Grant Bunker, Director

Report 3/2/99

The last couple of months we have been working hard to make the beamline an exemplary XAFS and solution scattering beamline - it already works very well indeed for fiber diffraction (tune in for some newsworthy fiber diffraction results in Nature March 4). We are about 2/3 the way through our first five year grant, and we intend to submit a competitive renewal application on June 1. Our short term priorities are to fully develop and demonstrate the capabilities of the beamline, expand our initial scientific productivity, and make the transition from "commissioning mode" to "operations mode" within the next few months. Recent experiments have focussed on getting our XAFS capabilities up to speed, and we have made significant strides in that direction. We still have to work on identifying and minimizing sources of non-statistical noise at high k in dilute fluorescence EXAFS data and to streamline our user interface.

The big news is that we have completed all the optical systems for the beamline, a significant milestone. In addition we have been making innumerable incremental improvements to the beamline.

The commissioning window, which used to keep our beamline vacuum separated from the ring vacuum, was removed several months ago to permit operation at lower energies, down to the undulator limit of 3.2 KeV. Beamline vacuum and ring vacuum are now contiguous but a differential pump allows us to have a slightly higher pressure in the beamline (say 10^-8 or 10^-9 torr upstream).

The liquid nitrogen cooled Si(111) crystals and Messer Griesheim liquid nitrogen cooler were installed; they have been working beautifully for the last month or so, except we have had one failure of the LN2 seals at the end of the last run. Fortunately the fast valves and equipment protection system prevented serious damage. Overall we have been very pleased with the performance of the crystal pressure seals, which have been a real nuisance (failure after repeated temperature cycling) in the earlier designs that are used by numerous CATs. This second generation crystal is a joint effort of the three CSRRI CATS (MR, IMCA, Bio) and SBC-CAT.

We are also very pleased with the performance of the Messer LN2 cooler. Very robust, doesn't ice up, easy to maintain. It just works. Its function is to pump pressurized LN2 through the crystals and dissipate the waste heat by boiling off LN2 in an ambient pressure bath.

The crystal was successfully tested to 4 KeV; we think it will work well down to 3.2 KeV once we have made minor modifications to the crystal mount. We also sciton polished the crystal to reduce some spatial structure in the beam we think was due to structure in the crystal surface. The beam is partially coherent and such surface structure could be a problem.

We also have a second monochromator with Si(400) crystals. This covers the range from 8 KeV to 35 KeV, presently using water cooled crystals. These can go up to 70 KeV in second harmonic, though we haven't tried it yet. The Si(111) crystals cover the range from 3.2 (undulator lowest energy) to 15 KeV, with about twice the flux of the Si(400). Once we have evaluated the performance of the LN2 crystal design, we will tweak it and build "final" versions of LN2 cooled crystals, both Si(111) and Si(400). Please let us know if you think Si(220) crystals are needed.

The beamline continues to work beautifully for fiber diffraction, and we are working to make it a great XAFS line, but sadly, we're not there yet. For garden variety XAFS, using a fluorescence ion chamber, it works well (>2*10^13 photons/sec) despite an intermittent problem with a monochromator trolley that is used to keep beam centered on the second crystal. We are shooting for a new level of position stability while scanning the monochromator, approximately 10 microns (we can focus vertically to less than 20 microns). At the time of this writing, we can keep the beam position fixed to within +- 50 microns over a 1 keV scan, without feedback. By implementing hardward feedback we should be able to reduce the beam position variation even more. Such a high degree of position stability is desirable for the multilayer analyzer and laue analyzers we're developing, as well as other demanding experiments. We have the goal of pushing to <100uM concentrations and the analyzers are needed, because conventional detectors saturate under the high fluxes available at the APS.

For ordinary use (e.g. XAFS) an improved GUI has been implemented for the MX beamcontrol software (which talks to EPICS for most functions); a CLI with scripting is also available. We have found this MX on top of EPICS architecture to be very flexible; weird hacks are easily accomplished. Primary limitations are the lack of QXAFS capability and rapid time slicing. These require additional low-level programming which is in our future plan.

There are many other things in the works, so stay tuned...