Silicon has been the leading material suited for the manufacture of a broad range of electronic, sensor, and actuator applications. However, it is limited in electronic device performance to temperatures below 250 degrees C and in mechanical device performance to below 600 degrees C. Its dim optical properties put silicon at a disadvantage with respect to the much acclaimed compound semiconductor rivals that have orders of magnitude higher optical emission. Consequently, for high-temperature MEMS applications, there is a need for semiconductors with good mechanical and thermal stability, and a wide bandgap for stable electronic and optoelectronic properties at elevated temperatures. This paper present a review of research activities on III-V nitrides and their experimentally established properties. We explore their suitability for microelectronics and microelectromechanical systems. Current efforts in developing III-N nitrides to extend the Si-based MEMS technology to applications in harsh environments is discussed. A summary is presented of the material properties that make them attractive for use in such environments. Challenges faced in crystal growth and development of processing techniques are also examined. Finally, a review is presented of the current state of novel optoelectronic devices made, and potential MEMS devices to be made from the proposed semiconductors, as well as an examination of issues facing future progress.

Reported modification of CMOS processes to incorporate MEM structure require the structure to be located outside the active area. This paper describes the fabrication of microstructures directly over active devices rather than alongside them. Several issues need to be addressed with this fabrication process. The first issue is the choice of the metal to be used for metalization. Aluminium is no longer a viable option as some of the post-metalization processing steps exceed its melting point. Tungsten and titanium are the alternative candidates with sufficiently high melting point and low resistivity. The second issue is the non-planar surface profile across the wafer after the fabrication of the active devices. The sacrificial layer must be greater than the surface profile across the wafer and furthermore planarization of the sacrificial layer is necessary before the structural layer is deposited or electroplated. The final issue is post-metalization process temperature considerations for stress relaxation where either rapid thermal annealing or a relatively low temperature annealing of the microstructure is carried out. In this paper, these issues will be addressed, using the processes developed for the fabrication of an opto- mechanical microaccelerator as an illustration.

Heart sounds can be utilized more efficiently by medical doctors when they are displayed visually, rather than through a conventional stethoscope. A system whereby a digital stethoscope interfaces directly to a PC will be directly along with signal processing algorithms, adopted. The sensor is based on a noise cancellation microphone, with a 450 Hz bandwidth and is sampled at 2250 samples/sec with 12-bit resolution. Further to this, we discuss for comparison a piezo-based sensor with a 1 kHz bandwidth. A major problem is that the recording of the heart sound into these devices is subject to unwanted background noise which can override the heart sound and results in a poor visual representation. This noise originates from various sources such as skin contact with the stethoscope diaphragm, lung sounds, and other surrounding sounds such as speech. Furthermore we demonstrate a solution using 'wavelet denoising'. The wavelet transform is used because of the similarity between the shape of wavelets and the time-domain shape of a heartbeat sound. Thus coding of the waveform into the wavelet domain is achieved with relatively few wavelet coefficients, in contrast to the many Fourier components that would result from conventional decomposition. We show that the background noise can be dramatically reduced by a thresholding operation in the wavelet domain. The principle is that the background noise codes into many small broadband wavelet coefficients that can be removed without significant degradation of the signal of interest.

This paper describes the development of a sensor for measuring acceleration-thresholds in the low-g range. The processing of the device is based on a new integration concept for MEMS called Additive Electroplating Technology (AET). This technology allows the integration of fixed and moveable electroplated microstructures on top of a standard ASIC by a low temperature back-end process. This permits the fabrication of the sensor on top of a CMOS signal processing circuit, where the development and the fabrication of the ASIC is absolutely independent of the following back-end process. In this context, the additive electroplating technology combines the advantages of monolithic- and hybrid-integrated sensor with respect to the reduction of interconnections, short development cycles, low tooling costs, and high design flexibility. A detailed technical concept for the fabrication of electroplated micromechanical acceleration switches is presented in this paper. It comprises the aspects of the interface technology between the ASIC and the sensor structure, the sensor fabrication and the back-end processing to capsulate the dies. The paper concludes with a characterization of the devices. The measured results are compared with the expected results of the simulation.

In this paper, a new micromachined tunneling accelerometer is described. Using the bulk-silicon fabrication technology and the silicon/glass electrostatic bonding process, a sandwich structure tunneling accelerometer has been fabricated and tested. In order to reduce the low frequency noise, the feedback circuit has been improved by adding oscillator and demodulator. The first prototype has a resolution of approximately 1e^-6g/ (root) Hz and can survive with at least 50g shocking.

