Promoting Polysulfide Redox Reactions through Electronic Spin Manipulation

Catalytic additives able to accelerate the lithium–sulfur redox reaction are a key component of sulfur cathodes in lithium–sulfur batteries (LSBs). Their design focuses on optimizing the charge distribution within the energy spectra, which involves refinement of the distribution and occupancy of the electronic density of states. Herein, beyond charge distribution, we explore the role of the electronic spin configuration on the polysulfide adsorption properties and catalytic activity of the additive. We showcase the importance of this electronic parameter by generating spin polarization through a defect engineering approach based on the introduction of Co vacancies on the surface of CoSe nanosheets. We show vacancies change the electron spin state distribution, increasing the number of unpaired electrons with aligned spins. This local electronic rearrangement enhances the polysulfide adsorption, reducing the activation energy of the Li–S redox reactions. As a result, more uniform nucleation and growth of Li2S and an accelerated liquid–solid conversion in LSB cathodes are obtained. These translate into LSB cathodes exhibiting capacities up to 1089 mA h g–1 at 1 C with 0.017% average capacity loss after 1500 cycles, and up to 5.2 mA h cm–2, with 0.16% decay per cycle after 200 cycles in high sulfur loading cells.

original state was analyzed through scanning electron microscopy (SEM), utilizing a field emission scanning electron microscope equipped with an electron column featuring a monochromator with UC (UniColore) Technology, specifically the FEI Magellan 400L.Atomic resolution aberrationcorrected high angle annular dark field scanning transmission electron microscopy (AC-HAADF-STEM) was performed in a double aberration-corrected Thermo Fisher Spectra 300 STEM operated at 200 KeV.The Spectra 300 is equipped with a Super X energy dispersive X-ray spectroscopy (EDS) detector.[3] The surface composition and chemical state of the host materials were confirmed by X-ray photoelectron spectroscopy (XPS).Thermogravimetric analysis (TGA) was used to measure the S content in the cathode from 50°C to 400°C under a 5°C min -1 heating rate in an N2 atmosphere.UV-vis absorption spectroscopy (Lambda 950 UV-Vis-NIR Spectrophotometer, Perkin Elmer) was used to analyze the adsorption performance of electrode materials on polysulfides.The X-ray absorption fine structure (XAFS) data were processed according to the standard procedures using the Athena module implemented in the IFEFFIT software package.The extended X-ray absorption fine structure (EXAFS) spectra were obtained by subtracting the post-edge background from the overall absorption and then normalizing with respect to the edge-jump step.Subsequently, the χ(k) data were Fourier transformed to real (R) space using Hanning windows (dk = 1.0 Å −1 ) to separate the EXAFS contributions from different coordination shells.To obtain the quantitative structural parameters around central atoms, least-squares curve parameter fitting was performed using the ARTEMIS module of the IFEFFIT software packages.Magnetization measurements were done using a superconducting quantum interferometer device (SQUID, Quantum Design, from ICMAB's scientific and technical services).The temperature-dependent magnetization (M) measurements were carried out with a magnetic property measurement system superconducting quantum interference device (MPMS SQUID) magnetometer and under magnetic field strength (H) of 50 Oe for all the samples.

Electrochemical Measurements:
The prepared electrode material, Super P and polyvinylidene fluoride (PVDF) binder were mixed and ground according to 8:1:1, and N-methylpyrrolidone (NMP) was added during the grinding process to prepare slurry.Then it was evenly coated on the current collector on the aluminum foil, and finally dried at 60 °C for 12 h and cut into 12 mm circular pieces to obtain the working electrode.The sulfur loading content of the cathodes were around 1 mg cm -2 .To further highlight the practical application of the electrode, the high-loading cathode was prepared by the same method.The sulfur electrode, primed for electrochemical processes, served as the cathode, while lithium foil functioned as the anode, with Celgard 2400 serving as the separator, and 1.0 M LiTFSI and 0.1 M LiNO3 were dissolved in DOL/DME (V: V=1:1) as the electrolyte to assemble a 2032-coin battery in the argon-filled glove box.The prepared battery is subjected to CV and GCD tests at a voltage window of 1.7 -2.8V, and an EIS test is performed at a frequency of 0.1 Hz -100K Hz.
Operando XRD: Operando XRD patterns were acquired on a Bruker D8 Advance A25 diffractometer in a Debye-Scherrer configuration equipped with a Mo Kα1 radiation source (λ = 0.7093 Å) using a costume holed-designed 2032 coin cells with a 5 mm diameter hole drilled and equipped with a 75 µm Kapton window.Diffraction patterns were measured upon continuous operation of the electrochemical cells between 8.5 and 13.5° with a step size of 0.02° and acquisition time of 60 min per pattern.The cells were cycled at C/25 rate with a Biologic SP-50 potentiostat in galvanostatic mode with potential limitation.
DFT calculations: All DFT calculations were performed using the Vienna ab-initio simulation package (VASP).The Perdew-Burke-Ernzerhof (PBE) functional for the exchange-correlation term was used with the projector augmented wave (PAW) potentials and a cutoff energy of 500 eV.
The convergence of energy and forces were set to 1×10 -5 eV and 0.05 eV/Å, respectively.The adsorption energy Ead was calculated as: where E(surf+ad) is the energy of the LiPS adsorbed on the surface, Esurf is the energy of the clean surface, and Ead is the energy of the free LiPS.
The formula for calculating the Gibbs free as fellow:

Pouch cell assembly and measurements:
The v-CoSe /S cathode and lithium anode were cut into 4×3 cm pieces.The sulfur loading of the cathode in the pouch cell was 1.8 mg cm −2 .The E/S ratio was about 20 μL mg −1 , and the thickness of the lithium belt anode was 0.4 mm.The separator and electrolyte were sandwiched between the tailored v-CoSe /S and lithium belt.

