調控表面電子態與電荷分佈以促進光催化二氧化碳還原:從單原子設計到反應物介導機制

Abstract

在人類人口成長以及科技的持續發展之下，能源需求日益增加，但是如果依靠使用傳統化石燃料會導致排放大量溫室氣體，加速全球暖化。因此尋求再生能源以及減少溫室氣體是目前全球的重要目標。其中，光催化二氧化碳還原反應(PC-CO2RR)能將溫室氣體轉化為高能量密度太陽能燃料的極具前景之途徑，然而其效率常受制於CO2分子的化學惰性以及遲緩的反應動力學。本論文結合實驗表徵與密度泛函理論(DFT)計算，從調控催化劑表面電子結構出發，系統性地探討了提高 CO₂ 活化與轉化效率的兩種創新策略。 第一部分研究成功透過濕浸漬法將過渡金屬單原子(Cr、Fe、Ni)錨定於氧化鎢(WO3)奈米片表面。擴展X射線吸收精細結構(EXAFS)證實了單原子的孤立分散狀態。研究提出以金屬的價電子數與電負度作為雙物理描述符的預測模型，指出 Fe 與 Mn 處於電荷轉化的協同最優點。光電子能譜(XPS、XAS)與計算皆證實，Fe 單原子能有效打破表面電荷分佈的對稱性，引導顯著的局部電荷積聚。這不僅增強了CO2的化學吸附(使其∠O-C-O鍵角劇烈彎曲至135.6o)，更大幅降低了速率決定步驟中*COOH 中間體的活化能障，使最初的質子化步驟轉變為自發的放熱反應。在 4 小時的光照測試下，最佳化的 SA-Fe-WO3 催化劑的一氧化碳(CO)產率達 12.1 µmol g⁻¹，較原始WO3提升了約三倍，表現出優異的穩定性與一氧化碳選擇性。 第二部分研究則聚焦於反應物吸附過程對催化劑電子態的動態反饋調控。我們以 20 奈米厚的二硫化鉬(MoS2)薄膜作為模型催化劑，並利用近常壓X射線光電子能譜(NAPXPS)進行原位觀測。研究首次發現反應物的吸附順序會劇烈影響能帶結構: H2O的先吸附會引導 MoS2 表面發生向下能帶彎曲(Downward band bending)，促使光生電子在MoS2表面大量積聚；而先吸附 CO2 則會引導微幅的向上能帶彎曲，反而增加了電子轉移的能壘。原位原創的環境依賴實驗與開爾文探針力顯微鏡(SKPM)及DFT計算相符，證實先通入H2O能大幅促進高活性彎曲態CO2 (b-CO2)中間體的形成，進而使隨後的光催化PC-CO2RR一氧化碳產率顯著提升20%。 綜上所述，本論文建立了電荷局域化、幾何活化與光催化性能之間的關聯，更展現了反應物調節能帶彎曲的動態行為，為設計下一代高效太陽能燃料催化表面提供了方向。

Rapid global population growth and industrialization have escalated energy demands, while reliance on conventional fossil fuels has accelerated global warming via greenhouse gas emissions. Developing renewable energy platforms and carbon mitigation strategies is therefore a paramount global objective. Photocatalytic CO2 reduction reaction (PC-CO2RR) offers a promising pathway to convert greenhouse gases into high-energy-density solar fuels; however, its efficiency remains severely constrained by the chemical inertness of the CO2 molecule and sluggish reaction kinetics. Here, by combining experimental characterizations with density functional theory (DFT) calculations, we systematically investigated two innovative strategies to modulate surface electronic structures for enhanced CO2 activation and conversion. First, isolated transition-metal single atoms (Cr, Fe, and Ni) were successfully anchored onto WO3 nanosheets via a wet-impregnation method, with their atomically dispersed states confirmed by extended X-ray absorption fine structure (EXAFS) spectroscopy. A predictive model utilizing the valence electron number and electronegativity of the metals as dual physical descriptors revealed that Fe and Mn occupy a synergistic optimum for charge transfer. XPS, XANES and DFT calculation demonstrated that Fe single atoms effectively broke the symmetry of the surface charge distribution, driving localized electron accumulation. This electronic localization significantly enhanced CO2 chemisorption—evidenced by a dramatic bending of the ∠O-C-O bond angle to 135.6°—and substantially lowered the activation barrier of the *COOH intermediate, rendering the initial protonation step a spontaneous, exothermic process. Consequently, the optimized SA-Fe-WO3 catalyst achieved a carbon monoxide (CO) yield of 12.1 µmol g−1 over 4 hours of illumination, representing a threefold enhancement relative to pristine WO3 while maintaining excellent stability and CO selectivity. Second, the dynamic feedback regulation of the catalyst electronic states induced by reactant adsorption sequences was explored. Utilizing a 20 nm thick MoS2 thin film as a model catalyst, in situ near-ambient pressure X-ray photoelectron spectroscopy (NAPXPS) revealed that the adsorption sequence dictates the surface band alignment. Specifically, the pre-adsorption of H2O induced a downward surface band bending on MoS2, promoting the accumulation of photogenerated electrons at the immediate surface. Conversely, pre-exposure to CO2 provoked a subtle upward band bending, which heightened the electron transfer barrier. This environment-dependent behavior, cross-validated by scanning Kelvin probe microscopy (SKPM) and DFT, confirmed that pre-introducing H2O significantly facilitated the formation of highly active b-CO2 intermediates, thereby boosting the subsequent CO yield by 20%. In summary, this thesis establishes a comprehensive correlation between charge localization, geometric activation, and photocatalytic performance, while unveiling the dynamic behavior of reactant-modulated band bending. These mechanistic insights provide a foundational design paradigm for engineering next-generation, highly efficient catalytic surfaces for solar fuel production.