Multifunctional polymer electrolyte improving stability of electrode-electrolyte interface in lithium metal battery under high voltage

Multifunctional polymer electrolyte improving stability of electrode-electrolyte interface in lithium metal battery under high voltage

Journal of Membrane Science 588 (2019) 117194 Contents lists available at ScienceDirect Journal of Membrane Science journal homepage: www.elsevier.c...

3MB Sizes 0 Downloads 13 Views

Journal of Membrane Science 588 (2019) 117194

Contents lists available at ScienceDirect

Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci

Multifunctional polymer electrolyte improving stability of electrodeelectrolyte interface in lithium metal battery under high voltage

T

Ming Zhua,1, Jiaxin Wua,1, Bingxue Liub, Wei-Hong Zhongc, Jinle Lana, Xiaoping Yanga, Gang Suia,∗ a

State Key Laboratory of Organic-Inorganic Composites, Beijing University of Chemical Technology, Beijing, 100029, China China Automotive Battery Research Institute Co., Ltd., China c School of Mechanical and Materials Engineering, Washington State University, Pullman, WA, 99164, USA b

A R T I C LE I N FO

A B S T R A C T

Keywords: Gel polymer electrolyte Cathode electrolyte interface Lithium metal High voltage

High-voltage lithium (Li) metal batteries are promising high-performance energy storage systems. However, their practical applications are troubled by unstable electrode-electrolyte interface, the narrow electrochemical window and unsafety of liquid electrolyte. Here, a multifunctional asymmetrical gel polymer electrolyte (a-GPE) with gradient structure is proposed to address the mentioned above issues. The host of a-GPE is consisted of the poly(vinylidene fluoride hexafluoropropylene) (PVDF-HFP) nanofiber membranes containing BaTiO3 (BTO) with gradually varying content along thickness. This design incorporates comprehensive consideration of ionic conductivity, Li-ion transference number and electrochemical stability window, as well as mechanical property, thus achieving smooth and steady Li-ion flux in interior of a-GPE. Significantly, the PVDF-HFP membranes without BTO face the Li-metal to accomplish homogeneous Li deposition/stripping, and the BTO-rich PVDF-HFP membranes are adjacent to cathode to generate an effective protection film for stabilizing interface owing to polarization effect on the migration kinetics of mobile Li ions, especially under high voltage and large current. Consequently, the LiNi0.5Mn0.3Co0.2O2/Li batteries based on a-GPE exhibit superior comprehensive performances, including dramatic cycling performance and excellent rate performance at high voltage, good electrochemical performance under low temperature and high safety, as well as favorable flexibility. These ideal comprehensive performances can be conveniently realized through ingenious design of conventional polymers, rather than expensive synthetic materials and complicated fabrication process. This work will initiate a new development of advanced polymer electrolyte for high-energy and safe Li batteries.

1. Introduction To meet rapidly growing requirements for energy storage technologies applied in the emerging electric vehicles and power supply, the rechargeable lithium batteries (LBs) are incessant in pursuit of highenergy/high-power density [1–3]. It is a promising strategy that a highvoltage cathode and a high-capacity anode are combined with suitable electrolytes to realize high-energy density of LBs. In the anode side, lithium (Li) metal anode is an intensively pursued candidate because metallic Li delivers high theoretical specific capacity of 3860 mA h g−1 and laudable redox potential of −3.04 V versus a standard hydrogen electrode [4–6]. In the high-voltage cathode side, layered lithium transition metal (TM) oxides, in particular, LiNixMnyCozO2 (NMC, for example, LiNi0.5Mn0.3Co0.2O2, often called NMC-532) represent a

family of prominent cathode materials with high operating voltage and large reversible capacity, which can dramatically upgrade energy density of Li metal batteries (LMBs) [7,8]. However, numerous fundamental challenges preclude the practical realization of LBs based on Li metal anode and NMC cathode with liquid electrolytes. For Li metal anode, the uncontrollable Li dendrite growth during the inhomogeneous Li deposition/stripping causes serious capacity fading and sporadic safety issues, hindering the development of LMBs [9,10]. For advanced NMC cathodes, they suffer from the micro-strain and micro-crack formation and unwanted side reactions with the electrolyte at the interface as well as the dissolution of TM ions, resulting in unstable cathode electrolyte interface (CEI) and severe capacity attenuation [11,12]. For liquid electrolytes, in addition to its the safety hazards, new challenges will be faced under high voltage that the



Corresponding author. E-mail address: [email protected] (G. Sui). 1 These authors contribute equally to this work. https://doi.org/10.1016/j.memsci.2019.117194 Received 5 March 2019; Received in revised form 29 April 2019; Accepted 15 June 2019 Available online 18 June 2019 0376-7388/ © 2019 Published by Elsevier B.V.

Journal of Membrane Science 588 (2019) 117194

M. Zhu, et al.

involving a-GPE also showed remarkable cycling performance at low temperature, reliable safety as well as highly flexible characteristic. These recommendable integrated performances suggested the multifunctional a-GPE can be applied to high-efficiency and safe high-voltage LMBs.

