Jump to Content

Recent projects (1997-2001)


An Intelligent Sensor System for Environmental Monitoring (1998)

PhD Candidate: Salam Maria Srie Rezeki
Supervisors: Y-P Xu, and MR Haskard

This research is under the TIL (Targeted Institutional Link) collaborative project which aims to build a distributed intelligent sensor system for environmental monitoring. In particular, this PhD-candidate is responsible for the intelligence section of the system that is designed to monitor above- and below-ground pollutants using miniature chemical and gas sensors. Up to eight different samples (gases or ions) are to be detected and measured at the same time. The information collected from the sensors is stored locally for a period and transferred to a central (host) computer when interrogated.

The system is to be equipped with self-diagnosis and self-calibration capabilities. It advises the central computer through the communication interface whenever a faulty sensor is located and needs to be replaced. Thus the system can be operated with minimum maintenance. The figure next page shows a simplified block diagram of the distributed intelligent sensor system.  

The shaded area is the intelligence section that controls the self-diagnosis and self-calibration. The self-diagnosis is to ascertain whether the sensor is functioning correctly. It uses the majority voting scheme to detect hard errors while the parameters extracted from the sensor output are used as the additional indicators for the soft-error check.

The implementation is intended to use a combination of conventional signal processing circuits and artificial (fuzzy) neural network. The calibration is needed to ensure that the sensor conforms to a known performance. It is performed at two stages, the primary and the pro-measurement calibrations. The former is mainly for the initial calibration. It consists of the calibration of signal processing circuits (such as offset and gain drift calibration) and the calibration of sensor output (for example, multi-point, temperature, and humidity calibrations) against the standard or reference sample. With the standard or reference sample provided on-site, the primary calibration is intended to be a self-operated procedure.

The pro-measurement calibration is performed when the drift condition is detected provided that the drift is within an acceptable range and can be corrected. This technique will be developed to suit the sampling methodology used by the sensors, in this particular project there are two possibilities, ie. chromatographic system (for gas sample) and flow-injection (for ion detection). The communication to the host computer can be made through I2C interface and sent via one of several mediums such as a two wire bus, telephone line or RF transmission.  


A Study and Fabrication of Field Emitter Arrays for Application in Field Emission Display (1997)

PhD Candidate: In-Jae Chung
Supervisor: A Hariz

The aim of this project is to develop an X-Y addressable, 2" in diagonal silicon based field emitter array having improved electrical performances for application in Field Emission Displays (FED). There have been increasing interest about FED (Field Emission Display) because of its capabilities of producing high quality colour image, thin profile, low power consumption, low manufacturing cost and etc.

Though FED is regarded as one of the flat panel displays which can replace CRT in the near future, there still remains some problems which have not been fully solved nor analysed yet, and have to be overcome. Namely, how to increase the fabrication uniformity of field emitters and improve stability in operation, how to reduce the driving voltages of the field emitters etc. The above mentioned problems are closely correlated with material design, field emitter design, manufacturing process and so on.

This second year's experimental work has been focused on the two subjects: First to investigate the electrical stability of a metal silicided silicon field emitter. Several silicide materials such as Cr, Ta, Ti, etc. have been examined in this study. Among the results, Cr silicided silicon field emitters showed more stabilised current emission characteristics and work at a lesser vacuum than a pure silicon field emitter. These results will be used for the building of a FED with longer life, stable operation and lower manufacturing cost.

The second subject was the optimisation of FED design. The electro-optical characteristics of a FED are mainly determined by the field emitter construction. However, because of a large number of field emitters in each pixel, through micron-size and process turbulence effects, it is very hard to fabricate a uniform field emitter array throughout the panel. For dealing with such a problem, a robust design concept was introduced. The Taguchi statistical method was used for it is simple to apply and very accurate. With the help of the field emission simulation program (SCALA), optimisation analysis have been executed under the multiple effects of various conditions.

The results of this work is represented in the figure as a signal to noise ratio of the proposed field emitter. It shows that the highest point in each graph represents the robust design point which have reduced insensitivity against outer variance caused by process, misalignment and so on.

These results were used to design a 1.8" diagonal, 120 x 64 pixels, X-Y matrix driven, silicon based field emitter array.

