Laser Spectroscopy of Isolated Molecules and Systems: Structure, Dynamics and Analysis

这项提案要求提供支持，以延长发现赠款计划，该计划在两个基本重点领域取得了实质性进展。(1)在化学和物理之间的一个重要的交叉学科前沿，我们发展了一种新的超音速光束方法，它利用脉冲激光在多共振激发序列中产生量子态选择的冻结分子里德堡气体。这种气体演化形成强耦合的超冷等离子体。(2)将激光光谱学与多变量信号处理相结合，设计了新颖的光学仪器和识别解释工具，为复杂材料的光谱分析和分类提供了广阔的应用前景。我们现在提出一个计划，在我们的超冷等离子体研究中开创新的实验和理论模型，将使用辐射发射和层析成像来绘制分子内弛豫的电子非绝热通道，并表征发生超冷等离子体状态驰豫的非弹性散射动力学。这些测量将为研究关联多体系统中的远程碰撞物理创造一个前所未有的实验室，与涉及热致密物质、非费米金属、耦合量子井和高亮度电子源的技术生产等重要问题直接相关。我们建议通过将新的仪器配置与强大的新数据挖掘和多变量分析策略相结合，着眼于复杂合成材料和生物材料的分类，显著推进我们在分析拉曼和红外光谱领域的工作。我们将使用机器学习策略，从将光谱与结构和功能属性联系起来的多变量协方差模型中提取有说服力的特征。通过目标定向的特征选择，我们将分离描述导致复杂材料性质偏移的基本化学变化的特征子谱。这项研究将产生新的实时光谱化学分类方法，具有重要的分析应用，从识别工业过程控制到临床诊断。

This proposal requests support to renew a Discovery Grant program that has made substantial progress in two areas of fundamental focus. (1) In work at an important interdisciplinary frontier between chemistry and physics, we have developed a novel supersonic beam methodology that employs pulsed lasers in multiresonant excitation sequences to create a quantum-state-selected frozen molecular Rydberg gas. This gas evolves to form a strongly coupled ultracold plasma. (2) Combining laser spectroscopy with multivariate signal processing, we have devised novel optical instrumentation and discerning interpretive tools that promise broad utility for the spectrochemical analysis and classification of complex materials. We now propose a program of pioneering new experiments and theoretical models in our ultracold plasma research that will use radiative emission and tomographic imaging to chart electronically nonadiabatic channels of intramolecular relaxation, and characterize the inelastic scattering dynamics by which the relaxation to the ultracold plasma state occurs. These measurements will create an unprecedented laboratory for the study of long-range collision physics in correlated many-body systems, with direct relevance to important questions concerned with warm dense matter, non-Fermi metals, coupled quantum wells and the technological production of high-brightness electron sources. We propose to advance significantly in our work at the frontier of analytical Raman and infrared spectroscopy by coupling novel instrument configurations to powerful new data mining and multivariate analysis strategies, with an eye to the classification of complex synthetic and biological materials. We will use machine learning strategies to draw telling characteristics from multivariate covariance models that link spectra to structural and functional properties. By means of target-directed feature selection, we will isolate signature sub-spectra that describe fundamental chemical variations responsible for excursions in the properties of complex materials. This research will yield new real-time spectrochemical classification methodologies with important analytical applications ranging from discerning industrial process-control to clinical diagnosis.