A vertically sensitive acceleration sensor has been fabricated by combining the low-cost UV-LIGA process with a sacrificial layer technique. The key structure of the sensor is an asymmetric proof-mass which is suspended over two stationary electrodes. Because of the asymmetric arrangement of the proof-mass, an acceleration in the Z-direction results in an opposite deflection of its large and small parts. Hence, the acceleration can be detected by the differential capacitor arrangement between the proof-mass and the two stationary electrodes. The fabrication of this vertically sensitive accelerometer is a planar batch procedure comprising only a few processing steps. The entire structures are first grown electrochemically within the UV- patterned thick AZ4562 photoresist on an electroplating base that composes of rigid and sacrificial layers. Moveable Ni- parts are then obtained by removing the underneath titanium sacrificial layer using wet etching. Sensor structures up to 30 nm with an aspect ratio of about 10:1 can be reliably manufactured. It is thought that this fabrication approach can be widely applied to economically realize other micromechanical components with oscillating structures. Design and realization of the vertically sensitive accelerometer is described in this paper.

The integration of MEMS, IDTs and required microelectronics and conformal antenna to realize a hybrid accelerometer and gyroscope device is presented in this paper. The inertial navigation system uses both gyroscopes and accelerometers to measure the state of motion of a target or a missile by sensing the changes in that state caused by accelerations. This unique combination of technologies results in conformal sensors that can be remotely sensed by a RF system with the advantage of no power requirements at the sensor site. The device presented here possess typical advantages of MEM sensors including the additional benefits of robustness, excellent sensitivity, surface conformability, durability and applicability to wireless antenna systems. Compared to conventional ones, these new sensor have no moving mechanical parts ultimately giving rise to inherent robustness and durability. Consequently, there is no electronics to balance or measure the movement of moving structures, which leads to even smaller micro device.

The synthetic jet actuator is a novel means of applying flow control that lends itself to incorporation into micro- electro-mechanical systems. Experimental and computational studies of the flow generated by this actuator are presented and discussed. The synthetic jet is seen to resemble a steady turbulent jet. Self-similarity is, however, achieved closer to the orifice in the case of the synthetic jet, and this is consistent with the observation from the numerical simulations that a main vortex is trapped close to the orifice. This main vortex merges with the secondary vortex that is generated at the circumference of the orifice during each cycle of oscillation of the actuator membrane. The effect of synthetic jets on wake flows and flames has also been examined. Preliminary results of these investigations suggest that synthetic jets can significantly reduce drag and enhance mixing.

Characteristics of a monolithic microvalve using a curved electrode electrostatic actuator and an integrable valve orifice are described. The valve is a normally open electrostatic microvalve with a 83 volts switching voltage. The curved electrode is a curled up cantilever structure with few hundred micrometers out of plane displacement. Lateral buckling of the actuator is overcome by using a mesh type or finger strip type metalization pattern. Cantilever with curvatures angle ranging 0 to more than 360 degrees have been achieved by changing the thickness ratio of composite layers of the cantilever. A valve channel consisting of three inverted pyramids in tandem have been sued. The flow rate of the valve linearly depends on the pressure difference across the valve. Thickness variation of the standard wafers has minimal effect on the final orifice sizes of the valve. The surface micromachining processes used to fabricate actuators and bulk micromachining process used to fabricate the valve channel, have been integrated in a set of processes to fabricate a monolithic normally open microvalve.

Microengineered open tubular (MOT) columns with semi rectangular cross-sections have been designed and fabricated using microengineering techniques. The creation of 100-micrometers wide, 20-micrometers deep, and 125-cm long columns employed isotropic etching on (100) silicon and anodic bonding with a Pyrex 7740 glass cover plate. Column geometry has been optimized to achieve maximum efficiency and allow extreme operating conditions. The walls of the microcolumns were coated with a non-polar liquid stationary phase. Performances of the MOT columns have been demonstrated by their ability to completely separate a series of hydrocarbon mixture in less than 1.25 min under isothermal condition of 150 degrees C. The achievable column efficiencies as measured in terms of theoretical plate height ranged from 0.57 to 1.45 mm, which agreed well with theoretical predictions.

