Once the experiment has been performed, the data stream can be processed to
reconstruct 2D images, if needed, or, for example, to count events in a region of
interest (e.g. a single diffraction peak). The key advantages of this readout method
are that (1) it eliminates detector readout time and (2) it allows the time resolution of
the experiment to be decided upon and adjusted after data collection, without
repeating the experiment.
It can be seen from this discussion that – with the possible exception of experiments using XFEL sources – HPC detectors are much better suited to TR experiments than conventional CCD and CMOS detectors. As well as providing better
quality data, including, importantly, the ability to accurately measure strong and
weak reflections in a single image, HPC detectors can also provide much better time
resolution, and this can be adjusted to suit different types of experiment. Most
synchrotron sources now use HPC detectors, and despite their higher cost they are
becoming more commonplace on laboratory instruments. The recent developments
towards continuous readout are even more interesting, since these effectively push
the time resolution close to the limit of the count rate and into the microseconds to
hundreds of nanoseconds range.
4.5 Sample Delivery
For many TR-SCXRD experiments, obtaining a complete time series from the one
crystal may not be feasible. Samples may degrade with exposure to high-intensity
pump light and/or high-flux X-ray beams, and structural changes over many repeated
cycles may lead to accumulation of strain and fatigue. Crystal damage is a particularly important consideration in experiments using high-brilliance X-ray sources
such as synchrotrons or XFELs and high-intensity pulsed lasers, though ideally
should be considered in any photocrystallographic experiment regardless of the
instrumentation and intended time resolution. Also, if an irreversible reaction is
being studied, then it is clearly not possible perform complete pump-(multi)probe
experiments on the same crystal.
These situations call for rapid replacement of the sample and what are commonly
termed “multi-crystal methods”. Multi-crystal methods have mainly been developed
to cater to serial femtosecond crystallography (SFX) experiments conducted at
XFEL sources but are also applicable to synchrotron studies. During XFEL experiments the high-intensity X-ray beam effectively destroys the crystal, and a new
sample is required for each pulse (a so-called “diffract-and-destroy” approach)
[78]. Multi-crystal sample delivery methods are therefore integral to these studies.
The numerous techniques that have become established in recent years can be
separated into two main categories, viz. injection and fixed-target methods.
Injection methods inject microcrystals suspended in a carrier medium, which in
many cases is the crystallisation liquor, into the path of X-ray beam. A variety of
injection systems have been developed [79, 80]. While relatively simple, this
technique suffers from a low hit rate – i.e. most X-ray pulses will fail to hit a
Watching Photochemistry Happen: Recent Developments in Dynamic Single-Crystal. . .
231
reconstruct 2D images, if needed, or, for example, to count events in a region of
interest (e.g. a single diffraction peak). The key advantages of this readout method
are that (1) it eliminates detector readout time and (2) it allows the time resolution of
the experiment to be decided upon and adjusted after data collection, without
repeating the experiment.
It can be seen from this discussion that – with the possible exception of experiments using XFEL sources – HPC detectors are much better suited to TR experiments than conventional CCD and CMOS detectors. As well as providing better
quality data, including, importantly, the ability to accurately measure strong and
weak reflections in a single image, HPC detectors can also provide much better time
resolution, and this can be adjusted to suit different types of experiment. Most
synchrotron sources now use HPC detectors, and despite their higher cost they are
becoming more commonplace on laboratory instruments. The recent developments
towards continuous readout are even more interesting, since these effectively push
the time resolution close to the limit of the count rate and into the microseconds to
hundreds of nanoseconds range.
4.5 Sample Delivery
For many TR-SCXRD experiments, obtaining a complete time series from the one
crystal may not be feasible. Samples may degrade with exposure to high-intensity
pump light and/or high-flux X-ray beams, and structural changes over many repeated
cycles may lead to accumulation of strain and fatigue. Crystal damage is a particularly important consideration in experiments using high-brilliance X-ray sources
such as synchrotrons or XFELs and high-intensity pulsed lasers, though ideally
should be considered in any photocrystallographic experiment regardless of the
instrumentation and intended time resolution. Also, if an irreversible reaction is
being studied, then it is clearly not possible perform complete pump-(multi)probe
experiments on the same crystal.
These situations call for rapid replacement of the sample and what are commonly
termed “multi-crystal methods”. Multi-crystal methods have mainly been developed
to cater to serial femtosecond crystallography (SFX) experiments conducted at
XFEL sources but are also applicable to synchrotron studies. During XFEL experiments the high-intensity X-ray beam effectively destroys the crystal, and a new
sample is required for each pulse (a so-called “diffract-and-destroy” approach)
[78]. Multi-crystal sample delivery methods are therefore integral to these studies.
The numerous techniques that have become established in recent years can be
separated into two main categories, viz. injection and fixed-target methods.
Injection methods inject microcrystals suspended in a carrier medium, which in
many cases is the crystallisation liquor, into the path of X-ray beam. A variety of
injection systems have been developed [79, 80]. While relatively simple, this
technique suffers from a low hit rate – i.e. most X-ray pulses will fail to hit a
Watching Photochemistry Happen: Recent Developments in Dynamic Single-Crystal. . .
231