Li2O:Li–Mn–O Disordered Rock‐Salt Nanocomposites as Cathode Prelithiation Additives for High‐Energy Density Li‐Ion Batteries

The irreversible loss of lithium from the cathode material during the first cycles of rechargeable Li‐ion batteries notably reduces the overall cell capacity. Here, a new family of sacrificial cathode additives based on Li2O:Li2/3Mn1/3O5/6 composites synthesized by mechanochemical alloying is reported. These nanocomposites display record (but irreversible) capacities within the Li–Mn–O systems studied, of up to 1157 mAh g−1, which represents an increase of over 300% of the originally reported capacity in Li2/3Mn1/3O5/6 disordered rock salts. Such a high irreversible capacity is achieved by the reaction between Li2O and Li2/3Mn1/3O5/6 during the first charge, where electrochemically active Li2O acts as a Li+ donor. A 13% increase of the LiFePO4 and LiCoO2 first charge gravimetric capacities is demonstrated by the addition of only 2 wt% of the nanosized composite in the cathode mixture. This result shows the great potential of these newly discovered sacrificial additives to counteract initial losses of Li+ ions and improve battery performance.

A recurring issue limiting the performance of Li-ion batteries is the formation of a solid electrolyte interface (SEI) during the first battery charge. 7-20% of lithium from the cathode material is irreversibly bound at the surface of graphite anodes, [1] and it can be as high as 30% for Si, [2] in order to form this passivation layer, which results in a loss of capacity. Several prelithiation routes incorporating sacrificial additives into the cell have been explored to alleviate irreversible capacity losses during the first charge. On the whole, sacrificial materials act as Li + donors and offset the loss of active Li + ions during the SEI formation.
weight. In spite of the promising improved potential ranges 3-4.5 V versus Li + /Li, the uncontrolled evolution of gases could potentially damage the battery.
Several compounds with high initial charge capacity, where nonreversible applications may be found (e.g., Li 2 Mn 2 O 4 , [11] Li 2 NiO 2 , [12] Li 6 CoO 4 , [13] and Li 2 CuO 2 [14] ), have shown promising results as sacrificial additives and although they offer effective compensation of Li + loss during the first cycle, these materials show low specific capacities of <300 mAh g −1 .
Reversible electrochemical conversion reactions between lithium and transition metal oxides have awoken interest as both positive and negative electrodes in Li-ion batteries, [15,16] and recently, Li 2 O:M, [17] Li 2 S:M, [18] and LiF:M [19] nanocomposites have been presented as attractive cathode prelithiation additives, able to store more than 4 times the theoretical specific capacity of existing cathodes ≈500-930 mAh g −1 . The best performance is given by Li 2 O:M (e.g., 724, 799, and 935 mAh g −1 for M = Co, Fe, and Mn, respectively), increasing the overall capacity of a LiFePO 4 (LFP) cathode with the Li 2 O:Co additive by 11%. [17] Such high capacity of sacrificial cathodes have only been improved in Li 3 N reaching 1399 mAh g −1 during the initial charge to 4.2 V. [20] However, Li 3 N is highly reactive and incompatible with most widely used solvents in lithium-ion batteries' manufacture.
