| 摘要: | 本計畫的目標是設計實作並理論論證前瞻高效率薄膜型矽鍺量子點/氮化氧矽奈米熱電材料系統與功能性元件，以建立與現行矽積體電路技術可以相容整合且具高熱穩定度的矽鍺奈米結構熱電材料的學理基礎與實作驗證平台，最後達到增進熱電致能的技術並實現致冷/廢熱發電元件之功效。我們擬發展實現高性能薄膜型矽鍺/奈米熱電材料系統和相關的奈米致冷元件所需的製作測量技術和理論模型基礎。這需要先進的材料生長和製造技術、高精確的熱電測量方法以及深入瞭解低維度熱電物理的基礎科學。以我們過去在矽鍺奈米結構製備、熱電測量以及低維度多體系統中理論模型建構的研究能量為基石，來優化矽鍺奈米熱電材料的熱電性能，並且設計鍺量子點奈米致冷元件的原型結構，以達最佳化之熱/電轉換效率。在此研究基礎上，我們或可進一步考量實際應用之需求，將鍺量子點奈米致冷元件整合於電子電路中，提供局部熱點的冷卻或溫度的調控。有鑑於能源短缺及全球暖化問題日趨嚴重，尋找無污染且可永續再生的天然能源與開發有效的能源轉換技術，已成為眾所矚目且積極投入的的世界性重要課題。太陽能電池即是將取之不絕的太陽光源轉換成方便使用的電源。然而，目前太陽能電池的直接轉換效率過低，導致絕大部份的太陽光源能量是以廢熱的形式平白損耗之。因此，如何能善加利用太陽光源並有效地將之轉換，是眾所努力的標的。熱電效應提供了一種直接熱(電力)轉換成電力(熱)的方法，若能有效地運用該效應，甚至能擴展至致冷/廢熱發電元件及系統(如:固態冰箱或熱泵)等應用。然而，傳統熱電塊材材料的能源轉換效能非常低，因此無法提供有效的實質應用。一個好的熱電材料必須具有高的導電率、高的Seebeck係數和非常低的導熱係數。但是，這種『電子晶體/聲子玻璃(Electron crystal phonon glass)』的理想境界是傳統的塊材材料無法實現的夢，因為三者物理參數彼此之間是互相影響的。令人振奮的是，奈米科技時代的來臨已為熱電材料與技術發展帶來革命性的影響，預料在未來能源技術中扮演越來越重要的角色。近年來，理論評估和實驗成果皆已論證了低維度奈米結構所致的量子侷限效應，不僅可以大幅地增進載子的光電傳輸能力，更可以有效地壓低熱傳導率，因此大幅地提昇了熱力係數與熱電轉換效率(ZT)。由此可知，研究開發前瞻高效率奈米結構熱電材料不僅可以增進熱電致能的技術與應用，更可以擴展致冷/廢熱發電元件及系統，以達到節能/ 致能之終極目標。雖然，PbSeTe 奈米結構材料如超晶格結構、量子點和奈米線已展現極佳的熱電效應(ZT ~ 1.5)。不過PbSeTe 超晶格系統的熱穩定性不高、製備時有環境污染的疑慮、不易大規模面積製作，和現有的矽材積體電路技術在整合上也非常困難。相較之下，矽鍺奈米結構材料是另一個吸引人且深具潛力的熱電材料系統。吾人不僅可以現行成熟的矽材積體電路技術為基石，來發展製備大面積規模的矽鍺奈米材料；而且對於電子/電洞在矽鍺奈米材料的電性傳輸機制與優化已有充分的了解與掌控。近來也有文獻報導矽鍺奈米超晶格材料的ZT 值在900 oC 的溫度下可達1.3 以上。因此，我們擬積極地研究，如何以塊材技術來開發具高度掌控能力的矽鍺奈米結構成長方法、有效地設計與調變矽鍺奈米材料與結構以增加載子的電導率與降低熱導值、優化矽鍺材料系統的熱電特性以及其熱穩定度，進而實作出高品質的奈米致冷功能性元件。基於產業實用以及發揮熱電轉換最大功效的考量，我們鎖定研究矽鍺薄膜型奈米結構 (點、線、柱、超晶格等)/介電質(矽/鍺、二氧化矽、氮化矽)材料系統的熱電特性如: 電導值、熱力係數、熱導值及熱電係數，進而設計開發製作高效率矽鍺薄膜型奈米結構以供致冷/廢熱發電元件應用。鍺量子結構的維度、大小、密度、所處之介質環境、與金屬電極之間的耦合強度、操作温度以及電子和聲子的交互作用都將顯著地影響材料系統的熱電係數。因此藉由量測不同熱電結構之熱電性能，將能夠幫助我們釐清各項參數與整體性能之相互關係，進而挑選出具有最佳熱電性能之材料結構。由於目前相關微奈米熱電性能之量測發展仍然未臻成熟(以熱傳導性能為甚，Seebeck 性能次之，有待突破之瓶頸包含量測理論方法與樣品製備)，仍然難以精確地量測出許多結構與材料之熱電性能，因此本計畫也將嘗試研究發展相關的量測技術，並建立相關的量測系統，以期達到理論與實驗交互驗證之效果。主要研究重點分述如下： (1)開發各式低維度薄膜型矽/鍺奈米結構(點、線、柱、超晶格等)/介質(矽/鍺、二氧化矽、氮化矽)材料系統的成長製作技術。奈米材料的熱電特性與材料的化學組成成分、維度、與嵌入環境等息息相關。我們將開發如:選擇性氧化矽鍺、奈米球搭配化學性蝕刻定義奈米孔洞、以及陽極氧化鋁等製程技術，來製備各式各樣的薄膜型矽/鍺量子點陣列結構，如: 嵌入於二氧化矽或氮化矽絕緣層中的高密度鍺量子點陣列、鍺量子點/矽超晶格、矽鍺/矽超晶格奈米線等。我們將系統性地利用掃瞄式電子顯微鏡、穿透式電子顯微鏡、 x 射線繞射、能量色散光譜、快速傅立葉轉換、電子能量損耗與微拉曼光譜等方法來分析與歸納各個製程參數對於鍺量子點內部結構特性(晶格結構、大小與分佈、化學組成、表面粗糙度、界面特性)之影響，從而建立穩定形成鍺量子點陣列的最佳化製程條件。此外將建立奈米量子點能階檢測技術，包含以光激發與橢圓儀等來檢測鍺量子點的電子能結構與介電係數譜線特性。期待能創建高效能矽鍺奈米結構，使其ZT > 3，以達實際應用之功效。 (2)研究適當的測試結構與開發測量技術來量化薄膜型奈米結構的載子與聲子的傳輸行為機制以及其熱力係數、熱導值、熱電係數與熱穩定性。實驗測量低維度材料的熱電性能有其技術上的瓶頸，也是發展新型低維度熱電材料的主要挑戰之一。吾人可以直接選用商業標準化的儀器設備來測量傳統塊材材料的熱電特性參數(如: 導電率、導熱率和 Seebeck 係數)。但是由於奈米結構材料的厚度有限，而且通常是以複合層的組態呈現之，因此目前一般現有標準化的量測設備並不適用於低維度奈米材料的熱電特性檢測。欲求得奈米結構的熱電特性，吾人必須發揮製程與學術創意，針對個別奈米結構「量身定作」適合的測試結構與方法。本計劃之研究主題為各式嵌入於絕緣層中的鍺量子點、線或超晶格結構，因此熱電參數測量分析的重點將著重於各向異性與溫度相依性，以確認其載子、聲子的傳輸行為機制與熱穩定性。我們最終的目標是發展出一有效的熱電性能測量技術，可以集三個熱電參數屬性的測量予一體，同時精準地量化低維度材料的熱電屬性。 (3)低維矽鍺量子點/氮氧化材料系統的熱電理論模型建構我們將利用多體理論與多能階安德森模型來建立系統模型，並使用非平衡態格林函數來計算系統的電流及熱流，進而決定在線性響應及非線性響應操作區間之熱電係數，以利實驗量測及元件佈局，達成ZT 大於3 之目標。在一個密集的量子點系統中，量子點之間強大的庫侖交互作用將會顯著地影響電子在其間的躍遷，因此我們需要在安德森模型中考量多個能階的交互效應。至今的理論工程中，多以單一顆量子點且單一個能階的最簡單的情況為考量基礎，從而探討在近藤效應作用下所引致的線性響應區間之熱電屬性。所模擬的情境與實際的奈米結構迴然不同。我們將利用多體理論來建立系統模型，以釐清鄰近效應如何影響奈米結構操作在庫侖阻斷區間的熱電屬性。將以多能階安德森模型為基礎，考量電子聲子交互作用和電子介面散射效應，來計算量子點奈米結構系統的熱電特性以及ZT。我們亦將延伸擴充所建構的理論模型，以計算與評估奈米結構系統在非線性響應操作區間的ZT 值。 (4)實現高效能的矽鍺量子點 p-n 微致冷器元件：我們將針對所開發的各式矽鍺量子結構，進行奈米致冷功能性元件應用化之可行性評估研究，以期最終達成實用化之功效。由於所研究開發的薄膜型矽鍺量子點/氮氧化矽熱電材料系統與現今的積體電路技術完全相容，因此非常具有潛力得以實作出高品質的微型冷卻元件，可整合於積體電路或在微機電系統中提供局部熱點的冷卻或溫度的調控。為達此目的，除了所須的鍺量子點奈米結構外，我們將開發製作元件所需的製程技術與工藝，並研究薄膜型p-n 微型致冷元件溫度與尺寸相依的熱電性能。本計畫的研究內容涵蓋創新的奈米量子點製程工藝、優化的奈米熱電元件結構設計、量身定作的熱電測量技術以及基礎的奈米熱電材料與物理認知。這正是學術界可以發揮創意與深入研究其物理意義的空間所在。研究成果可作為爾後熱電科技與永續再生能源技術的基石。 