but the speed and improved counting statistics of an ion chamber or pin diode. The
development of electronic area detectors follows this struggle – striving for the
highest possible contribution to spatial detail and angular coverage while also
producing statistically reproducible intensity measurements. A range of different
technologies was investigated. Multiwire proportional chamber (MWPC) gas detectors were capable of detecting many simultaneous diffraction spots but had limited
global count rates. They were superseded by the much more widely used image plate
detectors and CCD detectors. The image plate detector, which is still in use, has an
extremely large format with excellent resolution, low noise and high sensitivity. It is
ideal for studies requiring long exposure times, but the duty cycle, particularly the
readout time, is considerable for experiments requiring several images. In comparison CCD detectors have a similar pixel size to image plates but have a significantly
faster duty cycle, and, although they have greater noise than image plates (both dark
current and readout noise), the very short exposure times required for single-crystal
diffraction mean that the background counts generated by the electronics on each
exposure are negligible compared to the signal. With the advent of large-format
CCD detectors, the use of image plate-based systems declined rapidly.
CCD detectors have been in common use for over 20 years and are considered by
many to be reaching the technical limit of their capabilities. Arguably they provided
a step change for the diffraction experiment upon their introduction, and the CCD
detector, as known to X-ray crystallography, consisted most simply of a visible lightsensitive CCD sensor and an X-ray phosphor. Through the photoelectric effect, the
visible light produces a charge in the pixels, and voltage changes push the charges on
the pixels through neighbouring pixels out to a limited number of amplifier nodes.
Here the charges are converted to an analogue voltage, which then needs translating
into digital counts for the pixels. This readout process, where charges are shifted
through rows of pixels, requires the sensor to be dark, i.e. no X-ray photons
impinging on the sensor, to maintain the integrity of the intensity counts and
positions [21]. Advances in chip manufacture lead to the development of CMOS
chip-based detectors. The primary difference between CCD and CMOS [22] chips is
that each CMOS pixel has its own analogue to digital converter, so that an extended
dark period is not required for readout. The pixel readout is parallelised in CMOS
detectors, with pixels read out sequentially by multiplexers dedicated to a row or
region of the detector. With multiple multiplexers per detector, the readout of the
entire detector can be extremely quick. Monolithic CMOS detectors available today,
such as the Bruker PHOTON series, detect visible light created by an X-ray
phosphor and offer high-speed data collections due to their ability to collect continuous shutterless rotation data with no count rate-based saturation.
Along with the evolution of detector chip architecture, sensor materials have also
been evolving. Modern microchip fabrication techniques have allowed for another
detector advance – the technique of bump-bonding which joins a sensor module to
the CMOS chip with a microscopic solder bump. This technique allows the sensor
material and the chip material to be decoupled and optimised separately. These
Hybrid Photon Counting (HPC) detectors, originally developed for particle tracking
(e.g. at CERN), allow for faster and more sensitive data collection. These detectors
Leading Edge Chemical Crystallography Service Provision and Its Impact on. . .
75
development of electronic area detectors follows this struggle – striving for the
highest possible contribution to spatial detail and angular coverage while also
producing statistically reproducible intensity measurements. A range of different
technologies was investigated. Multiwire proportional chamber (MWPC) gas detectors were capable of detecting many simultaneous diffraction spots but had limited
global count rates. They were superseded by the much more widely used image plate
detectors and CCD detectors. The image plate detector, which is still in use, has an
extremely large format with excellent resolution, low noise and high sensitivity. It is
ideal for studies requiring long exposure times, but the duty cycle, particularly the
readout time, is considerable for experiments requiring several images. In comparison CCD detectors have a similar pixel size to image plates but have a significantly
faster duty cycle, and, although they have greater noise than image plates (both dark
current and readout noise), the very short exposure times required for single-crystal
diffraction mean that the background counts generated by the electronics on each
exposure are negligible compared to the signal. With the advent of large-format
CCD detectors, the use of image plate-based systems declined rapidly.
CCD detectors have been in common use for over 20 years and are considered by
many to be reaching the technical limit of their capabilities. Arguably they provided
a step change for the diffraction experiment upon their introduction, and the CCD
detector, as known to X-ray crystallography, consisted most simply of a visible lightsensitive CCD sensor and an X-ray phosphor. Through the photoelectric effect, the
visible light produces a charge in the pixels, and voltage changes push the charges on
the pixels through neighbouring pixels out to a limited number of amplifier nodes.
Here the charges are converted to an analogue voltage, which then needs translating
into digital counts for the pixels. This readout process, where charges are shifted
through rows of pixels, requires the sensor to be dark, i.e. no X-ray photons
impinging on the sensor, to maintain the integrity of the intensity counts and
positions [21]. Advances in chip manufacture lead to the development of CMOS
chip-based detectors. The primary difference between CCD and CMOS [22] chips is
that each CMOS pixel has its own analogue to digital converter, so that an extended
dark period is not required for readout. The pixel readout is parallelised in CMOS
detectors, with pixels read out sequentially by multiplexers dedicated to a row or
region of the detector. With multiple multiplexers per detector, the readout of the
entire detector can be extremely quick. Monolithic CMOS detectors available today,
such as the Bruker PHOTON series, detect visible light created by an X-ray
phosphor and offer high-speed data collections due to their ability to collect continuous shutterless rotation data with no count rate-based saturation.
Along with the evolution of detector chip architecture, sensor materials have also
been evolving. Modern microchip fabrication techniques have allowed for another
detector advance – the technique of bump-bonding which joins a sensor module to
the CMOS chip with a microscopic solder bump. This technique allows the sensor
material and the chip material to be decoupled and optimised separately. These
Hybrid Photon Counting (HPC) detectors, originally developed for particle tracking
(e.g. at CERN), allow for faster and more sensitive data collection. These detectors
Leading Edge Chemical Crystallography Service Provision and Its Impact on. . .
75