The micromotor is an extremely small device a few millimeters or less in size. Micromotors in the order of microns are realized by MEMS technology. Important applications in biomedicine include ultrasound probes for blood vessels, microrobots for colon intervention, smart pills and nanolitre pumps. Other uses include actuator for MOEMS and small variable capacitors. One exciting implication of micromotors is that they can be powered by rectifying mechanical vibrations. MEMS are playing an important role in our daily life as these systems are widely used in optics, communication and information systems, fluidics, biotechnology and medicine, scanning probe microscopes, automobiles and aerospace. There are a number technical challenges with micromotors, including the need to reduce stiction and increase torque. The precise geometry of the motor is usually tightly coupled to the stiction effect - the sticking of adjacent surfaces after release due to static friction. Piezoelectric, electrostatic and electromagnetic effects have been investigated to produce the electromotive force for the micromotor. However, we propose a micromotor design based on the Huber effect, as this will allow a new range of geometries and hence possibilities for managing stiction. To date, there have been no reported attempts at using the Huber effect, and this is possibly due to it being a poorly understood phenomenon. The reason for this is that large motors that utilize the Huber effect are self-destructive and hence have never been reliably characterized. Such motors are shown to be able to operate form a dc or ac source, and this property may be valuable in some MEMS applications.

One existing implication of micromotors is that they can be powered by rectifying non-equilibrium thermal fluctuations or mechanical vibrations via the so-called Feynman- micromotor. An example of mechanical rectification is found in the batteryless wristwatch. The original concept was described in as early as 1912 by Smoluchowski and was later revisited in 1963 by Feynman, in the context of rectifying thermal fluctuations to obtain useful motion. It has been shown that, although rectification is impossible at equilibrium, it is possible for the Feynman-micromotor to perform work under non-equilibrium conditions. These concepts can now be realized by MEMS technology and may have exciting implications in biomedicine - where the Feynman- micromotor can be used to power a smart pill, for example. Previously, Feynman's analysis of the motor's efficiency has been shown to be flawed by Parrondo and Espanol. We now show there are further problems in Feynman's treatment of detailed balance. In order to design and understand this device correctly, the equations of detailed balance must be found. Feynman's approach was to use probabilities based on energies and we show that this is problematic. In this paper, we demonstrate corrected equations using level crossing probabilities instead. A potential application of the Feynman-micromotor is a batteryless nanopump that consists of a small MEMS chip that adheres to the skin of a patient and dispense nanoliter quantities of medication. Either mechanical or thermal rectification via a Feynman- micromotor, as the power source, is open for possible investigation.

Most conventional eddy current sensors are used for position measurements where the target approaches the sensor coil along its axial direction. In this paper we investigate a versatile eddy current micro sensor that can not only be used for measuring displacements parallel to the axis of the sensor coil but also for transverse displacement measurement in micro magnetic bearings for micro machines. In the transverse mode, the sensor coil impedance variation has been typically found to be less than when used in the axial mode. A number of approaches and techniques have been investigated and will be discussed to overcome the problem of low sensitivity. As a result a versatile eddy current sensor has been developed which is suitable for transverse position measurement in micro magnetic bearings. Experimental results shown that micro sensor sensitivities up to 20mV/micrometers can be achieved when the target moves between two sensor coils. The performance of the micro sensor suggests that it will have considerable scope for application in the micro machine are.

A micro magnetic bearing actuator, with an outside diameter of 2.6 mm, is being fabricated at RMIT. The rotor of the bearing actuator is structured to operate with active control in the radial directions while being supported passively in the axial direction. This paper addresses the design problem for micro magnetic bearing actuators. The key issue in the design is the evaluation of magnetic forces. By applying the 'fringing range method' developed by the authors, a magnetic circuit model is developed for this actuator, which is used for calculating the flux densities at key locations and the magnetic forces on the rotor. Expressions for magnetic forces have been developed previously and a design procedure is proposed in this paper. It is found by using these expressions and the proposed procedure, that we can easily investigate the influence of design parameters on actuator performance. Design results for the micro magnetic bearing are given in detail.

A micro magnetic bearing actuator has been designed using elector-magnets. In this design, the rotor position is actively controlled in the radial direction and passively supported in the axial directions. A micro position sensor along with a Proportional Derivative control system constitutes the feed-back network, which ensures the rotor actively suspended in radial directions. The circular stator has four control coils which are sandwiched between two stator end plates. The diameter of the stator and rotor are 2.1 and 2.6 mm respectively, while the thickness is fixed at 250(mu) . The air gap between the stator and rotor has been fixed at 10(mu) . Stator and rotor plates were fabricated using permalloy electroplating, while the control coils of the stator were hand wound using conventional wire winding techniques. An enamel coated copper wire of 20(mu) diameter was used for this purpose. These parts are assembled using normal micro assembly techniques. The details of the fabrication and assembly techniques employed for the micro bearing are presented along with the preliminary test results.

