硼摻雜於矽奈米晶體鑲嵌於非晶二氧化矽的分布研究的模擬計算

Abstract

當減少碳排放已成為近年來面對全球氣候變遷共同的目標，再生能源之一的太陽能在過去十年間，每年都以40%的比例在持續成長。矽基太陽能電池由於其原料來源充足以及製程技術成熟，在目前的光伏電池市場中仍保有90%的市佔率。現今，許多追求低成本、高效率的第三代太陽能電池仍在持續發展，其中矽量子點薄膜太陽能電池被視為極具潛力的選擇之一。藉由量子侷限效應的特性，矽材料的能隙大小能夠隨著矽奈米晶體的尺寸來調變，因此不同尺寸的矽奈米晶體便能夠吸收不同的太陽光波段，而太陽能電池整體的光吸收係數也就能藉此提升。除了利用量子侷限效應來拓展吸收光譜，通常在製作過程中，還會加入摻雜來提升電流傳輸。事實上，從實驗觀察到在加入摻雜後，太陽能電池的電傳輸效率和開路電壓都有明顯提高。 因此，我們的研究主要針對兩個部分。首先，雖然從實驗製作出的薄膜TEM影像中，可以大致區別出矽奈米晶體和非晶二氧化矽的區域，但是對於介面的原子尺度細節仍然難以得知；再來，雖然在加入摻雜後，太陽能電池的效率有明顯提升，但是由於實驗上很難分析出太陽能電池中實際的摻雜分布，因此就不能確定摻雜實際造成的影響。針對上述兩項課題，我們藉由微觀尺度的模擬，希望能夠對矽量子點太陽能電池提升效率的機制有更多的了解。 我們首先建構了不同尺寸(直徑4, 6, 8 nm)的矽奈米球鑲嵌於非晶二氧化矽矩陣中，其中矽奈米球中央為一矽晶種(直徑2 nm)，並被非晶矽球殼所包覆的結構。藉由分子動力學，我們在實驗的退火溫度(1100 °C)下去模擬該系統，發現矽晶種在退火過程中會成長，並且在矽奈米晶體和非晶二氧化矽介面間有非晶矽層和富矽層的存在。另外，在相同的實驗退火溫度下，針對直徑8 nm的非晶矽奈米球，我們計算了不同硼摻雜濃度的系統，發現在結晶矽和二氧化矽矩陣的硼原子比較不能自由移動；而在非晶矽和富矽層的硼原子則能夠被觀察到明顯的遷移。最後，為了未來進一步的電傳輸研究，我們作了相關的初步計算和規劃，其所使用的計算方法為非平衡格林函數搭配密度泛函理論。 藉由分子動力學模擬的結果，我們對於矽量子點太陽能電池中的微觀結構和摻雜分布有了更深入的了解，期待該研究在矽量子點太陽能電池的開發工作上能夠有所幫助，並且有機會為未來的太陽能產業帶來新的突破。

The renewable solar energy has increased 40% every year in the past decade, as carbon emission reduction is becoming the common goal for fighting global climate change. Because of abundant materials and mature fabrication techniques, silicon-based solar cells still dominate 90% of the global photovoltaics market nowadays. The third-generation solar cells are being extensively developed in order to reduce the cost per watt. One of such underdeveloped solar cells consists of silicon quantum dots (Si QDs) embedded in a dielectric matrix, which have the Si-based advantages and follow the well-known QD fabrication process. Moreover, the Si QD band gap can be tuned by varying the nanocrystal sizes due to the quantum confinements. That is, the silicon nanocrystals with different sizes can absorb a wide range of solar spectra, and consequently the optical absorption coefficient is enhanced in such a QD design. Besides, one can further add dopants to improve the current transport. In fact, experimentalists have obtained both better carrier transport efficiency and higher open-circuit voltage by doping the Si QD solar cells. This work mainly focuses on two interests. First, experimentalists can roughly distinguish the silicon nanocrystals from its amorphous silicon-dioxide matrix in TEM images, but the atomistic details of their interface still remain unclear. Second, although better efficiency of solar cells is observed by adding dopants, one has no idea how dopants distribute in the solar cells. Both properties are important for further revealing the mechanism of efficiency improvements by quantum dots and dopants. We start by building spherical silicon nanoclusters (diameters 4, 6, 8 nm) embedded in the amorphous silicon dioxide matrix, where each silicon nanocluster contains a crystalline seed (diameter 2 nm) at its center, surrounded by an amorphous shell. We simulate such a structure by molecular dynamics under the experimental annealing temperature 1100 °C. We find that, after annealing simulation, the crystalline core grows, and the outer shell becomes a bilayer of amorphous silicon and silicon-rich oxide. We also dope the silicon nanoclusters (diameter 8 nm) with different numbers of boron atoms. We simulate such a doped system again under the experimental temperature. We find that the boron atoms hardly displace at all in both the crystalline silicon region and the silicon dioxide matrix, but significantly migrate within the bilayer shell of amorphous silicon and silicon-rich oxide. Finally, we have also done some preliminary calculations for preparing future electronic-transport study by the non-equilibrium Green’s function within the framework of density functional theory. In summary, we study the Si QD solar cells by molecular dynamics simulation to understand the microscopic mechanisms of the system. We expect that such understandings will help further improvements of the Si QD-based solar cells, and may eventually have impacts on future solar-energy industry.