Stable Photoemission from the Wehnelt Aperture Surface in 4D Ultrafast Electron Microscopy
Stable Photoemission from the Wehnelt Aperture Surface in 4D Ultrafast Electron Microscopy
复制标题
4D 超快电子显微镜中韦内尔特孔径表面的稳定光电发射
DOI：
10.1093/micmic/ozad067.1103
复制
发表时间：
2023
影响因子：
2.8
通讯作者：
Flannigan, David J
中科院分区：
文献类型：
--
作者：
Willis, Simon A;Flannigan, David J
基于激光的飞秒（fs）透射电子显微镜（TEM），被称为4D超快电子显微镜（4D- uem），由fs脉冲激光与TEM耦合组成[1-3]。为了研究皮秒到fs分辨率的化学和材料动力学，实验采用频闪泵和探针的方式进行[4-6]。通常情况下，紫外（UV）激光脉冲在枪区的电子源上训练，并通过光电效应产生离散的光电子包。虽然配置和要求各不相同，但配备(S) feg和teg的基本tem可以作为基于激光的UEMs运行[7-9]。由于相对较低的光束电流和时间结构，脉冲束透射电镜也被证明可以减轻试样损伤[10,11]。然而，与传统操作的情况一样，经常使用的电子源材料在长时间的光照射下具有固有的不稳定性[9,12]。这对寻求监测信号强度变化作为随时间变化的结构变化指标的长期收购构成了挑战。也许更有害的是，在获取单个数据点和整个数据集期间，电子包的时间持续时间和相干性会发生变化。在这里，我们证明了在基于teg的UEM中，可以从Wehnelt孔径表面产生具有高长期稳定性的即时和鲁棒性的光电发射。此外，我们表明，无论是在光或热离子操作下，所得到的光束质量至少可以与LaB6一样好。我们假设由于光子能量与孔径功函数的匹配更紧密，光束的能量分布和时间特性相对于LaB6有了改善。比较了两种结构的光发射稳定性和性能。第一种是传统的轴向配置，使用定制钝化的0.1 mm直径的LaB6尖端，周围环绕着石墨护套。对于这种结构，光发射在低于热离子阈值的情况下进行，热至值为0或20。第二种是非常规的离轴配置，使用Ni Wehnelt孔径（直径0.5 mm）的表面。利用安装在Tecnai Femto UEM探测潜望镜中的压电反射镜，对5个紫外激光脉冲（250 fs fwhm， 4.8 eV光子能量，估计光斑尺寸为50µm fwhm）进行定位。实验用激光脉冲完全在LaB6上或完全在Wehnelt孔径表面上进行。图1总结了稳定性实验结果。图1a显示了LaB6光发射的稳定性，从观察到的热离子发射和源饱和后立即降低热至值。t= 0的位置标志着到达较低的热至值的时刻，开始光电发射。测试的两个值都显示双指数衰减。热至20的设置比热至0的设置相对更稳定，在前30分钟内衰减40%，而不是95%。在热至20时的光电发射电流继续稳步下降，直至320分钟（测量停止）。该行为归因于冷却过程中LaB6表面的气态吸附，遵循hertz - knudsen型行为，并导致功函数增加。相比之下，孔径表面的光发射在长达70分钟的测量时间内明显更加稳定（注意，这种稳定性持续的时间更长），标准偏差仅为1.0%（图2）。1 b)。孔径-光电发射稳定性也立即…
Laser-based femtosecond (fs) transmission electron microscopy (TEM), dubbed 4D ultrafast electron microscopy (4D-UEM), consists of coupling a fs pulsed laser with a TEM [1-3]. Experiments are conducted in a stroboscopic pump-and-probe manner in order to study chemical and materials dynamics with picosecond to fs resolution [4-6]. Typically, fs ultraviolet (UV) laser pulses are trained on an electron source in the gun region, and discrete packets of photoelectrons are generated via the photoelectric effect. Though configurations and requirements vary, base TEMs equipped with (S) FEGs and TEGs can be operated as fs laser-based UEMs [7-9]. Owing to relatively low beam currents and to temporal structuring, pulsed-beam TEM has also been shown to mitigate specimen damage [10, 11]. However, as is the case for conventional operation, oft-used electron source materials are inherently unstable under photo-illumination over long periods [9, 12]. This poses challenges for long-time acquisitions seeking to monitor signal-intensity variations as an indicator of time-dependent structural changes. Perhaps even more detrimental is the resulting variation in electron-packet temporal duration and coherence during acquisition of both single data points and entire data sets. Here we show that photoemission with high long-term stability that is immediate and robust can be generated from the surface of the Wehnelt aperture in a TEG-based UEM. Further, we show that the resulting photobeam quality can be at least as good as that from LaB6, whether under photo or thermionic operation. We hypothesize that the energy distribution and the temporal properties of the beam are improved relative to LaB6 owing to the closer match of photon energy to aperture work function.Comparison of photoemission stability and performance was done using two configurations. The first was a conventional onaxis configuration using a custom blunted, 0.1-mm diameter LaB6 tip encircled with a graphite sheath. For this configuration, photoemission was carried out below the thermionic threshold at a heat-to value of either 0 or 20. The second was an unconventional off-axis configuration using the surface of the Ni Wehnelt aperture (0.5-mm diameter). Positioning of fs UV laser pulses (250 fs fwhm, 4.8 eV photon energy, estimated spot size of 50 µm fwhm) was done using a piezoelectric mirror mount housed in the probe periscope of the Tecnai Femto UEM. Experiments were done with laser pulses either entirely on the LaB6 or entirely on the Wehnelt aperture surface. Figure 1 summarizes the results of the stability experiments. Figure 1a shows the stability of LaB6 photoemission immediately after reducing the heat-to value from that which thermionic emission is observed and the source is saturated. The t= 0 position marks the moment when the lower heat-to value was reached, and photoemission was started. Both values tested display a biexponential decay. The heat-to 20 setting is relatively more stable than the heat-to 0 setting, decaying by 40% in the first 30 minutes compared to 95%, respectively. Photoemission current at heat-to 20 continued to steadily decline up to 320 minutes (where the measurement was stopped). The behavior is attributed to adsorption of gaseous species on the LaB6 surface during cooling, following a Hertz-Knudsen-type behavior, and a resulting increase in work function. By comparison, photoemission from the aperture surface was significantly more stable for measurement times up to 70 minutes (note that this stability persists for much longer times), with a standard deviation of only 1.0%(Fig. 1b). Aperture-photoemission stability is also immediate …