because they exceed the count rate of the detector [38]. A possible
solution is to collect data in two passes [39]: an initial
low-resolution pass using low exposure is aimed at recording only
those reflections that would cause overflows at higher exposure. A
subsequent high-resolution pass will then record the highresolution reflections using a higher exposure, without concern
for simultaneously preserving the integrity of the low-resolution
information.
An additional requirement on a MicroED camera is speed. The
absolute position of the stage in an electron microscope can usually
not be set accurately, which makes it impossible to arbitrarily orient
the crystal before each frame is recorded. Furthermore, accelerating
the stage to the desired rotation rate and precisely timing frame
capture to the appropriate angular interval requires a level of synchronization currently not available to electron microscopes. These
limitations are sidestepped by collecting data in “rolling shutter”
mode [40], such that the detector is constantly active while the
crystal is continuously rotated in the beam [14]. This is only
possible on fast cameras with short dead-times during readout,
lest reflections that pass through their diffractive conditions while
the camera is blind will be lost and introduce unwanted gaps in the
measurement of reciprocal space. Indeed, this shutterless mode of
data acquisition has become prevalent in synchrotron X-ray crystallography as well [41].
5 Conclusion and Outlook
Choosing a suitable exposure and rotation range for a given crystal
requires information not only about the lattice orientation of the
sample currently in the beam, but integrated results from all previously collected datasets. Only then can the experimenter make
informed decisions on how to distribute the probing electrons
across the rotation range accessible to the instrument, such that
data acquisition eventually leads to the best possible model. This is
the purpose of data collection strategy [42]. As integrated data
collection, processing, and automated strategy optimization tools
are not yet commonplace in MicroED, these decisions must currently be made by the experimenter.
Here, techniques to limit the total exposure, without unduly
compromising completeness or resolution have been outlined. The
direct relationship between exposure, dose, and radiation damage
suggest that exposure reduction is a sensible approach for structure
determination by MicroED. This is further motivated by the observation that the most interesting parts of a structure are often the
ones most susceptible to damage. A possible reason could be that
these parts are often in strained geometries that destabilize the
amino acid structure [4].
316
Johan Hattne
solution is to collect data in two passes [39]: an initial
low-resolution pass using low exposure is aimed at recording only
those reflections that would cause overflows at higher exposure. A
subsequent high-resolution pass will then record the highresolution reflections using a higher exposure, without concern
for simultaneously preserving the integrity of the low-resolution
information.
An additional requirement on a MicroED camera is speed. The
absolute position of the stage in an electron microscope can usually
not be set accurately, which makes it impossible to arbitrarily orient
the crystal before each frame is recorded. Furthermore, accelerating
the stage to the desired rotation rate and precisely timing frame
capture to the appropriate angular interval requires a level of synchronization currently not available to electron microscopes. These
limitations are sidestepped by collecting data in “rolling shutter”
mode [40], such that the detector is constantly active while the
crystal is continuously rotated in the beam [14]. This is only
possible on fast cameras with short dead-times during readout,
lest reflections that pass through their diffractive conditions while
the camera is blind will be lost and introduce unwanted gaps in the
measurement of reciprocal space. Indeed, this shutterless mode of
data acquisition has become prevalent in synchrotron X-ray crystallography as well [41].
5 Conclusion and Outlook
Choosing a suitable exposure and rotation range for a given crystal
requires information not only about the lattice orientation of the
sample currently in the beam, but integrated results from all previously collected datasets. Only then can the experimenter make
informed decisions on how to distribute the probing electrons
across the rotation range accessible to the instrument, such that
data acquisition eventually leads to the best possible model. This is
the purpose of data collection strategy [42]. As integrated data
collection, processing, and automated strategy optimization tools
are not yet commonplace in MicroED, these decisions must currently be made by the experimenter.
Here, techniques to limit the total exposure, without unduly
compromising completeness or resolution have been outlined. The
direct relationship between exposure, dose, and radiation damage
suggest that exposure reduction is a sensible approach for structure
determination by MicroED. This is further motivated by the observation that the most interesting parts of a structure are often the
ones most susceptible to damage. A possible reason could be that
these parts are often in strained geometries that destabilize the
amino acid structure [4].
316
Johan Hattne