Preparing for XRR

5. Preparing for XRR#

5.1. XRD mode of the photon delivery system#

The very first step for setting up the goniometer for XRR is to move the photo delivery system to the correct position for photo delivery to the goniometer position.

This involves moving

  • the monochromator to the specified energy

  • the focusing mirror to the correct pitch and bend

  • the harmonic rejection mirror out of the way

  • the hutch slit assembly (Section 2.5.5) to the correct height

  • the XAFS table to the correct height for supporting the flight path.

To set up the photon delivery system for scattering at 8600 eV:

RE(xrdmode())

or specify an energy:

RE(xrdmode(12000))

The single argument is the target energy in eV units. The default is 8600 eV, the normal operating energy for experiments on the goniometer.

This plan will look up the correct positions of all motors in the beamline lookup table and set all those axes moving to their correct positions.

Once all axes have arrived in position, a scan of the rocking curve of the monochromator will be performed and dcm_pitch will be moved to the peak of that scan.

Finally, the hutch slits will be opened wide, 7 mm wide by 1 mm tall, allowing the beam size to be determined by the gomiometer slits (see Section 2.2).

Future tech!

This plan will eventually be used to perform scattering measurements at any energy above 8000 eV without having to do a time-consuming realignment of the goniometer.

To obtain consistency in lateral position of the focused beam, Bruce is working with BLOP team in DSSI to optimize dcm_roll and the orientation of the focusing mirror to provide stable beam position over the energy range from 8 keV to 20 keV.

To obtain consistency in vertical position, a scan of the pitch of the focusing mirror into the goniometer slits will deliver consistent beam height.

5.2. Goniometer alignment strategy#

Note

A few things that are explicit steps in SPEC are handled differently in Bluesky. For example, the Mythen full_mca, ROI1, is set at Bluesky startup and does not need to be explicitly set. All alignment steps and associated data processing are discussed in detail in Section 4.

  1. Place the Mythen in the most downstream position on the (what is the arm called?). Measure and record the gap value – typically around 90 mm. See Figure 4.6 for a photo identifying what the gap is. To record the gap in a way that the data acquisition software can use, do:

    xrduser.gap = 90.0

  2. Using the YAG camera, center the pin under the beam.

    1. Open the slits wide

      RE(mv(slits.vsize, 4))
RE(mv(slits.hsize, 4))

    2. Adjust samplez to put the pin in the beam by seeing its shadow on the YAG.

      RE(mvr(samplez, <amount>))

    3. Mark the position of the pin in the beam

    4. Rotate phi stage by 180 degrees

    5. Mark pin again, then mark the geometric center of those two markings

    6. Move table.lateral so that the center of the two markings is in the center of the beam

    7. Rotate phi by -180 degrees to verfiy this alignment

    8. Rotate chi by -90 degrees:

      RE(mvr(chi, -90))

    9. Repeat steps (c) to (g) for this orientation

    10. Rotate chi back to 0 degrees:

      RE(mvr(chi, 90))

    Future Tech!

    Automate the pin centering procedure using a camera that is supported by AreaDetector. Automate the angle motions and determination of pin shadow positions. Compute and move to target position in each direction.

  3. Align the slits to be centered around the beam and define the 0 of each slit to be in the position that cuts the beam in half. This is done by:

    RE(align_slits())

    See Section 4.3 for more details.

  4. Set slit sizes:

    RE(mv(slits.vsize, 0.15, slits.hsize, 1.0))

    This vertical size – 150 μm – is considerably smaller than the focused beam, but appropriate for an XRR measurement.

  5. Align the table in the beam:

    RE(linescan(table.vertical, 'monitor', -1, 1, 51))
RE(linescan(table.lateral, 'monitor', -2, 2, 51))

  6. Do a linescan (Section 4.1) of the dethor motor to center the Mythen around the beam in the horizontal direction.

    RE(linescan(dethor, 'mythen', -3, 3, 61))

    See Section 4.4 for more details.

  7. Perform the Mythen calibration scan:

    RE(mythen_calibration(-4, 1, 1001))

    This will set the bounds of the dir and refl ROIs and write a calibration report to the proposal folder. It will also record the calibration parameters. See Section 4.5 for more details.

    Question

    What is the CHESS calibration? This needs to be written.

  8. Verify the alignment of beam, goniometer, and detector are acceptable by scanning the delta arm and plotting the signal from both dir and refl. The dir plot should be narrower than and well centered in the refl plot.

    RE(linescan(delta, 'mythen', -0.15, 0.15, 61))

You are now ready for sample alignment.

5.3. Sample alignment strategy#

A sample for XRR is usually a large, flat wafer. The correct alignment has the sample surface parallel to the beam path and at a height such that it blocks half the beam. With that alignment, the center of the beam will be on the center of the sample as the incident angle changes and the beam will spread symmetrically over the length of the sample as the angle changes.

Todo

Need example screenshots of the results of both sample alignment scans.

  1. Start by aligning the sample vertically.

    RE(sample_vertical())

    This will run a linescan (Section 4.1) of samplez against the signal in direct beam ROI then fit an error function to the measurement to find the position where the sample blocks half the beam. That position will be defined as 0 of samplez by setting the EPICS offset accordingly.

  2. Then align the pitch of the sample.

    RE(sample_eta())

    This will run a linescan (Section 4.1) of eta against the signal in direct beam ROI then do an appropriate analysis (more discussion below) to find the zero of eta. Move to that position and define it as 0 by setting the EPICS offset accordingly.

  3. Iterate those two steps as needed.

The interpretation of the pitch scan is a bit subtle. In the case of a very rough surface, the correct choice for eta will be very close to the peak of the measured scan.

However, in the case of a very smooth sample, the total external reflection will be intense enough that the structure near the peak will be such that the maximum intensity is not necessarily the proper 0 of eta. In that case, a more elaborate analysis is required.

Todo

Fully explain the smooth sample algorithm once it is implemented in code. Show the result of that analysis.


Contents