Most reactions in lithium-ion batteries occur at the solid/liquid interface, therefore the interfacial stability of the positive and negative electrodes determines the cycle stability of the battery. Temperature has a decisive influence on the reaction rate, and thus also significantly affects the electrode interfacial state.
Recently, Nicolas Gauthier (first author), Cecile Courreges (corresponding author), Herve Martinez (corresponding author), and others from the French National Centre for Scientific Research (CNRS) studied the interfacial characteristics of LiMn2O4/Li4Ti5O12 lithium batteries after 100 cycles at 25, 40, and 60 °C. The test results showed that the thickness of the interfacial film at the electrode interface increased with the increase of the cycling temperature, and a small amount of metallic Mn was also detected on the LTO surface after high-temperature cycling.
The electrode formulation used in the experiment consisted of 93% active material (LMO, LTO), 4% carbon black, and 3% 5130 adhesive, which were fully dispersed and coated onto aluminum foil. The coating amount for the positive electrode was 17 mg/cm², and for the negative electrode, it was 9 mg/cm². The theoretical capacity of the LTO negative electrode was less than that of the LMO positive electrode; therefore, this battery was negatively limited, allowing for a better reflection of the characteristics of LTO. The porosity of the rolled LTO electrode was 45%, and that of the LMO electrode was 40%. A full cell was assembled using coin cells and cycled at 25, 40, and 60°C.
Figures a, b, and c below show the charge-discharge curves of the LMO/LTO battery at different temperatures for the 1st, 99th, and 100th cycles, respectively. Figure a shows that during discharge, the charging capacity at 25℃ is 173 mAh/g, and at 40℃ it is 172 mAh/g, which is close to the theoretical value of 175 mAh/g. At 60℃, the LTO charging capacity is 178 mAh/g, slightly higher than the theoretical value. The irreversible capacities during the first charge-discharge cycle are 4 mAh/g (25℃), 8 mAh/g (40℃), and 19 mAh/g (60℃), indicating that some of the Li+ supplied by the cathode material is consumed on the LTO anode surface due to interfacial side reactions.
After 99 cycles at C/2, all batteries cycled at all temperatures exhibited capacity decay, with the battery cycled at 60°C showing the most severe degradation. We also noted that at C/2 rate, the charge and discharge capacities of the batteries were identical, meaning the coulombic efficiency was 100%. To reduce the impact of polarization on the electrodes, the authors reduced the charge and discharge current to C/10 at the 100th cycle. As shown in the figure, the battery cycled at 40°C had a discharge capacity 8 mAh/g greater than its charge capacity, and the battery cycled at 60°C had a discharge capacity 19 mAh/g greater than its charge capacity. However, there was no significant difference in charge and discharge capacity for the battery cycled at 25°C. This indicates that the batteries cycled at high temperatures exhibit greater polarization, and therefore, during discharge at C/2 rate, some of the battery's capacity was not released due to polarization. As can be seen from the cycling curves at different temperatures in Figure d below, the specific capacity of LTO material showed a rapid decline trend at 60℃, while the specific capacity of LTO cyclic at 25℃ did not show a significant decline after 5 cycles, and the battery cyclic at 40℃ did not show a significant decline in capacity after 30 cycles.
The authors analyzed the electrode interface characteristics after 100 cycles at different temperatures using AC impedance spectroscopy. The results showed a significant increase in the interfacial film impedance with increasing cycling temperature. This is likely due to the continuous growth of the interfacial film caused by side reactions during cycling at high temperatures. Therefore, the authors used XPS to analyze the surface characteristics of LTO.
The XPS analysis results of the LTO negative electrode and LMO positive electrode in the figure below show that the characteristic peaks of Ti2p and Mn2p of the new electrode split into two. Specifically, Ti2p splits into Ti2p3/2 (458.7 eV) and Ti2p1/2 (464.4 eV), and Mn2p splits into two characteristic peaks at 641.4 eV and 642.2 eV, corresponding to Mn3+ and Mn4+.
