Introduction
Rechargeable intercalated cells that convert between chemical and electrical energy through redox reactions will almost certainly undergo bulk structural and/or interface changes associated with ionic electron transfer. Therefore, the reversibility of the ionic electron transfer process and the evolution of the material structure determine the recharge capacity or cycle life of these batteries. Since ionic electron diffusion and structural evolution are primarily controlled by the intrinsic properties of the electrode material and the dynamic interfacial reactions between the electrolyte and the electrode material, maintaining structural integrity and interfacial stability is critical to achieving high-performance batteries.
Various types of surface coatings, material structure designs, and new electrolyte formulations and/or additives have been developed to control electrode electrolyte interactions. The underlying processes of the underlying mechanisms have also been extensively studied. However, research has focused mainly on surface phenomena, including the composition of the negative electrode "solid electrolyte interface", the "positive electrolyte interface", the surface transition metal dissolution, and the surface reconstruction layer (SRL) of layered transition metal oxides. The effect of electrolyte-electrode interactions on the breakdown of the cathode structure has been largely overlooked, not to mention the study of underlying molecular processes.
Early studies on protons Lithium-ion exchange has shown that protons from the electrolyte can bind to the body of layered lithium materials in the state of charge, but little is known about how the cathode host structure will evolve in successive cycles. Layered metal oxide cathodes, especially nickel-rich materials, have emerged as promising candidates for increasing the energy density of rechargeable alkaline ion batteriesTherefore, in order to improve performance, there is an urgent need to explain why the interface is highly unstable with the electrolyte. Recently, the interaction of nickel-rich cathodes with grain boundary electrolytes has been thought to be related to intergranular cracking in bulk spheres. It was also revealed that the nickel-rich cathode exhibited significantly different surface degradation rates in high-concentration electrolytes, local high-concentration electrolytes (LHCE), and conventional carbonate electrolytes.
Brief introduction of the results
Recently, researchers Biwei Xiao and Xiaolin Li of Pacific Northwest National Laboratory met with Perla B. of Texas A&M UniversityBalbuena et alUse nanio in electrolytes with different levels of proton productionAs a model system, a holistic picture of the effects of incorporation of protons is presented. The protonation of lattice oxygen stimulates the migration of transition metals to the alkaline layer and accelerates the phase transition of layered halite, resulting in the decomposition of the bulk structure and the formation of anisotropic surface reconstruction layers. A cathode that undergoes a vigorous protonation reaction acquires a porous structure corresponding to the decay of its multiple properties. This work reveals that the protonation-leading interaction between the electrolyte and the cathode can dominate the structural reversibility of the bulk cathode stability, an insight that provides clues for the development of future batteries. The study was published in the top international journal Advanced Materials with the title of "Protonation stimulates the layered to rock salt phase transition of ni-rich sodium cathodes".
Text guide
Electrochemical testing and cathode surface characterization
The Nanio2 used in this work is synthesized by a common solid-state reaction. The powder X-ray diffraction (XRD) profile in Figure S1A and the Rietveld refinement show that it is with C2M symmetry and 5Pure phase laminar material with 32 layer spacing. The results of scanning electron microscopy and high-angle annular darkfield scanning transmission electron microscopy (HAADF-STEM) showed that the size of the particles in the agglomeration was about 2 m, and the layer spacing was about 53 (Figure S1b-c). In 2 and 4The electrochemical performance of Nanio2 was tested between 2 V. Under similar conditions, LHCE [5.].1 M sodium bis(fluorosulfonyl)imide dissolved in dimethoxyethane (DME) diluted with bis(2,2,2-trifluoroethyl) ether (BTFE)] and carbonate-based electrolyte [1 M NAP6 dissolved in 1:1 vol% ethylene carbonate propylene carbonate (EC PC)] (denoted as EC PC. In almost all aspects of electrochemical testing, Nanio2 exhibits superior performance over EC PC in the LHCE, including specific capacity, cycling stability, and rate performance (Figures 1 and S2). nanio2 at 0LHCE at 1C (1C = 120 mA g) showed a higher performance than in EC PC electrolytes (152.)8 and 1089mAh g) with a much higher initial charge and discharge capacity (183., respectively1 and 1369mah/g）。In 0After 100 cycles at 3C, the specific discharge capacity of LHCE is about 113 mAh g (the capacity retention rate of the fourth cycle is about 91.)3%), which is almost 25-fold (volume retention for the fourth cycle is approximately 45 mAhg, 49%) (Figure 1A). A close comparison of the charging discharge curves (Figure 1B-C) shows that Nanio2 in LHCE exhibits a pair higher than 40v and about 3 during discharge8 V platform (Figure 1B), which is absent in the charge discharge curve obtained in the EC PC (Figure 1C). These platforms are assigned to O''3 to O""3 phase distortion and are highly reversible in LHCE, even after 100 cycles. Due to the fact that in nanio2||A similar pattern of electrochemical performance differences was observed in hard carbon whole cells (Figure S3), so it is reasonable to infer that the differences in performance in these electrolytes are due to the different interactions of the electrolyte with the Nanio2 cathode rather than the Na metal.
