DSpace community: 永續綠能科技碩/博士學位學程

金屬輔助結晶化技術於絕緣基板直接合成二硫化鉬薄膜之研究;Direct Growth of Crystalline MoS₂ Films on Insulating Substrates via Metal-Assisted Crystallization

Tue, 20 Oct 2026 07:15:55 GMT

title: 金屬輔助結晶化技術於絕緣基板直接合成二硫化鉬薄膜之研究;Direct Growth of Crystalline MoS₂ Films on Insulating Substrates via Metal-Assisted Crystallization abstract: 二硫化鉬 (MoS2) 因其優異的性能和原子級厚度，是一種極具潛力的下一代電子裝置通道材料。然而，目前的生長方法通常需要在藍寶石基板上沉積、生長，然後進行轉移工藝，這會使薄膜可能造成伸張、破損等缺陷，進而影響後續製程良率；並且傳統生長化學氣相沉積 (Chemical vapor deposition) 製程所需溫度在700 °C以上，於玻璃等低溫基板上限制了應用；此外，也與現代互補式氧化物金屬半導體 (Complementary Metal-Oxide-Semiconductor，CMOS) 半導體後段製程工藝 (Back-End of Line，BEOL) 的整合不相容。BEOL中其介電層和多層金屬對於溫度變化是敏感的，過高的溫度會使導線融熔、抑或是造成結構損壞。而於絕緣基板上直接生長的方法，從而提高了製程相容性；同時減少了轉印步驟，可以避免薄膜引入缺陷。然而，實現高結晶度和均勻性仍然面臨挑戰，限制了功能性元件的應用。本文提出了一種基於溶液化的 MoS2 薄膜合成方法，該方法使用複合雙功能前驅體，並以熱分解反應法將溶液還原成MoS2。本研究使用了四硫代鉬酸銨 ( (NH4)2MoS4) 和鉬酸鈉 (Na2MoO4) 的混合物，其中四硫代鉬酸銨提供鉬和硫，而鉬酸鈉提供額外的鉬，鈉則降低成核能障以及提升表面擴散於促進高結晶性。將前驅物塗佈成膜在絕緣基板上，並覆蓋一層薄金屬催化層，然後在 H₂/Ar 氣氛中以600 °C的溫度進行退火，藉由上層金屬驅動前驅物轉化為其結晶為MoS2。拉曼光譜和光致發光 (PL) 特性證實了高品質薄膜的形成。在優化的前驅物比例下，於600 ℃退火可獲得具有 MoS₂ 特徵訊號薄膜，其拉曼光譜 E2g1和A1g 峰間距為Δ=22.43 cm⁻¹、PL半高寬為82.94 meV。本研究也製備了背閘場效電晶體 (FET) 以評估其電性特徵。以證明該方法可望在絕緣基板上實現可擴展的、CMOS相容的晶體 MoS2 生長，具有應用於 BEOL 整合的巨大潛力。;Molybdenum disulfide (MoS2)， owing to its excellent properties and atomically thin structure， is considered a highly promising channel material for next-generation electronic devices. However， conventional growth methods generally require MoS₂ to be deposited and grown on sapphire substrates， followed by a transfer process. Such transfer procedures may introduce strain， cracks， and other defects into the films， thereby reducing the yield of subsequent device fabrication. In addition， conventional chemical vapor deposition (CVD) processes typically require growth temperatures above 700 °C， which limits their applicability to low-temperature substrates such as glass. These high-temperature processes are also incompatible with modern complementary metal–oxide–semiconductor (CMOS) back-end-of-line (BEOL) integration. The dielectric layers and multilayer metal interconnects used in BEOL structures are sensitive to thermal exposure， and excessive temperatures may cause metal interconnect melting or structural damage. Direct growth on insulating substrates can improve process compatibility while eliminating the transfer step and reducing transfer-induced defects. Nevertheless， achieving MoS₂ films with high crystallinity and uniformity remains challenging， thereby limiting their application in functional electronic devices. In this study， a solution-based method for synthesizing MoS₂ thin films was proposed. A mixed bifunctional precursor was employed and thermally decomposed and reduced to form MoS₂. The precursor solution consisted of ammonium tetrathiomolybdate ((NH4)2MoS4) and sodium molybdate (Na2MoO4) . Ammonium tetrathiomolybdate served as the molybdenum and sulfur source， whereas sodium molybdate supplied additional molybdenum. The introduced sodium species were expected to reduce the nucleation energy barrier and enhance surface diffusion， thereby promoting the formation of highly crystalline films. The precursor solution was coated onto an insulating substrate and subsequently covered with a thin metallic catalytic layer. The samples were then annealed at 600 °C under an H₂/Ar atmosphere， during which the upper metal layer facilitated the conversion and crystallization of the precursor into MoS₂. Raman spectroscopy and photoluminescence (PL) measurements confirmed the formation of high-quality MoS₂ films. Under the optimized precursor ratio and annealing conditions， a continuous bilayer MoS₂ film was obtained at 600 °C. The Raman peak separation between the E₂g¹ and A₁g modes was 22.43 cm⁻¹， and the full width at half maximum of the PL peak was 82.94 meV. Back-gated field-effect transistors (FETs) were fabricated to evaluate the electrical characteristics of the synthesized films. This method enables scalable and CMOS-compatible growth of crystalline MoS₂ directly on insulating substrates and therefore demonstrates considerable potential for BEOL-integrated electronic applications.