This work is a collaboration research project and receives a research grant from KIST in Korea . The author is supported by a DEET Targeted Institutional Links Scholarship.


Silicon Micromachined Diaphragm Micropump (1998)

PhD Candidate: In-Byeong Kang
Supervisors:
M R Haskard

The main objectives of this project are to design and fabricate a silicon micromachined micropump having a constant flow, low leakage and high pressure head. This involves silicon micromachining processes such as silicon bulk etching, electroplating through the etched silicon mould, wafer bonding and packaging technology. The project will culminate in the production of a working silicon micromachined micropump.

The work done over the 1st year of this project has involved literature search, the outcome being a new pump structure following the micromachining processes such as silicon bulk etching, wafer bonding using positive photo-resist and an electroplating process.

Literature search is being applied for the project proposal and new micropump structure. The cross-sectional view of the proposed silicon micromachined micropump is shown in the figure below. The micropump employs two floating passive valves (input valve & output valve), and a variable pressure chamber with a pneumatic actuator. The micropump will be expected to have a good pumping characteristics especially on low leakage and high pressure head.

Electroplated nickel has been used to fabricate the floating passive valves, due to its low adhesion characteristics to the silicon and easier fabrication process for a thick layer. An electroplating process through the etched silicon mould provides good reproducibility of floating microvalve, always having well defined shape. The floating microvalve structure has been successfully made by using silicon bulk etching with an EPW etchant and a nickel electroplating process through the etched silicon shape. The figure below shows fabricated microvalve structure.

In the next two years, an assembly technique using wafer bonding and a packaging process will be developed for the micropump. To evaluate the performance of the micropump, test procedures will be advised and implemented for all phases. Based on the characteristics of initial prototype, design parameters will be optimised to improve the overall performance.


A Micromachined Reaction Column for a Portable Modular Flow Injection Analysis System (2000)

PhD Candidate: M Son
Supervisors: F Peddie and D Davey

With the recent advance of electroanalytic chemistry, various sensors for metal ion and biochemical molecules have been developed. Those sensors are small enough to require very small volume of sample, promising the reduction of overall instrument size if they are combined with appropriate sample handling technique.

Flow injection analysis (FIA) is a field of analytical chemistry researching into the efficient methods to deliver a sample solution to a sensor surface. The FIA was developed in mid-70s and has grown rapidly to yield commercial automatic FIA system.

The fact that FIA enables the biochemical sensor to be used for automatic and reproducible monitoring, it has stimulated a lot of engineering research to produce the system cheaply because typical FIA system are currently very expensive.

One approach is to use micromachining since it allows mass production of small reproducible devices. In the late '80s, many prototypes of micromachined FIA system were reported which have a micromachined injector, pump, column or detector. However, to fabricated a practical FIA system using only micromachining technique is very difficult at this stage.

Individual micromachined module may still have problems such as a weak outlet pressure for the pump module, a complicated structure for the micromachined injector and a non-circular cross-section for the reaction column.

To compromise the demand for a cheap FIA system and the difficulty in micromachining of FIA modules, a modular approach is tried in this research. The resulting system is sufficiently small enough to enable portable use and easy maintenance.

This research to make a portable-modular FIA (PMFIA) system uses conventional fabrication techniques (namely, plastic machining) plus micromachining to make a reaction column module whose plastic counterpart is usually large in size.

Micromachining gives a reduction in the column module from a 2m x 1.6 mm tube to a 20mm x 20 mm plate. Bulk etching techniques were used to generate V-groove on silicon wafer followed by anodic bonding (electrostatic bonding under 1000V, 500oC for 14 minutes), the etched wafer to pyrex #7740 glass bond was tight without any noticeable defect. The inlet and outlet holes were created and trimmed after the bonding.

The micromachined column is currently being tested. The specifications of it are as follows:

Total length of channel : 80 cm
Number of fold: 62
Channel width: 250 micrometre


Thick film Multi-Ion Microsensor System for Waste-Water Analysis (1999)

PhD Candidate: Hiskia Sirait  
Supervisors:
Prof Malcolm Haskard, A/Prof D Mulcahy, Dr D Davey, and Dr S Soegidjoko

The aim of this project is to develop a sensitive, selective, robust and low cost thick film multi-ion microsensor system. It will be applied initially to the determination of a single solution species, H2S, and later to multiple parameters, for example inorganic sulphur speciation: H2S, pH, Redox, H2SO4/SO2. The detection of heavy metal such as copper (Cu2+), cadmium (Cd2+), lead (Pb2+) may also be considered. In the future the sensor system will be develop as part of a portable instrument for the detection of multiple species in aqueous streams above and below ground.  

