To overcome these problems, micro-technological based catalytic
gas sensors were developed [4–6]. The reduction of the
size improves the response time. Additionally, the design allows
different operating modes, e.g. programs with quickly varying temperatures,
thus leading to a broader application potential with more
options to analyze gas characteristics, e.g. gas quality or composition
based on the analysis of the frequency response [7]. To improve
the sensitivity and the response time, catalytic gas sensors are normally
equipped with membranes and porous structures as catalytic
layer, which also reduce the operating temperature and thereby the
power consumption [6]. Porous structures with sufficiently large
surfaces have been achieved by the use of nanostructured platinum
thin film catalysts [8–11] or palladium thin films [12] as well as
thick films of nanoparticles supported on aluminum oxide [13–16].
While these technologies lead to a decreasing operating temperature
in the range of 100 ◦C, the output signals are very small, in a
range of up to 15 mV/10,000 ppm hydrogen only.
However, membranes and porous structures can degrade the
stability notably because of inclusion of reaction products, agglomeration
and delamination of catalyst and electrical structures [6,17].
This results in a poor long-term stability [17,18]. First experiments
to increase the stability were conducted with platinum nanoparticles
as catalyst which are stabilized by a porous oxide support
coated on the sensor e.g. [18–20]. However, the spatial density of