We evaluate the efficiency of a new optoelectronic quenched avalanche sensor, with the potential of enabling high resolution imagin through turbid media with femtosecond lasers. Our target application is for imaging cancer in the human breast.

The rapidly developing field of terahertz (T-ray) imagin promises to provide a non-invasive method of identifying the composition of various objects. Biomedical diagnostics, semiconductor device diagnostics, trace gas analysis and moisture analysis for agriculture are among the growing number of important T-ray applications. We present results using an electro-optical sampling method and discuss how this can be simplified by sign a micro antenna array realized by MEMS technology. The challenges and advantages of both approaches are compared.

We have developed an optical fiber trapping system and verified that micro-sphere could be trapped and manipulated by a weakly focused single laser beam emerging from an otpical fiber end, which was inserted into a sample chamber at an angle of 35 degrees. In this paper, axial and transverse trapping forces exerted on a micro-sphere were experimentally and theoretically analyzed to corroborate the single-beam optical trap using an optical fiber. Experimental results were shown to be in excellent agreement with a numerical model that predicted the axial and transverse forces acting on a sphere gave a restoring force directed back to the stable point, and transverse force played a significant role in trapping and manipulation of a micro-sized object by means of an optical fiber. Furthermore, we experimentally verified that our proposed single-beam fiber optic trapping system was most economical, much simpler to operate, and required relatively low optical power to capture particle as compared with other embodiments.

In many applications, such as multimedia and on-chip camera, there is a need for the production of low power, low weight and low cost integrated circuits. Several CMOS vision chips have been proposed in the literature. Some limitations of conventional 2D architectures are discussed and a new 3D generation of vision chips is presented and reviewed in this paper. As a result of this analysis, some conclusions on the advantages and limitations of 2D vision chips and the feasibility of the 3D approach are explored.

A vertical spring-type micromirror was fabricated using a SiO₂ shielding screen structure. The fabricated vertical springs are well defined but the mirror plate is bent because the mirror plate is composed of SiO₂ and aluminum layers. To obtain a flat mirror plate, the stress of SiO₂ and aluminum is analyzed and the bending deflection of mirror plate's edge is measured. The measured deflection varies with the thickness of aluminum. Another SiO₂ layer on the mirror aluminum improved the flatness of mirror plate.

Semiconductor-polymer based sensors use silicon or compound semiconductors as inorganic parts with sensitive polymers as insulating, semiconducting or conductive materials. Organic thin film transistor has also been fabricated using this concept. These devices may allow control circuitry to be integrated with two and three dimensional MEMS. Interdigital type RF-MEMS can be designed and fabricated with Interdigital Electrodes (IDE) deposited on either polymer or an inorganic material such as Barium Strontium Titanate (BST). In the case of polymer based device, we study the capacitance change and calibrate it for desired sensing application. In the inorganic case, we make use of the change in dielectric properties of BST as a function of DC bias. IDE will act like a RF filter and oscillator just like the comb-type RF MEMS devices. These polymeric based devices can be integrated with organic thin film transistors.

With the future direction in satellite technology promising to be the production of nanosatellites, very small satellites no heavier than 10 kg or larger than 15 cm, operating in groups or clusters, the ability to detect and avoid collision is not known. Perhaps with the smaller size, the effects of solar, planetary and cosmic radiation in overwhelming the radiometric signal from another satellite is quite large, degrading the changes for collision avoidance. This paper predicts the sensitivities of our mm- wave collision avoidance sensor for the orbital environment. The effects of strong radiation sources and the dynamics of satellite heating and motions are explored. Methods and techniques for obtaining this information are discussed.

Phase-locked loop (PLL), as well known, is a very popular and cheap tool for pulse generation with almost constant frequency. On the other hand, a user-friendly and cheap tool for micro-displacement/capacitance servo control system still lacks in the market, although it is increasing the need in MEMS industries. In this article, a cheap and stable micro-displacement/capacitance measurement and control system is proposed. Simulations and experimental results show that, if the properties of the electronic components used in the PLL remain close to their nominal values, the proposed system can achieve good capacitance/ displacement measurement and/or servo control accuracy with high resolution. Application examples show that it is very useful in the micro-positioning system and MEMS industries.

