3.3 Small angle x-ray scattering
3.3.4 Background Subtraction
To subtract the background the pattern of a non-irradiated sample can be used. Alternat- ively a region on the tilted irradiated sample can be used, as the radial component of the track scattering is contracted in the thin oscillating streaks. Typically, the latter technique was employed in this work as it provides better results due to identical sample thickness and composition as well as subtracting out potential shadowing effects that would otherwise remain in the difference image.
Figure 3.9 shows different SAXS detector images for quartz. The SAXS image of a quartz sample without ion tracks (a) mainly consists of an isotropic scattering around the origin. The thin straight lines through the centre are a result of parasitic scattering within the sample and can also be observed in some quartz samples with ion tracks. For the image of a tilted ion track in quartz (b) a characteristic anisotropic streak is visible. Only the streak is isolated (orange area), allowing to transform the 2D SAXS image into a 1D-pattern. To subtract the background two different methods are discussed (c). Method #1 follows the area of the streak closely and provides the most accurate background in case of large variations of the background scattering. Method #2 is chosen perpendicular to the streak.
Figure 3.9(d) displays 1D-patterns for each background together with the isolated scat- tering pattern for the ion tracks (orange circles). The pattern from the non-irradiated sample (solid stars) provides an acceptable method for background subtraction. However, at higher q, where the intensity falls rapidly, the pattern is significantly lower than the
ion track data. This is a result of the lower background scattering in this particular non- irradiated sample and leads to less pronounced peaks after this background is subtracted. The other two backgrounds were taken from the irradiated sample and give a significant better agreement at highq. Additionally, they also remove potential isotropic components within the sample. Such effects can result from isotropic features, i.e. nano-sized voids that display ring-shaped oscillations at a distinct q-value. In direct comparison, the different
methods for background subtraction do not influence the results, but mainly increase the agreement with the fitting function. For a small number of samples, anisotropic shadowing effects required a scaling of the background between 0.8 and 1.1 to match the intensity at highq-values.
3.3. SMALL ANGLE X-RAY SCATTERING 43
Figure 3.9: SAXS detector images for (a) non-irradiated quartz sample, (b) ion irradiated quartz sample with a mask around the streak, (c) ion irradiated quartz sample with two different masks for background subtraction, (d) The 1D SAXS patterns corresponding to the masked areas in the images (a-c). Darker colours correspond to high scattering intensities.
Chapter 4
Experimental methods
This chapter describes the experimental techniques used for the present work. These range from the handling of the samples in the lab to the application of large-scale facilities for ion acceleration and x-ray and neutron characterisation.
Sect. 4.1 discusses the preparation of the samples, including the separation of the natural specimen from the rock and their thickness reduction through polishing.
Sect. 4.2 discusses the set-up and handling of diamond anvil cells to expose the samples to an environment of high pressure.
Sect. 4.3 discusses the experimental aspects of the sample irradiation with swift heavy ions. The three different ion accelerators used in this work are compared. Irradiation up to 185 MeV Au was carried out at the Heavy Ion Accerator Facility at the Australian National University in Canberra. Higher ion energies were provided by The Universal Linear Accelerator and the Heavy Ion Synchrotron, both at the Helmholtz Centre for Heavy Ion Research in Darmstadt, Germany.
Sect. 4.4 discusses the set-up of the Australian Synchrotron in Melbourne, Australia. The focus are small angle x-ray scattering (SAXS) experiments, the primary characterisa- tion technique for ion tracks in this work.
Finally, Sect. 4.5 briefly shows the generation of neutrons on two different types of neutron sources that were used for the small angle neutron scattering (SANS) experiments in the present work, at the Australian Nuclear Science and Technology Organisation (AN- STO) in Sydney, Australia and at Oak Ridge National Laboratory (ORNL), Tennessee.
4.1
Sample preparation techniques
Natural mineralogical samples
The sample preparation progress starts with separating the apatite crystal from the rock. For example, the Dashkesan-apatite sample shown in Figure 4.1(a), was surrounded with Co-Fe ore. The hexagonal shape of the apatite single-crystal and itsc-axis can easily be
identified. By using a diamond saw the apatite crystal was cut out of the other minerals 45
46 CHAPTER 4. EXPERIMENTAL METHODS
Figure 4.1: (a) Apatite crystal from Dashkesan (Azerbaijan) embedded in a Co-Fe rock before cutting with its hexagonal structure and c-axis clearly visible. (b) Apatite from
Durango (Mexico) after it was sliced perpendicular to the c-axis to approximately 1 mm
in and thickness and 15 mm in diameter using the diamond saw. (c) A variety of apatite samples, polished down to⇠50 µm thickness, after ion irradiation.
within the rock. This required the specimen to be mounted on a glass slide using crystal bond at a temperature of approximately 100 °C. Once the apatite crystal was isolated, small layers of approximately 1 mm thickness and 15 mm in diameter were cut from the mineral in two orientations, parallel and perpendicular to thec-axis as shown in Figure 4.1(b). The
layers were further cut into smaller, individual samples of⇠0.2cm2to prevent breaking or
crack formation of the fragile material during the polishing process. For polishing, the thin samples were mounted on a polishing tripod using epoxy glue. A rotating disk polishing machine with diamond polishing paper of 15 and 5µm grain size was used under constant water flow to reduce the sample thickness to between 30 and 200 µm. Upon reaching its
final thickness, the surface was polished again by using 1 µm diamond paper, removing
the larger scratches that occurred during the rough polishing process. Following this, the epoxy glue was removed using acetone for a duration of ⇠2−5 hrs. Finally, the thinned samples were mounted on an aluminium plate for ion radiation, as shown in Figure 4.1(c). In this work, only two different natural samples from Durango, Mexico and Dashkesan, Azerbaijan were investigated. The different compositions of the two single crystals was determined with microprobe analysis [SC99, War91]. They show the following halogen composition:
Durango apatite 3.4 wt% F and 0.4 wt% Cl Dashkesan apatite 3.5 wt% F and 0.5 wt% Cl
Synthetic wafers
The synthetic quartz for this work was purchased as crystalline wafers with thicknesses of 0.5 mm asx-, y− and z-cuts, as well as 0.050 mm for z-cuts only. The cut direction is a
technical term for the surface normals along the [11¯20]-direction (x-cut), [10¯10]-direction (y-cut) or [0001]-direction (z-cut). The thin wafers did not require any polishing and
4.2. HIGH-PRESSURE DIAMOND ANVIL CELLS 47