介面改質與多樣化金屬功函數調變於二維半導體通道的電性傳輸特性之研究;Electrical Transport Characteristics of Two-Dimensional Semiconductor with Interface Engineering and Metal Work Function Modulation

Tue, 20 Oct 2026 07:15:50 GMT

title: 介面改質與多樣化金屬功函數調變於二維半導體通道的電性傳輸特性之研究;Electrical Transport Characteristics of Two-Dimensional Semiconductor with Interface Engineering and Metal Work Function Modulation abstract: 為了延續摩爾定律（Moore′s Law），元件尺寸的微縮並帶來了許多嚴峻挑戰，例如短通道效應（Short-channel effects）、穿隧效應以及高接觸電阻，這些因素嚴重惡化了電晶體的效能。二維（2D）材料如二硫化鉬（MoS₂），憑藉其原子級薄、無懸鍵（Dangling bonds）的表面以及尺寸微縮下維持之高載子遷移率（Carrier mobility），已成為下一代電子元件極具潛力的候選材料。然而，當金屬與 MoS₂ 接觸時，由於費米能階釘扎（Fermi level pinning effect， FLP）和蕭特基能障（Schottky barrier）的形成，往往會限制載子的注入，這成為了元件優化中的主要瓶頸。因此，目前已開發出多種策略來應對這些挑戰並提升接觸效能，包括相位工程（Phase engineering）、邊緣接觸（Edge contact）、嵌入緩衝層（Inserting buffer layers）、凡得瓦接觸（Van der Waals contacts）、能隙態飽和（gap state saturation， GSS）以及分子或化學摻雜。然而這些方法目前仍遭遇較不穩定及製程較不相容等問題。在本研究中，採用了本實驗室過去所發展的低損傷電漿處理用於部分移除 MoS₂ 表面的硫（S）原子，從而部分顯露出底層的鉬（Mo）原子。這種界面改質使得在金屬沉積過程中，能夠形成直接的金屬－鉬（Metal–Mo）鍵結，可提升導通電流等電性提升。我們進一步研究了一系列具有不同功函數（Work functions）的接觸金屬釔（Yttrium， Y）和钪（Scandium， Sc），以探索它們與經電漿處理的MoS₂之間的界面交互作用，並評估它們對電晶體元件特徵電學效能的影響。結果表明，經過電漿處理後的元件，相較於未經電漿處理的參考元件，其導通電流增加了 4.5 倍，開關比（on/off ratio）提升了約 1.4 個數量級，載子遷移率提高了約 4.2 倍。推測是由於介面的N-type 摻雜及金屬不同功函數之調節所導致，並證明該製程在提升元件電學效能方面的有效性。;To sustain the continuation of Moore′s Law， continuous device scaling has introduced several critical challenges， including short-channel effects， quantum tunneling， and high contact resistance， all of which severely degrade transistor performance. Two-dimensional (2D) materials， such as molybdenum disulfide (MoS₂)， have emerged as promising candidates for next-generation electronic devices owing to their atomically thin structure， dangling-bond-free surface， and high carrier mobility that can be maintained under aggressive scaling. However， Fermi-level pinning (FLP) and Schottky barrier formation at the metal/MoS₂ interface impede efficient carrier injection and remain major bottlenecks in device optimization. Various contact-engineering strategies， including phase engineering， edge contacts， buffer-layer insertion， van der Waals contacts， gap-state saturation (GSS)， and molecular or chemical doping， have been proposed to address these challenges. Nevertheless， these approaches still suffer from limitations such as poor stability and limited process compatibility. In this study， a low-damage plasma treatment developed in our previous work was employed to partially remove sulfur atoms from the MoS₂ surface， thereby exposing the underlying molybdenum atoms. This interfacial modification promotes the formation of direct metal–Mo bonds during metal deposition， resulting in improved carrier injection and enhanced electrical performance. Furthermore， two contact metals with different work functions， Yttrium (Y) and Scandium (Sc)， were investigated to examine their interfacial interactions with plasma-treated MoS₂ and evaluate their influence on transistor characteristics. Compared with untreated reference devices， the plasma-treated devices exhibited a 4.5-fold increase in on-state current， a 1.4-order-of-magnitude improvement in the on/off current ratio， and a 4.2-fold enhancement in carrier mobility. These improvements are attributed to interfacial n-type doping induced by the plasma treatment together with modulation of the contact properties by different metal work functions， demonstrating the effectiveness of the proposed approach for enhancing the electrical performance of MoS₂ transistors.