A CMOS low noise, high sensitivity circuit suitable for detecting the change in capacitance along the sensing direction of a vibration gyroscope with interdigitized finger comb structure is presented. The circuit is sued in conjunction with a fully differential capacitor bridge, which is formed by the sensing port of the mechanical structure. A mathematical model of a gyroscope is established for electrical simulation with electrical equivalent circuit. Theoretical analysis and computer simulation using HSPICE show that as low as 0.2aF change in capacitance is detectable under ideal conditions. Two high frequency carriers of the same magnitude and frequency are used in the circuit. Simulation results review that the resolution of the technique is severely affected by the mismatch of the high frequency carries. The effect of parasitic capacitance, which is inherent in the silicon micromachining technology, can be greatly reduced by selecting appropriate circuit configuration.

A CMOS, self-biased transconductance amplifier has been designed to be associated and integrated with a silicon capacitive microphone. To meet requirements especially on gain sensitivity, power consumption, and minimization of parasite capacitance effect, we have proposed a cascode structure with the cascode transistor source used as signa input. Switched-capacitor techniques have been applied for realizing self-bias for the amplifier and ensuring its high- gain operation. The proposed amplifier has been designed and fabricated in a 0.8 micrometers CMOS process. It has a surface area of 210 micrometers by 170 micrometers . Experimental results obtained from measuring the fabricated chip show a high-gain sensitivity and a low power dissipation for the amplifier. Results of simulations and measurements have been discussed.

We have developed multi-sensing artificial touch devices for the recognition of object properties and its surface situation. In this paper we present the fabrication of a thin-plate micro tactile sensor with two sensing functions of capacitance and inductance, and introduce the results of comparison of characteristics among various types of micro sensor distinctly depend on each size of the sensors, them material of electrodes, and the depth of insulating layer in the sensor. Furthermore, by preparing the shielding layer in the sensor in the way of the fabrication process, it was found that the problem which occurred in miniaturizing the sensor would be solved.

Imaging laser radar systems that employ photodetector focal plane arrays with corresponding readout circuitry benefit from faster processing times and a reduction in weight due to their ability to be made into more compact units. This paper evaluates the concept of a 2D array of photodetectors and readout circuitry by performing a simple experiment on a custom-made four channel readout integrated circuit chip designed by the authors, with a single detector. Results from this initial experiment shows that though there are some problems, the concept is viable for a larger array of photodetectors.

We are attempting to develop a shape memory alloy (SMA) actuated micropump as a component for use in micro analysis or micro dosage systems. In this paper, we will discuss the fabrication process and dynamic actuation properties of an SMA actuator. TiNi thin film of about 6 micrometers in thickness was deposited onto a Si wafer with a square recess on its reverse side, and annealed at 500 degrees C for 1h in a vacuum to memorize an initial flat shape. The TiNi thin film and a Pyrex glass cap of 500 micrometers thickness with a square recess were then anodically bonded together in a vacuum to form a chamber to which a bias pressure was to be applied to deform the TiNi thin film. After removing the remaining 50 $mUm thick Si layer beneath the TiNi thin film by RIE in SF₆ plasma, a shape memory diaphragm of 5mm square in size was completed. The fabricated actuator was driven by resistive heating and air-cooling under a bias pressure of 200 kPa, which was applied by a nitrogen gas flow through a small hole pieced in the Pyrex glass cap. Under these conditions, the actuator gave displacements of about 95 micrometers at the center during repeated thermal cycles. A Si check valve structure, to be assembled with the actuator to construct a micropump, was also fabricated through a process of anisotropic etching and fusion bonding. The flow resistance for forward flow proved to be about one-three hundredths of that for backward flow.

We realized IC-compatible diamond AFM probe technology using our newly developed anodic bonding technique for diamond film to glass together with microfabrication techniques employing CVD diamond film. Using Al film as an intermediate layer, diamond film can be anodically bonded to Pyrex 7740 glass at a bonding temperature of 400-500 degrees C, with an electrostatic voltage of 600V, and a bonding time of 1-2 h in atmosphere. Based on the batch fabrication process proposed here, we demonstrate diamond AFM probes for contact mode measurements consisting of two kind of V-shaped cantilevers integrated with a pyramidal tip. They were optimally designed by FEA, giving spring constants of approximately 1 N/m and 5 N/m. The diamond cantilevers and diamond tips were fabricated by the selective deposition of diamond film and a Si mold technique, respectively. A diamond base, to which the diamond probes have been attached, was then anodically bonded to a glass backing plate to facilitate handling. This bonding technique was found to provide high accuracy and good repeatability. Finally, diamond proves attached with high accuracy to one end of the glass backing plate were obtained by removing the unnecessary Si substrate. They were then successfully mounted in a Si frame by means of four thin diamond brackets, making a chip array in the substrate. These results indicate that this fabrication process would allow the production of large quantities of diamond AFM probes, resulting in a high cost performance.