Recently, we reported the outstanding charge capacity of 350 mAh g −1 in nanostructured Li 4 Mn 2 O 5 [21] cathode material with a strongly disordered and nonstoichiometric MnO-type rock-salt (RS) structure, where 2/3 of Mn are substituted by Li and accommodating 1/6 oxygen vacancies. Thus, the chemical formula of Li 4 Mn 2 O 5 is henceforth given as Li 2/3 Mn 1/3 O 5/6 . Li 2/3 Mn 1/3 O 5/6 is part of an emerging family of Li-rich cathode materials based on disordered RS structures, displaying higher capacities (200-350 mAh g −1 ) than Li-rich layered oxides (150-250 mAh g −1 ). Besides their superior capacities, disordered RS offers a versatile chemical playground and several disordered RS compositions have been reported over the last few years that can present multivalent transition metals, mixed O 2− /F − anions, or oxygen redox. [22][23][24][25][26][27] Detailed compositional, structural, and electrochemical characterizations showed that the presence of 7 mol% excess of Li 2 O in the 0.07Li 2 O:0.93Li 2/3 Mn 1/3 O 5/6 composite [28] (previously reported as single phase Li 4 Mn 2 O 5 ) increased the capacity of the composite mixture by 100 mAh g −1 , while the capacity of single phase Li 2/3 Mn 1/3 O 5/6 was reduced to 250 mA h g −1 . In this work, we demonstrate that outstanding first charge capacities (>1150 mAh g −1 ) can be achieved by the increase of the Li 2 O content in the Li 2 O:Li 2/3 Mn 1/3 O 5/6 composite, thanks to the electrochemical activation of Li 2 O acting as a Li + donor to the RS.
We report for the first time 35 and 55 mol% Li 2 O-rich composites synthesized by mechanochemical routes [28] with exceptional 898 and 1157 mAh g −1 first charge capacities (see  and only the capacity of the active RS component is retained over the following cycles. Further increasing the Li 2 O proportion did not improve the capacity further in 0.75Li 2 O:0.25Li 2/3 Mn 1/3 O 5/6 which displayed 914 mAh g −1 first charge capacity (see Figure S1 in the Supporting Information), where only 55% of Li 2 O reacted during the first charge. Although no intermediate compositions between 55% and 75% of Li 2 O were studied in this work, an optimal capacity value could be found within this compositional range. The miscibility of Li 2 O and Li 2/3−x Mn 1/3 O 5/6 phases at the nanoscale was demonstrated by high-resolution transmission electron microscopy images of the 0.55Li 2 O:0.45Li 2/3−x Mn 1/3 O 5/6 and carbon black (30 wt%) composite in Figure S2 (Supporting Information), which shows an agglomeration of ball-shaped nanoparticles, with no clear cleavage planes, and no obvious surface layers of different compositions.
The reaction mechanisms between Li 2 O and Li 2/3 Mn 1/3 O 5/6 were studied by in situ total scattering. 7% and 35% Li 2 O composites were studied over the course of one and two charge/ discharge cycles in Figure 2 while for the 55 mol% Li 2 O-richest composite, only 20% of the first charge was recorded (see Figures S3 and S4 in the Supporting Information for in situ total scattering and ex situ diffraction and X-ray absorption near-edge spectroscopy (XANES) data for the pristine and charged at 4.5 V samples). Figure 2 shows a solid solution behavior of the RS with a continuous evolution of the lattice parameters over the 1.5-4.5 V potential window, in agreement with previous in situ XANES [29] and ex situ total scattering [28] studies. The lattice parameter evolution, quantified by the sequential Rietveld refinement of over 100 in situ data sets, mimics the shape of the electrochemical curve. A steeper change in the lattice parameter occurs at lower potential values 2-3.5 V, following by a more gradual increase in the range 3.5-4.5 V (see Figures S6-S8 in the Supporting Information for the evolution of all sequentially refined parameters).
The nonlinear evolution of Mn-Mn interatomic distances with Li concentration in 0.07Li 2 O:0.93Li 2/3 Mn 1/3 O 5/6 previously characterized by extended X-ray absorption fine structure, [29] is in good agreement with the refined lattice parameters in Figure 2a.  Figure 2b, disappear gradually until they are no longer observed after the first charge (see Figure S7 in the Supporting Information for the refined phase wt%), and no traces of other crystalline or amorphous secondary phases could be detected. After the first charge, Li 2/3−x Mn 1/3 O 5/6 cycles reversibly between charged and discharged states with an exchange of ≈0.4 Li per formula unit without the further participation of Li 2 O. In contrast to Li 2 O:MO nanocomposites with 100% capacity retention over several cycles, [15,30] the participation of Li 2 O in Li 2 O:Li 2/3−x Mn 1/3 O 5/6 is irreversible, which explains the large irreversible capacities in Figure 1f.
