Zintl CaSi₂矽奈米片之剝離與非晶化於高性能鋰離子電池負極研究

Abstract

矽因具有極高的理論容量，被視為下一世代鋰離子電池負極最具潛力的材料之一。然而，其在充放電過程中所產生的劇烈體積膨脹、機械破裂以及不穩定的固態電解質界面（SEI）形成，仍嚴重限制其實際應用。二維矽奈米片（silicon nanosheets, SiNS）與非晶矽（amorphous silicon, a-Si）因具備較短的鋰離子擴散路徑、受控體積變化及較佳機械韌性，被認為是改善上述問題的有效策略。本論文建立以層狀、具空氣穩定性且可商業取得之 Zintl 相二矽化鈣（CaSi₂）作為可擴展之固態前驅物，用於發展上述兩類高性能矽負極材料。首先，本研究利用無水乙醇中的軟化學拓撲轉換（soft topochemical conversion）結合超音波輔助剝離，成功製備矽奈米片材料，其初始比容量達 2390 mAh g⁻¹，並在 200 次循環後仍保有 96.6% 容量保持率。透過優化剝離條件，產率最高可達 10%，優於先前軟化學合成方法之文獻報導。所得之重疊式奈米片結構即使在面積負載量增加 800% 的情況下，仍可維持 90% 容量，同時將電極膨脹控制於僅 5%。為解決軟化學剝離法可擴展性不足的問題，本研究進一步探討熔鹽脫鈣（molten-salt decalcification）策略，並將產率提升至最高 80%。研究發現，在熔鹽循環過程中，水分為關鍵因素，其一方面促進材料剝離，另一方面亦加速不利的矽氧化反應。因此，本研究提出一種熱力學優化且受錯合物形成調控之脫鈣策略，以解耦上述競爭效應。此方法有效提高材料比表面積、穩定層間分離結構、改善鋰離子傳輸動力學，並使可逆容量保持率提升 51%。此外，本研究進一步發展全氣相剝離路徑，利用殘留 CaCl₂ 作為原位插層與膨脹媒介。透過控制 NH₃ 吸附–脫附循環，可實現五倍體積膨脹、五十倍 Brunauer–Emmett–Teller（BET）比表面積提升，以及近乎 100% 的剝離產率，且批次間差異極低。所得矽奈米片展現更佳鋰離子擴散能力、更低電荷轉移阻抗，並在 500 次循環後仍維持 99.38% 的庫倫效率。最後，本研究進一步將脫鈣後之 CaSi₂ 經由受限非晶化（constrained amorphization）途徑轉化為非晶矽，其中熔鹽環境主導脫鈣後之結構演化。所得非晶矽材料在高達 700 °C 下仍具抗再結晶能力，其初始庫倫效率達 85.6%，並在 20 A g⁻¹ 高電流密度下維持 94.1% 的容量保持率。總體而言，本研究建立了以 CaSi₂ 為基礎之可程式化材料平台，透過精準控制反應環境與後合成結構演化，實現高性能矽負極材料之設計與開發。

Silicon is a leading candidate for next-generation lithium-ion battery (LIB) anodes due to its exceptionally high theoretical capacity. However, its widespread implementation remains limited by severe volume expansion, mechanical fracture, and unstable solid-electrolyte interphase (SEI) formation during cycling. Two-dimensional silicon nanosheets (SiNS) and amorphous silicon (a-Si) represent promising approaches to mitigate these limitations through reduced diffusion barriers, controlled volume expansion, and improved mechanical resilience. This dissertation establishes calcium disilicide (CaSi₂), a layered, air-stable, and commercially available Zintl phase, as a scalable solid-state precursor for both approaches. Soft topochemical conversion of CaSi₂ in anhydrous ethanol combined with sonication-assisted exfoliation produced SiNS with an initial specific capacity of 2390 mAh g⁻¹ and 96.6% capacity retention over 200 cycles. Optimization of exfoliation conditions enabled yields of up to 10%, exceeding previously reported values for soft-chemical synthesis. The resulting overlapping nanosheet architecture retained 90% capacity despite an 800% increase in areal loading while limiting electrode expansion to only 5%. To address the limited scalability of soft-chemical exfoliation, molten-salt decalcification was investigated, achieving yields of up to 80%. During molten-salt recycling, moisture was identified as a critical factor that simultaneously promotes exfoliation and accelerates undesirable silicon oxidation. A thermodynamically optimized, complex-formation-regulated decalcification strategy was developed to decouple these competing effects, resulting in increased surface area, stable interlayer separation, enhanced lithium-ion transport kinetics, and a 51% improvement in reversible capacity retention. A fully gas-phase exfoliation route was further developed by repurposing residual CaCl₂ as an in-situ intercalation and expansion agent. Controlled NH₃ adsorption-desorption cycles produced fivefold volumetric expansion, a fifty-fold increase in Brunauer–Emmett–Teller (BET) surface area, and near-100% exfoliation yield with minimal batch variation. The resulting SiNS exhibited improved lithium diffusivity, reduced charge-transfer resistance, and a Coulombic efficiency of 99.38% over 500 cycles. Finally, decalcified CaSi₂ was transformed into amorphous silicon through a constrained amorphization pathway in which the molten-salt environment governed post-decalcification structural evolution. The resulting a-Si demonstrated resistance to recrystallization up to 700 °C, an initial Coulombic efficiency of 85.6%, and 94.1% capacity retention at 20 A g⁻¹. Collectively, this work establishes CaSi₂ as a programmable platform for engineering high-performance silicon anode materials through controlled reaction environment and post-synthesis structural evolution.