We propose to develop the fabrication and measurement technologies needed to implement high performance thin-film-like Si1-xGex quantum dots/oxynitride nanostructured thermoelectric (TE) material systems and the associated microcooler devices. To do this, cutting-edge material growth and device fabrication technologies, high accurate measurements, and a deep understanding of low-dimensional TE physics are crucial. Some of the techniques we will use have been demonstrated previously; but combining all of them to optimize the nanostructured TE performance and to test prototype structures for microcoolers will require unprecedented levels of nanostrcuctured material growth, electrical and thermal sensitivity in these complex materials, but also an extraordinary ingenious device and characterization designs. These are the critical issues on which we will focus. We do not propose to realize the integration of studied TE devices into the Si integration circuits (ICs) within this 3-year project. If we are successful, that goal will be undertaken in a follow-on program. Providing a sustainable supply and an efficient conversion of energy has become a major global societal problem. Sun power, which uses sunlight to generate electricity, is one promising clean and renewable energy source. Unfortunately, today’s solar energy conversion technologies are extremely inefficient (<32%), with the excess solar energy being lost to heat. TE materials provide a way to directly convert heat into electricity and are also capable of acting as solid state refrigerators or heat pumps. Still TE devices are not in common uses because of their low conversion efficiency. We desire good TE materials with high electrical conductivity, high Seebeck coefficient, and low thermal conductivity. However, maximizing the figure of merit ZT of TEs is challenging because optimizing one physical parameter often adversely affects another. Encouragingly the advent of nanotechnology era has shed light on the revolution of TEs, which is expected to play an increasingly important role in meeting the energy challenge. Low dimensionality have been theoretically predicted and experimentally proven not only to enhance the electrical transport properties, but also to decrease the thermal conductivity. This in turn has led to remarkable progress in boosting TE properties. Although a high ZT (~1.5) has been reported in PbTe-based nanostructured materials such as superlattices (SLs), quantum dots (QDs), and nanowires (NWs), these materials are not environmentally benign, thermal stable at high temperature, and practical for large-scale commercial use. In contrast, SiGe-based nanostructures are another attractive and promising TE material system. Superior electrical transport properties and bandgap engineering have made SiGe beneficially adopted into prevailing Si IC domain. In addition, SiGe materials have been the only proven TE materials in power generation devices operating in 600 oC < T < 1000 oC, and high peak ZTs of ~1.3 and ~1 have been reported in nanopowdered n-Si0.8Ge0.2 and Si NWs, respectively. This strongly motivates us to explore the feasibility of efficient thin-film-like SiGe nanostructured TE materials and the associated microcoolers. The principle of nanostructuring to enhance ZT have been demonstrated in SiGe SLs and NWs, although questions linger regarding the accuracy of the ZT reported due to experimental difficulties in measuring the properties accurately. Even so, the materials fabricated by atomic layer deposition techniques are not easily incorporated into commercial devices because they are expensive, slow and could not be fabricated in sufficient quantities. Challenges in creating the next generation of SiGe TE materials are not only a higher level of control in producing nanostructures using