Next: Measurement results. Up: Near edge structure of Previous: Sample preparation
X-ray transmission measurements of samples 1 and 3 at low energies (60 eV - 900 eV) were performed at the Advanced Light Source (Lawrence Berkeley National Lab), beamline number 6.2.3, which is equipped with a grating monochromator with superb energy resolution (resolving power 7000). Three different gratings are needed to cover the entire energy range. For each grating the interval is further divided into two or three subranges due to the necessity of using different filters to suppress higher orders. For this reason a dataset for each sample consists of 8 separate subsets. It should be noted that changing of the grating requires realignment of the beam.
Only samples 1 and 3 were characterized at ALS. Three other samples were too thick for the measurement in this energy range, being opaque at energies below 300 - 400 eV and above the O K edge (SiO2 sample). The energy step in the region from 360 to 580 eV containing nitrogen and oxygen edges was set to 0.25 eV, to 0.5 eV for energies below 180 eV (around the Si L edge), and 1 eV for the rest of the energy range.
Transmission at energies around the silicon K edge was measured at two other synchrotrons: at BESSY (Berlin) and at the Synchrotron Radiation Center (Madison, Wisconsin). The Canadian Double Crystal Beamline at the SRC has a double crystal monochromator with a resolution of 0.9 eV. It utilizes a pair of InSb(111) crystals (instead of the more commonly used in this range silicon), which is important for this measurement in order to avoid silicon edge ``steps'' in the incoming flux. Due to the low energy of electrons in the storage ring (800 MeV) and the double crystal arrangement of the monochromator, this beamline has very low higher order light penetration (below 0.2%). The data were taken in the range from 1770 to 2500 eV. The energy step was 0.25 eV around Si edge (from 1830 to 1860 eV), and 1 eV for the rest of the range. The accuracy and reproduceability of the SRC data is better than 0.1%, the noise being hardly noticeable.
At BESSY measurement were made over a wider range, from 1300 eV to 3000 eV. In general BESSY results are much noisier, but they are very valuable since they span a wider range and this helps to constrain the thickness of the films more reliably.
Since the X-ray flux emitted by a storage ring is continuously changing, it is necessary to normalize the output intensity to the input flux. This requires two detectors for any measurement - one in front, and one behind the sample. A solid state X-ray detector was installed downstream from the sample to register the output flux. At all three beamlines it was an uncovered Hamamatsu GaAsP photodiode G1127-02. Again, this was one of the precautions taken to avoid silicon absorption edges anywhere in the measurement system. Upstream from the sample a transparent beam normalization detector monitored the input flux. At ALS and BESSY this was a metal mesh whose electron yield was measured. At SRC a gas ionization chamber served as the beam normalization detector.
Each measurement consisted of two passes through the energy range - one with the sample in the beam, another one with the sample out. The ratio of the two normalized fluxes is the transmission of the film.