decomposition of electrolytes can lead to the loss of active Li and a large number of failure of electrode-electrolyte interface. Polymer electrolytes have been suggested to be the best candidate to substitute liquid electrolyte for providing the safety, flexibility and multi-functionality to energy-storage devices [13]. Moreover, some polymer electrolytes with designed structure are proposed to solve the above mentioned issues for either Li metal anode [14–16] or NMC cathode [17,18]. Instead of the traditional design, polymer electrolytes with asymmetric structures have more potential in practical applications of full batteries as they may meet the specific and different requirements of both the cathode and anode simultaneously. Duanu and co-workers [19] prepared a thin asymmetrical solid polymer electrolyte (SPE), in which a dense Li7La3Zr2O12-modified polymer electrolyte layer faced Li-metal to provide dendrite-suppression of Li anode, and a soft polymer electrolyte layer faced cathode to endow good interfacial connections. The prepared LMBs with the asymmetrical SPE exhibited high Coulombic efficiency. Unfortunately, the difference of ionic conductivity from the two electrolyte layers might trigger inhomogeneous Li ion flux in the entire electrolyte, which decayed long-time cycling stability of the batteries. To further improve the electrochemical performance of batteries, an electrolyte with asymmetric structure is needed to assure the smooth and steady transition of Li ions in its interior. Additionally, in the ideal asymmetrical electrolyte, a more superior Li ion conductive layer needs to be adjacent to the cathode than to the anode [20]. Compared with SPEs, gel polymer electrolytes (GPEs) demonstrate the comprehensive performance advantages, including high ionic conductivity, wide electrochemical window, satisfactory thermal stability and low leakage of the electrolyte solution [21–23]. But it is a pity, the reported GPEs were rarely designed with asymmetrical structure and hardly used to synchronously address the issues of Li dendrites and unstable CEI, especially when the LMBs worked under the high cut-off voltages. Furthermore, if incorporating inorganic nanofillers into GPEs, their physical and electrochemical performance may be prominently heightened [24–30]. Interestingly, among the inorganic nanofillers, the high-permittivity nanofillers present tremendous potential in improving properties of cathodes to achieve a breakthrough in energy and power densities of the batteries [31]. For example, Teranishia's group [32] prepared a cathode coated with ferroelectric barium titanate (BaTiO3, BTO) nanoparticles and acquired good high-rate performance of LIBs, owing to a polarization effect on the migration kinetics of mobile Li ions. Therefore, it is expected that an asymmetrical GPE modified by selected nanoparticles can synchronously deal with the issues for both Li anode and NMC cathode under high voltage without sacrificing reasonable Li-ion flux in the whole GPE interior. Here, we reported an asymmetrical multifunctional GPE (a-GPE) based on high dielectric BTO-hybrid poly(vinylidene fluoride hexafluoropropylene) (PVDF-HFP), which was designed and prepared for high-voltage LMBs. The host of a-GPE was constructed from a gradient structure consisting of PVDF-HFP nanofibers with gradually varied content of BTO nanoparticles along thickness through a simple one-step process. The PVDF-HFP nanofibers with high content BTO were adjacent to the cathode to form an effective protection film for stabilizing CEI owing to polarization effect on the Li ions migration kinetics, especially under high voltage and large current, and PVDF-HFP nanofibers without BTO faced the Li metal anode to enable a homogeneous Li deposition/stripping. Besides, the whole a-GPE would possess desirable transition distribution of Li ion in its interior due to the presence of gradient structure. This rational strategy for constructing a-GPE guaranteed a remarkable overall performance of the high-voltage LMBs. Consequently, the resulting a-GPE displayed a high ionic conductivity, an excellent Li ion transference number and a wide electrochemical window. More importantly, the Li metal cells with NMC using such aGPE exhibited a superior capacity retention (70% after 200 cycles) at high-rate of 5C and an outstanding rate capability when the dischargecharge voltage was up to 4.5 V. Furthermore, the NMC/Li cells

2. Experimental 2.1. Materials PVDF-HFP (Aldrich, Mw = 455000) and BTO (Aladdin) were dried under vacuum at 80 °C for 24 h before use. Acetone and N, N-dimethylformamide (DMF) were used without further treatment. 1 M lithium hexafluorophosphate (LiPF6)/ethylene carbonate (EC): dimethyl carbonate (DMC) (1:1, v/v) (Beijing Chemicals Co., China.) was used as the liquid electrolyte solution. 2.2. Preparation of samples 2.2.1. Preparation of polymer membranes Different weight ratio of BTO (0 wt%, 5 wt%, 10 wt%, 15 wt%, 20 wt%) powders were dispersed in the acetone and DMF mixed solution (1:1 wt%) with ultrasonication for 1 h. Then PVDF-HFP was added (17 wt%) into the mixture with magnetic stirring at the temperature of 60 °C for 8 h to prepare a homogeneous solution. The prepared spinning solution was poured into a 10 ml plastic syringe with a needle that was connected to a positive charge of 18 kV voltages. The distance between the collection place and the tip of the needle was 15 cm, and a constant flow rate of solution was set to 1.5 mL h−1. The electrospun nanofibers were collected on aluminum foil. According to the mentioned above method, the asymmetric gradient polymer composite membranes were prepared by using sequentially spinning solution involving 0 wt%, 5 wt % and 10 wt% gradient concentration of BTO. The membranes were further dried under vacuum at 100 °C for 24 h to remove the residual solvent prior to measurement and use. 2.2.2. Preparation of GPE, including neat GPEs (n-GPEs) and a-GPEs The nanofiber membranes were punched into several small discs with a diameter of 20 mm, and then transferred into an argon filled glove box (moisture level < 10 ppm). The membrane precursors were immersed in a liquid electrolyte solution of 1 M LiPF6/EC: DMC for about 24 h at room temperature, and the saturated GPEs were used after the removal of the excess liquid electrolyte by filter paper for the measurement of electrochemical properties. 2.3. Characterizations The surface morphology of the electrospinning composite nanofiber membranes were investigated using scanning electron microscopy (SEM, Supra55, Carl Zeiss) equipped with an energy dispersive spectroscopy maps (EDS). Structure and crystallization properties of the samples were investigated using an X-ray diffractometer (XRD, 2500VB2+PC, Rigaku Corporation, Japan). Differential scanning calorimetry (DSC) measurements were carried out under air atmosphere on a TA instruments equipment (Q5000IR) with a heating rate of 10 °C min−1 from 80 °C to 180 °C. The crystallinity degree of all membranes was calculated from the enthalpy of fusion from DSC measurements using the following Equation (1):

Xc (%) = ΔHm/(1 − x )ΔH0 × 100%

(1)

where Xc is the crystalline degree of all synthesized membranes, and ΔHm is the enthalpy of fusion of all membranes (J g−1), and x is the weight fraction of BTO in the nanocomposite membranes, and ΔH0 is standard enthalpy of fusion of 100% crystalline PVDF [104.7 J g−1] [33]. The thermogravimetric analysis (TGA) (Q50, USA) was performed under a nitrogen flow with a heating rate of 10 °C min−1 from 30 to 2

