| dc.description.abstract | The elevation of intracranial pressure (ICP) associated with traumatic brain injury (TBI),
hydrocephalus and other neurological conditions is a serious concern. If left untreated,
increased pressure in the brain will reduce cerebral blood flow (CBF) and can lead to brain
damage or early death. Currently, ICP is monitored through invasive catheters inserted into
the brain along with a shunt. However, insertion of catheters or shunts is an invasive procedure
that introduces vulnerability to infection. In principle, the risk of infection would be overcome
by a fully implantable pressure monitoring system. This would be particularly valuable
for hydrocephalus patients if lifetime monitoring was available. An implantable pressure
monitoring system relies on a thin flexible membrane as part of the pressure sensor. The
thin film membrane displaces under load and correspondingly induces a change in a relevant
electrical quantity (resistance, or capacitance). Micro-electro-mechanical system (MEMS)
is the technology that helps in creating micro/nano-mechanical structures integrated with
signal conditioning electronics. These micro structures can be inserted into the brain, where
the thin film is exposed to a corrosive fluid (saline/blood) at a temperature of approximately
37 ◦C. The miniaturization in MEMS permits examination, sensing and monitoring from
inside the patient for longer durations. However, the accuracy, particularly in terms of sensor
drift over long durations, is a key concern. In general, the issue of drift is attributed to the
aging and mechanical fatigue of thin film structures, particularly the thin flexible membrane.
Therefore, it is essential to analyze the thin film deflection and fatigue behaviour of MEMS
pressure sensors for achieving long-term reliability and accuracy. Thus, the high-level goal of this research is to identify a viable approach to producing a flexible membrane suitable for
deployment as a lifetime implantable pressure measuring system.
In this context, finite-element modelling (FEM) and finite-element analysis (FEA) of thin
film deflection and fatigue behaviour have been conducted. The FEM was implemented
in COMSOL Multiphysics with geometries resembling a capacitive type pressure sensor
with titanium (Ti) thin film membrane deposited onto the silicon substrate. The mechanical
behavior of thin film structures including stresses, strains, elastic strain energy density, and
thin film displacements of several thicknesses (50 μm, 25 μm, 4 μm, 1 μm, 500 nm, 200 nm)
have been studied. In addition, fatigue physics module has been added to the FEM to analyze
the fatigue life of thin film structures. The FEA results in the form of fatigue usage factors
have been plotted. Finally, to analyze the effect of fluid pressure transmission of the thin
film membrane inside the closed skull, fluid-structure interaction has been modelled. The
model represents a 2D fluid medium with the thin film membrane. The velocity magnitude,
displacement, shear rate (1/s) and kinetic energy density (J/m3) of 4 μm and 25 μm thick Ti
films has been plotted. From this analysis, 4 μm thin film membrane showed good tradeoff
for thickness, pressure transmission, and mechanical behaviour.
To validate the FEM, a custom designed acoustic-based thin film deflection and fatigue life
experiments have been set up. The experimental design comprised of: (1) A voice coil-based
multimedia speaker and subwoofer system to assist in displacing the thin film membranes,
(2) A laser displacement sensor to capture the displacements, (3) A spectrum analyzer palette
for generating random vibrations, (4) Dataloggers to record the input vibrations and thin
film displacements, and (5) Scanning electron microscopy (SEM) to visualize the surface
topography of thin film structures. Thin film titanium (Ti) foils of 4 μm and 25 μm thick were
obtained from William Gregor Ltd, Ti-shop, London. The thin-film specimens were clamped
to 3mm acrylic substrates and bonded to the subwoofer system. The Gaussian random
vibrations generated from the spectrum analyzer loaded the voice coil of the multi-media speaker system, which assists in displacing the thin films. The SEM surface observations
are divided into two regions: (1) Pre-cycle observation, where the thin film surfaces are
observed before the application of any load, and (2) Post-cycle observation, where the thin
films surfaces are observed after application of cyclic loadings.
Based on the understanding of the FEM and experimental studies, a conceptual framework
of MEMS pressure sensor has been developed. In this part of the work, initially, underlying
concepts of complementary-metal-oxide silicon (CMOS) circuit simulation, MEMS modelling,
and CMOS layout design have been discussed. Next the MEMS fabrication process
involving deposition (sputtering), etching, and final packaging have been discussed. Finally,
an optimized design process of the membrane-based sealed cavity MEMS pressure sensors
has been outlined. | en_US |
