比例差值谱技术在微尺度器件可靠性研究中的最新应用

本文应用比例差分(PDO,ProportionalDifferenceOperator)技术提出了一种新的表征微尺度MOS器件负偏置温度不稳定性(NegativeBiasTemperatureInstability,NBTI)的实验方法。和传统方法相比,这种新方法可以避免恢复效应对NBTI表征的影响。使用该方法,本文研究了不同直流应力电压和应力温度对NBTI退化的影响。

Evaluation of negative bias temperature instability in ultra-thin gate oxide pMOSFETs using a new on-line PDO method

A new on-line methodology is used to characterize the negative bias temperature instability （NBTI） without inherent recovery. Saturation drain voltage shift and mobility shift are extracted by ID-VD characterizations, which were measured before stress, and after every certain stress phase, using the proportional differential operator （PDO） method. The new on-line methodology avoids the mobility linearity assumption as compared with the previous onthe-fly method. It is found that both reaction-diffusion and charge-injection processes are important in NBTI effect under either DC or AC stress. A similar activation energy, 0.15 eV, occurred in both DC and AC NBTI processes. Also degradation rate factor is independent of temperature below 90℃ and sharply increases above it. The frequency dependence of NBTI degradation shows that NBTI degradation is independent of frequencies. The carrier tunnelling and reaction-diffusion mechanisms exist simultaneously in NBTI degradation of sub-micron pMOSFETs, and the carrier tunnelling dominates the earlier NBTI stage and the reaction-diffusion mechanism follows when the generation rate of traps caused by carrier tunnelling reaches its maximum.

面向复杂多陷阱的随机电报噪声提取技术

随着集成电路的发展,器件尺寸不断减小,导致MOSFET栅氧化层中的陷阱增多。低频噪声随之产生,尤其是随机电报噪声RTN变得愈发明显,对器件的可靠性提出了挑战。当器件中的陷阱个数多于1个时,陷阱间的耦合效应也对RTN信号的分析产生了重要影响。因此,开展针对复杂多陷阱情况下的RTN信号的提取技术的研究变得尤为迫切。目前现存的RTN提取技术在处理大数据量时迭代所耗费的时间成本大,且自动化程度不高,这些问题亟需解决。基于RTN信号非高斯性质,提出了自动化检测RTN信号的方法和自动化判定陷阱数方法,使得RTN信号的提取更加准确和高效。另外,还针对迭代过程中重要参数提出了自适应模型的方法,实现了大部分RTN信号提取的迭代加速。最后,对实测RTN信号应用上述方法进行参数提取,并分析了耦合效应对参数产生的影响。