Thin film stress has turned out to be a limiting factor for the fabrication of dielectric membrane based microsystem using commercial CMOS processes. Due to the compressive stress in the CMOS dielectrics, membranes tend to buckle and - in the extreme case - to break, i.e., the sensor yield is reduced and consequently production costs rise. In this paper we report the results of the stress analysis of a commercial 1 micrometers , single poly, double metal CMOS process. All dielectric layers from the field oxide to the passivation were analyzed. We investigate several methods to taylor the stresses in the dielectrics to the needs of microsensor fabrication. The most mature method was to change the stress in the passivation layer by adjusting the deposition parameters. We investigated two passivation layers in more detail. With the standard passivation the dielectric sandwich has a total thickness of nearly 4 micrometers and an average stress of -162 MPa. The second passivation layer is thicker and more tensile than the standard passivation. It reduces the average stress to -15 MPa. The critical buckling lengths of CMOS dielectric membranes were determined experimentally. We found the critical buckling lengths of such membranes containing these two passivation layers to be (194 + 26) micrometers and (967 + 103) micrometers , respectively.

Microactuator is one of the important parts in MEMS, various types of microactuators have been developed, which utilize different driving principles such as electrostatic force, piezoelectric effect, shape memory alloys and electromagnetic force. In this paper, the fabrication process of stator winding of (phi) 1mm electromagnetic micromotor is discussed. With high-aspect-ratio photolithography process and mask-plating process, the high- aspect-ratio coils can be fabricated, which greatly reduce the resistance and heat of stator winding. With thick sputtered Al₂O₃ films as isolation layers between coils, heat resistance of stator winding is improved. A stator winding has been fabricated, which is composed of 6 layer coils, 42 turns and 9 pairs, its diameter is 1mm and minimum line space is 1 micrometers , the maximum operation current of the coils is 180mA and resistor is 12-20 (Omega) . The output torque of the electromagnetic micromotor is 1.5 (mu) Nm. the rotational speed of the motor can be adjusted from several hundred to over ten thousand turns per minute and the rotational direction is reversible.

We have already developed a single-laser beam optical fiber trapping system. A biological cell dispersed in water solution could be easily trapped by a single laser beam emerging from an optical fiber, and the optically trapped micro-object could be freely transferred in 2D plain synchronized to the trapping fiber. In this paper, multibeam cell manipulation was successfully demonstrated. Experimental results were as follows: (i) Separation/coupling of an individual cell from the cell groups was easily achieved using plural optical fibers. (ii) Optical force acting on an elliptically shaped object was experimentally analyzed. From these result, it was found that the optical force acting on an elliptically shaped object tended to rotate the object about the center of mass G of the object. (iii) Biological cell laid at the bottom of the sample chamber could be levitated against gravity using plural optical fibers. Micro object was captured and held at equilibrium point, where the force of gravity, the total axial forces in the downward direction and the total transverse forces in the upward direction balanced. From these experimental investigations, we verified that transverse forces played an important role in levitating a microscopic object using plural optical fibers.

A novel micropump actuated by thin film shape memory alloy has been designed, fabricated and tested. It is one of the membrane reciprocating micropumps which is driven by shape recoverable force of NiTi thin film and biasing force of Si membrane. The micropump is composed of a deformable chamber and two Si flap check valves. The structure of pump is simple. The pump was manufactured by Si surface and bulk micromachining technologies, Au-Si eutectic and epoxy bonding. The outer dimension of the micropump is 6*6*1.5 mm/mm/mm. The NiTi/Si composite actuator has an area of 3*3mm/mm with a thickness of 20micrometers Testing result show that the pump has extremely good performance, such as high pumping yield, low power consumption high working frequency. All these advantages are attributed to the patterning of NiTi thin film used as self-heating element and optimizing design of NiTi/Si actuated structure.

