Electron microscopy: a new spin on electron beams.

We are used to electrons occupying states with quantized spin and orbital angular momenta in atoms, but existing methods for generating high-energy electron beams are very inefficient at either producing or transferring states with well-defined spin and/or orbital angular momentum. This failure stems from the dominance of the direct Coulomb interaction over the spin–spin and spin–orbit interactions for high-energy electrons. As a consequence, even though electron beams can be focused to subatomic dimensions and used to measure the local composition and bonding of a material, they cannot reveal magnetic and spin information on comparable length scales. However, recent work by groups in Japan1 and Europe2 points the way to filling this gap by showing that it might be possible to measure and manipulate orbital angular momentum on the atomic scale with electron microscopes. The key to creating an electron beam with quantized orbital angular momentum relies on techniques that were developed by John Nye and Michael Berry in 1974 to study dislocations in optical fields3. For example, if the pattern of the wave fronts resembles the pitchfork shape shown in Fig. 1a, then a dislocation core or vortex appears at the apex of the pitchfork, with the probability current (which is the energy flux or Poynting vector in optics) circulating about a point of indeterminate phase and zero amplitude. For electrons, the probability current associated with the vortex possesses an intrinsic orbital angular momentum and an associated magnetic moment that is determined by the shape of the wave packet4. The number of wave fronts added or removed at the core is an integer, l, and this results in the orbital angular momentum being quantized and equal to lħ, where ħ is Planck’s constant, h, divided by 2π, and the integer l is known as the topological charge of the vortex. The optics community has long exploited optical vortex beams to transport orbital angular momentum independent of its polarization (or spin) state in applications such as optical traps and quantum information. The first demonstration of orbital angular momentum in an electron microscope, reported by Masaya Uchida and Akira Tonomura of RIKEN earlier this year1, relied on a stack of thin films of graphite naturally (and fortuitously) forming a spiral phase plate. They produced an electron beam with a helical phase and confirmed the presence of a vortex with l = 1 by interfering the vortex beam with a reference plane wave to form an electron hologram. This groundbreaking work was a proof of concept rather than a general tool, however, because there were problems avoiding contamination of the beam and maintaining precise spatial control of the pattern. A few months later, Jo Verbeeck and He Tian of the University of Antwerp, and Peter Schattschneider of the Vienna University of Technology, overcame these problems by, essentially, running the RIKEN experiment in reverse2. Verbeeck and co-workers generated a hologram by digitally interfering a vortex beam with a reference plane wave in a computer. Then they carved a binary version of the hologram into a thin platinum foil using a focused ion beam. Finally, by illuminating the hologram with a coherent electron beam, they demonstrated that vortex beams with l = ±1 were produced in the far-field diffraction plane (Fig. 1a); beams elecTRoN MicRoScoPy