Production of Xrays

Historically, elapsed times to record one single XRD exposure have ranged from minutes to weeks. The reason for the long XRD exposure times has been the very weak X-ray beams available. The weak X-ray beams have resulted from the inability to use the bulk of the radiation arising in a conventional X-ray generator. Almost all of the X-rays arising inside a conventional generator go off in the wrong direction and are lost.

Note that X-rays cannot be focused with glass lenses in the way that light is: refraction of X-rays is poor and the absorption by glass is too strong. Mirrors employing grazing incidence are commonly used to reflect X-rays, but the maximum angle off the surface is a scant 1/4° (Guinier, 1963). Hence, not much of the total X-ray production can be gathered by such mirrors. Only about one part in one million, or even ten million, of the X-rays generated inside the X-ray tube are going in the right direction to reach the specimen. In addition, only a fraction of this radiation has the desired wavelength. Thus even though a kilowatts-worth of radiation can be generated easily, much less than a milliwatts worth of radiation is present in the finely collimated, monochromatic beam required for the XRD experiment.

The advent of synchrotron radiation sources is radically changing the situation. In the United States, there are currently synchrotron radiation laboratories at Stanford University (SSRL: Stanford Synchrotron Radiation Laboratory); Cornell University (CHESS: Cornell High Energy Synchrotron Source); Brookhaven National Laboratory (NSLS: National Synchrotron Light Source); the University of California at Berkeley (ALS: Advanced Light Source); and the University of Wisconsin. About one year from now a more technically advanced source will open at Argonne National Laboratory

(APS: Advanced Photon Source). There are also sources in various other countries, including England, France, Germany, Italy, and Japan. These sources generate X-rays with an efficiency similar to the efficiency with which light is generated in a laser.

The phenomenon of synchrotron radiation is not new. From the inception of high-energy accelerators, physicists carrying out experiments at the above laboratories have paid a high price: the continual loss of valuable energy as elementary particles, moving inside a vacuum tube, follow a circular path within "bending magnets". Particle energy is lost in the form of light, ultra-violet rays and, principally, X-rays.

The radiation process is analogous to the emission of radio waves from a radio tower, but with a twist: Because of a relativistic effect, the radiation from the high-energy particles cannot go off in all directions, as for a radio antenna. Instead, the radiation goes off in the near-forward direction. The result is a highly directional X-ray beam. The X-ray beam so generated has been compared to the light coming from the headlamp of a locomotive moving around a bend in the tracks.

Because of the directional emission, a much higher proportion of the X-rays in a bending magnet can be used in the XRD experiment. As a result, instead

01 iu photons, as in an experiment using a conventional generator, the beams now produced at these sources have up to 10 photons per second. Even higher fluxes are expected in the near future as "third-generation" sources start up at various laboratories around the world.

Further spectacular developments are under way. Devices now being reduced to practice at synchrotron sources will deliver X-ray beams with energy fluxes ranging from 10 to 350 watts per square millimeter. These fluxes are comparable to the heat and light radiating from the surfaces of meteoroids entering the earth's atmosphere -100 to 60^ watts per square millimeter. (In comparison, the Sun radiates a mere 60 watts/mm at its surface.)

2.2. Detector technology

Another key development is the technology for recording and storing XRD patterns. As might be expected, development of detector systems has been stimulated by the production of the intense synchrotron beams.

From the time of their discovery 100 years ago, X-rays have been detected both electronically and by photographic film. A relatively recent development are the position-sensitive X-ray detectors (PSD's). An outgrowth of detectors used in high-energy physics, PSD's use fast electronic processing techniques. They register, not only the presence, but also the location of X-ray photons, processing each photon in less than 1 microsecond.

PSD's offer several advantages over previous techniques for recording and processing XRD patterns. They record photons all but simultaneously at the various points in a diffraction pattern, whereas a scintillation counter is stepped sequentially from point to point. Film also records simultaneously at the various points, but PSD's have the advantage of recording photons individually against a low background of noise (cf. the ubiquitous photographic 'fog'). Also, natural background is low since X-ray sources are rare. Finally, PSD's avoid the time-consuming and tedious photographic development process: instead of physically replacing one piece of photographic film with another, these devices allow the rapid electronic transfer of each succeeding diffraction pattern out of the detector memory and into an electronic storage device. Thus the entire XRD pattern is stored away in 1 millisecond or less. Once the pattern is placed in storage, the detector is ready to start the next sub-second exposure. Other kinds of detectors with similar capabilities are being perfected.

Taken together, the developments in efficient X-ray production, recording of low-noise patterns, and rapid storage of multiple patterns mean that a new class of dynamic experiments becomes possible. Using these twin technologies ~ high-intensity synchrotron X-ray beams and fast electronic detector systems ~ allows one to do 'time-slice' exposures at a rate that was once inconceivable. After summarizing work in biological kinetics, two examples from the study of fat crystallization will be presented, that demonstrate second and even sub-second time resolution.

0 0

Post a comment