The CHESS capillary optics group designs a capillary profile and then draws hollow glass tubing into precision single-bounce x-ray optics for microbeam optics at CHESS and other laboratories.
The applications range over protein crystallography for biology, the study of fish ear stones to learn about their life history (environmental science), to solving problems in art history of paintings underneath paintings for museum masterpieces, etc.
We are continually trying to improve the quality of the optics by making more precisely-figured parts as we improved the hardware/software tools on our LabView operated, custom-built drawing tower consisting of an precision linear air bearing + small-bore electric furnace. The instrument allows us to soften the wall the borosilicate glass tubing of a few mm in diameter and stretch it to the proper shape (elliptical, parabolic, or linear) as determined in advance by an web-based mathematical design program.
The hope in the long run is to make higher quality capillaries with slope errors below 20 to 50 microradians - our present level of performance - and figure errors of less than 1 to 2 microns. This will help us to make smaller diameter microbeams. The puller mechanics are undergoing some performance testing at present as we continue to produce specialized parts for various x-ray experiments and optimize the shape for each application.
We continue to do "detective work" to try to figure out where the current slope errors are generated by the current drawing process and try to beat them down to even smaller levels.
For the last several years we have been operating a puller designed around an ABTech air bearing slide with on-board metrology obtained through a dual Keyence High Speed LED/CCD Optical Micrometer for profiling the glass before and after drawing.
We are presently reaching the limit of our measuring equipment. For some of the optics under 5 cm in length, we have achieved surface figure errors under 1 micron and slope errors under 10 microradians (we are at metrology limits so we don't know what the error bar is on these measurements). Some advancements in the way we pull the capillary has come from pulling at constant tension to pulling at constant pressure, an advance made by Dr. Heung-Soo Lee from the Pohang Accelerator Laboratory in Korea who helped us modify the capillary design program to do “reverse” pull (extension goes in opposite direction of furnace motion). He also has done some ray tracing calculations to try to model the inner surface of the capillary from the far-field pattern For more information click here. Sterling Cornaby, (shown here receiving an award at the Denver X-ray Conference in 2008), a PhD student in Applied Physics, graduated in 2008 and is now working at Mostek. His PhD thesis on capillary drawing and applications can be download for more detailed information on our level of practice.
We are currently trying to find a way to measure the inner profile of the capillary. One way is to use a small x-ray spot to move along the capillary inner wall and record the position of the focused beam in the far-field. With the help of an Applied Physics undergraduate student, Gavrielle Untracht, we are in the process of measuring the inner profile by ray tracing the output spot back to the source through the capillary. First preliminary results are just now be obtained.
We are moving forward in the reduction of noise during pulling and profile measurements on the puller. Both air noise from the air bearing and vibration noise from motor motion was examined. The furnace carrier was stiffened and stiff heavy braced mounts were added to the capillary pulling stage. (See figure at right).
Future challenges and upgrades:
- Many capillaries have a small mismatch between calculated and measured profiles near the tip. We need to modify the program that generates the "furnace " file to take this problem into account.
- Possibly use an optical laser microscope to measure outer and inner surface of capillary to submicron accuracy.
- Cut a capillary in half and attempt to use differential deposition to smooth out inner surface.