This subproject is one of many research subprojects utilizing theresources provided by a Center grant funded by NIH/NCRR. The subproject andinvestigator (PI) may have received primary funding from another NIH source,and thus could be represented in other CRISP entries. The institution listed isfor the Center, which is not necessarily the institution for the investigator. Objectives Our aim for X29 is to use it both as the venue for highly responsive production of excellent diffraction data, and as a workbench for creating new methods that improve speed and accuracy. We expected to explore use of the automounter and to develop improved software and methods to allow remote observation and operation, the interleaving of many projects, and improved handling of every experimental parameter from crystal centering to custom tuning of x-ray optical systems. Results The 'High Speed' aspect of beamline X29 has helped to drive our operations and development program in 2006. During the '06-'07 grant year there have been 647 scheduled data-collection sessions, 34 of which employed the automounter. There was a large Mail-in operation, and this employed about 167 visits of the beam time. During 0.5% of the time, the automounter was in operation, and about one quarter of this was local scientists doing Mail-in work. X29 is increasing in productivity, as measured by the standard metrics of the time: our records show that during 2006 there were 84 PDB depositions attributed to X29, and 55 publications, five each in Nature and Cell. This is a good record, and it's about twice the numbers reported in 2005. X29 is coming of age. An early feature of our operation of X29 has been to schedule short visits, keeping 'prime time,' that is 7AM-3PM on weekdays, for short experiments and development projects. Last year we also invented the 'quick ID project,' which was an opportunity for users at dipole beamlines, even those not our own (X4, X6), to ask for a few hours on X29 to take a seminal data set. During the course of this last year we have been able to employ the PX Operators more intensely in getting users started on the pre-scheduled runs and especially in moving them in to do quick projects. The new 'calendar' function in PXDB (see software) is extremely valuable in getting this to work well. Continual efforts to diagnose problems and improve performance have produced incremental progress. We got Crystal Logic to improve the synchronization of the omega/shutter operation, and now it appears that exposures down to 100ms can be made with essentially perfect timing. A common theme for most data-collection projects is the attempt to minimize x-ray damage to crystals. We reason that one can get the most out of a crystal by attempting to maximize the signal-to-noise ratio throughout the measurement. To help this, we wanted to be able to tailor the x-ray beam more precisely to the specimen and the mode of data collection. Obviously it pays to make certain the x-ray beam is no bigger than the actual physical size of the crystal, and the motorized slits accomplish this very nicely. In addition one might alter the convergence of the x-ray beam coming from the two focusing elements, the goal being to make the reflections as tight and sharp as possible on the detector. In the case of vertical focusing, we open up that focus to give a beam about 100 m high at the specimen position that is nearly parallel. In the case of horizontal focusing, we can be subtle. To obtain the maximum x-ray intensity, we focus the beam on the limiting slit to get the most light through. The horizontal divergence would be 1mrad; with a specimen-to-detector distance of 200mm the spots on the detector would be widened by 0.2mm. If we relax the bending on the crystal, putting the focal point downstream of the limiting slit, the edges of the beam are cut off by the slit and the result is that the beam hitting the crystal has a lower crossfire. Therefore our standard operating procedure is to produce a beam with a 0.4mrad convergence (which is about the maximum intensity that is possible). When the detector is far back, say <400mm, we use 0.2mrad convergence with a concomitantly lower intensity. The very clear advantage of this is that slightly longer exposure times are rewarded by smaller spots on the detector and much improved signal-to-noise ratios for the data, thereby improving the output from each crystal. X29 has also been a test bed for software development, especially driving the use of the dna automation system. Also we are exploring the possibility of automating the sequential use of specific parts of an individual large crystal, probing it with a small beam. These are described in the Software section. The beamline has experienced two significant mechanical failures during the last year. The main x-ray shutter, which is behind the x-ray ring shield wall and therefore difficult to get at, had a complex failure that involved damage to a water-cooling line in June '06. The technical staff from our partners at Case Western Reserve and the NSLS beamline staff mobilized and got it fixed in 60 days. Then more recently, in March, a complex failure of the cryogenic cooling system for the monocromator failed, and this ruptured the seal inside the vacuum chamber. The CWR team managed to fix this, and to add safety telemetry to prevent a recurrence, with only 1.5 days of missed data collection. Plans Things are running extremely well, but we have incremental improvements in mind. The plan to produce a queuing scheme that will allow the PXOps and dna to take data for remote users is badly needed here and we expect to drive that development. Users complain regularly about crystal visualization, and we plan to mount a very high quality digital video camera to help with that (see software). The plan to allow sequential probing of pre-selected regions on a crystal is really driven by the rapid damage crystals experience at X29, and we will work to get this running during the next grant year.
Showing the most recent 10 out of 167 publications
