In-situ GISAXS with 2D imaging ellipsometry at Bessy

Date: 4-10 October 2010 & 23-29 May 2011
Beamline: 7T-MPW-SAXS at Bessy
Team on site: Matt Gunn, Rudi Winter (Aber)
Dragomir Tatchev (Bulg. Acad. Sci., Sofia)
Sylvio Haas, Armin Hoell (HZB Berlin)
Back at Aber: Dave Langstaff (Labview)
Instrument: X-ray energy: 10keV
Detector: MarCCD
Camera length: 3.55m
Sample env.: Open flat ceramic heater stage,
mounted on ellipsometer,
mounted on Hexapod;
Laser and x-ray beam at right angles

The objective of this experiment was to obtain simultaneous in-situ data of a materials surface process in both real and reciprocal space. To this end, we took Aberystwyth's home-grown 2D imaging ellipsometer (real space) to Bessy and mounted it on the SAXS beamline (reciprocal space) in grazing-incidence geometry. The process we've chosen to study is one we have some prior experience in, the calcination of pure and yttrium-stabilised zirconia films on silicon wafers.

This required feeding the x-ray beam through a narrow vacuum tube through the goniometer of the ellipsometer, leaving us with an angular range of about 2o to tilt the whole instrument in order to adjust the x-ray grazing angle (which is only about 0.5o).

Fig.: Hot sample illuminated by laser.
Fig.: part of the team on the beamline.

The picture on the left shows Matt, Rudi and Dragomir (left to right) on the beamline. The flexible bellows of the camera tube can be seen on the right, with the laser and camera arm of the ellipsometer visible in front of the (unused, turquoise) goniometer.

The small picture shows a sample (a sol-gel coated piece of silicon wafer, approx. 6mm×20mm) sitting on the GISAXS hotplate and being illuminated by the green laser of the ellipsometer.

Photos: Matt Gunn

Correlated temperature-dependent data

Fig.: Movie showing temperature-dependent GISAXS and imaging ellipsometry data.
Montage: Dave Langstaff

The movie on the right shows the evolution of the GISAXS patterns (left) and the ellipsometric image (phase -or Δ- contrast) side by side. These are uncorrected raw data: the radial streak appearing in the scattering pattern at 475oC and the scratch and dominant circular feature in the ellipsometry image are artefacts (a reflection off the beamstop, a scratch and a grain on the surface).

As the temperature rises, the formation of an inorganic polymer network begins around 500oC as indicated by the increase of scattered intensity to the side of the beamstop. Around 800oC, a compact ceramic layer begins to form, resulting in another sudden increase in scattering intensity. At the same time, this results in the macroscopic surface becoming notably smoother as evident in the more uniform ellipsometry image.

What next?

We have demonstrated that combined in-situ experiments using GISAXS and 2D imaging ellipsometry are feasible. In the follow-up experiment, we have improved the interfacing of the two instruments (mutual triggering of data aquisition) to make the study of dynamic processes more efficient, leaving the experimental team to observe rather than to run the experiments manually. This also gave us an opportunity to widen our parameter space.

Ultimately, it will be interesting to use the real-space ellipsometry data to select specific regions of a sample surface and obtain GISAXS patterns specifically from those regions (which would have to be strips in forward direction of the x-ray beam). This will require the experiment to be mounted on a microfocus beamline.