Former projects (1996 - 1997)
- Novel Micromachined Acceleration Sensor
- ISFET Sensor Array for Biosensor Implementation
- Low Cost Micromachined Fail-Safe Accelerometers
-
Poly-Si - Poly- Si1-xGex Heterojunctions by Sputtering Depositions for
Thin Film MOSFET Applications
Novel Micromachined Acceleration Sensor (1997)
PhD Candidate: DB Murfett
Supervisors: MR Haskard
The aim of this research is to develop a silicon
micromachined acceleration sensor in the ±50g range for the automotive
market. This device, which will be primarily used for collision sensing in
car airbags, will be fully producible in a commercial silicon foundry and
will be able to include on chip circuitry for data processing and
environment compensation. The device incorporates both a novel sensing
technique, as well as a novel structure to implement this technique.
In 1992 the UniSA Microelectronics Centre formed a
cooperative agreement with the Korea Institute of Science and Technology
(KIST) to explore an accelerometer type micromechanical device. During that
year the team undertook an extensive research program into the field of
micromechanics, resulting in the development of several ideas for
micromachined accelerometers.
In 1993 our application for support from the ARC large
grant scheme was accepted, as was an application under the Targeted
Institutional Links (TIL) program. Using this funding, and based upon tests
on prototype sensors developed during that year, we moved into 1994 with a
clear target device structure. Extensive modelling and simulation of the
device were performed including first principles derivation of the Field
Effect Transistor (FET) equations for this structure, and both first
principles and computer aided Finite Element Analysis studies were conducted
on the mechanical structure.
Although based on a novel structure and production
technique, the style of construction methodology has now become one of the
most popular for micromechanical production. This methodology is a cross
between bulk and surface micromachining, in that the device is constructed
from a layer on the surface of the wafer, but that layer is a
mono-crystalline silicon epitaxial layer. In this way many of the advantages
of both bulk and surface techniques can be exploited. The mechanical device
itself is Reactive Ion Etched from the epitaxial layer, in this case forming
a rectangular seismic mass supported on four spring-like beams. Under
acceleration the mass will be displaced on the springs, and this
displacement can be measured by employing a field effect transistor in a
novel sensing technique.
Over the past 12 months, work on this design has
advanced extremely successfully leading to the device currently undergoing
full prototype production in a joint program between the
Highlights of the Year
-
Extensive exploration of production procedures resulting in two novel techniques (one presented at a conference in the
-
Completion of two "test bed" devices developed on which various production, mechanical and electrical parameters were evaluated.
-
Full novel accelerometer design completed.
-
Commencement of "commercial" prototype production of the design by Philips, DSTO and UniSA.
ISFET Sensor Array for Biosensor Implementation (1996)
PhD Candidate: T C
Supervisors: MR Haskard, D
The aim of this research is to develop a large pH-ISFET (Ion-Sensitive Field Effect Transistor) sensor array for biosensor implementation particularly for taste measurement. The PhD candidate will however only concentrate on the design and fabrication of the microelectro-chemical sensor.
The research work carried out is in close collaboration with the
The proposed sensor array will have the following
features:-
-
Using standard CMOS IC manufacturing technology
-
Using a base pH-ISFET sensor whereby enzymes or other biosensing materials could be placed on top of it to sense various biochemical materials
-
A large array of about 250 sensors
-
Integrated CMOS circuitry on-chip
-
Array signal read-out and processing by using the innovative threshold conversion technique
In order to demonstrate that the concept of threshold
conversion is a viable and innovative method of processing very large sensor
arrays, a thick-film thermistor array was fabricated by a final year student
in 1994 and tested. The results obtained demonstrated that the concept
works.
The chip design was successfully completed using a VLSI
layout editor, MAGIC, and simulated with SPICE. The main component of the
chip, the analog comparators which switches the output signal either high or
low depending on the comparison between the user defined threshold voltage,
Vth and the ISFET's output signal has been successfully designed.
The chip was fabricated at the Electronics and
Telecommunications Research Institute,
The project is funded by a UniSA's Pre-Competitive Grant Scheme and DEET's Targeted Institutional Links program. The PhD Candidature is funded by an Australian Postgraduate Research Award scholarship.
