Periodic Reporting for period 4 - ATMEN (Atomic precision materials engineering)
Periodo di rendicontazione: 2022-04-01 al 2022-09-30
Materials are defined by their chemical structure – which atoms are placed where. To control their properties, one can either exchange atoms, or change their arrangement. Until recently, there was only one way to manipulate individual atoms on surfaces: scanning probe microscopy. Scanning transmission electron microscopy, on the other hand, has been able to resolve atoms only more recently by focusing the electron beam with sub-atomic precision. This is especially useful in the two-dimensional form of hexagonally bonded carbon called graphene, which has superb electronic and mechanical properties that can be modified by doping the lattice with other elements.
The scattering of the energetic imaging electrons can cause silicon impurities to move through the graphene lattice, revealing a potential for an entirely new kind of atomically precise manipulation of atoms within crystal lattices. The capability for atom-scale engineering of strongly bound materials would open a new vista for nanotechnology, pushing back the boundaries of what has been so far possible with scanning probe techniques and allowing a plethora of materials science questions to be studied at the ultimate level of control.
However, to achieve these goal, improvements in the description of beam-induced displacements, advances in the implantation of heteroatoms into graphene, and a concerted effort towards the automation of manipulations are required. The overall objective of the project was to develop electron-beam manipulation into a practical technique available to the materials science community. The ERC project ATMEN tackled this in a multidisciplinary effort combining innovative computational techniques with pioneering experiments in a uniquely modified advanced scanning transmission electron microscope at the University of Vienna in Austria.
The scattering of the energetic imaging electrons can cause silicon impurities to move through the graphene lattice, revealing a potential for an entirely new kind of atomically precise manipulation of atoms within crystal lattices. The capability for atom-scale engineering of strongly bound materials would open a new vista for nanotechnology, pushing back the boundaries of what has been so far possible with scanning probe techniques and allowing a plethora of materials science questions to be studied at the ultimate level of control.
However, to achieve these goal, improvements in the description of beam-induced displacements, advances in the implantation of heteroatoms into graphene, and a concerted effort towards the automation of manipulations are required. The overall objective of the project was to develop electron-beam manipulation into a practical technique available to the materials science community. The ERC project ATMEN tackled this in a multidisciplinary effort combining innovative computational techniques with pioneering experiments in a uniquely modified advanced scanning transmission electron microscope at the University of Vienna in Austria.
We established that silicon impurities in graphene can be manipulated with essentially perfect control over more than a hundred lattice sites by tuning the electron beam energy and implementing active feedback of the scattering signal. Manipulation was found to also be quite effective in large-diameter single-walled carbon nanotubes, especially along the axis, though only impurities on the wall facing away from the beam can be affected.
Similar manipulation of phosphorus dopants in graphene was possible but much more difficult due to a competing chemical process whereby the heteroatoms are replaced by diffusing carbon adatoms. Temperature-dependent measurements indicate that this diffusion can be suppressed at cryogenic temperatures but would need an ultra-stable sample stage, which may become available in next-generation instrumentation.
To automate manipulations, we developed a neural network structure recognizer able to detect atom positions in real time, with path-finding as well as automatic positioning of the electron beam to move multiple impurities into a desired pattern. However, the preparation of ideal samples and these unwanted chemical interactions are challenging and hindered the creation of multi-atom manipulated structures by the end of the project.
Finally, we uncovered a new manipulation mechanism for group V impurity elements in bulk silicon: these can be non-destructively manipulated by a process we dubbed indirect exchange. Silicon is a highly interesting host material as single precisely positioned lattice dopants could act as solid-state cubits, and further work is envisioned there but requires instruments with higher primary beam energies.
Similar manipulation of phosphorus dopants in graphene was possible but much more difficult due to a competing chemical process whereby the heteroatoms are replaced by diffusing carbon adatoms. Temperature-dependent measurements indicate that this diffusion can be suppressed at cryogenic temperatures but would need an ultra-stable sample stage, which may become available in next-generation instrumentation.
To automate manipulations, we developed a neural network structure recognizer able to detect atom positions in real time, with path-finding as well as automatic positioning of the electron beam to move multiple impurities into a desired pattern. However, the preparation of ideal samples and these unwanted chemical interactions are challenging and hindered the creation of multi-atom manipulated structures by the end of the project.
Finally, we uncovered a new manipulation mechanism for group V impurity elements in bulk silicon: these can be non-destructively manipulated by a process we dubbed indirect exchange. Silicon is a highly interesting host material as single precisely positioned lattice dopants could act as solid-state cubits, and further work is envisioned there but requires instruments with higher primary beam energies.
During the project, electron-beam manipulation of covalently bound lattice impurities moved from a tentative possibility into a firm reality, while at the same time its scope expanded from one element and material to multiple model systems – moving far beyond the state of the art from when the project started.
We demonstrated excellent manipulation control of single silicon impurities in graphene and single-walled carbon nanotubes, the possibility to manipulate also phosphorus and aluminum dopants, and uncovered a novel mechanism for dopant manipulation in bulk silicon. At the same time, the unexpected hurdles we discovered revealed limitations that need to be overcome for electron-beam manipulation to reach its full potential in the future.
We demonstrated excellent manipulation control of single silicon impurities in graphene and single-walled carbon nanotubes, the possibility to manipulate also phosphorus and aluminum dopants, and uncovered a novel mechanism for dopant manipulation in bulk silicon. At the same time, the unexpected hurdles we discovered revealed limitations that need to be overcome for electron-beam manipulation to reach its full potential in the future.