[Fig. 1] Electrochemical properties and surface characterization of Nanio2. (A) Half-cell cycling performance in LHCE and EC PC (first cycle at C10, subsequent cycle at C3, 1C = 120 mAg). (b-c) Charge-discharge curves after the first and 100th cycles in LHCE and EC PC. (d-e) Non-situ Ni L3 side Xas of the Nanio2 cathode in various states of charge and discharge during the first cycle. (F-G)NANIO2 cathode at open-circuit voltage (OCV), full charge at cycle 10, and NI L3 side XAS at full discharge. (H-I) Haadf-STEM image of SRL on Nanio2 after 10 cycles in LHCE. (J-K) Haadf-STEM image of SRL on Nanio2 after 10 cycles in ECPC.
The evolution of the surface structure of the cathode in different electrolytes was studied by systematic soft X-ray absorption spectroscopy (SXAS) and transmission electron microscopy characterization. Ex-situ Ni L3 edge Sxas collected in total electron yield (TeY) mode generate signals primarily from the surface at several nanoprobe depths in this energy range;Therefore, this method is helpful to study the redox behavior of Ni on the surface of Nanio2. Figures 1D-1e show the spectra obtained during the first cycle in various charge-discharge states. Of the two electrolytes, Nanio2 shows a strong peak I (Ni2+) and a wide peak II (Ni3+ 4+) at OCV, which typically increases during charging and decreases during discharge. Peak II has been increasing to 42 V, reflecting the continuous oxidation of surface Ni to Ni3+4+. The oxidation tendency in EC PC lasts only until the positive electrode is charged to 3After 6 V, almost no change in peak II was observed. The results show that the surface reactivity and redox reversibility of Nanio2 are highly dependent on the electrolyte. After 10 cycles, the difference in surface redox reversibility in these electrolytes became more significant. The NI L3-side SXAS spectra in Figure 1F-1G show the repeated oscillations of the peak II signal during electrode charging and discharging in the LHCE, but with very little variation in the EC PC. The different Ni reactivity shows that the surface Ni in the sample circulating in EC PC becomes almost inactive after 10 cycles. The HAADF-STEM images in Figures 1H-1K and S4 show the surface structure of the cathode after 10 cycles in LHCE and EC PC. The particles exhibited anisotropic formation of NIO rock salts in both electrolytes. The SRLs formed in LHCE and EC PC, despite having different thicknesses, are thinner and denser on the faces parallel to the layers than on the others. In LHCE, the thickness of the SRL is 3 nm along the layer and 10 nm across the layer. In ECPC, the SRL thickness increases to 8 nm perpendicular to the layer and 28 nm along the layer;This essentially exceeds the detection depth of the SXAS in TEY mode and is therefore consistent with subtle peak II signal changes during charging and discharging. The formation of this anisotropic SRL also inferred that controlling the morphology of the primary particles into a rod-like structure with a higher (003) plane favors cyclic stability.
Learn about electrolyte breakdown and surface protonation
To investigate how electrolyte reactivity leads to the formation of SRL, we performed computational studies of bulk LHCE and EC PC decomposition as well as electrolyte interface reactions on the surface of Nanio2. To simulate the effects of the charged cathode interface, many electrons were removed from the simulation cell to simulate the electron-deficient environmental surface in the body and near the cathode. **The scheme is shown in Figure S5, and the calculation details are shown in the supporting information. Figure S6 summarizes the reaction mechanism and charge evolution of bulk electrolyte oxidative decomposition. Figures S7 and S8 show the instantaneous charge analysis, and the energetic values are listed in Table S1. In general, the dissociation of DME and BTFE is the main decomposition pathway for LHCE, while EC PC prefers to perform proton transfer reactions.
