First author: Jiang Bowen
Corresponding authors: Huang Yunhui, Xu Henghui
Communication unit: Huazhong University of Science and Technology
【Research Background】
The research on polymer electrolyte (SPE) and lithium metal anode began in the early days of lithium battery development, but decades of research have not yet achieved the industrialization of all-solid-state lithium metal battery (SSLMB). At present, the most serious problem in solid state is the growth of lithium dendrites. In SSLMB, lithium dendrites are easily generated during cycling even at small current densities due to the low ionic conductivity of SPE. At the same time, lithium (Li) has high reactivity with SPE, resulting in more organic components (such as ROCOOLi) in the solid electrolyte-electrolyte interphase (SEI). The unstable organic SEI layer continues to rupture, which continuously destroys the interface stability of SPE/Li. Under the action of these two factors, the sudden short circuit has become a typical failure form of many SPEs, such as the most studied polyethylene oxide (PEO) electrolyte, because it is easily penetrated by lithium dendrites, which greatly limits its industrial application.
【Introduction】
In order to solve this problem, Li0.46Mn0.77PS3 (LiMPS) was used to provide a physical barrier for PEO electrolyte and shield the growth and puncture of lithium dendrites. LiMPS has good Li+ conductivity and adsorption capacity for Li+, which can not only level the flow of Li+ through its two-dimensional nanomorphology, eliminate the source of lithium dendrite growth, but also physically inhibit the growth of lithium dendrites by using the rigidity of its two-dimensional sheet. In addition, LiMPS not only has a shielding effect, but also helps to generate SEI rich in inorganic components to enhance interface stability, so that Li/PEO-LiMPS/Li symmetrical cells can be stably cycled for more than 600 h at 0.2 mA cm-2. In addition, LiMPS significantly increased the ionic conductivity of the PEO electrolyte, and the PEO electrolyte added with 5 wt % LiMPS had an ionic conductivity of 2.6 × 10-4 S cm-1 at 45 °C, almost 10 times that of pure PEO. Due to these advantages, the full cells matched to the PEO-LiMPS electrolyte exhibit excellent cycling performance, with a capacity retention rate of more than 80 % after 200 cycles at 0.2 C. At the same time, in order to prove the universality of the two-dimensional shielding strategy, polyvinylidene fluoride (PVDF) electrolyte added with LiMPS can also effectively resist lithium dendrites. This work provides a unique idea for the design of protective strategies for polymer electrolytes. This article was published in the internationally renowned journal Energy Storage Materials under the title "Polymer electrolytes shielded by 2D Li0.46Mn0.77PS3 Li+-conductors for all-solid-state lithium-metal batteries".
【Content Details】
1. Physical and chemical properties and characterization of LiMPS and PEO-LiMPS composite electrolytes
Figure 1 (a) Schematic diagram of Li+ intercalated lithiation and preparation process of PEO-LiMPS electrolyte in bulk MPS. (b) SEM image of MPS. (c) Photograph of the Tyndall effect of LiMPS aqueous dispersion. (d-e) TEM image of a LiMPS nanosheet. (f) AFM image of a LiMPS nanosheet. (g) XRD patterns of MPS and LiMPS. (h) FTIR spectra of MPS and LiMPS.
The preparation of LiMPS is shown in Figure 1a. Prior to intercalation, the block MnPS3 (MPS) (Figure 1b) consists of a large number of tightly stacked MPS sheets. After Li+ intercalation, MPS is stripped into independent LiMPS nanosheets, as shown by the Tyndall effect of the LiMPS aqueous dispersion system (Figure 1c). Transmission electron microscopy (TEM) (Figures 1d and 1e) images show a distinct lamellar structure with folds, indicating that MPS was successfully peeled off. Figure 1f shows that the LiMPS sheet has clear edges and an average thickness of 1.3 nm. The two-dimensional structure and morphology of LiMPS ensure its shielding effect on lithium dendrites. X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR) further demonstrate the successful intercalation of Li+.
Figure 2 (a) Surface and cross-sectional SEM images of the PEO-5LiMPS electrolyte. (b) XRD spectra of PEO-xLiMPS (x = 0, 1, 3, 5, 10) electrolyte membranes. (c) DSC curve of PEO-xLiMPS (x = 0, 1, 3, 5, 10) electrolyte. Tm is the melting point and Tg is the glass transition temperature. (d) Solid-state NMR spectra of PEO and PEO-5LiMPS electrolytes. (e) EIS diagram of PEO-5LiMPS electrolyte at different temperatures. (f) Ionic conductivity of PEO-xLiMPS (x = 0, 1, 3, 5, 10) electrolyte at different temperatures.
