二維奈米薄膜熱傳導率量測

Abstract

熱電材料被運用於廢熱發電，然而目前主流熱電材的ZT值為1上下，無法與傳統熱機抗衡。而在提升ZT的研究中，主要的方法為提升賽貝克因子與電導率，或者降低熱傳導率。近年來低維度奈米材料 - 石墨烯備受矚目，他由單層之碳原子構成，並以蜂巢式的結構排列，具有優越的機械強度與導電性，未來有潛力成為下一代熱電原件的主要材料。另外有研究顯示，材料的奈米結構使聲子散射而降低熱傳導率的方法，包含單根奈米線、奈米孔洞薄膜與奈米線陣列等等，目前也被驗證在矽基材上，將有助於熱電材料的發展。本文回顧了目前熱電材料的現況，以及未來石墨烯做為熱電材料的優缺點與不同奈米結構之矽基材熱傳導率，在這樣的背景下，本研究以3法量測超薄多晶石墨與矽奈米線陣列薄膜的垂直薄膜方向熱傳導。 在室溫下吾人量測膜厚300 nm超薄多晶石墨之熱傳導率為0.26 W/m-K，此值遠低於塊狀尺度下石墨之熱傳導率，並與超薄石墨文獻提供之量測值比較後發現有所差異，推測其原因為製備方式不同，文獻中機械剝離法所製成之超薄石墨，其材料品質優於吾人採用轉印法。除此之外，吾人利用波茲曼傳輸方程式預測多孔超薄石墨之理論熱傳導率，並與超薄石墨文獻值比較後，發現奈米孔洞有助於降低熱傳導率。 吾人量測三種不同厚度之矽奈米線陣列薄膜，室在室溫下，1.3、2.91、5.71 μm之熱傳導率分別為1.2、1.5、2.08 W/m-K，除了遠低於矽塊狀尺度下之熱傳導率150 W/m-K，吾人發現矽奈米線陣列之熱傳導率隨薄膜厚增厚而上升。除此之外，吾人採用一高分子材料parylene填充6.81 μm矽奈米線陣列薄膜之孔洞，發現其熱傳導率明顯上升至2.81 W/m-K，原因在於高分子貢獻部分熱傳途徑。經由矽奈米線陣列薄膜之SEM圖得知，矽奈米線直徑為35至300 nm之間，無法獲得一精確值，故吾人定義有效熱傳線徑，帶入波茲曼傳輸方程式並預測矽奈米線陣列之熱傳導率，在室溫下長度1.3 μm、2.91 μm、5.71 μm、6.81μm矽奈米線陣列之有效熱傳線徑分別為78 nm、94 nm、125 nm、191 nm。 本研究量測之超薄石墨與矽奈米線陣列薄膜，兩者皆因奈米尺度的限制下，熱傳導率有顯著的下降，未來有可能因降低熱傳導率而提升ZT值之優點，被應用於熱電元件中。

Thermoelectric materials are used to perform thermal to electric energy conversion. Figure of merit (ZT) of current thermoelectric materials is closed to 1.In order to be competitive with other power generation systems, thermoelectric materials should have ZT value of 3 or more. In the study of ZT enhancement, the main method is to increase Seebeck coefficient and electrical conductivity, or reduce the thermal conductivity. Graphene, an allotrope of carbon in the form of a two-dimensional, atomic-scale, hexagonal lattice, have attracted significant attention recently. Many unusual properties, such as mechanical strength and electrical conductivity, make graphene promising for the next generation of thermoelectric elements. In addition, thermal conductivity reduction in nanostructures has been one of the major themes of thermoelectrics. Nanostructured materials, including single nanowire, silicon nanomesh and silicon nanowire array, have shown great relevance to thermal conductivity reduction and are desirable for thermoelectric applications. This study review the status quo of the thermoelectric materials, the pros and cons of thermoelectric material on graphene and thermal conductivity of different nanostructures of silicon. Based on these reseraches, we use the 3 method to measure the cross-plane thermal conductivity of ultrathin graphite and the silicon nanowire array. The thermal conductivity of ultrathin polygraphite with thickness of 300nm is measured to be 0.26 W/m-K at the room temperature which is much lower than the thermal conductivity of bulk graphite and is deviated from the literature. We speculate that the quality of mechanical exfoliated ultrathin graphite from the literature is better than the chemical-vapor-deposition transferred graphite we fabricated. Besides, Boltzmann transport equation is applied to predict the theoretical thermal conductivity of porous ultrathin graphite and compare to the literature value. The result shows that silicon nanostructures decrease the thermal conductivity. Three different thicknesses of silicon nanowire array thin films are measured in this study. At room temperature, the thermal conductivity increases with the thickness from 1.2 to 1.5 and 2.08 W/m-K for 1.3 to 2.91 and 5.71 um, respectively, which is much lower than the bulk value of 150 W/m-K. We also measure 6.81um SiNW with parylene coating. Thermal conductivity of this device has a great enhancement which is up to 6.81W/m-K because heat could transfer not only silicon nanowire but also polymer. From scanning electron microscope, we found that the diameter of silicon nanowire is random which sizes are from 35 to 300 nm. This phenomenon is difficult for us to define exact value. We define the effective diameter to represent real diameter for SiNW. Effective diameter is applied to the Boltzmann transport equation. The result of Boltzmann transport equation shows a good agreement to the experiment. The effective diameter increases with the thickness from 78, 94, 125 and 191 nm for 1.3, 2.91, 5.71 and 6.81 um, respectively. Thermal conductivities of ultrathin graphite and silicon nanowire array thin film are limited by nanostructure and thus lower than bulk value. This phenomenon is very promising for thermoelectric application in the future.