The aim of this work is to investigate the use of microchannels to concentrate pollutants present in the air. Devices with a length of 30 cm, a width of 100 micrometers and a depth of 30 micrometers , sealed by anodically bonded glass, were manufactured. Tests of adsorption characteristics were made using n-hexane. To reliably insert N₂ contaminated with 1000 ppm of n-hexane in the microstructure, a simple setup was manufactured. This setup allows to insert the reactant, remove the amount of reactant adsorbed and to detect it. An amount of 20 mg was inserted in the microstructure. In order to improve the adsorption characteristics, PS layers were manufactured in the microchannels using silicon nitride as mask. The sealing of the microstructure with anodic bonded glass showed to be feasible either if the surface present PS or silicon nitride. Samples of PS layers covered by a plasma polymerized film, produced using HMDS, were analyzed by Raman microscopy. It was noticed that the nanocrystals are completely fulfilled by the deposited material, indicating the high reactivity of the PS layer.

This paper presents a new cost-effective fiber-to-waveguide coupling method for self-aligning optical fibers on silicon platforms, and for achieving optical quality end-polished silicon-on-insulator (SOI) single-mode rib waveguide devices using wet chemical micromachining techniques. Through accurate alignment to the plane(s) of the device layer of a SOI wafer, rib waveguide devices with self-alignment features are fabricated with the ends of each waveguide wet etched and concurrently polished providing an optical quality facet or fiber-to-waveguide interface. Eliminating the need to saw cut and then mechanically polish the waveguide device ends, the overall fabrication process is simplified and provides a fiber alignment capability at the ends of the waveguide devices with an alignment accuracy limited by fiber size tolerance. Experimental measurements were carried out to verify the optical quality of the waveguide facets formed using this new technique, which proved excess facet losses of practically unmeasurable quantities. Both simulation and experimental results were obtained to verify the single-mode nature of the rib waveguides.

An electrostatic diaphragm micropump for fluids and gases has been designed and fabricated by silicon surface micromachining. Diaphragm structures are bridge-type, cantilever-type and fan-type polysilicon, and sacrificial layers are low-temperatures oxide on polysilicon substrates. The developed anhydrous HF gas-phase etching of sacrificial oxide on polysilicon substrates has been employed to release polysilicon microstructures. The fabricated micropump with size of the order of 1 mm² operates at square wave voltage of 10V and 2Hz under near room temperature and normal atmospheric pressure.

This paper describes reversed engineered biological peristaltics through the insightful evolution of MEMS technology. While the emphasis on pump applications is in the miniaturization of spacecraft bus and payload, terrestrial applications are evident. For spacecraft propulsion, the use of the peristaltic pump as a low leakage valve, a pressure regulator and a flow meter are discussed. Spacecraft stabilization is addressed through a fluidic reaction wheel while thermal management of spacecraft, flight instruments and lander is addressed through heat pumps and Joule-Thompson compressors.

The ultra-sensitive accelerometer, developed for NASA to monitor the microgravity environments of Space Shuttle, free orbitors and Space Station, needed to measure accelerations up to 10 mg with an absolute accuracy of 10 nano-g for at least two orbits to resolve accelerations associated with orbital drag. Also, the accelerometers needed to have less than 10^-9 F.S. off-axis sensitivity; to be thermally and magnetically inert; to be immune to quiescent shock, and to have an in-situ calibration capability. The utilization of these accelerometers in multi-axis compact seismometers designs that have twelve decades of dynamic range, density profilometers, precision gradiometers, gyros and vibration isolation designs and applications will be discussed. FInally, examples of the versatility of the proof mass suspension system will be demonstrated through the transformation of the basic accelerometer into sensitive anemometers and imaging spectrometers.

This paper presents and discusses experimental results which successfully demonstrate using flip-chip techniques, the bonding and operation of a GaAs edge emitting laser on a silicon substrate having recessed bonding pads. The flip chip alignment and bonding technique is discussed first, focusing on experimental results for both flux and flux-less forms of solder bump processing, the developed flux-less technique is shown to provide comparable solder bump integrity and yield. Subsequent to the newly developed solder bump fabrication methodology shown, a recessed bonding pad design is introduced and demonstrated experimentally to aid in the final coarse alignment and assembly of the GaAs laser device onto a silicon substrate. Together both the flux-less solder bump and bonding pad recess design and fabrication processes provide increased reliability for integration with MEMS and MOEMS based devices and systems over traditional methods for hybrid device attachment.