Journal of Membrane Science 588 (2019) 117194

M. Zhu, et al.

The microstructure of BTO and PVDF-HFP nanofiber membranes with 0 wt%, 5 wt%, 10 wt%, 15 wt% and 20 wt% BTO was characterized by using SEM, as shown in Figs. S1a–f. The BTO appeared to be uniform nanosphere with an average diameter of ≈100 nm, and the nanofibers interweaved with each other casually and created a 3D network structure, which was favored to enhance adsorption capacity of liquid electrolyte. Besides, compared to the pure PVDF-HFP samples, the diameter of nanofiber increased after adding a low content of BTO, and then the nanofibers became nonuniform with rising of BTO content, especially the mass ratio of BTO to PVDF-HFP reached to 20%. The magnified SEM images of PVDF-HFP membranes without BTO and samples with 10 wt% BTO are displayed in Fig. 1c and d, respectively, representing the top and bottom surface of composite membrane. In order to further investigate the structure of composite membrane, the SEM images and energy dispersive spectroscopy maps (EDS) of crosssection of composite membrane are shown in Fig. 1e–g. The signal of Ti and Ba element from BTO displayed gradient change, suggesting the composite membranes were successfully fabricated. As can be seen from Fig. S2a and Table S1, the degree of crystallinity of the PVDF-HFP decreased when the content of BTO increased, and the melting temperature of the corresponding samples slightly declined (see from Fig. S2b). It had been reported that the amorphous structure was favor to Li ion transmission and induced a high ionic conductivity [35,36]. This result suggested the distribution of Li ion was gradually changed in aGPE, which can facilitate uniform Li-ion flux in its interior. Moreover, the excellent thermal stability of membranes as separators or GPEs plays a significant role to ensure their safe application in LBs. The severe thermal shrinkage can cause internal short circuits of the batteries eventually resulting in explosion and fire hazard. The photographs of all the samples before and after exposure under 120 °C for 30 min are shown in Fig. S2c. The PVDF-HFP membranes and composite membranes with various BTO content maintained their dimensions under the high temperature, implying remarkable thermal stability of the host of a-GPE (the TGA curves of PVDF-HFP nanofiber membranes with different BTO content are shown in Fig. S3 in the Supporting Information). The various electrochemical measurements are tested to evaluate the feasibility of the a-GPE. Ionic conductivity as a crucial parameter in Li batteries was first measured. Impedance spectra of SS/GPE/SS cells obtained is shown in Fig. S4a, and the ionic conductivity of the GPEs can be obtained from bulk resistances, as given in Fig. 2a. The neat GPE (n-GPE) based on pure PVDF-HFP membranes showed an ionic conductivity of 4.1 × 10−3 S cm−1, and the ionic conductivity reached a maximum value of 6.5 × 10−3 S cm−1 at room temperature when the content of BTO was up to 10 wt%. This result was agreed with the high saturated uptake of liquid electrolyte, which was owing to high porosity (seen in Table S2) coming from the abundantly interconnected porous structures in the membranes, as shown in Fig. S4b. Through the design and integration, it can be found that the ionic conductivity of the a-GPE was as high as 5.2 × 10−3 S cm−1, which was higher than that of previous reported similar electrolyte membranes (shown in Table S3). It was needed to notice that the ion conductivity was gradually increased from anode side to cathode side in a-GPE, which accomplished smooth and steady Li ion transition. Additionally, the activation energy (Ea) for ionic conduction in all the electrolyte membranes can be calculated and shown in Fig. S4c according to the Arrhenius equation, δ = Aexp(-Ea/ RT), where R, T, A and δ are gas constant, temperature, pre-exponential factor and ionic conductivity, respectively. The ionic conductivity of all the GPEs increased in general with increase of temperature, and the GPE based on 10% BTO exhibited the lowest Ea, resulting in its high ionic conductivity. The Li-ion transference number (t+) of all the samples was obtained by using the method of chronoamperometry in the Li/GPE/Li cell, as shown in Fig. S4d. The t+ was simply calculated by the ratio of the initial current and steady current. The n-GPE, a-GPE and other GPEs possessed high t+ of above 0.7, which resulted from the remarkable

800 °C. The mechanical property of nanofiber membranes (40 mm × 10 mm × 0.2 mm in size) were investigated using a tensile testing machine (Instron 1121, UK) with a stretching speed of 2 mm min−1 at room temperature. To analyze the surface chemistry and elemental composition of LiNixMnyCozO2(NMC) cathode, X-ray photoelectron spectroscopy (XPS) measurements were carried out on an ESCALAB 250Xi spectrometer (Thermo Fisher Scientifc) using a monochromatic Al Kα X-ray (1486.6 eV) source. The porosity of different samples was measured by using n-butanol uptake method. The porosity can be calculated according to Equation (2):

Porosity(%) = (M − M0)/ρV0 × 100%

(2)

where M0 and M are the weights of the membranes before and after absorption of n-butanol, respectively, ρ is the density of n-butanol, and V0 represents the volume of the polymer membrane [34]. The saturated electrolyte uptake of stacked membranes was measured and calculated after the immersion process by using Equation (3)

A(%) = (W1 − W0)/ W0 × 100%

(3)

where and W1 and W0 were the mass of the saturated and the dry membrane, respectively. All the electrochemical properties of the samples were performed on Autolab PGSTAT 302 N (Metrohm) at room temperature. The ionic conductivity was calculated by AC impedance spectroscopy in the range from 0.1 Hz to 100 kHz using the sandwich cell of GPE into two parallel SS discs. The ionic conductivity could be calculated by Equation (4)

δ= L/(Rb⋅S)