基於Three-Stage Stackelberg 賽局的多充電站定價、能源調度與電動車充電決策之最佳化方法;Joint Optimization Based on Three-Stage Stackelberg Game for Multi-Charging Station Pricing， Energy Dispatch， and EV Charging Strategies

Tue, 20 Oct 2026 07:15:48 GMT

title: 基於Three-Stage Stackelberg 賽局的多充電站定價、能源調度與電動車充電決策之最佳化方法;Joint Optimization Based on Three-Stage Stackelberg Game for Multi-Charging Station Pricing， Energy Dispatch， and EV Charging Strategies abstract: 隨著電動車數量持續增加，充電需求逐漸呈現時段集中與區域分布不均等特性，使充電網路營運商在進行充電定價與能源調度時，需同時考量營運收益、電網購電成本與站點負載限制等因素。為提升各充電站的營運效率，本文提出一套基於三層階層式 Stackelberg 賽局之最佳化模型，用以描述充電網路營運商、電動車充電站與電動車使用者之間的互動決策關係。模型中，充電網路營運商負責制定各站點於不同時段之充電價格；充電站則根據使用者需求、時間電價、儲能系統與各站點需求進行能源調度；電動車使用者則依據充電價格、距離成本與充電效益選擇充電站。本文依據國內外實際案場資料去模擬實驗環境，分析儲能系統與能源交換對營運效益的影響，並與最佳化投標策略及多代理人強化學習式時間電價策略進行比較。實驗結果顯示，本文所提出之三層階層式最佳化方法能有效提升充電網路營運商的獲利，並降低部分尖峰時段的電網購電壓力，證明結合動態定價、儲能系統與站點間能源交換機制，能提升公共充電網路之能源管理效率與營運彈性。因此，本文可作為未來多站點電動車充電網路於智慧化定價與能源調度決策上之參考。;As the number of electric vehicles (EVs) continues to grow， charging demand has become increasingly concentrated during specific time periods and unevenly distributed across different regions. Consequently， charging network operators (CNOs) must simultaneously consider operational revenue， electricity procurement costs， and charging station capacity constraints when determining charging prices and energy dispatch strategies. To improve the operational efficiency of multi-charging station networks， this study proposes an optimization framework based on a tri-level hierarchical Stackelberg game to model the interactive decision-making process among the charging network operator， electric vehicle charging stations (EVCSs)， and EV users. In the proposed framework， the charging network operator determines time-varying charging prices for each charging station. Each charging station subsequently performs energy dispatch by considering user charging demand， time-of-use electricity prices， the operation of energy storage systems (ESSs)， and energy exchange among charging stations. Finally， EV users select charging stations according to charging prices， travel distance costs， and charging utility. The simulation environment is constructed based on real-world data collected from domestic and international charging infrastructures to evaluate the impacts of ESS deployment and inter-station energy exchange on operational performance. Furthermore， the proposed approach is compared with an optimal bidding strategy and a multi-agent reinforcement learning-based time-of-use pricing strategy. Simulation results demonstrate that the proposed tri-level hierarchical optimization framework effectively improves the profitability of the charging network operator while reducing electricity procurement from the main grid during peak-demand periods. The results further indicate that integrating dynamic pricing， energy storage systems， and inter-station energy exchange mechanisms can significantly enhance the energy management efficiency and operational flexibility of public EV charging networks. Therefore， the proposed framework provides a practical reference for future intelligent pricing and energy dispatch decision-making in multi-charging station EV charging networks.