Initial work will concentrate on the design of sensors, flow cells and the sensor interface electronics which is use to amplify the signal, reduce the noise, and transmit the signal for data processing. Sensor design specifications include type of membrane, electrode material, surface area, geometry, surface condition factors and interactions among electrode elements.  

The purpose was to develop Thick Film microsensor array to detect pH and H2S. The sensor is based on Thick Film metal-metal oxide electrodes.  

The multi-layer sensor structures consist of conductor, interfacial layer and membrane deposited sequentially on an alumna ceramic substrate by screen printing process.  

The chemically sensitive layer consists of a pressed pellet prepared by mixing sensitive powders such as Antimony (Sb) and Ag2S with organic binders. The pressed pellet size was about 2 mm and 1.5 mm. The pH effect was measured by Orion model 720A digital pH meter. Reference electrode is an Ag/AgCl electrode. Test device is shown in Figure 2. The sensitivity of the pH electrodes is good, +/- 50 mV/pH.  

R : the reference Electrode, C : the carrier or reagent solution, S : the samples solution, P1, P2 : peristaltic pumps, V : a rotary valve, L : the sample Loop, M : the mixing coil, MTR : the ISE Meter, FC : the flow cell with the Detector, PC : the Personal Computer, W : the outlet to waste.  

The flow cell was designed to achieve the following objectives : minimum dispersion, minimum dead volume, and efficient use of the sample plug by making the membranes in series in the cell. The cell, which is about 7.0 cm x 4.5 cm, consists of two perspex halves separated by a Teflon gasket, which controls the cell volume at 180 mm3.


Micromachined Gas Chromatography System for Detection of Pollutant Gases (1999)

PhD Candidate: Goib Wiranto
Supervisors:
Prof M R Haskard, A/Prof D E Mulcahy and
Dr D E Davey

The aim of this research is to develop a micromachined gas chromatography (GC) system for detection of pollutant gases, such as CO, CO2, CH4, NO2, etc. The system is expected to be capable of sensing parts per million (ppm) levels of pollutant gases and performing fast separation of gas mixtures (in the order of a few seconds). The micromachined system will also be small in size (portable) for on-site monitoring.  

Just as their conventional counterparts, the proposed miniaturised GC system (except the carrier gas) will consists of five major components: carrier gas, sample injection system, micromachined column, gas detector, and electronic circuitry. The prototype system will be constructed in three separate modules to allow flexibility in component failure analysis.  

The heart of the GC system is the micromachined capillary (open tubular) column where separation of sample mixtures takes place. It is designed as an interlocking spiral shape with 100 micron wide and at least 1 meter long. The column will be fabricated on borosilicate glass (Pyrex) by chemical etching and anodic bonding method in a sandwich structure.  

Column operation is determined by the coating of a stationary phase over the column walls. It will serve as a media of adsorption and chemisorption with the injected gas sample to produce propagation delay for each of the gas sample component. Candidates for stationary phase are many including carbon and metal-based phthalocyanine. Novel deposition technique will be used to deposit the stationary phase into the GC column.  

Sample injector is used to introduce a finite and precise amount of gas sample into the stream of carrier gas that flows continuously into the column. In addition, as a requirement of the GC operation, the volume of the injected gas sample has to be much smaller than that of the GC column. Therefor the injection system will be designed and constructed on silicon as a three-way microvalve to satisfy the above requirements. The valve will then allow instantaneous injection of the sample pulse.  

The gas detector(s) will be constructed along with the electronic circuitry in a single module. It is a multiple detector structure with individual sensor being sensitive to a specific gas, therefor compensating for lack of selectivity. However, the use of a multi purpose detector such as acoustic wave sensor has not been ruled out.  

This research is part of a multi project work in developing a smart sensor system for detection of pollutant species below and above ground. It is a collaborative work with Institute Teknologi Bandung (ITB) under Targeted Institutional Links (TIL) project.

top^