spectrometer (EDS) on the separator is 4.1% (Figure S16c and Table S3). These depositions further grow into bulky feather-like structure (Figure 6e and g) with a sulfur content of 14.57% after 50 cycles (Figure S16e and Table S3). The composition of the depositions on separator was further characterized by XPS, which is identified as a mixture of sulfate, sulfite, sulfur and Li2S with an amorphous structure (Figure S17 and Figure 6h). The long-chain polysulfides generated at the cathode with high mobility in LE would migrate to the lithium anode through the pores of separator in the S/LE/Li cell. Subsequently, these polysulfides are reduced to the short-chain soluble polysulfides and insoluble Li2S on the anode and

adhered to the surface of separator [45,46]. In contrast, it is clearly evident that the quantity of deposition on the separator in the S/GPE/Li cell is negligible during cycles (Figure 6d and f), and the sulfur content on the separator is greatly reduced (1.71% after 2 cycles and 4.96% after 50 cycles, see Table S3). In addition, the XPS of lithium anode was examined to investigate its surface composition after charge/discharge cycles. As shown in Figure S18, the S2p region spectra reveal that the peak intensity of Li2Sx species (161.2 and 162.4 eV) and oxidized sulfur species (sulfone/ sulfite, 165.2 and 166.3 eV) of the lithium electrode in S/ LE/Li cell are much stronger than that in S/GPE/Li cell, demonstrating that more side reaction products appear on

Novel gel polymer electrolyte

287

Figure 6 Morphologies of the separators employed in the S/LE/Li cell (a, c, e, g, h) and the S/GPE/Li cell (b, d, f) after charge/ discharge cycles: in the initial state without charge and discharge (a, b), after 2 cycles (c, d), and after 50 cycles (e, f) at 0.1 C. All the images are taken from the side where the separator is in contact with the lithium electrode. TEM image (g) and selected area electron diffraction (SAED) (h) of the depositions shown in Fig. 6e.

288

M. Liu et al. Bent

Specific capaicity / mAhg

-1

Flat

1400 1200 1000 800 600 400 200 0

Clustered

LE discharge GPE discharge

Flat

Bent

Flat

0.1C discharge 0

10

20 30 Cycling number

40

Figure 7 Comparisons of the implementation of LE and PETEA-based GPE for flexible Li–S batteries. (a) Digital images of red LEDs powered by flexible Li–S batteries based on a LE (upper panels; S/LE/Li cell) or PETEA-based GPE (lower panels; S/GPE/Li cell) under various deformed states (i.e., flat, bent and clustered states). (b) Cyclic performances of S/LE/Li and S/GPE/Li cells under flat and bent states at 0.1 C.

the lithium electrode in S/LE/Li cell. These observations suggest that the so-called “shuttle effect” in Li–S battery has been successfully alleviated by judiciously exploiting the PETEA-based GPE that acts as a polysulfide obstacle.

Implementation of PETEA-based GPE for highperformance flexible Li–S batteries Developing high-performance flexible GPEs has been one of the key challenges for flexible batteries [47,48]. In this work, to explore the potential application of PETEA-based GPE in flexible devices, a soft packed Li–S battery was assembled in-situ with bare sulfur as the cathode, PETEAbased GPE as the electrolyte, and lithium strip as the anode (Figure 7). Clearly, the S/LE/Li battery can only normally power a red light-emitting diode (LED) lamp under a static flat state but suffers from a power supply failure under a bent or clustered condition (Figure 7a), due primarily to the internal disconnection caused by the detachment of the electrode/LE interface under the serious shape deformation. However, because of the excellent adhesion between PETEA-based GPE and electrodes, the S/GPE/Li battery can readily light up the LED lamp no matter under flatted, bent, or even clustered states. As seen from Figure 7b, after 45 flat–bent–flat cycles at 0.1 C, the S/LE/Li retains a capacity of 151.6 mA h g  1 with a capacity retention of only 14.62% (Figure 7b). Remarkably, the S/GPE/Li battery can, however, retain a capacity of 803.2 mA h g  1 with a capacity retention of 91.78% after 45 cycles, exhibiting excellent cycling stability under both flat and bent states. This result demonstrates that the electrode/PETA-based GPE interface can sustain tight adhesion and maintain excellent stability under a significant shape deformation.

