高通量、高亮度、超短90 keV电子束，频率达40 kHz
High-throughput, high-brightness, ultrashort 90 keV electrons at 40 kHz
AI总结：
本研究提出了一种40 kHz高通量高亮度90 keV电子源，通过直接电子检测实现了超快电子衍射中低电荷与高流率的平衡，显著提升了检测效率和应用范围。
摘要：
用于超快电子衍射（UED）的射频压缩千电子伏特（keV）电子源面临相互竞争的需求：短脉冲需要低电荷，但弱散射体需要高流率；高重复率可以实现信号平均，但大多数系统在1 kHz以下运行且检测效率低。本文展示了一种在40 kHz下运行的90 keV直流-射频（DC-RF）源，具有直接电子检测功能，同时解决了这些挑战。太赫兹 streaking 从压缩脉冲持续时间的测量中获得了97 ± 3 fs（FWHM）在370 aC和114 ± 47 fs（FWHM）在2.8 fC。长期的t0时间漂移，通过卷积分析压缩数据和直接太赫兹 streaking 测量独立表征，介于65-95 fs（FWHM）之间，是已报告的射频系统中最低的。在低电荷（17 aC）时，我们报告了从模拟与测量压缩数据比较得到的固有脉冲持续时间为56 fs（FWHM），是大于16 aC的keV UED中最短的。此外，2.8 fC的束流，结合40 kHz的重复率和直接检测，产生了一个可检测的归一化通量，比现有的keV（MeV）源高一个（三到四）数量级。这使得能够实际研究弱散射样品和过程，这些过程由于低截面和长采集时间而以前难以实现。
英文摘要：
Radiofrequency-compressed keV electron sources for ultrafast electron diffraction (UED) face competing demands: short pulses require low charge, yet weak scatterers demand high flux; high repetition rates enable signal averaging, yet most systems operate $\lesssim$1 kHz with low detection efficiency. Here, we demonstrate a 90 keV DC-RF source operating at 40 kHz with direct electron detection that address these challenges simultaneously. THz streaking retrieves compressed pulse durations of 97 $\pm$ 3 fs (FWHM) at 370 aC and 114 $\pm$ 47 fs (FWHM) at 2.8 fC. Long-term $t_0$ timing drifts, characterized independently both by convolution analysis of compression data and direct THz streaking measurements, lie between 65 - 95 fs (FWHM), among the lowest reported for RF-based systems. At low charge (17 aC), we report an intrinsic pulse duration of 56 fs (FWHM) from comparison of simulations to measured compression data, among the shortest for keV UED at $>$16 aC. Moreover, 2.8 fC bunches, combined with 40 kHz repetition rate and direct detection, produce a detectable normalized throughput that is one (three-to-four) orders of magnitude higher than existing keV (MeV) sources. This enables practical UED studies of weakly scattering samples and processes previously impractical due to low cross-sections and long acquisition times.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
