Understanding the electrochemical reaction mechanism to achieve excellent performance of the conversion-alloying Zn 2 SnO 4 anode for Li-ion batteries

New insights into the (de-)lithiation mechanism of the Zn2SnO4 conversion-alloying anode material obtained by an industry-scalable method allowed preparing fully operational anodes for Li-ion full-cells through controlling the anode's working range.


Supplementary Note 1: Charge transfer during (de-)lithiation studied by EIS measurements and DRT analysis
To get an insight into the influence of different lithiation levels on charge transfer EIS measurements were conducted.The voltage profile for the Zn2SnO4-based half-cell is presented in Figure S2a.The spectra were collected at the marked points.All the measured EIS data are presented collectively in Figures S2b-d.To analyze the data, the distribution of relaxation times (DRT) approach was used, as described in the Experimental section.It allows resolving various processes from the total impedance, especially the ones characterized by similar time constants, which otherwise would be hard to analyze unambiguously using the conventional equivalent circuit fitting.The area under each observed peak corresponds to the resistance related to a given polarization effect.The results of DRT calculations are presented in Figure S2e-g, for the 1 st lithiation, 1 st delithiation, and 2 nd lithiation, respectively.The data are analyzed in the 1-10 6 Hz frequency range, to exclude the Warburg diffusion part, as the DRT method can be used only for the spectra converging toward the real axis at the low frequency part.The character of the DRT spectra is analogous to a high-entropy CAM spinel from our previous study 29 , which appears to be a typical character for CAMs.0][31] The high-frequency peak (ca. 10 5 Hz), not varying significantly during cycling, is labeled as Pcontact and corresponds to the contact impedance.The two intermediatefrequency peaks (ca. 10 3 -10 5 Hz), labeled as PSEI, are assigned to the polarization related to the SEI layer and charge transfer in the Li counter electrode (higher frequency peak, less variable with the lithiation level), and to the polarization effects from the SEI layer formed on the investigated electrode (peak centered at ca. 5‧10 3 Hz).Similarly as for the high-entropy CAM, the SEI resistance related to the Zn2SnO4-based electrode decreases in the 1 st cycle, as the film gets stabilized.It remains rather stable during the 2 nd lithiation, except for the last point (full lithiation) where the SEI resistance noticeably increases.It can be interpreted as SEI cracking during deep lithiation due to big volume changes, with a new, thicker film formation at the fresh surfaces exposed to the electrolyte.The peaks located in the range between 1 and 10 3 Hz, with a couple of maxima (the number depends on the lithiation level), are assigned to the polarization connected with charge transfer (PCT) phenomena in the studied electrode.During the 1 st lithiation, the properties of the pristine Zn2SnO4 change significantly, which is in line with the observed in operando XRD decomposition of the initial spinel structure (points 1-2).Then, upon further conversion and alloying processes, the charge transfer (CT) resistance continuously decreases.In the 1 st delithiation and 2 nd lithiation, two distinct regions related to the CT polarization can be named, with the clearly visible separation points between 12-14 and 18-20.This confirms the conversion-alloying character of the electrode, with the regions characterized by a higher resistance corresponding to the regime of the voltage profiles assigned to the conversion reaction (worse transport properties of oxide phase compared to metallic counterparts).The separation point between conversion and alloying is in agreement with the phase analysis based on operando XRD.It is worth noting that the resistance increase during 1 st delithiation when moving to the conversion regime is much more pronounced than for the previously studied high-entropy spinel 29 , which can be correlated with the unfavorable high polarization related to the Li6ZnO4/Li4ZnO3 phases formation.Particular attention should be drawn to the differences in charge transfer (PCT) at the points on the voltage profiles related to the proposed in this work formation of α-Sn and β-Sn (point 12, 1 st delithiation, and point 24, 2 nd lithiation, respectively, according to the phase composition study by operando XRD, Figure 2a-d), with the EIS data and corresponding DRT presented in Figure S2h-i.First of all, during the 1 st delithiation, when the Sn phase starts to emerge (points 9-11), the CT resistance decreases.Therefore, even if the semiconducting α-Sn is formed 32,33 , it is not a predominant factor influencing CT in this case.On the other hand, when the DRT spectra are directly compared for different Sn allotropes (according to operando XRD, while the remaining phases are similar between these two points), the total resistance in the whole frequency range is lower for the 2 nd lithiation (including SEI and possible electronic transfer processes occurring at higher frequencies than PCT).As α-Sn is semiconducting and β-Sn has a metallic character, this indirectly supports the claim about the formation of two different polymorphs of Sn in the 1 st and 2 nd discharge/charge cycles.Table S2.Phases assigned in the SAED pattern with the measured and expected d-spacings for different phases.The differences between measured and expected values are below 4.3%, which lies within the experimental errors, especially considering the nanoscale dimensions of the crystallites in the sample. 33The following phases were assigned: LZO = Li6ZnO4 (PDF 00-040-0202) / Li4ZnO3; LiZn (PDF 04-003-6868); β-Sn (PDF 00-004-0673); Li2O (PDF 00-012-0254); Zn (PDF 00-004-0831).The full-cell with the anode prelithiated to 0.8 V was disassembled after 300 cycles, used to make a fresh half-cell, charged to 4.3 V and discharged to 2.75 V under a current of 0.05C.

