Nitridation of Silicon and Oxidized‐Silicon
Nitridation of Silicon and Oxidized‐Silicon
复制标题
硅的氮化和氧化硅
DOI：
10.1149/1.2124388
复制
发表时间：
1982
影响因子：
3.9
通讯作者：
K. Kajiwara
中科院分区：
文献类型：
--
作者：
Y. Hayafuji;K. Kajiwara
对硅和氧化硅的氮化进行了研究。氮化温度为900~176℃，氮气中氨气分压为10-3~5 kg/cm2，用椭偏仪和俄歇电子能谱对氮化膜进行了分析。对于硅氮化形成的薄膜，我们发现其生长动力学以及化学成分、刻蚀速率和抗氧化性等性能与氨分压无关。硅的氮化可以用修正的Ritchie-Hunt理论来解释，该理论假设氨氮化物界面上非常缓慢的表面反应是速率决定因素，使用对数速率定律。根据这一修正的Ritchie-Hunt理论，硅的氮化主要是在恒定电场下的阳离子迁移进行的。另一方面，发现氧化硅的氮化反应强烈依赖于氨气分压。这种依赖可能是由于氨或其衍生物通过氧化物扩散造成的。二氧化硅转化为氮氧化硅的过程贯穿整个氧化物过程。目前，二氧化硅薄膜最常用作栅极绝缘体。然而，不久的将来，硅集成电路技术将需要厚度小于200A的薄非晶介质膜，用作金属绝缘体半导体场效应晶体管(MISFET)的栅绝缘体和VLSI电路中的电容器。可能有比二氧化硅更好的材料用于VLSI电路，因为二氧化硅对杂质(如钠离子)扩散的电阻率很低，它很容易与电极反应，而且它在薄膜中的稳定性程度也不确定。我们可以预计，要确定二氧化硅是否可以用作200A以下的栅氧化物，还需要做很多工作。
The nitridation of silicon and oxidized-silicon has been studied. The nitrided films were prepared at 900~ 176 under ammonia partial pressures of 10-3 to 5 kg/cm 2 in nitrogen and were analyzed by ellipsometry and Auger electron spectroscopy. For films formed by nitridation of silicon, we found that the growth kinetics and properties such as chemical composition, etching rate, and oxidation resistance were independent of the ammonia partial pressure. The nitridation of silicon can be explained by a modified Ritchie-Hunt theory, which assumes that a very slow surface reaction at the ammonianitride interface is the rate-determining factor, using the logarithmic rate law. According to this modified Ritchie-Hunt theory, the nitridation of silicon proceeds mainly by cation migration under a constant electric field. On the other hand, it was found that the nitridation of oxidized-silicon depended strongly on the ammonia parital pressure. This dependence may be caused by diffusion of ammonia or its derivatives through the oxide. The conversion of silicon dioxide to silicon oxynitride occurred throughout the oxide.Presently, silicon dioxide film is most commonly used as a gate insulator. Soon, however, thin amorphous dielectric films less than 200A thick will be required for silicon integrated circuit technology for use as gate insulators in metal-insulator-semiconductor field-effecttransistors (MISFET's) and as capacitors in VLSI circuits. There may be better materials than silicon dioxide for use in VLSI circuits, because of silicon dioxide's low resistivity to the diffusion of impurities (as, for example, sodium ions), its strong tendency to react with electrodes, and the uncertainty of the degree of its stability in thin films. We can expect that much work will be done to determine if silicon dioxide can be used as gate oxides below 200A.