(4)

where δ was the ionic conductivity, Rb was the bulk resistance obtained from AC impedance plot, L and S was the thickness and area of the electrolyte membrane, respectively. The electrochemical stability window was measured by the system of linear sweep voltammetry (LSV) in the cell of Li/GPE/SS at a scan rate of 10 mV s−1 in the potential voltage ranging of 2.0–8.0 V at 25 °C. The lithium-ion transference number (t+) was evaluated at 25 °C with the method of direct current polarization/alternating current impedance in a symmetrical Li/GPE/Li cell. The interface stabilities of the electrolyte membranes against lithium electrode were measured by a galvanostatic cycling of a symmetric cell with charging for 2 h and discharging for 2 h at a current density of 0.5 mA cm−2. The NMC electrode were prepared by mixing NMC-532, carbon black and poly(vinylidene difluoride) in the weight ratio of 8:1:1. The active material loading of the NMC-532 cathode was controlled to be ∼5.72 mg cm−2 and the electrode plate used in the cell was a wafer with diameter of 14.00 mm. The electrochemical performances of assembled CR2025-type coin cells were tested at various current densities in the voltage range of 2.5 V–4.3 V/4.5 V on a Land cell test instrument at different temperature. 3. Results and discussion In this work, a multifunctional a-GPE with gradient structure was proposed to realize stable CEI in the cathode and effective Li dendrites suppression in the anode for high-voltage LMBs under guaranteeing smooth and steady Li-ion flux in interior of the whole a-GPE. Fig. 1a shows the schematic diagram for the construction process of the host materials in a-GPE. A gradient composite membrane skeleton was composed of three layers of PVDF-HFP nanofiber membranes with different content of BTO. The top layer was pure PVDF-HFP membrane, the interlayer was PVDF-HFP membrane with 5 wt% BTO and the bottom layer was PVDF-HFP membrane with 10 wt% BTO. After gelation with liquid electrolyte, the top layer in a-GPE was close to Li metal anode and the bottom layer was near to NMC-532 cathode. The optical photographs of the composite membrane are shown in Fig. 1b. It can be observed that the composite membrane was folding and flexible, and the bottom surface of composite membrane was much grayer than the top surface, verifying the designed structure from Fig. 1a. 3

Journal of Membrane Science 588 (2019) 117194

M. Zhu, et al.

Fig. 1. (a) The schematic diagram for the fabrication process of asymmetric gradient gel polymer electrolyte (a-GPE); (b) The digital image of composite membrane skeleton of a-GPE, PVDF-HFP membrane without BTO (left) and sample with 10 wt% BTO (right); The SEM images of PVDF-HFP nanofiber membranes without BTO (c) and sample with 10 wt% BTO (d); (e) Cross section SEM images of the skeleton of a-GPE and its EDS maps of Ti element (f) and Ba element (g).

polymer chain. In the meantime, the tensile strength of PVDF-HFP membranes would be reduced when the BTO content was higher than 10 wt%, which originated from the inhomogeneous dispersion of excess BTO in nanofiber membranes breaking the continuity of the fibers. It should be noticed that the PVDF-HFP nanofiber membranes with 5 wt% and 10 wt% BTO were chosen to construct a-GPE under comprehensively considering mentioned above ionic conductivity, Li-ion transference number and electrochemical stability window, as well as mechanical properties. The resulting a-GPE possessed smooth and steady transition distribution of Li ion in its interior, and the Li ion became more and more richer from anode to cathode. To evaluate mechanical performance of the GPEs, 0 wt% BTO and composite membrane were infiltrated with liquid electrolyte and then the redundant liquid electrolyte was removed. Interestingly, the tensile strength of a-GPE membrane was reinforced to 7.4 MPa after completely absorbing the organic electrolyte, as shown in Fig. 2b. The improvement of mechanical properties was aroused by further swelling of the composite membranes and forming gel state, which endowed the a-GPEs with a more uniform and compacted structure, and thus a high mechanical characterization can be obtained [37,38]. The mechanical strength of aGPE was higher than some reported solid polymer electrolyte to be used in LBs [39–41].

polarity of the PVDF-HFP matrix and high dielectric constant of BTO fillers, hindering the movement of large PF6− anions to assist the movement of Li ions and providing a continuous Li ion transport channel. The high Li-ion transference number is beneficial for enriching the rate performance and reducing the overpotential or polarization of cells. The wide electrochemical stability window is significant to GPEs for the application in high-voltage LMBs. A linear sweep voltammogram curve is displayed in Fig. S4e, and the corresponding electrochemical window value was obtained, as shown in Fig. 2a. The n-GPE presented a secure electrochemical working window of 5.3 V versus Li/Li+, and the electrochemical stability of GPE slightly decreased as the BTO content increased. Anyhow, the a-GPE would not decomposed under 5.1 V, which can maintain good electrochemical stability during the chargedischarge process and ensure the safety of the cell at a high voltage. Good mechanical properties of GPE will promote its practical application in LBs. In this study, the representative stress-strain curves of various nanofiber membranes are shown in Fig. S5. It can be observed that the incorporation of low content BTO into PVDF-HFP membrane can be conducive to improve the tensile strength of nanofiber membrane because inorganic nanoparticles can act as the cross-link point of the 4

Journal of Membrane Science 588 (2019) 117194

M. Zhu, et al.

Fig. 2. (a) The ionic conductivity at room temperature (black line) and electrochemical stability window (blue bar) of GPEs with different BTO contents and a-GPE; (b) Representative stress-strain curves of nanofiber membranes and GPEs; (c) Galvanostatic cycling curves of Li/Li cell based on GPEs with different BTO contents. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

5C rate, as illustrated in Fig. 3a. The NMC/a-GPE/Li cells delivered high reversible discharge capacity (147 mA h g−1) at 5.0C rate and high capacity retention (70%) after 200 cycles. In contrast, the NMC/Li cell with the n-GPE displayed only a capacity retention ratio of 62%. It worthy to be noticed that the average active material mass loading of the NMC-532 was extremely higher than that of other cathodes reported in the literature involving GPEs, as shown in Table S5. Furthermore, Fig. 3b shows the rate performance of the 4.5 V-class NMC/Li cells with a-GPE and n-GPE after cycling 200 at 5C. The NMC/a-GPE/Li cell maintained a higher discharge capacity upon increasing the current rate from 1.0 to 10C than the NMC/n-GPE/Li cell. A satisfactory discharge capacity of 41 mA h g−1 was realized even at a rate of 10C from the cell involving a-GPE, while the corresponding capacity of the cell involving n-GPE was only 19.8 mA h g−1. In addition, from Fig. 3c, the charging/discharging profiles were observed and the voltage polarization of the NMC/a-GPE/Li cell was lower than that of the NMC/n-GPE/ Li cell. Meanwhile, the charging/discharging voltage platform of the cell with n-GPE became more unclear with increasing current rate than that of the cell involving a-GPE. The remarkable cycling performance and superior rate capability of a-GPE was probably ascribed to the outstanding interface compatibility between the asymmetrical polymer electrolyte and the electrode, especially the cathode, as well as the

