Nanobeam Electron Diffraction
# Nanobeam Electron Diffraction Strain Mapping
Nanobeam electron diffraction solves a problem that standard TEM imaging can't: measuring lattice strain at the few-nanometer scale inside an individual transistor channel, rather than just seeing that a structure is there. A conventional transmission electron microscope image shows contrast and shape — where the silicon is, where the metal gate sits, where a defect lives — but it doesn't directly report how far the crystal lattice has been stretched or compressed at a specific point. Strain is exactly what modern transistors depend on: stressed silicon channels switch faster because strain changes carrier mobility, and knowing the strain profile at the nanometer scale is the only way to confirm a process is actually delivering the mobility boost it was designed for.
The technique works by rastering a narrow, nearly-parallel electron beam across the sample and recording a diffraction pattern at every single point, rather than forming one image the way standard TEM does. Each diffraction pattern is a map of where the electrons scattered off the crystal lattice, and the exact spacing between diffraction spots encodes the local lattice spacing at that one nanometer-scale spot on the sample. By comparing the spot spacing at a point inside a strained region against the spacing in a known-unstrained reference region of the same crystal, software reconstructs a strain value for that point — repeat across thousands of raster positions and the result is a full two-dimensional strain map overlaid directly on the device cross-section.
The main tradeoff against competing strain-measurement methods is spatial resolution versus sample preparation burden: nanobeam diffraction resolves strain at a scale finer than most alternatives, but it requires an electron-transparent thin-foil sample prepared by focused ion beam milling, and the raster-and-fit acquisition is slow compared to a single wide-field image. Alternatives like Raman spectroscopy or X-ray diffraction can measure strain non-destructively and faster, but their spatial resolution is limited to hundreds of nanometers or microns — too coarse to resolve strain gradients inside a single transistor channel a few tens of nanometers wide. Nanobeam diffraction trades that convenience for resolution fine enough to see exactly where inside the channel the strain is highest, which is often not uniform and not where a simpler bulk measurement would suggest.
Process engineers validating a strained-silicon or strain-engineered device use nanobeam diffraction specifically to confirm the strain profile matches simulation at the one location that matters — directly inside the active channel — turning a design assumption about carrier mobility into a measured, nanometer-resolved fact rather than an inference from a process recipe.