/
Grazing Incidence (extra information)

Grazing Incidence (extra information)

Grazing incidence is a powerful technique for surfaces and interfaces, but please be aware it comes with considerable constraints on sample requirements, plus additional complexities in data processing and interpretation.

Sample Requirements

To obtain the most highly surface specific information (i.e. measured at or around the critical angle of reflection), you need to have very flat, smooth samples on X-ray opaque, rigid substrates. There are a few key reasons for this:

  • Critical angles are typically around 0.1 to 0.2 degrees (depending on photon energy and sample density). Roughness or flatness problems can easily result in local surface slopes varying considerably more than the critical angle. Non-flat samples can’t yield the same kind of surface sensitivity.

  • The core automated sample alignment algorithm is highly optimized for flat, smooth samples. The geometry and the way such samples interact with the beam is quantitatively built into the standard alignment algorithm itself. That alignment process will fail if there are excessive issues with roughness, curvature, or X-ray transparency from the samples. Users will need to be aware of this when planning experiments and know the technique well to know what to do when running experiments.

    • There are alternative alignment options available that will assist running non-ideal samples, but these cases are much harder, and it's imperative Users have sufficient expertise and experience in the technique to attempt non-ideal samples.

Ideal Sample Types

The ideal samples for grazing incidence are smooth and flat, and the right size. In many cases, samples are thin films deposited (eg. spin coating, vapour phase coating) on smooth, flat substrates . Typical substrates are polished silicon, glass etc, This gives you a feel for how flat ideal substrates are.

Other good samples are solid materials that have been fabricated flat, or cut, then carefully ground and preferably polished to a flat surface plane. If you are mechanically preparing solid samples (eg for metals) consider polishing a relatively large area section first, then cutting to size because polishing always rounds out surfaces and especially edges.

Non-Ideal Sample Types

Flexible substrates

Anything on a flexible substrate is not ideal for standard grazing incidence scattering:

  • You can try to run those kinds of samples stuck to the sample mount with double-sided tape (to make them kind of-flat) but they won’t produce the same surface sensitivity as flat solid substrates, and automated sample alignment is rather challenging.

  • Consider running in low-angle transmission geometry. This is a slightly different holder (sample suspended over a grooved metal support) mounted on the same stages (hexapod or goniometers) and doesn’t typically involve the full, complex alignment used for grazing incidence scattering.

Blobs on flat substrates:

Sometimes Users try to deposit a small smear of material (such as a gel, solvent-deposited droplet, or even powder) on a flat silicon or glass substrate and try to run that as a grazing incidence sample. Please bear a few things in mind:

  • Critical angles of ~0.1 to 0.2 degrees can easily be lower than the local slopes of such samples, so it's not true grazing, just low angle with limited surface sensitivity

  • The height of the standard focus beam is ~25 µm, and the sample is aligned at the height it blocks 50% of the beam within ± 1 micron. It's very easy for the thickness of a blob of material sitting on top of a substrate to be of the order or (or much larger than) half the height of the beam, and so this can severely challenge the rotation alignment component of the sample alignment system. The worst case is a thick blob deposited only on a small part of the substrate, and/or non-uniformly deposited.

  • If you do need to make samples with blobs of material deposited on a substrate:

    • Make the deposit as thin as practicable.

    • Deposit as uniformly as possible, and over the whole area of the substrate

    • Expect alignments to be challenging and bring enough personnel and expertise to manage the challenges.

    • Consider an alternative geometry (do in transmission, do in low-angle transmission).

 

Grazing Sample dimensions

1 cm (lateral width) x 1 cm (long) broadly speaking is a good general staring point grazing incidence samples for most setups. There are a few further considerations:

  • For samples 20 mm long, the effect can be quite noticeable. Generally, you do not need 20 mm long samples for routine grazing incidence. For in-air GIWAXS acquisition we definitely recommend NOT using samples 20mm long: long samples increase the air-gap and the distance between the rotation centre and the entrance window of the detector vessel, restricting the maximum scattering angle due to the limited size of the X-ray transparent window. For anything but samples run in the flow-thru cell, we recommend keeping sample lengths around 10-12 mm.

  • For GIWAXS, angular resolution is limited by the length of the sample. For 10 - 12 mm long samples, resolution broadening is modest, and most experiments will be OK. This also gives a good sample cross section (how much beam cross section the sample intercepts as a function of angle).

  • If you need highest resolution GIWAXS, ~ 5 mm sample length is a good working length. The sample alignment system generally runs 5mm samples OK, and resolution broadening is low. It's possible to run as low as 3 mm length samples, but alignment needs quite a lot of care, and scattered intensity is reduced because only part of the sample intercepts beam (which is what improves angular resolution).

  • For solid-liquid interfaces, geometry, absorption and flux considerations favour collection at around 18 keV, on samples around 18 mm long. The standard flow-through grazing cell was designed for this geometry and runs best with samples 18mm long. To keep samples sitting flat (in the roll direction) also recommend at least 10 mm sample width. It's likely in 2026 we will also offer a thinner (12 mm) version of that cell.

  • Sample width: ~10 mm wide samples (lateral dimension) are great for most situations. There is not generally much need to make samples wider than ~ 10 mm. For the high throughput modes (in vacuum or in-air) 10 mm is a good compromise between samples sitting flat (in roll direction) and being able to load lots of samples on mounts (to minimise sample exchange, vacuum pumpdown time). You can consider smaller width (minimum recommended is 5 mm) to squeeze on more samples but be sure to mount them carefully and they sit flat.

Labeling

  • Label the sample container (e.g. every wafer holder, every well in a well plate)

  • label the bottom of every sample if at all possible. Unlike other types of samples, hand written labels written in permanent marker are OK (stick-on labels on the bottom of samples could cause problems). Labeling the bottom of the sample is important to help avoid mix-ups, and to confirm identity when loading AND unloading samples.

  • If direction of measurement in relation to the X-ray beam matters, label the sample holder and the bottom of the sample clearly, with arrow pointing in X-ray direction(s). 

Important notes (mail-in samples)!

  • It is possible for us to consider mail-in access where justified, but ONLY measurements at ambient temperature, in-vacuum and on ideal substrates are possible. NO high or low temperature measurements can be realised by mail-in.

  • Samples MUST be compatible with vacuum conditions; ALL samples will be measured in our vacuum sample chamber. We cannot support solid-liquid GISAXS/GIWAXS by mail-in.

  • Beamline scientist must be on site to exchange samples. Runtime is limited to standard working hours. Please consider sample amounts accordingly! Your mail-in beamtime allocation has been based around samples being run during day-shift hours.