Calculation of spin-state:
We name the low spin ratio of Co 2+ in CoSe as R 2+ low, and the high spin ratio of Co 2+ in CoSe as R 2+ high, so we could perform a formula (1-2) below.
To be more specific, in the case of v-CoSe supposed to take into account the presence of 20% of Co 3+ ions, which has the three potential spin configurations (Figure S23), low spin (0 muB), medium spin (2 muB), and high spin 4 (muB), then the calculation formula should apply the following formula (3): In this scenario, since there are too many uncertain values, we assume two extreme cases which is if all Co 3+ is with low spin( R 3+ low=1, R 3+ med=0, R 3+ high=0), then R 2+ high=100%; Or all Co 3+ is with high spin, ( R 3+ low=0, R 3+ med=0, R 3+ high=1), then R 2+ high=75%.

Thickness calculation:
To calculate the thickness of the nanosheets composing our sample, we utilize low-loss EELS spectra.The thickness (t) is computed using the logarithmic-ratio formula (3).
Where  represents the inelastic mean free path,   denotes the total area under the low-loss EELS spectrum and  0 is the area under the zero-loss peak.The logarithmic term is computed using DigitalMicrograph 3.0 software.We designate a specific region of the image and determine a value of 0.09 for this term.
To calculate the , we employ Equation (4) as proposed by T. Malis et al. [4] :

Figure S2 .
Figure S2.XRD patterns of v10-CoSe and hexagonal and cubic CoSe and CoSe2 references.The unmatched peak intensities are related to the asymmetric geometry of the nanosheet.[5]

Figure S5 .Figure S6 .
Figure S5.Atomic models and the corresponding simulated AC-HAADF-STEM images for (a-c) CoSe, and (e-f) v-CoSe along its [0001] zone axis, evidencing the different image contrast dependence on the number of Co atoms vacancy within the same atomic column.

Figure S14 .Figure S15 .
Figure S14.(a) EXANES spectra of Se K-edge.(b) EXAFS oscillation extracted from K-edge spectra of the composites in k space of CoSe and v-CoSe.

Figure S16 .
Figure S16.Wavelet-transform contour plots of the EXAFS signal of v-CoSe, CoSe, and Co foil as reference.

Figure S21 .Figure S22 .
Figure S21.Se L1 XMCD spectra of CoSe and v-CoSe at 50 K under an applied field B = 6 T.

Figure S24 .
Figure S24.Partial density of states (PDOS) of Co at d-orbital for (a) CoSe and (b) v-CoSe.

Figure S25 .
Figure S25.(a) Charge density distribution of CoSe.(b) Charge density distribution of v-CoSe.(c) Differential charge density map, between CoSe and v-CoSe.To obtain the differential map, a Co atom and its associated charge density were removed from the CoSe charge density distribution.(d) 2D spin density map of CoSe.(e) 2D spin density map of v-CoSe.(f)3D spin density model of CoSe.(g) 3D spin density model of v-CoSe.

Figure 29 .
Figure 29.Top view of CoSe and v-CoSe spin density before and after Li2S6 adsorption.

Figure 30 .
Figure 30.(a)Top view and (b) side view of the spin density of CoSe and v-CoSe before and after Li2S4 adsorption.

Figure S33 .
Figure S33.Nyquist plots of the EIS spectra of the CoSe/S electrodes at (a) 2.8 V, (b) 2.6 V, (c) 2.4 V, (d) 2.2 V, (e) 2.0 V with the enlarged figure and (f) 1.7 V during discharging at different temperatures (300 K, 310 K, and 320 K).The raw impedance data and the fitted data are shown as symbols and lines, respectively.

Figure S34 .
Figure S34.Nyquist plots of the EIS spectra of the v-CoSe/S electrodes at (a) 2.8 V, (b) 2.6 V, (c) 2.4 V, (d) 2.2 V, (e) 2.0 V with the enlarged figure inserted, and (f) 1.7 V during discharging at different temperatures (300 K, 310 K, and 320 K).The raw impedance data and the fitted data are shown as symbols and lines, respectively.

Figure S35 .
Figure S35.SEM images and EDX elemental maps of lithium sulfide particles deposited on a carbon paper electrode of (a) CoSe/S and (b) v-CoSe/S.

Figure S36 .
Figure S36.(a) PDOS of bonding state orbital hybridization for Li2S4 of CoSe and v-CoSe, (b) Enlarged view of the blue region of the panel (a).

Figure S37 .
Figure S37.Linear sweep voltammetry (LSV) curves of (a) reduction I and (b) reduction II processes from the Figure 5a pink regions.

Figure S38 .
Figure S38.CV curves of (a) CoS/S and (b) v-CoSe/S at different scanning rates.

Figure S39 .
Figure S39.Galvanostatic charge and discharge curves of (a) CoSe/S and (b) v-CoSe/S at different current densities.

Figure S40 .
Figure S40.(a) Fitted Rct values at different discharge voltage during the first discharge process at 300 K. (b) EIS spectra and fitting results after 1500 cycles of CoSe and v-CoSe.

Figure S41 .
Figure S41.XRD patterns of CoSe/S and v-CoSe/S electrodes before and after cycling.

Figure S42 .
Figure S42.High-resolution Co 2p XPS spectra of CoSe/S and v-CoSe/S electrodes before and after cycling.

Figure S43 .
Figure S43.High-resolution Se 3d XPS spectra of CoSe/S and v-CoSe/S electrodes before and after cycling.

Table S3 .
Atomic fraction from EDS spectra of CoSe and v-CoSe.

Table S4 .
Overpotential and square area of symmetrical cells CV curves of different catalysts.

Table S5 .
Structural parameters extracted from the Se K-edge EXAFS fitting.