Physical Biology Center for Ultrafast Science and Technology, Arthur Amos Noyes Laboratory of Chemical Physics, California Institute of Technology, Pasadena, 91125, United States.
Acc Chem Res. 2012 Oct 16;45(10):1828-39. doi: 10.1021/ar3001684. Epub 2012 Sep 11.
The transmission electron microscope (TEM) is a powerful tool enabling the visualization of atoms with length scales smaller than the Bohr radius at a factor of only 20 larger than the relativistic electron wavelength of 2.5 pm at 200 keV. The ability to visualize matter at these scales in a TEM is largely due to the efforts made in correcting for the imperfections in the lens systems which introduce aberrations and ultimately limit the achievable spatial resolution. In addition to the progress made in increasing the spatial resolution, the TEM has become an all-in-one characterization tool. Indeed, most of the properties of a material can be directly mapped in the TEM, including the composition, structure, bonding, morphology, and defects. The scope of applications spans essentially all of the physical sciences and includes biology. Until recently, however, high resolution visualization of structural changes occurring on sub-millisecond time scales was not possible. In order to reach the ultrashort temporal domain within which fundamental atomic motions take place, while simultaneously retaining high spatial resolution, an entirely new approach from that of millisecond-limited TEM cameras had to be conceived. As shown below, the approach is also different from that of nanosecond-limited TEM, whose resolution cannot offer the ultrafast regimes of dynamics. For this reason "ultrafast electron microscopy" is reserved for the field which is concerned with femtosecond to picosecond resolution capability of structural dynamics. In conventional TEMs, electrons are produced by heating a source or by applying a strong extraction field. Both methods result in the stochastic emission of electrons, with no control over temporal spacing or relative arrival time at the specimen. The timing issue can be overcome by exploiting the photoelectric effect and using pulsed lasers to generate precisely timed electron packets of ultrashort duration. The spatial and temporal resolutions achievable with short intense pulses containing a large number of electrons, however, are limited to tens of nanometers and nanoseconds, respectively. This is because Coulomb repulsion is significant in such a pulse, and the electrons spread in space and time, thus limiting the beam coherence. It is therefore not possible to image the ultrafast elementary dynamics of complex transformations. The challenge was to retain the high spatial resolution of a conventional TEM while simultaneously enabling the temporal resolution required to visualize atomic-scale motions. In this Account, we discuss the development of four-dimensional ultrafast electron microscopy (4D UEM) and summarize techniques and applications that illustrate the power of the approach. In UEM, images are obtained either stroboscopically with coherent single-electron packets or with a single electron bunch. Coulomb repulsion is absent under the single-electron condition, thus permitting imaging, diffraction, and spectroscopy, all with high spatiotemporal resolution, the atomic scale (sub-nanometer and femtosecond). The time resolution is limited only by the laser pulse duration and energy carried by the electron packets; the CCD camera has no bearing on the temporal resolution. In the regime of single pulses of electrons, the temporal resolution of picoseconds can be attained when hundreds of electrons are in the bunch. The applications given here are selected to highlight phenomena of different length and time scales, from atomic motions during structural dynamics to phase transitions and nanomechanical oscillations. We conclude with a brief discussion of emerging methods, which include scanning ultrafast electron microscopy (S-UEM), scanning transmission ultrafast electron microscopy (ST-UEM) with convergent beams, and time-resolved imaging of biological structures at ambient conditions with environmental cells.
透射电子显微镜(TEM)是一种强大的工具,能够以仅比 200keV 时 2.5pm 的相对论电子波长大 20 倍的放大倍数观察小于玻尔半径的原子。在 TEM 中能够以这些比例观察物质,在很大程度上是由于为校正透镜系统的不完善而做出的努力,这些不完善会引入像差,最终限制可实现的空间分辨率。除了在提高空间分辨率方面取得的进展外,TEM 已成为一种多功能的表征工具。实际上,大多数材料的特性都可以直接在 TEM 中映射,包括组成、结构、键合、形态和缺陷。其应用范围几乎涵盖了所有的物理科学,包括生物学。然而,直到最近,在亚毫秒时间尺度上发生的结构变化的高分辨率可视化才成为可能。为了达到原子运动发生的超短时间域,同时保持高空间分辨率,必须采用与毫秒级 TEM 相机完全不同的方法。如下面所示,该方法也与纳秒级 TEM 不同,纳秒级 TEM 的分辨率无法提供超快动力学。出于这个原因,“超快电子显微镜”被保留给关注结构动力学纳秒到皮秒分辨率能力的领域。在传统 TEM 中,电子通过加热源或施加强提取场产生。这两种方法都会导致电子的随机发射,对时间间隔或相对到达样品的时间没有控制。时间问题可以通过利用光电效应并用脉冲激光器产生精确定时的超短持续时间的电子包来克服。然而,用包含大量电子的短而强的脉冲实现的空间和时间分辨率分别限于几十纳米和纳秒。这是因为在这样的脉冲中库仑斥力很重要,电子在空间和时间上扩散,从而限制了光束的相干性。因此,不可能对复杂转换的超快基本动力学进行成像。挑战是在保持传统 TEM 的高空间分辨率的同时,同时实现可视化原子尺度运动所需的时间分辨率。在本报告中,我们讨论了四维超快电子显微镜(4D UEM)的发展,并总结了说明该方法的威力的技术和应用。在 UEM 中,图像要么通过相干单电子包或单电子束进行频闪观测获得,要么通过相干单电子包或单电子束获得。在单电子条件下不存在库仑斥力,因此允许进行成像、衍射和光谱学,所有这些都具有高时空分辨率,达到原子尺度(亚纳米和飞秒)。时间分辨率仅受激光脉冲持续时间和电子束包携带的能量限制;CCD 相机对时间分辨率没有影响。在单电子脉冲的情况下,当数百个电子在束中时,可以达到皮秒级的时间分辨率。此处给出的应用程序是为了突出不同长度和时间尺度的现象而选择的,从结构动力学过程中的原子运动到相变和纳米机械振荡。我们最后简要讨论了新兴方法,包括扫描超快电子显微镜(S-UEM)、会聚束扫描透射超快电子显微镜(ST-UEM)和在环境室中对环境条件下的生物结构进行时间分辨成像。