Conclusions In summary, we crafted an intriguing pentaerythritol tetraacrylate (PETEA)-based gel polymer electrolyte (GPE) with an extremely high ionic conductivity (1.13  10  2 S cm  1) for Li–S battery. Interestingly, compared with liquid electrolyte (LE), even with a bare sulfur cathode, the incorporation of PETEA-based GPE in Li–S battery dramatically enhances the rate capacity and cycle stability. Such exciting performance improvements can be attributed to the strong immobilization of polysulfides by PETEA-based GPE and the formation of stable passivation layer on the sulfur electrode surface induced by the polymer matrix of GPE. Notably, the passivation layer with high flexibility imparts the formation of an integrated structure against the volume change during the charge/discharge process, and thus strongly inhibits the dissolution of polysulfides and alleviates the irreversible sulfur electrode/electrolyte interfacial reaction. The massive dendrites continuously growing on the surface of the sulfur electrode and the feather-like deposits forming on the separators are significantly restrained in the S/GPE/Li system, unambiguously signifying that the shuttle effect in Li–S batteries is successfully suppressed by the PETEA-based GPE that effectively acts as a polysulfide obstacle. Furthermore, due to the tight adhesion between PETEA-based GPE and electrodes, a highperformance flexible PETEA-based polymer Li–S battery that works under large mechanical deformation is successfully developed. Clearly, such simple, low-cost and easily controlled in-situ synthetic route to PETEA-based GPE in conjunction with the excellent performance of the assembled polymer Li–S battery render this novel GPE as one of the most promising candidates to replace

Novel gel polymer electrolyte conventional liquid electrolytes and revolutionize the development of Li–S battery.

Acknowledgments This work was supported by National Key Basic Research Program of China (No. 2014CB932400), National Natural Science Foundation of China (No. 51232005), the Key Project for Basic Research for three main areas of Shenzhen (No. JCYJ20120831165730900 and JCYJ20140417115840246), Guangdong Province Innovation R&D Team Plan for Energy and Environmental Materials (No. 2009010025). M. Liu and D. Zhou contributed equally to this work.

Appendix A.

Supplementary material

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/ j.nanoen.2016.02.008.

References [1] B. Dunn, H. Kamath, J.-M. Tarascon, Science 334 (2011) 928–935. [2] D. Larcher, J. Tarascon, Nat. Chem. 7 (2015) 19–29. [3] P.G. Bruce, S.A. Freunberger, L.J. Hardwick, J.-M. Tarascon, Nat. Mater. 11 (2012) 19–29. [4] Y. Yang, G. Zheng, Y. Cui, Chem. Soc. Rev. 42 (2013) 3018–3032. [5] W.J. Chung, J.J. Griebel, E.T. Kim, H. Yoon, A.G. Simmonds, H.J. Ji, P.T. Dirlam, R.S. Glass, J.J. Wie, N.A. Nguyen, Nat. Chem. 5 (2013) 518–524. [6] A. Manthiram, Y. Fu, S.-H. Chung, C. Zu, Y.-S. Su, Chem. Rev. 114 (2014) 11751–11787. [7] C. Huang, J. Xiao, Y. Shao, J. Zheng, W.D. Bennett, D. Lu, L.V. Saraf, M. Engelhard, L. Ji, J. Zhang, Nat. Commun. 5 (2014) 3015. [8] A. Manthiram, Y. Fu, Y.-S. Su, Acc. Chem. Res. 46 (2012) 1125–1134. [9] Y. Wang, W.H. Zhong, ChemElectroChem 2 (2015) 22–36. [10] K. Xu, Chem. Rev. 114 (2014) 11503–11618. [11] J. Hassoun, B. Scrosati, Angew. Chem. Int. Ed. 49 (2010) 2371–2374. [12] Y.M. Lee, N.-S. Choi, J.H. Park, J.-K. Park, J. Power Sources 119 (2003) 964–972. [13] J.-W. Choi, J.-K. Kim, G. Cheruvally, J.-H. Ahn, H.-J. Ahn, K.-W. Kim, Electrochim. Acta 52 (2007) 2075–2082. [14] M. Rao, X. Geng, X. Li, S. Hu, W. Li, J. Power Sources 212 (2012) 179–185. [15] D. Zhou, Y.-B. He, R. Liu, M. Liu, H. Du, B. Li, Q. Cai, Q.-H. Yang, F. Kang, Adv. Energy Mater. 5 (2015) 1500353. [16] D. Zhou, Y.-B. He, Q. Cai, X. Qin, B. Li, H. Du, Q.-H. Yang, F. Kang, J. Mater. Chem. A 2 (2014) 20059–20066. [17] X. Ji, K.T. Lee, L.F. Nazar, Nat. Mater. 8 (2009) 500–506. [18] D. Zhou, L.-Z. Fan, H. Fan, Q. Shi, Electrochim. Acta 89 (2013) 334–338.

