Zhengkang Xu, Jin Niu, Huixiang Lin, Yongjun Ma, Hao Xu, Yutong Wang, Yunsong Liang, Changjun Ke, Fucai Zhang, Xiaoshi Zhang, Jie Li, Zhongwei Fan，“Extended numerical monochromatization for broadband phase retrieval”，Optics & Laser Technology 198, 114894（2026）

【原文】

Extended numerical monochromatization for broadband phase retrieval

ABSTRACT

With the advancement of attosecond science, probing ultrafast processes at unprecedented scales has become possible, opening new frontiers in ultrafast imaging. The spectrum of ultrafast attosecond pulses typically exhibits a broad bandwidth rather than monochromatic radiation, presenting significant challenges for image reconstruction. In this study, we introduce a novel algorithm, Extended Numerical Monochromatization (ENM), which imposes a prior constraint on the solution of numerical monochromatization. This approach enables precise numerical monochromatization of broadband diffraction patterns, achieving a substantial improvement in resolution by approximately 2 times, as demonstrated in experimental results with 15% bandwidth light source. Notably, ENM maintains robust reconstruction performance even with a substantial reduction in the spectral sampling rate. The method is applicable to various broadband sources, including both continuous and discrete spectra, under the assumption of investigating non-dispersive specimens. Additionally, we introduce an improved method for constructing the key matrix in the numerical monochromatization process, resulting in a significant speedup. For example, the computation time was reduced by up to 18 times in selected cases, and the speedup increases even further as the data size grows. With its increased speed, high accuracy, and robustness, ENM provides a new perspective and powerful tool for ultrafast imaging at the attosecond scale.