以造孔劑和二氧化鈦覆蓋層改善建立在碳球殼堆疊層裝載白金奈米顆粒的質子交換膜燃料電池觸媒層之研究;Improvement of PEMFC catalyst layer based on Pt nanoparticle-loaded carbon shell stack by adding porogen and TiO₂ overcoat

Tue, 20 Oct 2026 07:15:41 GMT

title: 以造孔劑和二氧化鈦覆蓋層改善建立在碳球殼堆疊層裝載白金奈米顆粒的質子交換膜燃料電池觸媒層之研究;Improvement of PEMFC catalyst layer based on Pt nanoparticle-loaded carbon shell stack by adding porogen and TiO₂ overcoat abstract: 本研究分別從質子交換膜燃料電池觸媒層之界面工程與結構工程兩大方向進行優化，以提升觸媒層之電化學耐久性與單電池輸出性能。在界面工程方面，利用化學浴沉積與電沉積於多孔碳球殼表面導入 TiO2 保護層，並系統性調控熱處理順序與氣氛（Air、Ar、Ar/H2）。實驗結果顯示，適量的 TiO2 可提供良好的 Pt 錨定位點：採用乙醇輔助電沉積並結合 Ar 氣氛預熱處理，ECSA 保持率可達 57.8%；經 2.5 min CBD 改質後，再經由 Ar/H2 還原熱處理，推測誘發強金屬－載體交互作用，使 ECSA 保持率達到最高的 63%。在結構工程方面，採用 PMMA與SiO2 雙模板法建立連通型多孔碳球殼載體。研究發現，當 SiO2 與 PMMA 之體積比為 7:3 時，可建立最佳化之開放性孔隙骨架，使 Pt 奈米顆粒均勻分散，平均粒徑維持於 3.06 nm。進一步針對Nafion功能層進行參數優化，最佳配方為：無退火、7.8 wt% 正丙醇前驅液及 2000 nm 等效厚度。在此條件，單電池於 60 °C、一大氣壓之常壓操作下，0.6 V 條件之電流密度達 1228 mA/cm2。綜上所述，TiO2 界面改質可提升觸媒之電化學耐久性，而 PMMA 雙模板孔隙工程與 Nafion 於微結構分布調控則可提升單電池性能，本研究結果可作為後續開發 low-Pt 觸媒層之參考。;This study optimizes the PEMFC catalyst layer via interfacial and structural engineering to enhance Pt durability and single-cell performance. For interfacial engineering， a TiO2 layer was deposited onto porous carbon shells via chemical bath deposition (CBD) and electrodeposition， with thermal sequence and atmosphere (Air， Ar， Ar/H2 ) systematically controlled. Results show that TiO2 provides effective Pt anchoring sites: ethanol-assisted electrodeposition with Ar pre-thermal treatment achieved 57.8% ECSA retention， while 2.5 min CBD with Ar/H2 reduction reached 63% retention， likely induced by strong metal–support interactions (SMSI). For structural engineering， a PMMA/ SiO2 dual-template method constructed interconnected porous carbon shells. A SiO2/PMMA ratio of 7:3 yielded an optimal porous framework with an average Pt particle size of 3.06 nm. Optimizing Nafion (unannealed， 7.8 wt% n-propanol， 2000 nm thickness) achieved a single-cell current density of 1228 mA/cm² at 0.6 V (60 °C， 1 atm). In summary， TiO2 modification improves durability， whereas PMMA pore engineering and Nafion control enhance performance， offering a valuable reference for low-Pt catalyst layer design.