[Figure 2] Electrolytes (LHCE and EC PC) in Nanio2 and Na0End configuration (top) and interfacial reaction (bottom) of 75NiO2 surface decomposition. （a-d）na1.0nio2（0 0 1）；(e-j) Na0 after 10 ps AIMD simulation in an electron-deficient environment75nio2 (10-1) surface. Color**: Sodium (surface), purple;Nickel, dark blue;Carbon, grey;Oxygen, red;Hydrogen, light pink;Fluorine, light blue;Sodium (from salt), green;Sulfur, yellow;Nitrogen, blue;Phosphorus, light purple;Helium, white. The green arrows in each simulation unit highlight the location of surface protonation.
We incorporated the Nanio2 cathode surface into our model to study the molecular processes of the cathode electrolyte interface reaction. Two facets, (0 0 1) and (1 0-1) were selected to represent the possible orientation of Na+ migration on the Nanio2 cathode. As shown in Figures S5C-5D, the (1 0-1) surface passes through an inclined Na+ moving channel, while the (0 0 1) surface is parallel to the Na layer, with the NiO2(O-Ni-O) unit exposed on top, which hinders Na+ migration. Figures 2, S9, and S10 illustrate the initial and final configurations and reaction mechanisms of the decomposition of LHCE and EC PC on the original and charged NanIO2 surfaces. Our interface simulations of the fully discharged or primitive Nanio2 surfaces show that the reactivity of LHCE and EC PC on the (0 0 1) surface (Figures 2A and 2C) is still higher than on the (1 0-1) surface (Figures S9b and S10b): no interface reaction is detected on the (1 0-1) surface (Figures S9C and S10C), while surface protonation is detected on the (0 0 1) surface (Figure S11). Deprotonation of DME is manifested by the loss of terminal hydrogen atoms to the surface O(OS) atoms, which results in the formation of hydroxyl groups (-OH) on the (0 0 1) surface, as shown in Figure 2b. When the BTFE is initially close to the surface, it undergoes deprotonation, or when it is in the middle of the analog cell, it undergoes dissociation. EC and PC follow a similar mechanism by transferring vinyl protons to the surface, as shown in Figure 2D. Compared to the proton transfer reaction (vinyl proton to carbonyl oxygen (OC) or fluorine anion) in the bulk EC PC electrolyte (Figure S6E), the formation of surface protonation indicates that the OS atom is a stronger oxidizing agent than the OC atom of carbonate;As a result, protons are more likely to be extracted by the oxide surface. Taking into account the effect of positive desorption during charging, partially desorption of Na0 was also studied75 NiO2(001) (Figs. 2e and 2i), Na075 NiO2 (1 0-1) (Figures S9e and S10E) and Na0Electrolyte decomposition on the surface of 5NiO2 (1 0-1) (Figures S9h and S10h).
When the surface is partially charged, the LHCE and EC PC are significantly more reactive, as evidenced by the identification of the deprotonation of the two electrolytes shown in Figures 2, S9, and S10. Since the dissolution of Na+ (different sodiumization states) promotes electrolyte deprotonation, it is necessary to compare Na10nio2 and na05NiO2(1-10) reaction energy of electrolyte deprotonation on the surface. Unlike AIMD, precise energetic calculations are performed in an electron-neutral environment (no electron removal in the simulated battery). Figure 3 shows the schematic diagram and Table S2 shows the detailed charge analysis details. As shown in Figure 3a, Na1All proton transfer reactions on the surface of 0nio2(1 0-1) are thermodynamically unfavorable, which explains why the proton transfer reactions on the surface of 0nio2 (1 0-1) are thermodynamically unfavorableNo deprotonation reaction was detected in the AIMD simulations of the 0nio2(1 0-1) surface.
However, when the surface is partially charged, the reactivity of the electrolyte deproton increases significantly, which is consistent with our AIMD simulation results that the electrolyte is more reactive on the deproton charged surface. Calculate na0The reaction energy of the 5NiO2(1 0-1) surface is negative (as shown in Figure 3B), PC>EC>DME>BTFE. Since EC PC is more reactive than LHCE, and given the results in Figure 1, there is also reason to believe that the lower reactivity of deprotonation behavior in LHCE is the reason why SRL formation is largely inhibited, resulting in a more stable cycle.
