3.2 Characterization techniques 47
oxidize until there is an excess of oxygen that is free to react with the target
surface. Once that happens the sputtering yield goes down and less sputtered
atoms are available, causing even more oxygen to react with the target. When
decreasing the partial pressure from a high level the target will be completely
passivated and since there is a low sputter yield it will take a lower partial
pressure before the oxidised target surface is back to a metallic state. This type
of behavior is important to realize since a high rate combined with a complete
oxidation of the prepared lm is often desirable. This means operating at the
balance point between excess oxygen and excess sputtered atoms. For reactive
co-sputtering the issue of target passivation becomes more complicated since two
dierent materials may passivate at very dierent partial pressures of oxygen.
In that case, obtaining the desired composition requires careful calibration and
possibly using either very low rates or having incomplete oxidation of the lms.
3.1.1.2 Equipment and deposition rate calibration
The sputtering equipment used throughout this thesis was a magnetron system
from AJA International, with two DC power sources, two RF power sources
and substrate heater. With all power sources in use up to four dierent materials
could be sputtered simultaneously. The deposition rates were calibrated
frequently using an in-chamber Quartz Crystal Microbalance. Calibration runs
were made by depositing 15-20 Å three times and noting the time. For reactive
sputtering it is also important to note down the time it takes for the rate to
stabilize. For MnOx the measured rates were further conrmed by lm thickness
measurements with a Prolometer. The Prolometer method comprises a
needle scanning the surface and for a suitable edge the height can be measured
down to a few nm in accuracy.
3.2 Characterization techniques
3.2.1 X-ray Photoelectron Spectroscopy
X-ray Photoelectron Spectroscopy, XPS, is a widely used method for investigating
surface composition 190. The technique is based on the ejection of
electrons from a surface caused by radiation, known as the photoelectric eect.
These electrons will have an energy dependent on the photon energy, h, individual
binding energy with reference to the Fermi level of the sample, Eb and
work function of the spectrometer, . This can be formulated as the following
equation.
Ek = h �� Eb �� (3.2)
Since the electronic conguration of an element is unique the knowledge of these
binding energies, Eb, allows us to identify the elemental composition of a sample.