289 [19] P. Petrov, M. Bozukov, C.B. Tsvetanov, J. Mater. Chem. 15 (2005) 1481–1486. [20] H.-J. Ha, E.-H. Kil, Y.H. Kwon, J.Y. Kim, C.K. Lee, S.-Y. Lee, Energy Environ. Sci. 5 (2012) 6491–6499. [21] E. Quartarone, P. Mustarelli, Chem. Soc. Rev. 40 (2011) 2525–2540. [22] F. Croce, M.L. Focarete, J. Hassoun, I. Meschini, B. Scrosati, Energy Environ. Sci. 4 (2011) 921–927. [23] H. Fan, H. Li, L.-Z. Fan, Q. Shi, J. Power Sources 249 (2014) 392–396. [24] Y. Seino, T. Ota, K. Takada, A. Hayashi, M. Tatsumisago, Energy Environ. Sci. 7 (2014) 627–631. [25] K. Xu, Chem. Rev. 104 (2004) 4303–4418. [26] G. Appetecchi, F. Croce, B. Scrosati, Electrochim. Acta 40 (1995) 991–997. [27] A. Abouimrane, Y. Abu-Lebdeh, P.-J. Alarco, M. Armand, J. Electrochem. Soc. 151 (2004) A1028–A1031. [28] Z.W. Seh, W. Li, J.J. Cha, G. Zheng, Y. Yang, M.T. McDowell, P.C. Hsu, Y. Cui, Nat. Commun. 4 (2013) 1331. [29] Y.-S. Su, A. Manthiram, Nat. Commun. 3 (2012) 1166. [30] L. Ji, M. Rao, H. Zheng, L. Zhang, Y. Li, W. Duan, J. Guo, E.J. Cairns, Y. Zhang, J. Am. Chem. Soc. 133 (2011) 18522–18525. [31] S. Risse, S. Angioletti-Uberti, J. Dzubiella, M. Ballauff, J. Power Sources 267 (2014) 648–654. [32] D.-R. Chang, S.-H. Lee, S.-W. Kim, H.-T. Kim, J. Power Sources 112 (2002) 452–460. [33] Y.X. Yin, S. Xin, Y.G. Guo, L.J. Wan, Angew. Chem. Int. Ed. 52 (2013) 13186–13200. [34] D. Aurbach, J. Power Sources 89 (2000) 206–218. [35] S.-H. Chung, A. Manthiram, Chem. Commun. 50 (2014) 4184–4187. [36] K.H. Kim, Y.-S. Jun, J.A. Gerbec, K.A. See, G.D. Stucky, H.-T. Jung, Carbon 69 (2014) 543–551. [37] G.C. Li, G.R. Li, S.H. Ye, X.P. Gao, Adv. Energy Mater. 2 (2012) 1238–1245. [38] Z.W. Seh, Q. Zhang, W. Li, G. Zheng, H. Yao, Y. Cui, Chem. Sci. 4 (2013) 3673–3677. [39] J.-T. Yeon, J.-Y. Jang, J.-G. Han, J. Cho, K.T. Lee, N.-S. Choi, J. Electrochem. Soc. 159 (2012) A1308–A1314. [40] J. Song, H. Noh, H. Lee, J.-N. Lee, D.J. Lee, Y. Lee, C.H. Kim, Y.M. Lee, J.-K. Park, H.-T. Kim, J. Mater. Chem. A 3 (2014) 323–330. [41] J. Zheng, M. Gu, H. Chen, P. Meduri, M.H. Engelhard, J.-G. Zhang, J. Liu, J. Xiao, J. Mater. Chem. A 1 (2013) 8464–8470. [42] S.S. Zhang, J. Power Sources 164 (2007) 351–364. [43] L. Suo, Y.-S. Hu, H. Li, M. Armand, L. Chen, Nat. Commun. 4 (2013) 1481. [44] G. Zheng, S.W. Lee, Z. Liang, H.-W. Lee, K. Yan, H. Yao, H. Wang, W. Li, S. Chu, Y. Cui, Nat. Nanotechnol. 9 (2014) 618–623. [45] Q. Wang, J. Jin, X. Wu, G. Ma, J. Yang, Z. Wen, Phys. Chem. Chem. Phys. 16 (2014) 21225–21229. [46] Y.V. Mikhaylik, J.R. Akridge, J. Electrochem. Soc. 151 (2004) A1969–A1976. [47] G. Zhou, F. Li, H.-M. Cheng, Energy Environ. Sci. 7 (2014) 1307–1338. [48] K.H. Choi, S.J. Cho, S.H. Kim, Y.H. Kwon, J.Y. Kim, S.Y. Lee, Adv. Funct. Mater. 24 (2014) 44–52.