Fabrication of semiconductor nano-scale structures and control of its electrical characteristics by means of quantum doping

首次尝试通过单离子注入(SII)注入少量掺杂原子到半导体器件中的有源区，从而使我们能够一个接一个地将掺杂离子注入到半导体细区，从而修整与半导体器件中的有源区相对应的精细电阻的电学性质的固有波动。首先，在亚μm量级的硅电阻器中测得每个掺杂原子的电导增量为18 ns/原子。其次，采用传统的器件制作工艺制作了尺寸为亚μm的超精细测试电阻，得到了测试器件中电导的统计分布。初始电导与平均值有较大偏差，估计为63%。最后，基于每个离子电导的实验增量，在初始电导分布…的较高侧将电导值修整到某个特定值所需的单个离子的数量更多的被植入到每个测试电阻中。结果表明，初始电导波动由63%降至13%。这一结果得出的结论是，通过SII逐个注入必要数量的掺杂原子，可以控制精细半导体区域的电学性质。通过对各种波动因素的分析，发现这一数值大于实验值13%。产生较大差异的原因是单离子注入间隔为0.3μm的样品中杂质位置的起伏被高估，且杂质位置实际上并不是随机的。我们预计，单离子入射SII的较高瞄准精度将导致较小的位置波动，从而实现更好的杂质位置之间的关联。本研究首次从实验上阐明了不仅要控制杂质原子的数目，还要控制它们的位置的基本问题。较少

Trimming of the inherent fluctuation in electrical properties of a fine resistor which corresponds to an active region in semiconductor devices has been tried for the tirst time by implanting a small number of dopant atoms by means of single ion implantation (SII) which enables us to implant dopant ions one by one into a fine semiconductor region until the necessary number is reached. Firstly the conductance increase per one dopant atom in a sub-μm scale Si resistor was measured to be 18 nS/atom. Secondly very fine test resistors with a size of sub-μm were made by conventional device fabrication technology and the statistical distribution of conductance in the test devices was obtained. The initial conductance deviated considerably from the mean value, evaluated to be 63%. Finally based on the experimental increment of conductance per an ion, the number of single ions necessary to trim the conductance value into a certain value in the higher side of the initial conductance distribution … More was implanted to each test resistor. It was foud that the initial conductance fluctuation of 63% has been successfully reduced to only 13%. This result leads to the conclusion that the electrical properties in a fine semiconductor region can be controlled by implanting a necessary number of dopant atoms by SII in one-by-one manner. It was discovered that the value based on the analysis of various factors of fluctuation was larger than the experimental one, 13%. The reason for the large discrepancy was believed to be due to the overstimation of impurity position fluctuation in the specimens in which single ions were implanted at an interval of 0.3μm and the impurity position was not random in reality. We expect that a better correlation between impurity positions to be achieved by higher aiming precision of single ion incidence with SII would cause a smaller position fluctuation. This study clarifies an essential issue of controlling not only the impurity atom number but also their positions experimentally for the first time. Less