Therefore,Minimizing electrolyte deprotonation by using less reactive LHCE is an effective strategy to achieve better capacity retention for nickel-rich cathodes. In contrast, electrolyte deprotonation on the surface of Nanio2 (0 0 1) is thermodynamically favorable, however, as shown in Figures 3c and 3d, there is little difference in the energy values of the charging (deradiation) process. This can be explained by the change in the exposed layer (where the proton transfer reaction takes place) on the surface of Nanio2(1 0-1) and (0 0 1) during charging. o atoms are the only exposed layer on the surface of (0 0 1) and the deradiated layer can hardly change the structural composition of the exposed layer. In contrast, the exposed layer on the surface of Nanio2(1 0-1) is composed of Na, O, and Ni atoms, and when they are charged, Na atoms are removed from the exposed layer, leaving more exposed O atoms on the surface as protonation sites. This is very consistent with the HAADF-STEM results in Figure 1, which show that thinner SRLs are formed on the plane parallel to the layer, and the surface reactivity does not increase during charging.
【Fig. 3】Deprotonation reaction energy (in EV) of a single electrolyte molecule (from left to right: DME, BTFE, EC, PC) on (A) Nanio2(10-1);（b） na0.5nio2（1 0-1）；（c） nanio2（0 0 1）；（d） na0.5nio2(0 0 1) surface. Color ** Same as in Figure 2.
Finally, we calculated na05niO2(1 0-1) reaction barrier for electrolyte deprotonation on the surface to determine the kinetic favorability of the reaction. As shown in Figure S12, the deprotonation of DME is spontaneous, as shown by the negative energy value in state 1. BTFE deprotonation has 2The highest kinetic barrier of 58 EV, which suggests that BTFE deprotonation is kinetically disadvantageous. The increased stability of LHCE due to the contribution of BTFE to the main component of the LHCE solution (due to the large diluent ratio) confirms our observation of thinner SRLs formed in LHCE, as shown in Figure 1. The kinetic barrier of EC deprotonation is 082EV, PC deprotonation is spontaneous (negative energy value for each transition state), which means that PC is more reactive than EC.
Proton-induced NI reduction and migration
The above results of LHCE and EC PC decomposition show that the electrolyte, which is more stable to deprotonation, is less reactive on the surface of Nanio2. However, the underlying mechanisms underlying protonation-induced degradation of surface and bulk structures are still poorly understood. To elucidate them, we first study the oxidation state of Ni atoms during proton transfer using the example of deprotonation of dimethyl ether on the surface of Nanio2(1 0-1). Analyzing the oxidation state by a change in the local magnetic moment (Table S2), it was shown that the Ni atom (highlighted in yellow in Figure S13a) was reduced from the oxidation state of +3 to +2 when the proton transferred to its adjacent O atom, forming an OH group on the surface.
To verify this, we use partially desorbed na0A model was established for the 25NiO2(10-1) surface with some degree of surface protonation (H content = 0.).25), as shown in Figures 4a-4c. Since during the heat treatment, at temperatures above 243 °C, Nanio2 will undergo a wide phase transition from monoclinic (C2 m symmetry) to rhombohedral (R-3 m symmetry) structure, and the 25 C2M structure is used to study Na025h0.Structural transformation of 25NiO2(1 0-1) surface at low temperatures. As shown in Figure 4a, protons are observed to jump between the OS atoms when they are initially attached to the OS atoms on either side of the NiO2 cell (green circles are highlighted). Perform 1. at t = 450kAfter the AIMD simulation of 7 PS, some protons were transferred to the OS atoms of the adjacent NiO2 unit and the OS of the subsurface, which indicates the mobility of surface protons. At the same time, the migration of NI from its original position (transition metal layer) to the NA vacancy (de-radiated channel) was found (highlighted in blue triangle), as shown in Figure 4b. A similar Ni migration in Linio2 is thought to provide a nucleus for the phase transition of layered NiO2 to spin NiO2, which disrupts the active Li-site and causes the capacity to gradually decrease with cycling.
