In plain words
Electrons are fired through a sample thin enough to be transparent to them, forming an image that can show individual atoms and which element each one is. A 2D crystal is the ideal specimen, because it is already thin enough.
Going deeper
Two ways of imaging
In conventional transmission electron microscopy a broad, parallel beam passes through the specimen and an image is formed from the transmitted wave; contrast comes from interference and depends delicately on focus, which makes single-atom images beautiful but hard to interpret directly.
In scanning transmission electron microscopy a focused probe is scanned across the sample and the electrons scattered to high angles are collected by an annular detector. This dark-field signal grows roughly as the square of the atomic number, so heavier atoms appear brighter, and each column can be identified chemically. On a , with only one atom in each column, that turns an image into a map of which element sits where – how molybdenum, sulfur pairs, single sulfur atoms and are arranged.
What it shows in 2D materials
Because the specimen is a single layer, everything is in the field of view at once: point defects and their species, the atomic structure of and , edges and their terminations, stacking sequences seen in cross-section, and in twisted stacks. Electron diffraction from the same region gives the orientation of each grain, and dark-field imaging maps grains over micrometres.
Spectroscopy in the same instrument adds chemistry: electron energy-loss spectroscopy identifies elements and bonding at a point, and energy-dispersive X-ray analysis gives composition. Defects can even be followed as they move under the beam, which is how vacancy migration and the growth of holes have been watched directly.
Damage, and what it costs
The same electrons that image the sample also damage it. Above a threshold energy, an electron can transfer enough momentum to knock an atom out of the lattice, and the threshold depends on the element: light atoms such as carbon are knocked out more easily. This is why are usually imaged at 60–80 kV rather than the 200–300 kV common for thicker specimens, at a cost in resolution that aberration correctors recover.
Specimens also need to be clean and suspended, and hydrocarbon contamination polymerises under the beam into a layer that hides the structure. Vacancies seen in an image may have been made by the beam rather than by growth, so defect densities need a dose-dependence check before they are quoted as intrinsic.
For specialists
Imaging with transmitted electrons, either parallel-beam (TEM) or scanned-probe (STEM), where annular dark-field contrast scales roughly with Z2 and identifies single atoms chemically. It resolves point defects, grain boundaries, stacking and moiré patterns directly, at the cost of – which is why 2D samples are imaged at 60–80 kV – and of a clean, suspended specimen.
Where this comes from
- The structure of suspended graphene sheets cited by 5,050
- Intrinsic structural defects in monolayer molybdenum disulfide cited by 2,231