沉基氧化鋁絕緣層之前的不同表面處理之研究

Abstract

隨著半導體元件不斷的微縮，傳統以二氧化矽為介電層將面臨到物理的極限。當閘極氧化層薄到3奈米時，直接穿遂電流開始增加。當漏電流大到無法接受的地步，高介電係數材料將用來替換二氧化矽，高介電係數材料在相同的等效二氧化矽厚度之下，能有較大的實際物理厚度以阻擋直接穿遂電流。 不過當高介電係數材料直接沈積接觸在矽晶圓表面上時，在介面的地方會有很多介面缺陷產生，有些材料因為本身的化性與熱穩定性的關係，會與矽反應或者因後續的高溫製程而產生介面層，這些現象都是我們要極力避免的。所以我們想要成長一層像二氧化矽一般，有很少的介面缺陷特性極薄的氧化層，來改善介面特性與遲滯現象。從一些文獻當中，用臭氧所成長的二氧化矽具有緻密性、自我成長限制、較少的suboxide、layer by layer的成長、較好的品質等特性。所以我們比較了沒做表面處理、氮化處理、快速熱氧化法與用臭氧水成長氧化層等不同表面處理方法，探討對特性的影響。 在論文中，我們發現先經臭氧水成長薄氧化層，再沉積氧化鋁之後經過900°C的退火，擁有好的介面與抑制漏電流的特性，有效減低遲滯現象，並且有較高的電容值的表現

The aggressive scaling of semiconductor device has led silicon dioxide gate dielectric to meet the physical limitation. The direct tunneling current occurs as gate oxide thickness thinner than 30A°. When the leakage current is more than 1A/cm2, high-k materials are investigated to replace silicon dioxide. The high-k materials show thicker physical thickness to achieve leakage current reduction at the same effective silicon dioxide thickness.

There are many interfacial states as high-k materials directly contact silicon. Some high-k materials possible react with silicon or interlayer formation during subsequent high temperature process. The phenomenon must be avoided. Therefore we want to grow a thin SiO2 or SiO2-like interfacial layer between silicon and high-k materials. The SiO2-like layer demonstrates fewer interface states and hysteresis reduction. The ultrathin oxide grown by ozone water shows that density, self-limited growth, less suboxide, layer by layer growth, and superior film quality in literature. The various surface treatments such as without treatment, nitridation, RTO, and ozone water are compared and studied. In the thesis, we demonstrate that a thin oxide grown by ozone water treatment prior Al2O3 deposition and post Al2O3 deposition annealing at 900°C are the better choice which possessed characteristics of fewer interface states, leakage current and hysteresis reduction, and higher capacitance value.