Photographs and SEM images of the PEO-xLiMPS electrolyte (x is the weight ratio of LiMPS wt %, x = 0, 1, 3, 5, 10) are shown in Figure 2a. At the same time, the crystal structure of the PEO-xLiMPS electrolyte was studied by XRD (Figure 2b). The PEO membrane has two typical crystalline peaks at ~19° and ~23° (inside the red wireframe). After the addition of LiMPS, the two peaks at 13.4° and 27.3° gradually enhanced with the increase of the amount added. At the same time, with the addition of LiMPS, the crystallinity of PEO decreases, which is conducive to segment movement and ion transport.
The Tm of the PEO-xLiMPS electrolyte (x = 0, 1, 3, 5, 10) were 48.48, 48.23, 47.02, 44.57, and 46.70 °C, respectively (Figure 2c). With the addition of LiMPS (x≤5), Tm has a significant tendency to decrease. At the same time, studies of the charge environment of Li+ by solid-state nuclear magnetic resonance (NMR) (Figure 2d) showed that there is a low field shift in the Li spectrum of the PEO-5LiMPS electrolyte compared to PEO. This indicates that the charge environment around Li+ is thinner, which is conducive to the migration of Li+ and improves ion conductivity.
The electrochemical impedance spectroscopy (EIS) of the PEO-5LiMPS electrolyte at different temperatures is shown in Figure 2e, and the ionic conductivity of the PEO-xLiMPS (x = 0, 1, 3, 5, 10) electrolyte was tested (Figure 2f). The ionic conductivity of PEO-5LiMPS at 45 °C is 2.6 × 10−4 S cm−1, and its ionic conductivity is mainly due to two aspects: 1) the crystallinity of PEO decreases, and 2) LiMPS provides additional ionic conductivity.
2. LiMPS shielding effect mechanism
Figure 3 (a) EIS pattern and fitted equivalent circuit of Li/PEO/Li and Li/PEO-5LiMPS/Li symmetrical cells. (b) CCD testing of Li/PEO/Li and Li/PEO-5LiMPS/Li symmetrical cells. (c) Cycling performance of Li/PEO/Li and Li/PEO-5LiMPS/Li symmetrical batteries at a current density of 0.2 mA cm−2 and an area capacity of 0.1 mAh cm−2. (d-i) SEM image of lithium metal acquired from a cycled Li-Li symmetrical cell with a number of cycles of (d, e) 20 turns, (f-i) 50 turns. wherein (d, f, g) lithium metal obtained from Li/PEO-5LiMPS/Li symmetrical cells. (e, h, i) Lithium metal obtained from Li/PEO/Li symmetrical cells. The battery cycle has a current density of 0.1 mA cm−2 and an area capacity of 0.1 mAh cm−2. (j - m) Computational simulation of DFT of Li+ adsorption on MPS and LiMPS surfaces. (j, k) MPS (l, m) LiMPS。 (j, l) is the initial state and (k, m) is the final state. Et is the total energy of the system and Ead is the adsorption energy.
In order to verify the LiMPS shielding effect, PEO-5LiMPS was used to assemble a lithium-lithium symmetric battery with PEO electrolyte and test its cycle stability. The EIS pattern and fitting data of Li/PEO-5LiMPS/Li-symmetric cells are shown in Figure 3a, and the overall impedance is significantly reduced compared with Li/PEO/Li-symmetric cells. Furthermore, the limit current density (CCD) of Li/PEO-5LiMPS/Li symmetrical cells increased significantly, from 0.6 mA cm−2 to 1.6 mA cm−2 (Figure 3b). In the symmetrical cell cycle test, the Li/PEO-5LiMPS/Li cell cycled stably for more than 600 hours at a current density of 0.2 mA cm−2 with an overpotential of 100 mV (Figure 3c). In contrast, due to the low ionic conductivity and unstable interface of the PEO electrolyte, the current density reaches 0.2 mA cm−2 and is immediately short-circuited. The excellent performance of Li/PEO-5LiMPS/Li symmetrical batteries reflects the advantages of LiMPS shielding effect.
