Focus Documentation

a boolean switch allowing to activate a special SpotScan unbending.

SpotScan image recording was introduced by Ken Downing ~1992. Instead of exposing the micrograph with a flood beam illumination that covers the entire sample and records the entire micrograph in one go, the beam is condensed into a smaller spot that “jumps” over the sample, while several such jumps are recorded onto the same micrograph. One local spot could have a diameter of 100 nm on the sample, and an exposure time of 100 ms, while 100 such SpotScan spots then cover the micrography in a hexagonally arranged pattern, and the entire exposure of the micrograph takes 1.5 minutes. Usually this is done by leaving the shutter of the microscope open for two minutes, while some X/Y deflector coils let the beam jump over the sample, advancing every 100 ms the beam position.

This SpotScan procedure significantly reduces the beam-induced specimen movement for tilted samples, presumably reducing the effect of sample charging to within the tiny SpotScan spot.

This procedure bears a risk: If the sample is defocused and the illumination is not parallel, but convergent or divergent, then the SpotScan spots have locally a different magnification than the entire micrograph. (Defocusing then zooms locally within each SpotScan spot). This may mean for the entire image that the 2D crystal from the edge of one SpotScan spot to the beginning of the next SpotScan spot has a ruptured crystal structure, even though the sample was perfectly monocrystalline. For samples with a simple unit cell this may not be of any effect. If, however, the sample has a rather similar looking repetition within the unit cell, as might be the case for a non-crystallographic symmetry (NCS), then the SpotScan artifact may make it impossible to obtain a reasonable unbending over the entire image.

We have therefore introduced the switch treatspotscan, which tries to perform instead of the usual two-step unbending now a three-step unbending: In the first step, the Unbend I and the Unbend II script uses a very large reference boxa2 or boxb2, which is used to move the entire SpotScan spot into its correct location. Only then would the following two-step unbending try to do the unbending of the local crystal deformations, without too much risk of loosing the 2D crystal order from one SpotScan spot to the next one. To this end, the QUADSERCHK_2DX program also can extrapolate over signal-free “black” gaps between the SpotScan spots, without resetting the ERROR field: The script forces QUADSERCHK_2DX to not allow interruption of the 2D crystal pattern.

Activation of this switch is recommended in cases of SpotScan imaging with samples that have a NCS or other higher-order crystal symmetry.