Quantum systems are described by wavefunctions, which have an amplitude and a phase: the square of the amplitude describes the probability of finding a particle in a given region of space-time, whereas the phase describes the sign (plus or minus) of the wavefunction. The ability to control the phase of systems of electrons would open up opportunities for the development of quantum devices, and the first step in achieving such control is to ascertain what the phase is in the first place. Unfortunately, the phase of an object’s wavefunction is not directly observable. It is, however, possible to work out the relative phase by observing interference patterns formed from the superposition (summation) of coherent electron waves (those between which there is a constant phase difference), by borrowing schemes from classic experiments that observed interference patterns in light, such as Thomas Young’s ‘double-slit’ experiment1 or Dennis Gabor’s demonstration of holography2. Writing in Nature, Esat et al.3 report a tabletop experiment that allows the phase of a molecular orbital to be determined from an interference pattern that arises as a result of electron emission from the molecule concerned.
Esat and colleagues began their investigation by assembling a molecule on a silver surface, using the tip of a scanning tunnelling microscope at cryogenic temperatures (5 kelvin) to manipulate atoms and molecules with subnanometre precision. More specifically, they attached two silver atoms to one end of a flat pigment molecule known as 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA) that was lying on the surface. They then lifted up the metal–molecule complex using the tip of their microscope, so that it stood upright on the surface (Fig. 1).
The authors find that the complex is stable in this upright position — which might seem surprising to those in the know, because organic molecules preferentially lie flat on metallic surfaces. It is not known which conformation of Esat and colleagues’ complex (flat or upright) is the more stable. However, their experimental finding casts light on how molecules such as PTCDA can be stacked on metal surfaces, knowledge of which is essential for constructing nanoscale devices in which molecules are in electrical contact with metals.
Erecting the molecule into this upright conformation allows it to perform a peculiar new function: it can emit electrons in the presence of an electric field. When the authors positioned the microscope tip 7 nanometres above the standing molecule and applied a voltage of about 25 volts, they detected an electron current of 100 picoamps (1 pA is 10–12 amps). Almost all of the electrons in the current pass across the sharp peak formed by the standing molecule. The electric field at the molecule’s apexes is much greater than it would be between flat electrodes, because it is enhanced by the curvature of the molecule. Esat and co-workers show that the field enhancement is sufficiently high to allow electrons on the molecule to ‘tunnel’ into the surrounding vacuum, as measured in the field-emission current.
The authors report that the electrons undergo a two-step tunnelling process to pass from the silver metal surface to the vacuum. First, a single electron tunnels from the surface into the lowest unoccupied molecular orbital (LUMO), where it adopts the orbital’s phase. In the second step, the electron is emitted at the edges of the molecule. The spatial distribution of the emitted current contains patterns caused by the interference of each electron with itself. The existence of these features indicates that the emitted electrons ‘remember’ the phase adopted from the part of the LUMO from which they were emitted — the patterns wouldn’t form unless the emissions had retained the orbital’s phase.
Esat and colleagues’ experiment is reminiscent of Young’s double-slit experiment1, in which the patterns formed by the interference of light proved that light is a wave. But, in contrast to Young’s experiment, the emission patterns observed by Esat et al. can be explained only if the electron wavefunction has a different sign depending on whether it is emitted from the top right or top left corners of the molecule. The relative phases of the electrons emitted from different sites of the molecule can thus be worked out from the spatial distribution patterns of the emission current.
The authors used an established method for moving atoms and molecules4 to produce their device. A complementary approach has previously been reported5 in which electrons are coherently emitted from carbon nanotubes. The physics underpinning the emission process is the same in both systems, but the approaches used to realize it are completely different: Esat and colleagues’ method can be thought of as a ‘bottom-up’ approach, in which the emitter is constructed from scratch, whereas the nanotube method was a ‘top-down’ approach in which nanotubes were painstakingly processed to allow the interference patterns to be observed and studied. The structures of Esat and colleagues’ emitters are therefore much more precisely defined and reproducible.
The emission of electrons from a molecular device could, in principle, be triggered and steered using a laser, as was recently demonstrated for larger emitters6. This would require the stability of the molecular emitters to be improved, but would be another step towards the development of phase control. Molecular emitters might eventually find applications in devices such as electron microscopes, detectors that identify the phase or spin of electrons, or even quantum computers.
Nature 558, 525-526 (2018)
Young, T. Phil. Trans. R. Soc. Lond. 92, 12–48 (1802).
Gabor, D. Nature 161, 777–778 (1948).
Esat, T., Friedrich, N., Tautz, F. S. & Temirov, R. Nature 558, 573–576 (2018).
Eigler, D. M. & Schweizer, E. K. Nature 344, 524–526 (1990).
Oshima, C. et al. Rev. Lett. 88, 038301 (2002).
Yanagisawa, H. et al. Sci. Rep. 7, 12661 (2017).