MEMS devices have evolved in the last 10 years from an academic and laboratory curiosity to a viable technology with potential widespread application. Significant advancements are now being made to package and manufacture MEMS devices for the commercial and consumer marketplace. In MEMS devices that are silicon-based, integrated circuit fabrication techniques are used to reduce the size of mechanical features to the miniature dimensions of microelectronics. Integrating these mechanical features with electronic circuits on the same chip offers a wealth of new product opportunities not possible with conventional design and manufacturing approaches. Initially MEMS referred to silicon-based devices but today refers to a complete miniature device that may or may not be silicon based. Also MEMS devices have expanded beyond the integration of electronic and mechanical functions. Today devices can have mechanical, optical, fluidic, and thermal functions in some combination with each other and generally integrated with electronic functions. MEMS technology can be used to fabricate extremely sensitive sensor for a variety of applications including pressure, sound, mass, motion, acceleration, rotation, flow, chemicals, vibration, and temperature. The technology can also be used to fabricate field-emission tips and probes for use in high-brightness specialized displays, electron-beam sources, information storage applications and precision temperature sensing. Actuators can be fabricated for use in optical switching, flow control, micropositioning, adaptive optics, and displays. Finally, a variety of passive, 3D, micromechanical structures can be fabricated with applications in telecommunications, electronic components, and sensors.

Liquid-crystal over silicon is an established technology for reflective spatial light modulators and microdisplays. This paper reviews their development to date, highlighting in particular the micromachining of the mirror array and the associated packaging issues.

Ovonyx, Inc. is developing a nonvolatile memory that is potentially denser, faster, and easier to make than Dynamic RAM. It relies on phase transitions induced by nanosecond heating and cooling of small region of the memory cell. Initial target markets are FLASH memory, Embedded memory, and DRAM.

When developing new MEMS devices there are several decisions which have to be made. The first of these is which technology to use. Today there are a range of silicon based micromachining technologies available. In addition to backside bulk micromachining and surface micromachining, new processes available include epi-micromachining and deep plasma etching. These are all silicon based technologies. New processes using alternative materials have also been developed and these include processes such as LIGA and SU-8. The decision of which process to use is often based on the equipment available and the type of structure to be fabricated. The second question is often more difficult, and that is whether or not to integrate. The integration of sensor with electronics can present a number of advantages, in particular when dealing with small signals or noisy environments. However, there are many applications where environmental factor ensure that the electronics will not function correctly. It is also important to ensure that the processing used for the MEMS devices does not adversely affect the electronics. MEMS device scan be fabricated as pre- or post-processing modules, or integrated within standard processing. In each of these optics on the compatibility issues such as non-contamination of the processing line and thermal budget have to be considered. Furthermore the potential drop in yield of the total devices has to be weighed up against the advantages. This paper will examine these decision and how the different aspects of technology and applications effect the choices concerning whether or not to integrate.

CMOS has been used for advanced LSIs so many years because of its excellent low-power and high-speed characteristics for highly-integrated digital circuits. For high frequency analog applications such as RF, it has not been used popularly because CMOS has been assumed to have poor capability compared with silicon bipolar and compound semiconductor devices. However, recent miniaturization of CMOS devices has significantly improved the CMOS RF characteristics. For example, typical values of f_T and fmax for 0.25 micrometers n-MOSFETs already exceed 40 GHz, and those for 0.1 micrometers n-MOSFETs are more than 100 GHz. RF noise of the MOSFETs is also as god as less than 1 dB at 2 GHz operations. In this paper, CMOS technology for RF front- end circuit for mobile telecommunication devices is explained.

It has been close to 60 years since the first operational silicon solar cell was demonstrated. However, the last 15 years have seen large improvements in the technology, with the best confirmed cell efficiency improved by over 50 percent. The main drivers have been improved electrical and optical design of the cells. Improvements in the first area include improved passivation of contact and surface regions of the cells and a reduction in the volume of heavily doped material within the cell. Optically, reduced reflection and improved trapping of the light within the cell have had a large impact. Features have increased silicon cell efficiency to a recently confirmed value of 24.7 percent. Over recent years, good progress has been made in transferring some of the corresponding design improvements into commercial product with commercial cells of 17-18 percent efficiency now commercially available, record values of a mere 15 years ago. The theory supporting these improvements in bulk cell efficiency shows that thin layers of silicon, only a micron or so in thickness, should be capable of comparably high efficiency.