bulk processes and accuracy in TE characterization, but also a profound understanding of carrier transport processes in these complex materials. We have been granted by “National research program for nanoscience and technology NSC” to study “high-temperature Ge QD single electron transistors (SETs)” and “Ge QD functional optoelectronic devices”, respectively. We have developed a simple, low cost, and IC-compatible method for forming SiGe QDs in a self-organized manner using selective oxidation of SiGe/Si-on-insulator (SGOI). Instead of being determined by the resolutions of electron-beam lithography and etching, the physical properties of SiGe QDs such as diameter, shape, spatial density, crystallinity, and internal structures are well controlled by conditions of thermal oxidation, SGOI layer structure, and Ge content in Si1-xGex. Furthermore, the realization of effective room-temperature Ge QD SETs has enabled us to gain insight into carrier transportation in Ge QDs/SiO2 system, and thereby we have established a solid foundation on theoretical modeling and experimental characterization of low-dimensional carrier transports. In this project, we propose to explore the feasibility and the fundamental physics of Si1-xGex TE nanostructures and microcoolers. We are aiming to (1) develop bulk fabrication processes to produce high efficient TE Si1-xGex nanostructures (such as QDs, NWs, and SLs) in SiO2 or Si3N4 ambient and pn coupled microcoolers, (2) to develop the measurement technology to effectively characterize the TE properties of Si/Ge nanostructures, and (3) to study the fundamental TE physics of low dimensional systems. The research topics are planned as follows: (1) Fabrication of high-performance thin-film-like Si1-xGex QDs/oxynitride thermoelectric materials: TE properties strongly depend on the nanostructured material’s chemical composition, dimensionality, embedded ambient, and so on. We would like to develop Si1-xGex zero-dimensional QDs and one-dimensional NWs embedded in various dielectric matrices using bulk processes, such as thermal oxidation of Si1-xGex, Au-assisted chemical etching combined with nanosphere lithography, and anodic aluminum oxide as a patterning mask. Nanostructures include but not limit to (1) dense Ge QDs array in SiO2 or Si3N4 ambient, (2) Ge QDs/Si SLs, and (3) Si1-xGex/Si-SL NWs. Effects of QD/NW internal structure properties, chemical composition, surface roughness, interface property, density, size and SLs periods (from several hundred to sub-twenty nanometers) on TE properties would be systematically investigated. We expect to create efficient Si/Ge nanostructures with ZT > 3 for practical applications. (2) Thermoelectric characterization of low-dimensional Si1-xGex /oxynitride materials: The experimental measurement of TE properties of low-dimensional materials is one of the major challenges that impede the development of novel low-dimensional TE materials. To determine the dimensionless ZT, electrical conductivity, thermal conductivity, and Seebeck coefficient are to be measured. Unlike bulk materials, whose TE properties can be directly measured with standard characterization techniques or commercially available instruments, TE characterization of low-dimensional materials requires extraordinary ingenuity on design and fabrication of testing structures depending on the corresponding TE material systems. The low-dimensional materials we will study are typically in a form of thin films incorporated with different nanostructures, such as SLs, NWs or QDs embedded in dielectric matrices. The anisotropy and the temperature dependence of these TE materials