2008-08-012024-05-18https://scholars.lib.ntu.edu.tw/handle/123456789/707317摘要：熱電致冷器(thermoelectric cooler)是很好的冷卻及溫控設備，因其具有體積小、可 靠度高、維護性高等優點；此外，面對日趨嚴重的地球溫室效應，利用熱電效應來進 行廢熱回收是一減緩溫室效應之可行方法。傳統的熱電元件之熱效率甚差，但近年來 拜奈米科技的突破，發現在尺寸效應影響下的半導體材料可以有很好的熱電性質。許 多相關的理論也不斷被提出來，用來理解尺寸效應如何影響材料的熱電性質，包括本 計畫主持人也有數篇SCI期刊論文探討這方面的物理機制。然而，理論分析僅能釐清 物理機制，或用於預測簡單幾何之材料，對於複雜材料的最佳設計與分析，數值計算 還是最好的工具，而數值計算方法中又以分子動力學模擬最接近真實材料。而到目前 為止，分子動力學模擬在建立模型系統、計算方法、邊界條件、以及複雜系統內部的 熱流計算等問題上，學界中的見解仍有很多分歧。 因此，本計畫將整合前人的研究結果，截長補短，並利用主持人過去已累積之豐 富學理基礎，建立一套符合物理意義且完整的分子動力學模擬工具，來模擬奈米尺度 下的熱傳問題，特別是不同材料之間的介面熱阻。此外，也將企圖整合分子動力學模 擬工具與理論模型，利用模擬數據與理論預測值之比對，相互校正彼此不足之處；完 備之理論模型可用以針對簡單幾何結構之材料進行完整的參數探討，得到最佳的結構 設計；完備之模擬工具則可用以針對較複雜幾何結構之材料進行重要的參數探討，得 到最佳的結構設計。為克服MD模擬的龐大計算量，本計畫將發展平行化之模擬工具。 計畫預定用三年時間完成此項研究計畫。<br> Abstract: The thermoelectric devices are good as temperature-control devices and good for a recovery of waste heat. However, their practical applications are limited because of the associated poor cooling efficiency, which in turn is limited by the material thermoelectric properties. It has been found recently that semiconductor materials at nanoscale can have largely improved thermoelectric properties. Although many studies have been conducted for a qualitative understanding of this improvement, many unreal assumptions are needed and the associated physical mechanisms are still not very clear. A quantitative analysis for recognizing a nanostructure design and its optimization is thus still not available. A numerical study via the molecular dynamic simulations (MD) is probably the best solution because of its solving the fundamental motion of atoms and involving least assumptions. We thus propose herein to simulate the transport phenomenon of heat in a nanostructured thermoelectric material via MD simulation tool in this project. For the time being, many controversies still exit in modeling the system, potential function, boundary conditions, interfacial conditions, etc associated with MD tool. In this proposal, we intend to integrate and revise the existing MD tools based on the plentiful related knowledge and experiences of the principal investigator. The numerical parameters will be calibrated by taking advantage of predictions from the analytical model; model parameters will be confirmed also through a comparison with the simulation results. The complete and accurate MD tool will be then applied to the exploration of the thermal property of nanostructured thermoelectric materials and an optimum design of the nanostructures will be shot. To overcome the tremendously large amount of calculations, the parallel computation technique will be employed. This project is arranged to be finished in three years.熱電材料尺寸效應奈米結構分子動力學模擬平行計算thermoelectric materialssize effectnanostructuresmolecular dynamical simulations,