Therefore, there is reason to believe that protons can also induce similar phase transitions of Nanio2, as this phenomenon is not found on proton-free surfaces even at high temperatures at temperatures t = 670 K, as shown in Figure S14. When NI migrates, the NA+ migration pathway is blocked;As a result, the number of active Na+ ions is also reduced. Considering that protons are mobile, we set up another na025h0.25NiO2(10-1) surface model in which protons are attached to OS atoms on adjacent NiO2 cells, as shown in Figure S15. Ni migration was also observed after 1 ps of AIMD simulation at t = 450K;At the same time, the jump of protons to other O atoms is also observed, resulting in the formation of H-O-H bonds, in which O bonds with Ni atoms on the surface. When the simulated temperature is elevated to t=600K (Figure 4C), it is found that H2O is released from the surface by breaking the Ni-O bond, triggering structural degradation of Nanio2. It has been reported that at high temperatures of t = 600K, the bulk Nanio2 adopts R3M symmetry, and oxygen precipitates on the surface of R-3M Linio2 (0 1 2), and we also studied R-3M Na0Thermal stability of 25NiO2(1 2). As shown in Figure 4d, at t = 600k, only 0After an AIMD simulation of 5PS, the formation of O-O bonds and the release of O2 were observed. This indicates that Nanio2 is unstable when heated, and its structural degradation can be summarized by the following equation:
Therefore, we thinkProtonation leads to nanionanioRapid degradation of the surface of the cathode material. Because there are many other protons produced in addition to the vigorous decomposition of the electrolyte solvent**, and it is already known that protonation can occur on layered metal oxides in the body, we then investigated the effect of protons on the crystal structure of Nanio2. Figure 4E shows the bulk structure of Nanio2 during de-radiation and proton intercalation. The proton is bonded to the O atom and is close to the Na vacancy. The relationship between H content and the C-axis lattice constant is shown in Figure 4f. It is evident that as more and more protons are aligned with the Nanio2 crystal, the C lattice parameters decrease. With the deradiation process, it is expected that protonated Nanio2 will undergo lattice contraction. Once the lattice shrinks, the ion transfer capacity deteriorates further, which explains the more severe volume loss in the ECPC solution, as shown in Figure 1A, as ECPC involves more proton generation.
Fig. 4: Effect of protonation on structural degradation of Nanio2. (a) 1. at t = 450 kAfter 7 PS AIMD simulation, C2 M Na025h0.25NiO2 (1 0-1) surface proton jump between the O atoms. (d) 0. at t = 600 kAfter AIMD simulation of 5 PS, R-3M Na025NiO2 (0 1 2) O2 release from the surface. (E) Proton intercalation in a partially desorbed Nanio2 crystal structure** (C2 m symmetry). (f) Relationship between Na content (X) and C lattice parameters in desorption Na1-XNiO2 and protonated Na1-Xhynio2 (Y:H content) crystals. The color ** is the same as in Figure 2.
Protonation and bulk phase structure transition
On the first charge up to 4After 2 V, the Na1-XNIO2 sample was subjected to mass spectrometry coupled with in-situ heating of high-energy XRD. v in the same two electrolytes to verify the presence of lattice protons and their effects. Figure 5a shows two water release peaks;The peak at 100-200 °C corresponds to the embedded water molecules, and the second peak is centered at 250 °C and is recognized as reflecting the removal of lattice protons. (See the support information for a detailed discussion of water peaks.) Thermogravimetric analysis of layered Ni(OH)2 and NiOOH also showed that the temperature was in the same temperature range as these materials losing lattice protons in the form of water (Figure S16A-C).
The peaks of the EC PC samples were significantly stronger compared to the LHCE samples, which is also consistent with our simulations of electrolyte decomposition and NanIO2 protonation: the EC PC was deprotonation to a greater extent on both the bulk electrolyte and the cathode surface (Figure S6 as well as Figures 2 and 3). The possible hydrolysis of NAP6 will also produce more protons, resulting in a large number of protonation of the Nanio2 cathode. Since lattice oxygen, once protonated, can also be unstable, leading to potential oxygen loss behavior, oxygen release has been studied. Figure 5b shows the oxygen released after heating both samples. The initial oxygen release temperature of the sample in EC PC was higher than that of the sample in LHCE (210 °C vs.).200°C), which may be due to the fact that oxygen is lost in the EC PC at the beginning, mainly in the form of water, delaying the release of O2 gas. The oxygen release peak at around 300°C is associated with the phase transition, which will be discussed further later in the discussion on in-situ heating XRD. We also collected CO2 emissions, as shown in Figure S16D, the LHCE samples still showed less CO2 emissions, indicating a more structurally stable structure compared to the EC PC samples.
