Grant-Jacob James A, Zervas Michalis N, Mills Ben
Optoelectronics Research Centre, University of Southampton, Southampton, UK.
Sci Rep. 2024 Dec 5;14(1):30301. doi: 10.1038/s41598-024-81389-9.
The appearance of plasma generated during femtosecond laser machining depends strongly on the features present on the sample before machining occurs. However, the complexity of femtosecond light-matter interaction means that development of a theoretical understanding of plasma generation is challenging. In this work, principal component analysis is applied to experimental images of plasma generated during femtosecond laser machining of silicon to calculate the orthogonal spatial patterns of the plasma variance (plasma modes), and to identify which sample variance (sample modes) are associated with these plasma modes. The results demonstrate the potential of principal component analysis for data-driven scientific discovery in the field of femtosecond light-matter interactions.
飞秒激光加工过程中产生的等离子体的外观很大程度上取决于加工发生前样品上存在的特征。然而，飞秒光与物质相互作用的复杂性意味着对等离子体产生进行理论理解的发展具有挑战性。在这项工作中，主成分分析被应用于硅的飞秒激光加工过程中产生的等离子体的实验图像，以计算等离子体方差的正交空间模式（等离子体模式），并确定哪些样品方差（样品模式）与这些等离子体模式相关。结果证明了主成分分析在飞秒光与物质相互作用领域中进行数据驱动的科学发现的潜力。
