Outboard branch order sorting mirrors

For running near the carbon edge, order sorting mirrors are available. These suppress the considerable harmonic contamination in the beam. The theory is that two nearly parallel mirrors are adjusted in such a way that they displace the beam without changing the angle. The displacement is about 2 mm upwards. The first diffraction order from the monochromator is reflected with a modest loss, while harmonics are beyond the cutoff (see Fig. 3.28).

Figure 3.27: Micrometer settings for tilting the order sorting mirrors on the outboard branch.

In going from the mirrors to STXM, the beam passes through the slot in the middle of a quadrant photodiode (see Fig. 3.29 for a picture of what the assembly looks like). This slot needs to be re-positioned when the mirror is inserted or removed.

Figure 3.28: Calculated reflectivity for the order sorting mirrors. At each single-bounce angle in mrad, the reflectivity for two bounces is shown.

Figure 3.29: Photograph of the quadrant photodiode assembly. At top is a solid-area silicon photodiode, at middle is a phosphor screen, and at bottom is the quadrant photodiode with a hole in its middle.

To read the quadrant photodiode, the absolute numbers are given here based on the readings on the barrell scale (.05'' divisions on right side), since the digital read-out can be reset. In the ``mirror out'' position the quadrant photodiode should be at about the 0.15'' mark on scale, with top and bottom outputs roughly balanced, reading a few nA. If you instead move one quadrant fully into the beam, you should see a signal of several $\mu$A (see e.g., Fig. 3.3). With the order sorting mirros in position, the top output of the quadrant photodiode should read about 1 $\mu$A (varies with gap, slits, wavelength and beam current by large factor!), with the bottom output having dropped significantly from the previous reading. This verifies that the beam is now raised by the mirrors. (A significant reading on the bottom photodiode may still be present due to visible light through uncovered windows).

Raise the quadrant photodiode by 2 mm. At this point top and bottom outputs should once again be balanced, but less than before inserting the mirrors: (remember: the throughput is only about 35% at 4.1 nm).

Now STXM needs to be raised to bring it in line with the beam.