To investigate the dynamical stability of the Li/electrolyte interface and ensure the long cycle stability of Li battery, a galvanostatic Li deposition/stripping electrochemical cycling was tested. As shown in Fig. 2c, the symmetric Li/n-GPE/Li cell exhibited steady lithium plating/stripping at a current density of 0.5 mA cm−2 when the cell was cycled for over 300 h. This result revealed GPE with 0 wt% BTO could generate a favorable solid electrolyte interface (SEI) film to achieve a stable Li deposition-stripping process because the electronegative C–F functions groups provided strong affinity to Li, Li ions and possibly generated favorable component of LiF in SEI via interfacial reaction. Nevertheless, the voltage polarization of the Li/Li cell with GPE showed slight fluctuation as the BTO content was higher than 5 wt%, and the symmetric Li/Li cell involving GPE containing 15 wt% and 20 wt% BTO exhibited scattered high overpotential in the voltage hysteresis because of the reaction between Ti4+ and Li [42,43]. The reaction can deteriorate electrode-electrolyte interface, and cause cracking of SEI, then eventually induce the depletion of electrolyte and production of excessive SEI (see Figs. S6a–f, Supporting Information). Therefore, the construction of a-GPE involving moderate content of BTO was reasonable and effective. The electrochemical performance of the a-GPEs at high voltage was further evaluated by using a NMC-532/Li cell tested from 2.5 to 4.5 V at 5

Journal of Membrane Science 588 (2019) 117194

M. Zhu, et al.

Fig. 3. (a) Cycle performance of NMC/n-GPE/Li and NMC/a-GPE/Li cells with 4.5 V cut-off at 5C at 25 °C; (b) Rate capability of 4.5 V-class NMC/n-GPE/Li and NMC/ a-GPE/Li cells after cycling 200 at 5C; (c) Charge and discharge curves of cells at varied current rates at 25 °C; (d) The schematic illustration of the mechanism of action of a-GPE in the 4.5 V-class NMC/Li cell.

homogeneous Li-ion flux in interior of a-GPE. Here, the multi-functionality mechanism of a-GPE in a cell with high-voltage cathode and Li anode was proposed and its schematic illustration is shown in Fig. 3d. For anode, a homogeneous Li deposition/stripping process and stable SEI were realized because of the presence of electronegative C–F functions groups from PVDF-HFP polymer [44]. More significantly, the

a-GPE could generate a more favorable CEI protection layer to suppress deterioration of the NMC structure and unwanted side reactions on the surface of NMC, whereas the n-GPE only produced an unstable, thin and multi-defect CEI. This result was because the Li deficient layer was abolished at the interface between the polymer electrolyte and the cathode, and the rich BTO layer in a-GPE could induce rearrangement 6

Journal of Membrane Science 588 (2019) 117194

M. Zhu, et al.

Fig. 4. (a) XPS spectra of Ni 2p, F 1s, O 1s and C 1s of the pristine NMC cathode and NMC cathode disassembled from the NMC/Li battery using n-GPE and c-GPE after rate performance test; Typical SEM images of the surface of (b) the pristine NMC electrode and (c) the cycled NMC electrode obtained from the cell with n-GPE and (d) with a-GPE; The surface SEM images of (e) pristine Li electrode and (f) the cycled Li electrode obtained from the NMC/n-GPE/Li cell and (g) from the NMC/aGPE/Li cell after rate performance test. The dotted circles represent byproducts.

of the Li ion distribution to balance polarization and form “Li ion pathways” [31]. To verify the above-mentioned stable CEI and obtain more insightful information about the effect of a-GPE on the cathode, XPS was further conducted to investigate the components of the CEI layer generated on a cycled high-voltage NMC cathode, as shown in Fig. 4a. In Ni 2p spectra, the peaks of cycled NCM cathode with a-GPE were in good agreement with the pristine NMC cathode, and the prominent peaks assigned to Ni2+ (854.7/873.4 eV), Ni3+ (856.5 eV) and NiF2 (859/ 877.8 eV) can be observed [45]. Nevertheless, the Ni spectra of cycled NMC based on n-GPE presented different peaks because of metal dissolution leading to cathode surface degradation and then coverage by electrolyte decomposition products. In the F 1s XPS, the broad and strong peaks, corresponding to LiF/MFx signal (685.4 eV) and LixPOyFz signal (686.4 eV) were found in the cycled cathode with n-GPE, indicating the much severe corrosion by the trace amount of hydrofluoric acid (HF) in LiPF6 electrolyte [46]. The presence of highly resistive LiF on the cathode surface can produce sluggish Li ion transport kinetics and consequent capacity fading [47]. In contrast, the few electrolyte degradation products were generated in the cycled NMC involving aGPE, and the peak of C–F from PVDF binder still appeared, suggesting the cathode was well protected by a-GPE. The O 1s spectra of the cycled

