论文标题
部分可观测时空混沌系统的无模型预测
Origin and Quantitative Description of the NESSIAS Effect at Si Nanostructures
论文作者
论文摘要
储层计算是预测湍流的有力工具,其简单的架构具有处理大型系统的计算效率。然而,其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间,然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用,该系统模拟了湍流的若干特征。我们表明,使用径向基函数作为非线性投影器,即使只有部分观测并且不知道控制方程,也能稳健地捕捉复杂的系统非线性。最后,我们表明,当测量稀疏、不完整且带有噪声,甚至控制方程变得不准确时,我们的网络仍然可以产生相当准确的预测,从而为实际湍流系统的无模型预测铺平了道路。
The electronic structure of low nanoscale (LNS) intrinsic silicon (i-Si) embedded in SiO2 vs. Si3N4 shifts away from vs. towards the vacuum level Evac, as described by the Nanoscale Electronic Structure Shift Induced by Anions at Surfaces (NESSIAS). Here, we fully explain the NESSIAS based on the quantum chemical properties of the elements involved. Deriving an analytic parameter Lambda to predict the highest occupied molecular orbital energy of Si nanocrystals (NCs), we use various hybrid-DFT methods and NC sizes to verify the accuracy of Lambda. We report on first experimental data of Si nanowells (NWells) embedded in SiO2 vs. Si3N4 by X-ray absorption spectroscopy in total fluorescence yield mode (XAS-TFY) which are complemented by ultraviolet photoelectron spectroscopy (UPS), characterizing their conduction band and valence band edge energies E_C and E_V, respectively. Scanning the valence band sub-structure by UPS over NWell thickness, we derive an accurate estimate of EV shifted purely by spatial confinement, and thus the actual E_V shift due to NESSIAS. For 1.9 nm thick NWells in SiO2 vs. Si3N4, we get offsets of Delta E_C = 0.56 eV and Delta E_V = 0.89 eV, demonstrating a type II homojunction in LNS i-Si. This p/n junction generated by the NESSIAS eliminates any deteriorating impact of impurity dopants, offering undoped ultrasmall Si electronic devices with much reduced physical gate lengths and CMOS-compatible materials.