In February 2015, G1 achieved 40% more flux at 50 mA than its previous high number obtained with the wiggler source operating at 200 mA, under otherwise identical conditions. At G3 in December 2014, the improvement was even more pronounced. A flux of over 1014 photons/sec was measured in a 1 mm2 at 11.2 keV with 120 mA positron current, approximately 2.5 times the prior G3 record. The larger gain at G3 results from the larger current for the G3 measurement, and from different focusing conditions employed for these two measurements. In both cases, the undulator upgrade represents substantial gains for G-line users.
In addition to these initial gains, a significant optics upgrade at G-line promises even further increases in flux. Even before the undulator upgrade, it was well-known that G-line’s maximum flux was limited by so-called “heat bump” problem — thermal deformations induced on the upstream monochromator by the high (several-kW) heat-load incident beam.
During the December 2014 measurements at G3, and thanks to unique detector capability brought by the user group at that time (see caption below), CHESS staff obtained a unique view of the effect of this thermal bump. The image in figure 1 was generated by a prototype, 1024-pixel imaging detector fabricated from a 68-micron thick diamond single crystal. The detector is 96.4% transparent to the x-ray beam, and electrodes deposited onto the front and back surfaces of the diamond allow the photocurrent in each pixel to be independently measured. The image clearly shows three lobes in the vertical direction, resulting from thermal-induced deformation of the upstream multilayer. (That this is the case is known from many previous tests at varying positron current.) Vertical-focusing mirrors are often used at G-line to make these lobes coincident and hence maximize flux on the sample. But this solution is far from ideal, since focusing introduces vertical divergence. Also, since these different lobes differ from one another slightly in photon wavelength, they cannot all diffract optimally from the 2nd multilayer monochromator, reducing the total transmitted intensity.
In metallic x-ray mirrors made specifically for high heat-load applications, the heat load problem is solved by fabricating water-circulating channels into the mirror itself, located just below the reflecting surface. Until recently, this solution was not available for multilayers, which generally require lower surface roughness than x-ray mirrors, and are therefore fabricated from silicon or silicon carbide substrates.
Now, the situation has changed. Figure 2 shows a model of an internally-cooled silicon substrate, recently delivered to CHESS by InSync, Inc. These substrates are predicted to substantially mitigate the heat-bump problem suffered by the current substrates, and hence contribute to significant, additional flux gains at G-line. The substrates are slated to receive the multilayer coatings in the coming months, and installed during CLASSE’s 2015 summer down. If so, users in the fall of 2015 will experience the best beams at G-line to date.
*See: http://news.chess.cornell.edu/articles/2014/Fontes141117.html, http://news.chess.cornell.edu/articles/2014/Woll141105.html