After cycling, the anode of lithium metal has a flat morphology and no dendrites, which directly proves the effectiveness of the two-dimensional shielding strategy (3D - 3i). Lithium metal paired with PEO-5LiMPS is flat and smooth after 20 cycles (Figure 3d), while lithium metal paired with PEO is filled with lithium dendrites (Figure 3e). The difference is even more pronounced after 50 laps. Lithium metals matched to PEO-5LiMPS still have a flat surface topography (Figure 3f) and a dense lithium deposition layer (Figure 3g). In stark contrast, lithium metal paired with the PEO electrolyte has a thicker layer of dead lithium after cycling and a large number of dendrites (Figures 3h and 3i). Therefore, the LiMPS two-dimensional sheet can effectively protect PEO and inhibit lithium dendrites, so that the lithium deposition morphology is uniform.
In addition to the suppression of lithium dendrites, the second mechanism of the LiMPS shielding effect is the leveling effect on the Li+ flow rate caused by the two-dimensional topography of LiMPS, the Li+ adsorption energy and the high Li+ conductivity (Figure 3j-3m). Density functional theory (DFT) calculates the interaction of LiMPS with Li+. The adsorption energy of Li+ on the MPS surface is -0.529 eV/Li+ (Figures 3j and 3k), compared to -1.483 eV/Li+ on the LiMPS surface (Figures 3l and 3m). This strong Li+ adsorption energy can promote the leveling effect of Li+ flow. The adsorption-conduction synergy maximizes the role of LiMPS as a two-dimensional leveling agent and eliminates the uneven distribution of Li+ flow, the source of dendrite growth.
Figure 4 (a-c) Cyclical lithium metal acquired in Li/PEO/Li (top row) and Li/PEO-5LiMPS/Li (bottom row) symmetrical cells. XPS spectrum of (a) F1s, (b) O1s and (c) N1s, with a battery cycle of 20 turns. (d) XPS spectrum of lithium metal S 2p obtained after 50 cycles in Li/PEO-5LiMPS/Li symmetrical cells. XPS spectra of (e) Mn 2p, (f) P 2p and (g) S 2p of PEO-5LiMPS membrane after 50 cycles of Li/PEO-5LiMPS/Li symmetrical cell cycle.
3. Electrochemical performance of LFP/NCM811-Li battery
Figure 5 (a) Cycling performance of LFP/PEO-5LiMPS/Li (red) and LFP/PEO/Li batteries (blue) with a discharge rate of 0.2 C. (b) Rate performance of LFP/PEO-5LiMPS/Li (red) and LFP/PEO/Li batteries (blue) with discharge rates of 0.1, 0.2, 0.5, 0.7 and 1 C. (c) Cycling performance of NCM811/PEO-5LiMPS/Li (red) and NCM811/PEO/Li batteries (blue) with a charge-discharge ratio of 0.2 C. (d) Rate performance of NCM811/PEO-5LiMPS/Li (red) and NCM811/PEO/Li (blue) batteries. (e) Voltage curve of NCM811/PEO-5LiMPS/Li battery at different discharge rates. (f) Cycling performance of NCM811/PVDF-5LiMPS/Li batteries at 0.5 C. (g) Voltage curve of LFP/PEO-5LiMPS/Li pouch battery during cycling. (h) EIS pattern of LFP/PEO-5LiMPS/Li pouch battery after 0, 50 and 100 cycles, respectively. (i) Cycling performance of LFP/PEO-5LiMPS/Li pouch batteries at a discharge rate of 0.2 C. (j) Ultrasound of LFP/PEO-5LiMPS/Li pouch cells before and after 200 cycles. All coulomb efficiencies are assigned to PEO-5LiMPS or PVDF-5LiMPS cells.
