Biomedical Engineering Reference
In-Depth Information
the different components of a mass spectrometer are aligned and confined in one
mask layer. For its fabrication well established micromachining processes are
applied, which allow to transfer the fabrication directly into an industrial environ-
ment without substantial engineering efforts and time delay. This approach also
allows to get around the cited shortcomings of a hybrid assembly and adjustment to
a “macro”-periphery.
To follow this approach subsystem layout and functions as well as microfabrica-
tion processes have to be selected, which are apt to generate and integrate the vari-
ous subsystems of a mass spectrometer into one substrate in one batch process, i.e.,
all the components are defined in their geometry and orientation among each on one
photo mask, which is then transferred into features in the substrate. Thus, submi-
cron accuracy of device features and their orientation and adjustment can reproduc-
ibly be guaranteed, i.e., many identical devices can be fabricated simultaneously in
one run. This, however, also means that all the subsystems are fabricated in the same
material within a narrow range of the necessarily different dimensions.
Such an approach is evidently inherently rather challenging, since mostly due to
the restriction to a 2 ½-D geometry in a single material at least part of the subsys-
tems will need a complete redesign as compared to established mass spectrometer
geometries or even the introduction of alternative physical principles with com-
pletely new structures. Furthermore, all subsystems must be compatible with respect
to fabrication, size, function, and pressure regime as well as electrical interfaces.
Finally such a micro-mass spectrometer will need a modified and adapted hardware
and software of the electronics—which actually means a completely new one.
Another difficulty is that all these subsystems have to function at once, i.e., the
development, characterization, and optimization of the individual subsystems is
limited, and there is no way of a mechanical fine tuning. This means that the total
system has to be designed on very reliable simulations and pattern transfer, which
comprise not only the geometrical features for potentials and trajectories of elec-
trons and ions, but also the behavior of the mass separator or pressure regimes in
different areas of the system. Such an approach may appear rather adventurous and
risky; at least it may and will take much more time to generate a working system
than the demonstration of a single subsystem implemented into a standard mass
spectrometer environment.
When successful, however, the full advantage of microsystem design can be uti-
lized, i.e., low power consumption, small volume and weight, low cost, cheap, and
reproducible mass production, as well as low vacuum requirements, portability, and
even long-term application on battery or even energy harvesting power supply can
be envisioned.
Figure
1
shows the actual planar integrated micro-mass spectrometer (PIMMS).
It is fabricated as a three wafer glass-silicon-glass sandwich with all the relevant
structures realized in a highly doped silicon wafer via a deep reactive ion etch
(DRIE) process, which generates 2 ½-D structures, i.e., deliberate structures in
the plane and vertical walls with depths of several 100 mm. All features of the
mass spectrometer are contained in one photo mask layer. The thickness of the
Pyrex glass wafers is 500 mm, that of the double side polished silicon wafer
