image: The degradation of Li metal batteries is primarily attributed to active Li loss, encompassing isolated Li, also known as “dead Li”, and solid electrolyte interphase (SEI). Under practical lean electrolyte conditions, the formation of dead Li occurs through two distinct mechanisms: either via disconnection from the electron percolation network, traditionally identified as electronically isolated Li (E-iLi) due to SEI encapsulation, or through disconnection from the Li-ion percolation network as a consequence of electrolyte dry-out, termed as ionically isolated Li (I-iLi).
Credit: ©Science China Press
Batteries that exploit the Li plating/stripping electrochemistry of Li metal anode harbor the potential to attain the ambitious target of 600 Wh kg-1 at the cell level, which could respond to the burgeoning energy requirements of electric vehicles, grid-scale energy storage, and even battery-powered aviation. However, this energy-maximizing architecture suffers from aggravated Li loss and shortened lifespan due to inferior reversibility/stability of Li growth. During repeated Li plating/stripping process, the Li metal anode experiences cyclic and substantial volume expansion/contraction. This phenomenon is particularly pronounced during the operation of high-energy Li metal batteries with high Li capacities. Throughout this process, undesired (electro)chemical reactions continuously consume limited active Li and electrolyte, generating solid electrolyte interphase (SEI-Li) and electronically isolated dead Li (E-iLi). As a result of the persistent consumption of electrolyte and the formation of a loose electrode structure, the wetting state of the electrode progressively deteriorates. This degradation leads to the inevitable occurrence of the electrolyte dry-out phenomenon at the anode side upon cycling. This disrupts the Li-ion conduction pathways, resulting in the formation of ionically isolated Li (I-iLi), an alternative form of dead Li that serves as a crucial complement to E-iLi, further exacerbating capacity degradation. After electrolyte refilling, the I-iLi can be reactivated due to the reconnected Li-ion conduction pathway, whereas the E-iLi remains trapped by SEI aggregation and isolated from the electron percolation network. Therefore, to advance the practical application of Li metal batteries, it is imperative to conduct a quantitative analysis of the I-iLi content, along with the combined content of SEI-Li and E-iLi.
To confirm the presence of I-iLi and quantify its content, an electrolyte-dependent capacity recovery experiment was executed. This experiment utilized a high-energy anode-free pouch cell configuration and electrochemical performance measurement, involving the reinjection of an abundant electrolyte (6 g Ah-1) when the cell attained a 50% state of health (SOH) during cycling at 0.3 C. The anode-free architecture employed an NMC (4 mAh cm-2) cathode and a bare Cu foil anode, with an electrolyte dosage of 1.6 g Ah-1, delivering a cell-level energy density of 506 Wh kg-1. Within this context, the cell underwent aggressive capacity decline during cycling and dropped to 50% SOH after merely 41 cycles, accompanied by rapid electrode/electrolyte interface deterioration. The Rs value, indexing solution resistance, increased over threefold from 0.5 Ω at the 5th cycle to 2.2 Ω at the 41st cycle, signifying severe electrolyte decomposition and weakened solid-liquid interfacial contact at the anode. Subsequent to electrolyte reinjection at 50% SOH, a remarkable capacity recovery of 17.6 mAh was observed at the 42nd cycle, responding to 15% of the total capacity loss. Meanwhile, the solution impedance decreased to approximately 0.5 Ω by the 45th cycle, indicating a restored wetting state of the electrode. These recovery phenomena firmly corroborated the existence of I-iLi and its reactivation through electrolyte-mediated reconnection of the ion conduction pathway.
Enlightened by the profound impact of stack pressure in hindering E-iLi formation, the evolution of the I-iLi content and its effect on electrochemical performance were investigated under external pressures ranging from 0.1 to 1 MPa. The 240 mAh-level NMC||Cu pouch cells tested at 0.3 C under 0.1, 0.5, and 1 MPa showed different cycling stability, with their capacities decaying to 50% of the initial values after 14, 41, and 71 cycles, respectively. The upward trend of lifespan and Coulombic efficiency (CE) in tandem with increasing external pressure indicates significantly alleviated Li loss and battery degradation through effective pressure regulation. Electrolyte refilling was performed for these cells after cycling to 50% SOH, and the capacity recovery of these cells decreased from 30.4 mAh to 1.2 mAh as the mechanical restriction elevated from 0.1 to 1 MPa. Meanwhile, the cell cycled under 0.1 MPa showed the highest CE of 116% in the first cycle following electrolyte refilling, which was 30.1% higher than that observed before electrolyte refilling. Similarly, the cell cycled under 0.5 MPa showed a 17.7% increase in CE, reaching over 112% in the first cycle following electrolyte refilling. These elevated values of CE above 100% confirm that I-iLi generation is an important contributor to capacity decay under practical conditions, and can be rejuvenated by reconnecting the ion channel. By comparison, the variation in CE before and after electrolyte refilling was reduced to 2.5% for the cells operated under 1 MPa, cohering with the minimal capacity recovery, and manifesting significantly mitigated electrolyte decay. This resulted in prolonged cycle life and a decreased proportion of I-iLi to total Li loss from 25% at 0.1 MPa to 1% at 1 MPa, indicating well-preserved ion/electron percolation networks enabled by pressure modulation during repeated Li plating/stripping.