will also be characterized. We aim at developing a technique that integrates all the measurements of three TE properties or allows us to determine these properties simultaneously. (3) Theoretical modeling of low-dimensional Si1-xGex/oxynitride thermoelectric materials Theoretical study involves the simulation of TE properties of multiple QDs embedded into an insulator matrix, which is proposed for realizing solid state refrigerators or electrical generators. The solid state refrigerators (electrical generators) need to remove (generate) large amounts of heat (current). Consequently, a dense QD system is required for realistic applications. For a dense QD system, we need to consider the multiple energy level cases in Anderson model due to strong interdot Coulomb interactions and electron hopping effects. So far, most theoretical works focused on the thermal properties of a single QD in the linear response regime arising from the Kondo effect, based on the simplest case of a single level in Anderson model. We attempt to clarify how the proximity effect influences the ZT values in the Coulomb blockade regime. The electrical conductivity, thermal conductivity, thermal power, and figure of merit (ZT) of QDs or nanostructured junction system will be calculated based on the multiple energy level Anderson model, including the electron-phonon interactions and electron interface scattering effects. We will employ the Keldysh Green function technique to calculate the charge and heat currents in the sequential tunneling process including inelastic scatterings. We will also extend the ZT calculations into the nonlinear response regime of ΔT/T ≈ 1, where ΔT and T are the temperature difference and the equilibrium temperature, respectively. The thermal properties in nonlinear response regime are more close to the realistic operation regime of thermal devices and materials. Notably the nonlinear response has crucial impacts on thermal devices such as thermal diodes and transistors, we would investigate if multiple QDs junction systems act as effective thermal rectifiers for solar energy storage applications. (4) Realization of Si1-xGex QDs/oxynitride p-n couple thermoelectric microcoolers: Thin-film TE material is one of the promising materials for miniature cooling devices, such as localized cooling for hot spots in advanced electronic devices and temperature control in a micro-total analysis system and MEMS devices. For the proposed Si1-xGex QDs/oxynitride TE material system, it is particularly suitable for the aforementioned applications due to its high degree of fabrication compatibility. To demonstrate the feasibility of TE devices using the developed low-dimensional materials, we will develop the fabrication process and study the performance of a thin-film p-n couple TE microcooler. Thin-film microcoolers studies have been extensively reported, however, most of them are with single polarity and in turn have limited cooling area. A p-n serial structure module is a good platform to have larger cooling area without applying large current. Temperature and size dependence of device performance will be also evaluated. Our goal is to experimentally demonstrate an innovative and high efficient Si/Ge nanostructured TE material system for practical TE microcoolers applications. There are scientific and technologic significances involved in this proposed project. The implementation of SiGe nanostructured TE microcoolers demands smart device design, nanostructures growth, and process integration, which is a frontier research field and will become the key technologies for the nano-TEs industry. 研究期間：9908 ~ 10007 |