NCM111
Zn2SnO4 120 c (300 cycles) 367 this work *The capacity based on anode mass was marked "a", the capacity based on cathode mass was marked "c".TM-HEO: transition metal-based high-entropy oxide

Figure S1 .
Figure S1.Particle size distribution for the Zn2SnO4 powder used to prepare electrodes, measured using DLS method.The distribution shows two maxima: the less intense at ca. 0.7 µm, and the more intense at ca. 6 µm.Sizes above 10 µm likely correspond to the agglomerated primary particles.

Figure S2 .
Figure S2.Charge transfer study at different (de-)lithiation stages for the Zn2SnO4 anode (CMC/SBR binder and LiPF6 in EC:DEC electrolyte).(a) Normalized capacity voltage profiles for the 1 st lithiation, 1 st delithiation, and 2 nd lithiation in the voltage range of 0.01-2.5 V under a specific current of 50 mA g -1 .The marked points correspond to the stages when EIS measurements were conducted; points 8 and 26 are overlapped.(b-d) EIS spectra measured at the marked points during the 1 st lithiation (b), 1 st delithiation (c), and 2 nd lithiation (d) with the insets showing zoomed regions for clarity.(e-g) Distribution function of relaxation times as a function of frequency for the 1 st lithiation (e), 1 st delithiation (f), and 2 nd lithiation (g) with the insets showing zoomed region corresponding to peaks assigned to the charge transfer polarization; Pcontact -peaks from the contact impedance; PSEI -peaks related to the polarization from SEI layer and Li electrode; PCT -peaks from the electrode's charge transfer polarization.(h-i) Selected EIS spectra measured at points corresponding to the formation of α-Sn (point 12) and β-Sn (point 24), according to the operando XRD study (h) with the corresponding DRT (i).

Figure S3 .
Figure S3.Analysis of the pseudo-capacitive influence on the Li-storage for the Zn2SnO4-based anode.(a) CV collected in the 0.01-2.5 V range with 3 cycles for each scan rate: 0.05, 0.1, 0.5, 1.0 mV s -1 .For clarity, the 2 nd cycle for each scan rate is presented.The peaks taken for the analysis are marked.(b) Log(peak current) -log (scan rate) plot used to determine b-values according to the relationship between the scan rate and peak current i = av b (i -peak current; v -scan rate; a, b -parameters from a linear fitting of log-log plot) 34,35 .The b-value equal to 0.5 indicates the diffusion controlled storage, while b equal to 1 indicates the pseudo-capacitive controlled storage.The values obtained for the analyzed peaks indicate mixed, hybrid pseudo-capacitive and diffusion storage, where the alloying reaction is shifted more toward diffusion domination and the conversion reaction more toward pseudo-capacitive domination.

Figure S4 .
Figure S4.Selected nanoparticle showing a crystallite of the Li6ZnO4 phase.The sample corresponds to the XRD pattern at x = 36 % (0.52 V) in Figure 2c.(a) HR-TEM image with the FTT (b) with the selected peaks indexed as the tetragonal Li6ZnO4 phase.The remaining spots correspond to β-Sn phase (as visible in Figure 3d) and are not indexed here for clarity.(c) Inverse FTT for the selected spots marked with circles in (b).

Figure S5 .
Figure S5.Comparison of the XRD pattern for the electrodes measured in the ex-situ and operando modes.The XRD pattern for the operando electrode corresponds to Figure 2c, x = 36%; 0.52 V.The patterns are in very good agreement, except for the Be-related and cellcasing-related additional peaks for the operando measurement.For the ex-situ sample all the phases could be assigned in the same way.

Figure S6 .
Figure S6.GDC curves for Zn2SnO4 half-cells tested with different binder/electrolyte setups.(a,b) Data corresponding to the long-term stability for 100 cycles presented in Figure 4b.(c) Data corresponding to the rate capability tests from Figure 4c for the best performing cell with CMC/SBR binder and LiPF6 in EC:DEC:FEC:VC electrolyte.

Figure S7 .
Figure S7.SEI formation studies.(a) Comparison of the 1 st cycle cyclic voltammetry curves for Zn2SnO4 electrode with CMC/SBR and PVDF binders and with different electrolytes: 1M LiPF6 in EC:DEC and 1M LiPF6 EC:DEC:FEC:VC.(b,c) Ex-situ XAS studies of O K-edge for Zn2SnO4 anode with CMC/SBR binder in half-cells with different electrolytes: (a) measured in TEY mode (surface information) and (b) measured in PFY mode (bulk information, up to ca. 100 nm).

Figure S8 .
Figure S8.High magnification SEM image for the pristine electrode with CMC/SBR binder in BSE mode.The pristine particles are not agglomerated but distributed within carbon-binder matrix (the particle size is significantly higher than for the agglomerated nanoparticles visible in Figure4e).

Figure S9 .
Figure S9.Long-term stability tests for the full-cells with anodes prelithiated to different voltages in the 2.25-4.2V range under the current of 0.5C for 200 cycles.

Figure S10 .
Figure S10.Assessment of the performance degradation in the full-cell from the cathode.The full-cell with the anode prelithiated to 0.8 V was disassembled after 300 cycles, used to make a fresh half-cell, charged to 4.3 V and discharged to 2.75 V under a current of 0.05C.
Electronic Supplementary Material (ESI) for Journal of Materials Chemistry A. This journal is © The Royal Society of Chemistry 2023 3

Table S3 .
Full-cell performance comparison with different electrode materials.