NMC based on n-GPE were similar to that of the cycled NMC-containing a-GPE, except that the peak width of the latter was smaller than that of the former, confirming the capability of a-GPE on inhibiting electrolyte decomposition and formation of stable CEI in the cell with a-GPE. The C 1s spectra of NMC after cycling mainly reflected CF2 from PVDF binder and C]O, C–O and C–C ascribed to the decomposition products of electrolyte. Moreover, the peak values of 290.5 eV belonging to Li2CO3 appeared in the cycled cathode with a-GPE [46]. It was known that the appropriate content of Li2CO3 as an electronic-insulating but ionicconductive compound in CEI layer could obviously enhance the interface stability due to boosting the kinetics of Li ion transportation [48]. Therefore, a-GPE was beneficial for maintaining the integrity of the cathode structure and generating an effective CEI, owing to preventing electrolyte attacks and decomposition, and then forming satisfied component of CEI, which enabled sustainable operation of LMBs using NMC at a high charge cut-off voltage. The surface morphology of NMC cathodes and Li anodes disassembled from NMC/GPE/Li after rate cycling is shown in Fig. 4b–g. The cycled NMC cathode from the cell with a-GPE demonstrated a relatively untainted surface similar to the pristine NMC, while the cycled NMC from the cell involving n-GPE displayed irregular structures resulting from pulverization and byproducts (can also see from Fig. S7). 7

Journal of Membrane Science 588 (2019) 117194

M. Zhu, et al.

commercial LED, as illustrated in Fig. 5b. The red LED could be easily lit, even though the flexible battery was bent into a 180°. Besides, the green and blue LEDs were also successively lit, which proved that the LBs based on a-GPE showed superior safety as well as highly flexible characteristic.

Meanwhile, it can be observed that the surface of the cycled n-GPE adjacent to cathode was polluted by some byproducts from dissolution of transition metal, while corresponding surface of a-GPE was clean, as shown in Figs. S8a–b. This result indicated the a-GPE could generate a protection film to prevent notorious structural degradation of NMC cathode and suppress the perishing side reactions occurring at the electrode-electrolyte interface. In addition, the cycled Li anode in NMC/Li with a-GPE maintained a smooth and compact surface as same as the fresh Li anode. Although the obvious Li dendrites were not found on the surface of the cycled Li anode from the cell based on n-GPE, its surface was covered by little byproducts from deposition of transition metal ion from cathode. Figs. S8c–d shows that the surfaces of both cycled n-GPE and a-GPE closed to anode were unpolluted by dead Li, which revealed the PVDF-HFP was beneficial to uniform Li deposition leading to Li dendrites suppression. Therefore, the cell with a-GPE delivered an effective and stable CEI layer and enabled an enhanced cycle performance and extraordinary rate performance. So far, there are scarcely any researches to investigate the performance of the cell with GPEs at relatively low temperature. In this study, the cycling performances of the cell involving a-GPE were characterized at 25 °C, 0 °C and −10 °C, as shown in Fig. 5a. It can be seen that the discharge capacities of cell at 0.5C rate showed a slight downtrend with reducing temperature from 25 °C to −10 °C. The cell with the a-GPE delivered an initial discharge capacity of 114.7 mA h g−1 at −10 °C, which was corresponding to 81.8% of the initial discharge capacity of the cell at room temperature (see Table S6). The high initial capacity retention of the cell involving a-GPE indicated that the polarization in the cell was effectively mitigated because of the existence of BTO in aGPE. In addition, the NMC/a-GPE/Li cell at 0 °C and −10 °C exhibited good cycling stability demonstrating the a-GPE could be used at low temperature, which endowed the LMBs involving a-GPE with fascinating application prospect. Furthermore, to better understand the potential application of flexible LMBs based on a-GPE, the as-fabricated flexible NMC/a-GPE/graphite battery was assembled to various

4. Conclusion A novel multifunctional a-GPE was successfully designed and fabricated by one-step for 4.5 V-class NMC/Li cells. The skeleton of a-GPE presented a gradient structure, which was collocated elaborately from PVDF-HFP nanofiber membranes with different content of BTO. The gradient structure could guarantee smooth and steady Li-ion flux in interior of the entire a-GPE, enhancing long-time cycling stability of the batteries using a-GPE. More significantly, the PVDF-HFP membrane without BTO in a-GPE was close to anode, and a homogeneous Li deposition/stripping was achieved due to the presence of the electronegative C–F functions groups and possibly generating favorable LiF from interfacial reaction. The PVDF-HFP membrane with 10 wt% BTO was adjacent to cathode to form an effective protection film for stabilizing CEI and enable discharging/charging in cells owing to polarization effect on the migration kinetics of mobile Li ions, especially under high cut-off voltage and high rate. The resulting a-GPE involving the composite membrane skeleton exhibited high ionic conductivity, Li-ion transference number and wide electrochemical window. Furthermore, the 4.5 V-class NMC/Li cells based on a-GPE delivered superior capacity retention (70% after 200 cycles) at high rate of 5C and outstanding rate capability. The NMC/Li cells involving a-GPE also displayed remarkable electrochemical performance at low temperature, high safety and favorable flexibility. These superior comprehensive performances validated the multifunctional a-GPE could be used in high-efficiency and safe high-voltage LMBs. This work provides a promising strategy to design multifunctional polymer electrolyte and meet the functional requirements of next generation high-performance energy storage

Fig. 5. (a) Cycle performance of NMC/a-GPE/Li cell with 4.3 V cut-off at 0.5C under 25 °C, 0 °C and −10 °C; (b) Photographs of different LED lighted by flexible NMC/graphite battery based on a-GPE. 8

Journal of Membrane Science 588 (2019) 117194

M. Zhu, et al.

systems.Acknowledgements The authors acknowledge the financial support from the National Natural Science Foundation of China (No. 51873011 and No. U1664251).