To explore the underlying mechanism behind the pressure-dependent battery performance, a parameter of electrolyte infilling ratio (EFR) was introduced to directly evaluate the health of the electrode/electrolyte interface. EFR, which depends on the Li deposit porosity and electrolyte uptake, can be quantitatively represented by the ratio of wetted pore volume to total pore volume within the electrode. It reflects the integrated outcome of the Li plating/stripping structure and the extent of electrolyte degradation, where a porous architecture and diminished electrolyte amount result in a low EFR value. After the initial charging at 0.3 C, Li deposits in the NMC||Cu pouch cells under 0.1 MPa reached a thickness of up to 104.8 μm (calculated based on statistical measurements taken at six different locations), five times thicker than that of dense metallic Li counterpart with same Li capacity (20 μm, 4 mAh cm-2). This substantial discrepancy is attributed to low-density Li plating, resulting in a high porosity of 81.1% and a low EFR value of 25.4%. These findings indicate a poorly wetted electrode interface, consistent with the pre-observed high solution resistance. Low-quality Li plating and inferior interfacial state deteriorate the subsequent cycling, leading to a significantly thickened Li deposit layer (172 μm), increased porosity (95.6%), and lowered EFR value (5.2%) after 30 cycles. Such a highly porous structure and poor electrode-electrolyte interfacial contact jointly destroyed the ion/electron conduction network, leading to rapid battery failure. Contrastingly, the thickness of the Li deposit (27.5 μm) following the initial charge under 1 MPa is analogous to that of a fully dense metallic Li foil possessing an identical Li capacity (21 μm, 4.2 mAh cm-2). This indicates high-density Li plating behavior, resulting in a low porosity of 24.7% and an advantageous interfacial state with a high EFR value of 87.8%. The compact electrode structure successfully hinders undesired reactions between active Li and electrolyte, and preserves efficient conduction for both ions and electrons, enabling enhanced electrochemical cycling stability. Subsequently, after undergoing 30 cycles at 1 MPa, the Li deposits (36.5 μm) exhibited a significantly diminished porosity level (53.3%) and an elevated EFR value (68.9%).
To highlight the significance of the EFR criterion in practical battery systems, ambitious parameters of zero-excess Li, high cathode loading of 4 mAh cm-2, and lean electrolyte dosage of 1.4 g Ah-1 were harnessed to fabricate a prototype 1.4 Ah NMC||Cu pouch cell with extremely high energy density. Under 0.1 MPa, the anode-free battery's capacity retention fell to 67% after just 40 cycles at 0.2 C, accompanied by increased voltage hysteresis and internal resistance. This decline is ascribed to rapid electrolyte degradation and the resultant low EFR environment, which exacerbates the accumulation of SEI and dead Li during battery operation. This observation coincides with the substantial I-iLi content and Li loss detected under 0.1 MPa. In sharp contrast, the cell cycled under 1 MPa afforded capacities of 1.49 and 1.45 Ah at 0.1 and 0.2 C, respectively, corresponding to exceptionally high cell-level energy densities of 551 and 530 Wh kg-1. Synchronously, the Ah-level NMC||Cu pouch cell exhibited stable cycling, retaining 70% of its capacity after 100 cycles at 0.2 C under these extremely aggressive conditions. As a result, an impressive energy density exceeding 500 Wh kg-1 was preserved in the cell after 30 cycles, outperforming the counterpart of 405 Wh kg-1 under 0.1 MPa. Meanwhile, the impedances of both the solution and the interface, along with the solid-liquid interfacial contact status, showed negligible degradation during operation, underscoring the crucial importance of maintaining a high EFR environment and minimizing I-iLi generation to enhance the stability of high-energy rechargeable battery systems. In this case, an extremely high cell-level energy density of 551 Wh kg-1 and stable cycling were simultaneously accomplished in Ah-level pouch cells, representing the best performance in reported anode-free battery technology to date.