Low Cost Micromachined Fail-Safe Accelerometers (1997)
Masters Candidate: L Truong
Supervisors: MR Haskard , and A Hariz
In recent years, an additional safety feature of
automobiles is the inclusion of airbags which, as a product, has attracted a
large industrial market. The operation of such a device is based on
accelerometers which are fabricated using either conventional mechanical
techniques or using micromachined technology. Such systems are required to
be highly reliable, posses high performance repeatability, manufactured at
low cost and to operate flawlessly where, ideally, the system should allow
no false triggering.
One approach implemented is introducing minor
structural changes to the well known double cantilever beam accelerometers.
The two beams which support the proof-mass, include notches etched to
control the fracturing force for airbag deployment. Hence the size of these
notches is a crucial factor on determining the force of impact.
The design of the proposed structure is based on simple
circuitry aiming at minimising non-desirable triggering. The surrounding
substrate is etched into two halves allowing only a single electronic path
through the beams of the seismic mass. The supporting beams of the proof
mass posses accurately machined fracture points allowing one or both beams
to fracture once the threshold acceleration (typically 50g) is achieved. The
acceleration will exert a force resulting in the deflection of the mass
placing the beams under stress, and eventually snap off.
One of the several ways of weakening the beams is to employ the "one-step" etching approach which uses wet anisotropic etching technique to "sculpt" a V-groove cavity into the two cantilever beams.
Two prototypes of such sensor have been fabricated. The
first prototype had an overall dimension of 2mm ´ 2mm. The etched grooves
are 100mm wide with a depth of 70mm.
Using the "one-step" etching process, a total
of 55 chips have been fabricated on a quarter of a two inch wafer with a
yield of more than 85% which is considered cost-effective.
The second prototype had an overall size of 2mm x 4mm,
the principle of operation has not been changed except for the geometry of
the beams. Improvement have been made for handling of the device and its
natural frequency of vibration.
Finite Element analysis have also been carried out to
predict the performance of the device, this includes natural frequency
simulation of various modes and crack growth simulation at the notch
(currently under development).
Future work will incorporate further computer simulations aiming at controlling the fracture points on the two beams targeting a ±5g deviation from the desired g-force.
Poly-Si - Poly- Si1-xGex Heterojunctions by Sputtering Depositions for
Thin Film MOSFET Applications
Masters Candidate:
M
Supervisor: A Hariz
The aims of this research are to fabricate Poly-Si-
Poly-SiGe junctions and to apply the junctions in thin film MOSFETS.
There is a new trend in microelectronics beside the
large scale integration (LSI). It is the concept of "the giant
microelectronics" in which an increase of the substrate size is more
important than size reduction of its devices. Display technology is an
example application of this field.
Thin film (TF) poly or amorphous Si MOSFETs has been
widely used to address LCD matrix and image sensing arrays. The main problem
of using poly and amorphous material is the mobility degradation and higher
leakage current due to higher trap state density. For that reason, its
application for high definition and very dynamic displays has not been
realised. However TF MOSFETs technology is now gaining new grounds in
applications, such as multi-layer devices and low cost ICs.
In this research, poly-Si1-xGex and poly-Si junctions
formed by sputtering depositions and annealing will be studied. These
junctions can be applied to TF MOSFETs which does not need an ion
implantation step for its process, as it is difficult and expensive to
implant impurities into a large area. Another major consideration in
choosing the sputtering method to deposit semiconductor materials (Si and Ge)
is that the method is simpler and safer than CVD (chemical vapour
depositions) which involves toxic gases. In this research, the Si1-xGex
alloys are deposited by a single target.
It is well known that Si1-xGex alloys in a crystalline
strained structure have higher carrier mobility than that of silicon but the
alloys in polycrystalline form have not been well explored. The
poly-Si1-xGex layers will be used as channels in TF MOSFETS. The speed of TF
MOSFET is strongly determined by the hole mobility in the channel.
A new technique using a surface treatment during solid-phase crystallisation by furnace annealing is introduced in this research. The technique can dramatically enlarge the grain size of poly-Si thin layers. The average grain size is more than 50m.