Synchrotrons were originally developed for research using the electron beams that they contain, and the x-ray radiation they generated was considered a nuisance. Eventually, scientists started using larger synchrotrons for the highly collimated x-ray beams that they produce during hours when the beams were not being used for research. The second generation of synchrotrons were built specifically as x-ray sources, primarily using bending magnet sources, which produce a broad wavelength spectrum with a fairly large source effective source size. The third generation of synchrotrons were built around use of insertion devices, that produce a strongly-peaked x-ray spectrum and have a much smaller source size. The fourth generation of synchrotron sources reduce the size of the x-ray beam considerably generating even smaller effective source sizes from insertion devices and much higher coherence in the x-ray beams. The coherence is of no advantage for powder diffraction, but the small source size of insertion devices, particularly at 4th generation sources, is definitely beneficial to design of powder diffraction instrumentation.
There is a plethora of designs for synchrotron powder instruments, but nearly all use a double-bounce incident-beam monochromator, where the x-ray beam is diffracted from two perfect crystals to remove all but one precise wavelength of x-rays from the beam. Since perfect crystals are used, the wavelength spread is very small, which in turn allows for very high resolution instruments. While a survey of the different types of synchrotron instruments across the world would itself make for an interesting book, I will attempt to categorize them here.
These instruments place a perfect-crystal monochromator between the sample and the detector. This analyzer creates an instrument that has excellent signal-to-noise as the monochromator will only transmit x-rays that have precisely the right wavelength and originate from the right location it will also have very narrow peak widths, limited usually by the performance of the sample. The analyzer detector is superb at rejecting fluorescence and the very limited angular acceptance of the perfect crystal creates a parallel-beam optic that means that a larger sample size does not degrade resolution. The advantages of a perfect-crystal analyzer were demonstrated by Dave Cox at Brookhaven National Lab in the 1980’s. and it was he who constructed the first dedicated powder instrument of that type at Brookhaven’s NSLS source. The downside of use of three crystals in the beamline optics is that only a very small fraction of the photons from the source can be used for measurements and most ultra-high resolution powder diffractometers use multiplexed detectors to speed their measurements. In such instruments data collection speeds are typically in the range of 10 minutes to an hour. The monochromator-detector assembly is quite massive and these instruments are quite large and expensive.
It is possible to make detectors that will record the position of an incident photon along the length of the detector. Hence they are known as linear position-sensitive detectors. They may be flat or curved. It is now possible to build an instrument that covers the entire angular range for measurement with no gaps, but older designs did leave gaps and the detector would need to be rotated in order to collect an entire pattern.
Linear PSDs can be constructed with very high angular detection precision, so their angular resolution can come close to that of the ultra-high resolution instruments with perfect-crystal analyzers, but the resolution depends on having a very small sample. The larger the sample, the more the uncertainty in where the x-ray originated, which directly translates to wider peaks. Linear detectors may have the ability to electronically filter out some fluorescence but are not as effective as a perfect-crystal analyzer and will also count any photon that reaches the detector regardless of the angle of origin, so they are not as good at rejecting air scattering and other sources of stray photons, such as scattering from slits, furnaces and refrigerators, so linear PSD instruments will have significantly higher backgrounds than ultra-high resolution measurements, but will be much faster: seconds to minutes. The as the resolution of the linear detector dictates the ultimate resolution, there becomes a tradeoff between angular range coverage, resolution and cost for the instrument. The size is quite variable across institutions but these instruments are in general quite expensive.
Initially, electronic area detectors for x-rays were developed as a replacement for x-ray film in medical procedures and for this we can be very thankful as it significantly reduced the amount of x-rays that we needed to be exposed to for medical and dental diagnosis. These detectors were valuable for diffraction instrumentation, but since that time, detectors that are more specialized for x-ray science have been developed. These detectors are faster in their readout times, have progressed towards smaller pixels and to larger sizes. The fastest detectors with the largest numbers of pixels can alone cost millions of dollars.
Area detection has revolutionized powder diffraction measurements for pair distribution function (PDF) determination. These require a short x-ray wavelength to obtain patterns collected over a very wide Q range. The work of Pete Chupas and others showed that with an area detector placed close to a sample, one could collect a full pattern for a PDF in seconds rather than take scans that took a large fraction of a day for a single PDF. Likewise, while I used a point detector for single-crystal diffraction in my student days, I do not think that anyone would consider using a point detector for this in the current century. Area detectors provide the most efficient diffraction measurements as the capture the greatest solid angle of scattering from the sample, so count rates are exceedingly high; measurements can be as quick as the detector electronics allow. There is typically no collimation possible between the sample and detector, so backgrounds can be quite high.
The resolution possible with an area detector is determined by the size of the sample, the pixel size of the detector, relative to the distance between the sample and the detector so there is a playoff between angular range and resolution for a particular size of detector, which is why larger and higher pixel density detectors are preferred. Powder diffraction usually does not demand the highest detector readout rates needed for many other techniques. My personal thinking is that while PSD-based instruments provide the most convenient data collection, an area detector instrument can offer equivalent resolution and much faster data collection. Having been involved with ultra-high resolution diffraction since my days with Dave Cox, this is my favorite choice for x-ray diffraction.