After cycling, the Ti3+ signal, characteristic of the lithium intercalation state of LTO, remained consistently present, and the proportion of Ti3+ increased with increasing cycling temperature. At 60℃, Ti3+ accounted for 25% of the total Ti content, at 40℃ for 20%, and at 25℃ for 7%. This is primarily due to the increased interfacial film thickness, which significantly affects both electron and ion diffusion rates. Regarding the LMO electrode, after 100 cycles, the characteristic Mn2p peak remained almost unchanged at all temperatures, indicating good reversibility of the LMO electrode. Furthermore, we observed a similar decrease in Mn content in all LMO electrodes after 100 cycles (from 4.6% to approximately 3.0%), suggesting that the LMO electrode surface was also covered by a surface film.
To analyze the composition of the electrode interface film, the authors further analyzed the O1s and F1s spectra of the LTO electrode surface before and after cycling. In the new LTO electrode O1s spectrum, the characteristic peak near 530.1 eV corresponds to O2- in LTO, while the other two slightly higher energy characteristic peaks correspond to O in some compounds adsorbed on the LTO particle surface (e.g., OH, O=C, and COC). After 100 cycles, two new characteristic peaks appeared in the O1s spectrum at 531.5 eV and 532.5 eV, which are mainly the decomposition products of the electrolyte on the LTO surface (e.g., ROCO2Li). At the same time, the characteristic peak appearing near 533.5 eV corresponds to PF compounds (e.g., PO3-4, PO(RO)3, or LixPOyFz), which mainly comes from the decomposition of LiPF6. We can also observe that the O content decreases significantly with increasing temperature. For example, after 100 cycles at 25°C, the O content is 9.2%, but after 100 cycles at 60°C, only about 2.7% remains. This indicates that the thickness of the LTO surface coating continues to increase as the cycling temperature increases.
The F1s spectra show that the F element in the new LTO electrode mainly originates from the PVDF binder. The two characteristic peaks at 687.9 eV and 689.5 eV correspond to CF2 and CF3/CF2-CF2 in the PVDF binder molecules. The F content in PVDF decreases after cycling, primarily due to the thickening of electrolyte decomposition products on the LTO electrode surface. We also found two new characteristic peaks after cycling, in addition to F in PVDF, corresponding to LiF (685 eV) and LixPOyFz (686.5 eV). These new F-containing decomposition products increase with increasing cycling temperature; for example, at 25°C, these F-containing products account for 2.0%, at 40°C for 4.1%, and at 60°C for 8.7%. Furthermore, we note that the cycling temperature has no significant effect on the LiF content on the LTO electrode surface.
The figure below compares the proportion of surface layer and active material in LTO and LMO electrodes according to the number of atoms after cycling at different temperatures. After cycling at 25℃, 40℃ and 60℃, the thickness of the LTO surface film also increases continuously, but for LMO, the cycling temperature has a relatively small effect on the thickness of its surface film.
To analyze whether Mn elements dissolve from the cathode and migrate to the LTO electrode surface, the authors analyzed the LTO electrode surface after cycling at 40°C and 60°C using XPS. As shown in the figure below, two characteristic peaks of Mn2p appeared on the LTO electrode surface after cycling at 40°C, near 641.3 eV and 642 eV, corresponding to Mn2+. Studies indicate that these may be products such as MnF2 or MnO, which would lead to an increase in battery impedance.
A significant increase in Mn content was observed on the LTO electrode surface after cycling at 60℃. A new characteristic peak appeared near 638 eV, corresponding to Mn in the 0-valence state, i.e., metallic Mn. This is likely because the high temperature intensified the decomposition of LiPF6, resulting in more HF and thus accelerating Mn dissolution, leading to the presence of metallic Mn on the negative electrode surface. The presence of Mn on the negative electrode surface can damage the SEI film, thereby accelerating electrolyte decomposition on the negative electrode surface and ultimately accelerating electrolyte decomposition on the LTO surface, causing a continuous decline in battery capacity.