ions trapping, Adv. Energy Mater. 8 (2018) 1702561. [23] E.M. Masoud, Citrated porous gel copolymer electrolyte composite for lithium ion batteries application: an investigation of ionic conduction in an optimized crystalline and porous structure, J. Alloy. Comp. 651 (2015) 157–163. [24] M. Yanilmaz, Y. Lu, J. Zhu, X. Zhang, Silica/polyacrylonitrile hybrid nanofiber membrane separators via sol-gel and electrospinning techniques for lithium-ion batteries, J. Power Sources 313 (2016) 205–212. [25] W. Zheng, Y. Zhu, B. Na, R. Lv, H. Liu, W. Li, H. Zhou, Hybrid silica membranes with a polymer nanofiber skeleton and their application as lithium-ion battery separators, Compos. Sci. Technol. 144 (2017) 178–184. [26] W.-K. Shin, J. Cho, A.G. Kannan, Y.-S. Lee, D.-W. Kim, Cross-linked composite gel polymer electrolyte using mesoporous methacrylate-functionalized SiO2 nanoparticles for lithium-ion polymer batteries, Sci. Rep. 6 (2016) 26332. [27] E.M. Masoud, Nano lithium aluminate filler incorporating gel lithium triflate polymer composite: preparation, characterization and application as an electrolyte in lithium ion batteries, Polym. Test. 56 (2016) 65–73. [28] E.M. Masoud, A.A. El-Bellihi, W.A. Bayoumy, E.A. Mohamed, Polymer composite containing nano magnesium oxide filler and lithiumtriflate salt: an efficient polymer electrolyte for lithium ion batteries application, J. Mol. Liq. 260 (2018) 237–244. [29] E.M. Masoud, A.A. El-Bellihi, W.A. Bayoumy, M.A. Mousa, Effect of LiAlO2 nanoparticle filler concentration on the electrical properties of PEO-LiClO4 composite, Mater. Res. Bull. 48 (2013) 1148–1154. [30] E.M. Masoud, M.E. Hassan, S.E. Wahdaan, S.R. Elsayed, S.A. Elsayed, Gel P(VdF/ HFP)/PVAc/lithium hexafluorophosphate composite electrolyte containing nano ZnO filler for lithium ion batteries application: effect of nano filler concentration on structure, thermal stability and transport properties, Polym. Test. 56 (2016) 277–286. [31] C. Yada, A. Ohmori, K. Ide, H. Yamasaki, T. Kato, T. Saito, F. Sagane, Y. Iriyama, Dielectric modification of 5V‐class cathodes for high‐voltage all‐solid‐state lithium batteries, Adv. Energy Mater. 4 (2014) 1301416. [32] T. Teranishia, Y. Yoshikawa, R. Sakuma, H. Hashimoto, H. Hayashi, A. Kishimoto, T. Fujii, High-rate performance of ferroelectric BaTiO3-coated LiCoO2 for Li-ion batteries, Appl. Phys. Lett. 105 (2014) 143904. [33] B. Dutta, D. Deb, S. Bhattacharya, Electroactive phase nucleation and isothermal crystallization kinetics in ionic liquid-functionalized ZnS nanoparticle-ingrained P (VDF-HFP) copolymer nanocomposites, J. Mater. Sci. 54 (2019) 2990–3008. [34] H. Lee, M. Yanilmaz, O. Toprakci, K. Fu, X. Zhang, A review of recent developments in membrane separators for rechargeable lithium-ion batteries, Energy Environ. Sci. 7 (2014) 3857–3886. [35] P. Zhang, L.C. Yang, L.L. Li, M.L. Ding, Y.P. Wu, R. Holze, Enhanced electrochemical and mechanical properties of P(VDF-HFP)-based composite polymer electrolytes with SiO2 nanowires, J. Membr. Sci. 379 (2011) 80–85. [36] W. Li, Y. Wu, J. Wang, D. Huang, L. Chen, G. Yang, Hybrid gel polymer electrolyte fabricated by electrospinning technology for polymer lithium-ion battery, Eur. Polym. J. 67 (2015) 365–372. [37] P. Grandgeorge, N. Krins, A. Hourlier-Fargette, C. Laberty-Robert, S. Neukirch, A. Antkowiak, Capillarity-induced folds fuel extreme shape changes in thin wicked membranes, Science 360 (2018) 296–299. [38] M. Zhu, J. Lan, C. Tan, G. Sui, X. Yang, Degradable cellulose acetate/poly-L-lactic acid/halloysite nanotube composite nanofiber membranes with outstanding performance for gel polymer electrolytes, J. Mater. Chem. A 4 (2016) 12136–12143. [39] X. Zhang, T. Liu, S. Zhang, X. Huang, B. Xu, Y. Lin, B. Xu, L. Li, C.-W. Nan, Y. Shen, Synergistic coupling between Li6.75La3Zr1.75Ta0.25O12 and poly(vinylidene fluoride) induces high ionic conductivity, mechanical strength, and thermal stability of solid composite electrolytes, J. Am. Chem. Soc. 139 (2017) 13779–13785. [40] W. Zhang, J. Nie, F. Li, Z.L. Wang, C. Sun, A durable and safe solid-state lithium battery with a hybrid electrolyte membrane, Nanomater. Energy 45 (2018) 413–419. [41] Z. He, L. Chen, B. Zhang, Y. Liu, L.-Z. Fan, Flexible poly (ethylene carbonate)/garnet composite solid electrolyte reinforced by poly (vinylidene fluoride-hexafluoropropylene) for lithium metal batteries, J. Power Sources 392 (2018) 232–238. [42] Y.J. Shan, Y. Inaguma, M. Itoh, The effect of electrostatic potentials on lithium insertion for perovskite oxides, Solid State Ionics 79 (1995) 245–251. [43] J. Yan, J. Yu, B. Ding, Mixed ionic and electronic conductor for Li‐metal anode protection, Adv. Mater. 30 (2018) 1705105. [44] J. Luo, C.-C. Fang, N.-L. Wu, High polarity poly (vinylidene difluoride) thin coating for dendrite‐Free and high‐performance lithium metal anodes, Adv. Energy Mater. 8 (2018) 1701482. [45] W. Li, A. Dolocan, P. Oh, H. Celio, S. Park, J. Cho, A. Manthiram, Dynamic behaviour of interphases and its implication on high-energy-density cathode materials in lithium-ion batteries, Nat. Commun. 8 (2017) 14589. [46] W. Zhao, J. Zheng, L. Zou, H. Jia, B. Liu, H. Wang, M.H. Engelhard, C. Wang, W. Xu, Y. Yang, J.-G. Zhang, High voltage operation of Ni‐rich NMC cathodes enabled by stable electrode/electrolyte interphases, Adv. Energy Mater. 8 (2018) 1800297. [47] Y. Ji, P. Zhang, M. Lin, W. Zhao, Z. Zhang, Y. Zhao, Y. Yang, Toward a stable electrochemical interphase with enhanced safety on high-voltage LiCoO2 cathode: a case of phosphazene additives, J. Power Sources 359 (2017) 391–399. [48] X. Dai, A. Zhou, J. Xu, Y. Lu, L. Wang, C. Fan, J. Li, Extending the high-voltage capacity of LiCoO2 cathode by direct coating of the composite electrode with Li2CO3 via magnetron sputtering, J. Phys. Chem. C 120 (2015) 422–430.

