If silicon anodes are already approaching their conventional theoretical capacity, can lithium-ion batteries still move to higher energy density? Graphite anodes have limited capacity, while silicon is attractive because of its much higher lithium-storage potential and relatively low cost. Yet silicon also suffers from large volume expansion, unstable interfaces, and cycling degradation. A new study published in Research by Prof. Shisheng Lin group from Zhejiang University and collaborating institutions reports a defect-rich silicon–carbon composite anode with an initial discharge capacity of 6,694.21 mAh g⁻¹, well above the commonly cited theoretical capacity of silicon.
The central message of the study is not only that the capacity is unusually high, but that defect engineering may open additional lithium-storage behavior in silicon–carbon systems. The researchers synthesized silicon–carbon composites by growing defect-rich silicon nanostructures on a graphene-based carbon matrix. The carbon framework provides electrical connectivity and helps buffer mechanical strain, while internal defects, grain boundaries, and partially disordered silicon domains may create extra lithium-storage sites and transport pathways. This architecture offers a possible explanation for why the composite behaves differently from conventional crystalline silicon anodes.
Electrochemical testing showed that samples with lower carbon content reached the highest initial capacities. One sample delivered 6,694.21 mAh g⁻¹ at 0.1 C, with an initial Coulombic efficiency of 74.71%. Additional samples in the same group also exceeded 6,000 mAh g⁻¹, supporting the reproducibility of the high-capacity behavior. Increasing the carbon content reduced capacity but improved initial Coulombic efficiency. After electrolyte optimization, one sample reached 5,294.88 mAh g⁻¹ with an initial Coulombic efficiency of 90.96%, indicating that capacity and efficiency can be partially balanced through formulation and electrolyte design.
Structural and chemical analyses support the proposed role of defects. Before cycling, silicon nanodots were distributed on graphene sheets and showed abundant crystal defects and twin boundaries. After cycling, silicon lattice fringes were still partially preserved, suggesting that the composite does not undergo complete structural collapse. In situ Raman measurements and surface chemical analysis further indicated that lithiation disrupts crystalline order, while delithiation can lead to nanoscale silicon domains confined within the carbon framework. Lithium was detected mainly in chemically bonded states rather than as metallic lithium, supporting the interpretation that the ultrahigh capacity arises from lithium storage within the silicon–carbon system and its interfacial products.
The study also examined whether the material could move beyond half-cell testing. When a small fraction of the high-capacity silicon–carbon material was blended with graphite and paired with an NCM811 cathode in a pouch-type full cell, the cell retained about 80% capacity after 256 cycles. This result suggests a possible pathway toward practical electrode integration, although the highest-capacity materials still face major hurdles, especially low initial Coulombic efficiency, long-term cycling stability, and volume-change management.
AI-assisted design added a further layer to the study. The researchers trained a multilayer perceptron model using 96 experimental synthesis–performance pairs and then combined it with a constrained genetic algorithm to search for optimized formulations. The model predicted that the maximum initial discharge capacity could be further enhanced. It also showed that the formulation maximizing capacity differs from the formulation maximizing initial Coulombic efficiency, highlighting an intrinsic trade-off that future anode design must address.
Overall, the work provides experimental evidence that defect-engineered silicon–carbon composites may push lithium storage beyond conventional silicon-anode expectations. The next challenge is to translate this high-capacity behavior into cells that combine high initial efficiency, long cycle life, controlled expansion, scalable manufacturing, and verified full-cell energy-density gains.
Sources: https://spj.science.org/doi/10.34133/research.1179
Research
News article
Not applicable
Super High Capacity of Lithium Battery Silicon–Carbon Anode over 6,500 mAh g−1
13-Mar-2026