Figure S17 shows desorption to 4D-spacing of 2 V initial Na1-XNIO2. (003) The location of the peak is lower in EC PC than in LHCE, confirming Figure 4F, as the greater the number of protons in the structure, the smaller the D spacing along the C axis. Figures 5C and 5D show the contour plots of the in-situ heated XRD, and the line plot is shown in Figure S18. Although intercalated water was observed in the sample at room temperature, it was not from the electrolyte, as the in-situ XRD of Nanio2 (Figure S19) was previously cycled in situ 50 times in situ in LHCE at 7Above 0 does not show any bloat. During the preparation of samples for heating, trace amounts of embedded water are unavoidable. Karl Fischer titration showed that the water concentration in LHCE and EC PC was 24., respectively2ppm and 161ppm, which is considered to be a similar level of the average water content of the liquid electrolyte. The fact that similar Nanio2 electrodes exhibit different fading behavior in electrolytes with similar water content supports that the protonation effect comes primarily from electrolyte breakdown rather than trace H2O in air, electrodes, or electrolytes. As shown in Figures 5C and 5D, samples loaded in LHCE and EC PC showed a similar layered structure before heating and a similar halite structure after heating to >300°C. In both cases, the initial heating process also shows invariant lamellar phase diffraction.
However, several different phase transition behaviors were observed, suggesting that different degrees of protonation affect the structural degradation behavior upon heating. First, at a very low temperature of 145 °C in the EC PC sample, but at a temperature of about 200 °C in the LHCE sample (marked with a red arrow in Figures 5C-5D), the (111) layered peak intensity and the (04 0) peak of spinel-like NiO2 at 1The beginning of around 98. In the EC PC sample, neither lattice protons nor oxygen begin to release at such low temperatures (Figure 5A-B). Second, in the EC PC sample, the layered and spinel facies coexisted over a wide range until the temperature reached 200°C and the (111) layered peak disappeared. The LHCE samples show a rapid decrease in the peak intensity of the (111) layered at around 200°C, accompanied by the presence of spinel phases. Third, the transition from spinel facies to halite facies begins at about 300 °C, corresponding to the second oxygen release event tip in Figure 5b;Its relative diffraction intensity is significantly stronger than that of LHCE in EC PC.
[Fig. 5] (A) Mass spectrometry of H2O release (mass-to-charge ratio M Z=18) from Na1-XNiO2 filled with LHCE and EC PC when heated. (b) Mass spectrometry of O2 release (Mz 32) of Na1-XNIO2 filled with LHCE and EC PC when heated. (c) In-situ heating of Na1-XniO2 in LHCE for high-energy XRD. (d) Loading of EC PCs with in-situ heating of Na1-XNIO2 for high-energy XRD. (e) Schematic diagram showing the phase transition of charged Na1-XNIO2 when heated.
