Diamond Radiator R&D Phase I
Phase I of diamond radiator R&D took place during the period September 2009 - May 2011. The goals of diamond radiator R&D in Phase I were to optimize diamond radiator crystal quality assessment techniques using X-rays, and using these techniques to diagnose the issues that, up to the present, have prevented the effective use of diamond crystals as thin as 20 microns in polarized gamma ray sources. A sequence of three experimental runs were conducted at CHESS, mapping the rocking curves of a variety of diamond samples, ranging thickness from 300 microns down to as thin as 9 microns. Two factors were identified during these studies as the main performance limiters.
- vibrations -- This is particularly a problem for diamonds mounted on stretched wires, which is the usual way to hold diamond radiators in a target ladder. As a general principle, the way to reduce vibrations is to employ a mount whose fundamental vibration frequency is high, while at the same time presenting a negligible scattering cross section in the vicinity of the crystal.
- warping -- This is increasingly a problem as diamond thickness is reduced. Whole-crystal rocking curve broadening on the order of several mrad (compare with GlueX requirement of 20 μrad RMS) were seen in all 3 crystals less than 20 microns thick that were examined at CHESS. Two of these had never been exposed to an electron beam, so the warping cannot be attributed to radiation damage. A technique must be found to thin diamonds down to 20 microns that does not result in a free-standing crystal that is this badly warped, or else a mounting technique must be found which can hold the crystal flat once it has been thinned to 20 microns.
A Vibration-Free Sample Mount for CHESS
As an intermediate step in the above investigations, we needed to find a way to hold diamond samples in the target holder at CHESS, and allow samples to be moved in and out of the mount without risk of breaking the samples or damaging the mount. The first design was a stretched wire mount similar to the one used to hold diamonds in the Hall B target ladder at Jlab.
However, this proved not to work, due to excessive vibrations. An elegant solution to this problem was found in the form of a pair of tightly-stretched mylar films attached to aluminum rings. The sample is placed in the middle of one of the mylar films, then the other film is placed over it like a sandwich and the two rings bolted together. Van der Waals forces between the sample and the diamond are sufficient to prevent the sample from moving between the mylar sheets while the target is rotated in the goniometer. The fabrication technique described below is capable of producing stretched films whose flatness in the central 1 cm region is better than 20 μrad RMS, and whose fundamental resonance is in the vicinity of 1 kHz.