To fully demonstrate the shielding strategy, the full battery was assembled using the actual positive electrode. Figure 5a shows the cycling performance of LFP/PEO-5LiMPS/Li and LFP/PEO/Li cells at 0.2 C discharge rate. Consistent with Li/PEO-5LiMPS/Li symmetrical cells, the battery matched with the lithium iron phosphate (LFP) cathode has good cycle stability, and the coulombic efficiency (CE) exceeds 99%. Due to the increase of ionic conductivity and good interfacial stability, LFP/PEO-5LiMPS/Li cells still maintain a specific capacity of 113.3 mAh g−1 after 200 cycles. In contrast, pure PEO full batteries have a low specific capacity retention rate. LFP/PEO-5LiMPS/Li batteries benefit from the high ionic conductivity of PEO-5LiMPS electrolyte, and their discharge specific capacities are 143.0, 132.6, 134.2, 126.9 and 123.2 mAh g−1 at discharge ratios of 0.1, 0.2, 0.5, 0.7 and 1 C, respectively (Figure 5b). In contrast, LFP/PEO/Li batteries have poor rate performance and greater specific capacity attenuation.
In order to prove the high voltage resistance of the PEO-5LiMPS electrolyte, a full battery with the positive electrode of LiNi0.8Co0.1Mn0.1O2 (NCM811) was assembled. As shown in Figure 5c, the specific capacity of the NCM811/PEO-5LiMPS/Li battery at 0.2 C is 156 mAh g−1, and the capacity retention rate of 200 turns is 80%. In contrast, the cycle performance of NCM811/PEO/Li batteries is poor, halving the specific capacity after only 30 cycles. Figure 5d shows the rate performance of NCM811/PEO-5LiMPS/Li and NCM811/PEO/Li batteries, and Figure 5e shows the voltage curve of NCM811/PEO-5LiMPS/Li battery rate test. The discharge specific capacities of NCM811/PEO-5LiMPS/Li batteries were 171.9, 161.5, 154.9 and 136.3 mAh g−1 at discharge ratios of 0.2, 0.5, 0.7 and 1 C, respectively, showing excellent electrochemical performance. The specific capacity of a battery matched with pure PEO electrolyte is only 112.1 mAh g−1 at 0.2 C and 50 mAh g−1 at 1 C. In addition, PVDF-5LiMPS electrolyte was prepared, and the NCM811/PVDF-5LiMPS/Li cell had an initial capacity of 156.6 mAh g-1 at a discharge rate of 0.5 C, and the cycle performance was good (Figure 5f).
Further, an all-solid-state LFP/PEO-5LiMPS/Li pouch battery was assembled to study the practicality of the PEO-LiMPS electrolyte. As shown in Figure 5g, the LFP/PEO-5LiMPS/Li pouch cell has excellent cycling performance at 0.2 C, and the interface impedance gradually decreases with cycling (Figure 5h). Figure 5i shows that the initial specific capacity of the pouch cell is 149.1 mAh g−1, and the capacity retention rate of 235 cycles reaches 93%, which proves the excellent interface stability of PEO-5LiMPS. In addition, the battery passed the abuse test shown in the illustration in Figure 5i, which still lights up the light panel even when the battery is cut into small pieces. The ability of pouch batteries to operate under harsh conditions proves the safety of PEO-LiMPS electrolytes.
Lithium battery ultrasonic imaging technology is a very sensitive technology for interface signals, and the stronger the transmitted signal, the better the interface contact of the solid electrolyte. Figure 5j shows the pseudocolor diagram of the transmitted signal of the LFP/PEO-5LiMPS/Li pouch battery, red indicates strong transmission signal (1 V), good interface contact, and blue (0 V) is the opposite. Pouch batteries are blue before cycling, indicating poor interface contact inside the battery. After 200 cycles, the peripheral area turns green, indicating better interface contact after cycling. This suggests that the shielding effect of LiMPS facilitates tight lithium deposition and makes the interface contact more compact.
【Conclusion】
In this work, a two-dimensional shielding strategy implemented by fast ion conductor LiMPS two-dimensional nanosheets is reported to protect SPE from lithium dendrite erosion and realize dendritic-free cycling of SSLMB. Due to the nanometer size effect of LiMPS two-dimensional nanosheets, the flow rate of Li+ is uniformly flattened, and it has strong adsorption energy and high conductivity for Li+, thereby clearing the source of lithium dendrite growth. In addition, the two-dimensional sheet layer of LiMPS also has a good physical inhibition effect on lithium dendrites. In addition, the Li+ conductivity of the PEO-LiMPS composite electrolyte was significantly increased (2.6 × 10−4 S cm−1 at 45 °C). SSLMB using LFP and NCM cathodes has a good cycle life, with a capacity retention rate of more than 90 % after 230 cycles. Therefore, this shielding strategy can effectively protect SPE and give SPE good physicochemical and electrochemical properties.