The Nanio2 cathode was characterized, which showed significantly different capacity retention rates after 100 cycles in the LHCE and EC PC electrolytes. Figure 6 shows the significant differences in SRL evolution and bulk structural degradation between the two electrolytes. Figures 6A-6D are STEM images and electron energy loss spectra taken on the surface of some Nanio2 particles circulating in LHCE and EC PC. Low-magnification STEM images and representative EES spectra are shown in Figure S20. The LHCE sample (Figure 6A) shows a much thinner SRL layer than the EC PC sample (Figure 6C). The EELS plots in Figures 6B and 6D show that the Ni2+ layer on the surface of the LHCE sample is about 20 nm thick, but there are hundreds of nanometers deep in the EC PC sample. After long-term cycling in LHCE, the anisotropic SRL feature remains. Figure S21 shows a typical STEM image with 2 nm SRL along the layer and 20 nm SRL perpendicular to the layer. Comparing the results of the original electrode and the electrode after 10 cycles, the SRL thickness along the layer varies little, but doubles in the direction perpendicular to the layer. For the evolution of bulk structures, the LHCE samples mainly show crack formation along the layer direction (Figure 6E). It resembles many layered lithium and sodium materials and can also be attributed to the slippage of the layers during ion deintercalation and reintercalation. HAADF-STEM is used to examine the crystal structure of particles. Carefully inspect the *** regions of the cracked particles using HAADF-STEM and selective region electron diffraction (SAED) (marked with yellow, red, and purple boxes in Figure 6E). Figures 6f-6h show the crystallites on the surface of the halite particles, the layered knots of the rock salt on the surface of the crack interface, and the regular layered structure in the block. The crystal structure is indexed in the insert SAED pattern. Intragranular cracking introduces new surfaces to further react with the electrolyte and is thought to be the primary capacity attenuation mechanism for the multilayer oxide cathode of lithium-ion batteries, producing a layer of rock salt of only 2 nm on the Nanio2 cathode circulating in the LHCE. Considering the good capacity retention shown in Figure 1, it does not cause significant capacity decay. EC PC samples experienced severe bulk structural disruption after 100 cycles. Figure S22 shows the formation of randomly oriented fractures and thick rock salt layers composed of discrete nanodomains.
It has been established that lattice oxygen oxide shows fingerprint signatures in O-K-MRIs, particularly by sharp signatures of 531EV (excitation energy) and 524EV (emission energy), while very weak or no signal is observed at the electrodes circulating in the LHCE and EC PC electrolytes. Nanio2 shows more tetravalent Ni in samples charged in LHCE (Figures 1D and 1F). It is thought that the porosity is related to proton-induced oxygen loss after long-term cycling, which leads to NI migration and halite phase development. In addition, water is produced in the process, which exacerbates the phase change. Extensive protonation of Nanio2 and subsequent lamellar structure destruction and particle pulverization are the main reasons for the low specific capacity and poor cycling stability in EC PC electrolytes. This is not uncommon;The doping of protons into lithium transitions, metal oxides, and the exchange of protons for lithium has been known for many years. However, no studies have observed bulk structural breakdown of layered lithium cathodes. In one respect, this is not the focus of previous studies. On the other hand, Nanio2 has more phase transitions and larger interlayer spacing than lithium cathodes, which facilitates the exchange of protons and NA at triangular prismatic positions. For layered lithium cathodes, the formation of a triangular prismatic position requires additional layer slip, which makes this less likely to happen. Still, once combined, these protons may still act as defect sites, triggering local structural changes, even though they are only likely to occur locally.
[Fig. 6] Structural characterization of long-term circulating Nanio2. (A-D) Haadf-STEM images of Nanio2 cycled 100 cycles in LHCE and EC PC and corresponding EELS mappings. (e) STEM images of particles with intragranular cracks at low magnification, and three different regions examined with HAADF-STEM and SAED. (f-h) Haadf-stem image showing the surface region of the latrite structure (yellow), the internal crack region showing the mixed lasalt layered structure (purple), and the main region showing the layered structure (red) (Inset: SAED pattern) (i) Haadfstem image of the typical region of Nanio2 circulating 100 times in the EC PC, showing the discrete halite nanostructure domain (some holes are marked with yellow dotted circles). (j) Low-magnification HAADF-STEM image showing the location of the 3D tomography reconstruction area in (K). (k) Electron tomography 3D surface mapping of pores in an EC PC sample.
Summary and outlook
In summary, the basic principle of protonation of nickel-rich sodium cathode and its influence on surface phase reconstruction and bulk structure breakdown are elucidated by using carbonate-based electrolytes and the Nanio2 model system in LHCE. Although traditional studies have shown thatProtonation leads to the dissolution of transition metals, and protons can be exchanged with lithium ions in the lithium cathode, but our study clearly shows that its role is greatly underestimated for the nickel-rich sodium cathode
It also calls for more detailed studies of proton-related structural transitions in layered materials and engineering of nickel-rich cathodes and electrolytes for better cycling performance.
References
b. xiao, y. zheng, m. song, x. liu, g.-h. lee, f. omenya, x. yang, m. h. engelhard, d. reed, w. yang, k. amine, g.-l. xu, p. b. balbuena, x. li, protonation stimulates the layered to rock salt phase transition of ni-rich sodium cathodes. adv. mater. 2023, 2308380.