The figure below shows the surface elemental distribution of LTO electrodes after cycling at different temperatures. The P element distribution map reveals the presence of FP compounds on the LTO particle surface, particularly high levels of P on the LTO electrode surfaces after cycling at 40℃ and 60℃, indicating more severe electrolyte decomposition on the LTO electrode surface at higher temperatures. Mn elemental analysis shows that Mn is only visible on the LTO surface after cycling at 60℃.
The authors used mass spectrometry to analyze the molecular structure of the inert film layer on the negative electrode surface. The results show that after 100 cycles, the main secondary ions detected included carbon-based ions: 12 (C-), 13 (CH-), 24 (C2-), and 25 (C2H-), which are mainly related to the decomposition of the electrolyte solvent. F- and P-based ions, including 45 (LiF2-), 63 (PO2-), 79 (PO3-), and 101 (PO2F2-), mainly originated from the decomposition of LiPF6. The strongest ion, 17 (OH), originated from both the decomposition of the electrolyte solvent and LTO. The number of P-related secondary ions on the LTO negative electrode surface cycled at 40 and 60 °C was significantly higher than that cycled at 25 °C, indicating that the increase in SEI film thickness during LTO cycling at high temperatures is mainly due to the decomposition of LiPF6. In addition, we observed a 71 (MnO) peak on the surface of the LTO electrode after cycling at 60 °C, indicating that Mn elements dissolved from the positive electrode were deposited on the surface of the LTO negative electrode.
To further analyze the SEI film structure on the LTO surface, the authors used sputtering to etch away the SEI film at different depths layer by layer, analyzing the elemental information at different depths. As shown in the figure below, the Ti+ content is higher closer to the LTO particle surface. However, the Ti+ content increases more rapidly on LTO electrodes cyclic at low temperatures, indicating that the SEI film thickness is thicker on LTO electrodes cyclic at high temperatures.
After cycling at 25℃, the intensities of C3H6O+ and PO2- on the LTO electrode surface almost dropped to zero after the first sputtering (30s), while the intensity of Li2F- showed a trend of first increasing and then decreasing. This indicates that organic components and phosphorus-containing compounds in the SEI film are mainly concentrated in the surface layer, while Li2F is mainly concentrated in the lower layer of the SEI film. After cycling at 40℃, the intensities of C3H6O+ and PO2- on the LTO electrode surface also decreased significantly during the first sputtering process, but the intensity of PO2- was slightly higher than that of the LTO electrode cycled at 25℃, indicating that the decomposition of LiPF6 is more severe at high temperatures than at room temperature.
Meanwhile, no phenomenon of Li2F- intensity first increasing and then decreasing was observed on the LTO electrode surface after cycling at 40℃, indicating that organic components, P-containing compounds, and LiF are mixed together at this temperature. On the LTO anode surface after cycling at 60℃, the intensities of C3H6O+, PO2-, and Li2F- were initially relatively high, indicating that LiPF6 decomposes more severely at high temperatures. The intensities of C3H6O+ and PO2- also showed a significant decrease in the first sputtering, indicating that these components are mainly concentrated on the surface of the SEI film. However, the intensities of these components remained at a relatively high level in subsequent sputterings, indicating that these components have a higher proportion in the SEI film and are distributed deeper. The intensity of Li2F- decreased continuously from the beginning of sputtering. Finally, regarding the Mn element, the intensity of MO- on the LTO anode surface after cycling at 60℃ was significantly higher than that after cycling at 40℃, indicating that Mn in the cathode is more easily dissolved and deposited on the surface of the LTO electrode at 60℃. The intensity of MnO- first increases and then decreases indicates that the Mn element is deposited below the organic matter and P-containing compound layer.
Nicolas Gauthier's research shows that high-temperature cycling significantly increases the thickness of the SEI film on the LTO electrode surface and alters its composition. At 60°C, it leads to more severe LiPF6 decomposition, resulting in an SEI film with a high FP compound content on the LTO electrode surface. It also accelerates the dissolution of Mn in LMO and allows it to migrate to the LTO electrode surface, embedding itself in the lower layer of the SEI film, thus disrupting the SEI film's composition and causing a continuous decrease in the battery's reversible capacity.