The local structural evolution of the RS phase remaining after the initial charge is investigated in more detail in Figure 2c, where the effects of the lattice parameter changes were removed by multiplying the r-scale by the ratio of the lattice parameters determined through Rietveld refinement. The narrow Mn-Mn distributions centered around the expected values for the average RS structure indicate that the Mn framework is well ordered, while the broader, asymmetric, and shifted Mn-O distributions indicate a high degree of disorder within the oxygen site (since Mn is well ordered, the broadness of these peaks could only be ascribed to disorder within the oxygen sites). Thus, Li 2/3−x Mn 1/3 O 5/6 is able to accommodate varying concentrations of lithium, thanks to the breathing of the cubic Mn framework that isotropically contracts and expands to extract and incorporate lithium, accompanied by displacements of oxygen atoms. It is worth noting that the evolution of the RS phase is identical to all composites, whichever the initial concentration of During the experiments in Figure 2 and Figure S3 (Supporting Information) performed in transparent cells made of quartz glass, no significant bubbling of electrolyte was observed that could result from an uncontrolled evolution of O 2 (g).Thus, if O 2 release occurs, it must be formed at a slow rate through the course of the first charge over a wide potential window lapsing several hours.
As for O 2 •− , the formation of these more reactive radicals is frequently associated to decomposition reactions with the carbonate solvent, [31] leading to the formation of CO 2 , H 2 O, and crystalline Li 2 CO 3 or Li 2 O 2 decomposition products (among other phases). However, neither crystalline nor amorphous secondary phases were detected by total scattering (see Figure S9 in the Supporting Information). Given that non-negligible amounts of Li 2 CO 3   The electrochemical curves of the pristine cathodes and with 0.55Li 2 O:0.45Li 2/3−x Mn 1/3 O 5/6 additive appear more different at 3.5-4 V due to a more predominant contribution to the charge capacity from the additive at the higher potential. The addition of 2 wt% of 0.55Li 2 O:0.45Li 2/3−x Mn 1/3 O 5/6 resulted in a 13% increase of the first charge capacity, and matching discharge capacities to that of pristine LFP and LiCoO 2 . Thus, while increasing the initial charge capacity, the sacrificial additive did not interfere with the electrochemical performance of the pristine cathode materials (see Figure S10 in the Supporting Information for cycling stability over 7-10 cycles). Note that the large initial charge capacity of 0.55Li 2 O:0.45Li 2/3−x Mn 1/3 O 5/6 allows for the use of such small amount of additive of only 2 wt% (vs more routinely used 5-10 wt% [11][12][13]19] to compensate for similar capacity losses). Moreover, ≈55 mol% of the sacrificial is consumed after the first charge and only <1 wt% remains in the cathode, and the smaller volumes of released gas versus sacrificial salts [9] mitigate potentially detrimental effects related to gas evolution during battery cycling.
In summary, we propose a design principle for cathode prelithiation to compensate the first-cycle Li loss in Li-ion batteries based on nanoscale mixtures of Li 2/3−x Mn 1/3 O 5/6 and Li 2 O. The high prelithiation efficacy demonstrated exploits the irreversible electrochemical activation of Li 2 O during the first charge. With such a cathode prelithiation additive, the first charge capacity of LFP and LiCoO 2 has been improved by 13%, while the subsequent discharge capacity matches that of the pristine cathode materials. Due to their low cost, ease of preparation, and potential compatibility with industrial battery fabrication, Li 2 O:Li 2/3−x Mn 1/3 O 5/6 nanocomposites are highly promising additives for the prelithiation of cathode materials. The reported pretreatment could be applied to other Li-cathode materials, and seems extrapolable for the presodiation of Na-ion batteries based on initial tests on Na 2 O:Na-Mn-O composites that showed an analogous irreversible electrochemical activation of Na 2 O with improved first charge capacities.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.