在全球追求碳中和的背景下，醇類催化分解被視為利用液態能量載體實現永續氫氣生產的潛力途徑之一。為了連結基礎表面科學與實際應用之間的材料差距，並深入探討氫氣生成的反應機制，本研究由模型系統逐步延伸至較接近實際催化條件的系統，進行三項系統性的研究。 首先，本研究以Ir(100)及共吸附原子氧O*的Ir(100)表面作為模型系統，探討甲醇(CH₃OH)的分解反應，以釐清氧物種對結構明確之金屬單晶表面的影響。吸附於Ir(100)表面的甲醇優先進行初始O–H鍵斷裂，形成甲氧基(CH3O*)作為主要中間物種。隨後，甲氧基分解可沿兩條相互競爭的反應路徑進行：脫氫形成 CO*，或進行 C–O 鍵斷裂生成甲基(CH3*)與原子碳(C*)。最終氣相產物為 H₂、H2O 及 CO，整體反應機率為 60%。Ir(100) 表面 O* 的存在改變了反應路徑，並顯著提升反應活性。藉由降低脫氫反應的活化能、抑制 CH3OH 的再結合脫附，以及限制 C–O 鍵斷裂（進而降低主要毒化物種 C*的生成），甲醇的分解機率可提升至接近 100%。然而，儘管O*能促進甲醇分解，其對H*的捕捉會使產物生成由 H2 轉向 H2O。因此，在半飽和O*覆蓋度下，每個吸附甲醇所對應的 H2 產量達到最大值。 其次，本研究探討吸附於Al2O3/NiAl(100)上的Rh奈米團簇之乙醇(CH₃CH₂OH)分解反應，作為接近實際催化劑的模型系統，以釐清負載型催化劑中的結構–反應活性關聯。結果顯示，隨著Rh奈米團簇尺寸減小，乙醇的分解機率顯著提升。反應始於O–H鍵斷裂，接著主要經由Cα–H鍵斷裂，形成表面乙醛(CH3OH*；*表示吸附物種)作為主要中間物種。隨後，經由進一步的Cα–H及C–C鍵斷裂，最終生成氣相CO、CH4及H₂。當奈米團簇直徑由3.5 nm減小至1.1 nm 時，每個吸附乙醇所產生的氣相產物量提升了3.5–8.8倍，而在1.1 nm的奈米團簇上，反應機率達到接近100%。密度泛函理論(DFT)計算顯示，此反應活性的提升可歸因於小尺寸奈米團簇中的Rh邊角位點；在這些位點上，O–H鍵斷裂的活化能低於乙醇脫附所需的能障。此外，小尺寸奈米團簇上氣相CH4產量的增加，則可歸因於CO*吸附物種所造成的位點阻塞效應。 最後，本研究結合離位(ex situ)表面分析技術，包括掃描穿隧顯微鏡(STM)、反射式高能電子繞射(RHEED)及同步輻射光源光電子能譜(PES)，以及結合循環伏安法(CV)，探討PtTe2上的甲醇電氧化反應，以模擬直接甲醇燃料電池(DMFC)陽極的反應。研究結果顯示，在0.4–0.7 V的電位範圍內，電流密度會隨著連續掃描次數增加而持續提升。此增強現象可歸因於Te 的移除所產生的表面未配位Pt位點(Ptuc)因此，藉由Ar+離子轟擊控制此類活性位點的濃度，甲醇電氧化所產生的總電荷亦隨Ptuc濃度增加而同步提升。雖然單一Ptuc位點所對應的電荷與 Pt 奈米團簇相當，但Ptuc位點展現出較低的碳毒化效應，顯示其具有較佳的抗碳毒化特性。進一步研究顯示，甲醇電氧化過程中所形成的額外Ptuc原子，源自PtTe2表面轉變為三維多晶Pt奈米團簇；這些新形成的Pt位點在經過25次反應掃描後仍維持活性。 綜合而言，本研究透過由理想化金屬單晶表面、負載型金屬奈米團簇至二維材料電極的系統性研究，建立了醇類分解與電氧化反應中材料結構、活性位點及反應路徑之間的關聯，並提供了由基礎表面科學理解醇類催化轉化與氫氣生成機制的重要見解。;Driven by global goals of carbon neutrality， the catalytic decomposition of alcohols represents a promising avenue for sustainable hydrogen production from liquid energy carriers. To bridge the material gap between the fundamental surface science and the realistic applications， while shedding light on the mechanisms of hydrogen production， three systematic studies from the model-system to realistic conditions are presented herein. First， the decomposition of methanol (CH3OH) on Ir(100) and atomic oxygen (O*) co-adsorbed Ir(100) surfaces， as a model system， was investigated to elucidate the effect of oxygen on a well-defined metallic single-crystal surface. The adsorbed methanol on Ir(100) preferentially cleaved the initial O–H bond to form methoxy (CH3O*) as the major intermediate species. Its decomposition was followed by two competing pathways: dehydrogenation to CO* or bond scission of C–O to yield methyl group (CH3*) and atomic carbon (C*). The ultimate gaseous products were H2， H2O and CO with the overall reaction probability of 60 %. The presence of O* on Ir(100) altered the reaction pathway and significantly enhanced the reactivity. The decomposition probability was promoted up to nearly 100 % via decreasing the dehydrogenation barriers， suppressing the re-combinative desorption of CH3OH， and limiting the scission of C–O bond (which reduced the major poisoning species C*). Although the O* promoted the reaction， the capture of H* shifted the production from H2 to H2O， leading to the maximal production of H2 per adsorbed methanol at the half-saturated O* coverage. Secondly， the decomposition of ethanol (CH3CH2OH) on Rh nanoclusters supported by Al2O3/NiAl(100)， serving as a model for realistic catalysts， was explored to elucidate the structure – reactivity correlation on the supported catalysts. The results indicated that the decomposition probability was remarkably promoted with a decreased size of the clusters. The reaction initiated with the O–H bond scission， followed by the Cα–H bond scission predominantly to produce surface acetaldehyde (CH3CHO*; * denotes adspecies) as a major intermediate. Subsequently by the further Cα–H and C–C bond cleavage， the reaction ultimately produced gaseous CO， CH4， and H2. The gaseous production per adsorbed ethanol increased by 3.5–8.8 times when decreasing the cluster diameter from 3.5 to 1.1 nm and the reaction probability attained nearly 100% at 1.1 nm. DFT calculations attribute the promoted reactivity to the corner Rh sites on small clusters， where the activation energy for the scission of the O–H bond became smaller over the competing ethanol desorption. Moreover， the production of gaseous CH4 on small clusters was ascribed to the site-effect of CO* blocking. Finally， the electrooxidation of methanol on PtTe2 – mimicking the reaction at the anode of direct methanol fuel cells (DMFCs) – was studied by combining ex-situ surface characterization (including STM， RHEED， and synchrotron-based PES) and cyclic voltammetry (CV). Our results revealed that the current density from 0.4 to 0.7 V exhibited a continuous enhancement with further sweeps. This enhancement was attributed to the generation of surface Ptuc sites via the removal of Te; therefore， by controlling the concentration of these active sites through Ar+ bombardments， the total charge from methanol electrooxidation exhibited concomitant increase with the abundance Ptuc concentration. While the charge per Ptuc sites comparable to that of Pt nanoclusters， the less carbon poisoning effect was revealed， attributed to the active Ptuc sites. The additional Ptuc atoms， created from electrooxidation of methanol， was ascribed to the transformation of PtTe2 surface into three-dimensional polycrystalline Pt clusters; they still remained active after 25 scans of the reaction.
