Janus Quasi-Solid Electrolyte Membranes with Asymmetric Porous Structure for High-Performance Lithium-Metal Batteries

Highlights Janus quasi-solid electrolyte membranes with asymmetric porous structure were constructed, showing a high σLi+ of 1.5 × 10-4 S cm-1 and a high t+ of 0.71. The solvation structures and ion transport dynamics in nanopores have been deciphered, manifesting a concentrated electrolyte-like structure and regulated transport behaviors. Quasi-solid NCM 622||Li cells have been demonstrated to stably cycle for 200 cycles at 1 C, and pouch cell has shown high tolerance for abuse. Supplementary Information The online version contains supplementary material available at 10.1007/s40820-024-01325-4.


S1.3 Characterization
X-ray diffraction patterns were acquired by an X-ray diffractometer (EMPYREAN PANalytical) with Cu-Kα radiation (λ=1.54Å).The morphologies of the samples were observed by field-emission scanning electron microscopy (SEM) (Phenom, PW-100-060).Nitrogen adsorption-desorption measurements at 77 K were performed on a TriStar II 3020 surface area analyzer.Before analysis, samples were degassed at 120 °C for 24 h.The pore sizes of MOF and MS were determined by density function theory (DFT) model and Barret-Joyner-Halenda (BJH) model, respectively.Raman spectra was obtained on a HORIBA LabRAM HR Evolution Raman spectrometer with a 473 nm laser.Thermal gravimetric analysis (TGA) profiles were obtained on a SDT Q600 V8.2 Build 100 with a heating rate of 10 °C min -1 in air.The atomic force microscopy (AFM, MFP-3D Origin, Asylum Research, Oxford Instruments) was used to analyze Young's modulus of the samples.Uniformly distributed 100 points were sampled for Yang's Modulus testing.The surface composition and valence state were analyzed by X-ray photoelectron spectroscopy (XPS, Thermo SCIENTIFIC ESCALAB XQI) with Al Kα radiation.All XPS spectra were calibrated by shifting the detected adventitious carbon C 1s peak to 284.8 eV.Contact angle measurements were performed on JY-82C.The porosity was assessed by immersing the porous matrix in a liquid electrolyte solution (1M LiTFSI in PC) and calculating the volume ratio between the absorbed electrolytes and the porous matrix.

S1.4 Electrochemical Measurements
Ionic conductivity was tested with an electrochemical workstation (BioLogic).The ionic conductivity was determined using electrochemical impedance spectroscopy (EIS) with pellets or membranes between two stainless steel (SS) electrodes in 2032-type coin cells.The frequency range was from 100 kHz to 0.1 Hz and the AC amplitude was 10 mV.The ionic conductivity (σ, S cm -1 ) was calculated by using the endpoint of the semicircle as the ion resistivity (R, Ohm), thickness (L, cm), and area (S, cm 2 ) of the pellets or membranes based on σ = L/(R×S).Symmetric cells were fabricated using two pieces of the Janus electrolyte to ensure MOF side was placed toward lithium metal electrode.Electrochemical window was measured by linear sweep voltammetry (LSV) using Li|electrolyte|SS cells with a scan rate of 1 mV s -1 range from open voltage to 5.0 V.
For the measurement of average CE, 5 mA h cm -2 Li was first plated on Cu foil and stripped to 1.0 V at a current density of 0.5 mA cm -2 .Then plate quantitative Li reservoir (Qt = 5 mA h cm -2 ) on Cu foil, repeatedly strip/plate Li with an area capacity of 0.5 mA h cm -2 (Qc) for n cycles (n=10), and finally trip all the residual Li (Qs) to a cutoff voltage of 1.0 V at a current density of 0.5 mA cm - 2 .The average CE over the n cycle was calculated based on CEave = (Qs+nQc)/(Qt+nQc).

S1.5 Simulation of Current Density and Lithium Dendrite
Li-ion flux and deposition were simulated by using the finite element solver of COMSOL Multiphysics.For the simulation of Li-ion flux when using different electrolytes, simplified 2D models were first built (Fig. S9), where the bottom of the model represented the Li anode.An SEI layer with a thickness of 1 nm was constructed on the surface of the anode.The geometry parameters were presented in Fig. S12.The diffusion coefficients of Li + in PE, MS, MOF and SEI layer were set as 5 × 10 -7 , 1.5 × 10 -6 , 1 × 10 -6 , and 5 × 10 -9 cm 2 s -1 , respectively.The diffusion coefficients of anions were set as 2 × 10 -6 , 3 × 10 -6 , 7 × 10 -7 , and 2 × 10 -8 cm 2 s -1 , respectively.The initial concentration of Li + was set to 1 M and the average current density was set as 2 mA cm -2 .
As for the simulation of the growth of lithium dendrite under different densities, simplified models with spherical cap-shaped nucleus were built, whose geometry parameters were illustrated in Fig. S13.The diffusion coefficient of Li + in electrolyte and SEI layer were set as 5 × 10 -7 cm 2 s - 1 , 5 × 10 -9 cm 2 s -1 , respectively, while that of anions were 2 × 10 -6 cm 2 s -1 and 2 × 10 -8 cm 2 s -1 , respectively.The average current density was set as 0.5 mA cm -2 , 2 mA cm -2 and 5 mA cm -2 , respectively.Note that all models were ideal and cannot fully represent the real situation.

S1.6 Molecular Dynamic (MD) Simulations
MD simulations were performed based on the Large Scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) code [S3].Three models of different types of electrolytes were built for simulation: 1) MOFLi QSEs: UiO-66 crystal with size of 42.0 Å × 42.0 Å × 42.0 Å was first built, followed by the adding of 32 LiTFSI and 386 PC based on the mass ratio of the experiment; 2) MSLi QSEs: SiO2 crystal with size of 73.65 Å × 76.5 Å × 54.02 Å was built, following by deleting the atoms inside the cylinder with a diameter of 5.0 nm to build MS.Liquid electrolyte (1 M LiTFSI in PC) was then filled the channel of MS to get MSLi.3) Liquid electrolytes: A box with a size of 33.9 Å × 33.9 Å × 33.9 Å was filled with 20 Li + , 20 TFSI -, and 240 PC to simulate 1 M LiTFSI/PC solution.
The OPLS-AA force field was used to describe the energy potential of Li + [S4, S5] and UFF force field was used for the residual part [S6].The bonded and non-bonded parameters for Li + and the remaining part were obtained from Jensen et al. [S7].and Gouveia et al. [S8], respectively.The partial charges of UiO-66 were fitted from DFT calculations result.A cutoff of 12 Å was used for both van der Waals interactions and long-range correction (particle-particle-particle-mesh) of Coulombic interactions.
The initial configurations were first minimized by conjugated-gradient energy minimization scheme employing a convergence criterion of 1.0 × 10 -4 .The systems were then equilibrated in NPT ensemble using the Parrinello-Rahman barostat for 2 ns to maintain a temperature of 1000 K and a pressure of 1 atm.Another 5 ns production run in NPT ensemble at 1000 K was conducted finally.A time step of 0.2 fs was used for all simulations.Only the final 5 ns was sampled for radical distribution function (RDF) calculation and mean square displacement (MSD) calculation.

Fig. S3 Fig. S4
Fig. S1 TG profiles of MOF and MS particles

Fig. S27
Fig. S27 Performance of LCO||Li batteries.a, b) Rate performance of LCO|MOFLi/MSLi|Li batteries and the corresponding voltage profile.c) Cycling performance of LCO||Li batteries at 1 C

Table S2
Porosity measurement of PE and MOF or MS pellets