Conflicts of interest The authors declare no conflict of interest. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.memsci.2019.117194. References [1] A. Manthiram, An outlook on lithium ion battery technology, ACS Cent. Sci. 3 (2017) 1063–1069. [2] L. Zhou, K. Zhang, Z. Hu, Z. Tao, L. Mai, Y.-M. Kang, S.-L. Chou, J. Chen, Recent developments on and prospects for electrode materials with hierarchical structures for lithium-ion batteries, Adv. Energy Mater. 8 (2017) 1701415. [3] R. Schmuch, R. Wagner, G. Hörpel, T. Placke, M. Winter, Performance and cost of materials for lithium-based rechargeable automotive batteries, Nat. Energy 3 (2018) 267–278. [4] D. Lin, Y. Liu, Y. Cui, Reviving the lithium metal anode for high-energy batteries, Nat. Nanotechnol. 12 (2017) 194–206. [5] X.-B. Cheng, R. Zhang, C.-Z. Zhao, Q. Zhang, Toward safe lithium metal anode in rechargeable batteries: a review, Chem. Rev. 117 (2017) 10403–10473. [6] P. Albertus, S. Babinec, S. Litzelman, A. Newman, Status and challenges in enabling the lithium metal electrode for high-energy and low-cost rechargeable batteries, Nat. Energy 3 (2018) 16–21. [7] C. Tian, F. Lin, M.M. Doeff, Electrochemical characteristics of layered transition metal oxide cathode materials for lithium ion batteries: surface, bulk behavior, and thermal properties, Acc. Chem. Res. 51 (2018) 89–96. [8] W. Li, B. Song, A. Manthiram, High-voltage positive electrode materials for lithiumion batteries, Chem. Soc. Rev. 46 (2017) 3006–3059. [9] X.-Q. Zhang, X.-B. Cheng, Q. Zhang, Advances in interfaces between Li metal anode and electrolyte, Adv. Mater. Interfaces 5 (2018) 1701097. [10] B. Liu, J.-G. Zhang, W. Xu, Advancing lithium metal batteries, Joule 2 (2018) 833–845. [11] S. Nowak, M. Winter, The role of cations on the performance of lithium ion batteries: a quantitative analytical approach, Acc. Chem. Res. 51 (2018) 265–272. [12] Y. Xia, J. Zheng, C. Wang, M. Gu, Designing principle for Ni-rich cathode materials with high energy density for practical applications, Nanomater. Energy 49 (2018) 434–452. [13] Y. Wang, W.-H. Zhong, Development of electrolytes towards achieving safe and high‐performance energy‐storage devices: a review, Chem. Electro. Chem. 2 (2015) 22–36. [14] S.-J. Tan, X.-X. Zeng, Q. Ma, X.-W. Wu, Y.-G. Guo, Recent advancements in polymer-based composite electrolytes for rechargeable lithium batteries, Electrochemical Energy Reviews 1 (2018) 113–138. [15] D. Lin, P.Y. Yuen, Y. Liu, W. Liu, N. Liu, R.H. Dauskardt, A silica‐aerogel‐reinforced composite polymer electrolyte with high ionic conductivity and high modulus, Y. Cui, Adv. Mater. 30 (2018) 1802661. [16] W. Fan, N.‐W. Li, X. Zhang, S. Zhao, R. Cao, Y. Yin, Y. Xing, J. Wang, Y.‐G. Guo, C. Li, A dual‐salt gel polymer electrolyte with 3D cross‐linked polymer network for dendrite‐free lithium metal batteries, Adv. Sci. (2018), https://doi.org/10.1002/ advs.201800559. [17] W. Dong, X.-X. Zeng, X.-D. Zhang, J.-Y. Li, J.-L. Shi, Y. Xiao, Y. Shi, R. Wen, Y.X. Yin, T.-S. Wang, C.-R. Wang, Y.-G. Guo, Gradiently polymerized solid electrolyte meets with micro-/nanostructured cathode array, ACS Appl. Mater. Interfaces 10 (2018) 18005–18011. [18] Q. Zheng, L. Xing, X. Yan, X. Li, C. Ye, K. Wang, Q. Huang, W. Li, N-Allyl-N,N-bis (trimethylsilyl)amine as a novel electrolyte additive to enhance the interfacial stability of a Ni-rich electrode for lithium-ion batteries, ACS Appl. Mater. Interfaces 10 (2018) 16843–16851. [19] H. Duan, Y.-X. Yin, Y. Shi, P.-F. Wang, X.-D. Zhang, C.-P. Yang, J.-L. Shi, R. Wen, Y.G. Guo, L.-J. Wan, Dendrite-free Li-metal battery enabled by a thin asymmetric solid electrolyte with engineered layers, J. Am. Chem. Soc. 140 (2017) 82–85. [20] J.-Y. Liang, X.-X. Zeng, X.-D. Zhang, P.-F. Wang, J.-Y. Ma, Y.-X. Yin, X.-W. Wu, Y.G. Guo, L.-J. Wan, Mitigating interfacial potential drop of cathode-solid electrolyte via ionic conductor layer to enhance interface dynamics for solid batteries, J. Am. Chem. Soc. 140 (2018) 6767–6770. [21] X. Cheng, J. Pan, Y. Zhao, M. Liao, H. Peng, Gel polymer electrolytes for electrochemical energy storage, Adv. Energy Mater. 8 (2018) 1702184. [22] M. Zhu, J. Wu, W.-H. Zhong, J. Lan, G. Sui, X. Yang, A biobased composite gel polymer electrolyte with functions of lithium dendrites suppressing and manganese

9