In addition to being small, simple and robust, surface acoustic wave (SAW) devices have the advantage of being passive (batteryless), remotely requestable (wireless) and inexpensive if fabricated on a large scale. The use of SAW devices as passive and wireless sensors allows them to operate in extreme conditions such as those with high levels of radiation, high temperatures up to 1000°C, or electromagnetic interference, in which no other wireless sensor can operate. This is obviously conditioned by the fact that the materials constituting the device can withstand these harsh conditions. General principle of the SAW sensor in wireless configuration is developed and a review of recent works concerning the field of high temperature applications is presented with a specific attention given to the characterization of materials constituting the SAW device: piezoelectric substrate and metallic electrodes.

For future measurements while depth drilling, commercial sensors are required for a temperature range from -40 up to 300 °C. Conventional piezoresistive silicon sensors cannot be used at higher temperatures due to an exponential increase of leakage currents which results in a drop of the bridge voltage. A well-known procedure to expand the temperature range of silicon sensors and to reduce leakage currents is to employ Silicon-On-Insulator (SOI) instead of standard wafer material. Diffused resistors can be operated up to 200 °C, but show the same problems beyond due to leakage of the p-njunction. Our approach is to use p-SOI where resistors as well as interconnects are defined by etching down to the oxide layer. Leakage is suppressed and the temperature dependence of the bridges is very low (TCR = (2.6 ± 0.1) μV/K@1 mA up to 400 °C). The design and process flow will be presented in detail. The characteristics of Wheatstone bridges made of silicon, n- SOI, and p-SOI will be shown for temperatures up to 300 °C. Besides, thermal FEM-simulations will be described revealing the effect of stress between silicon and the silicon-oxide layer during temperature cycling.

We have fabricated microthruster chip pairs - one chip with microthruster structures such as injection capillaries, combustion chamber and nozzle, the other chip with platinum thin film devices such as resistance temperature detectors (RTDs) and a heater. The platinum thin film was sputtered on thermally oxidized silicon wafers WITHOUT adhesion layer. The effects of anneal up to 1050°C on the surface morphology of platinum thin films with varied geometry as well as with / without PECVD-SiO₂ coating were investigated in air and N₂ and results will also be presented. Electrical characterization of sensors was carried out in a furnace tube in which the sensors' temperature was varied between room temperature and 1000°C with a ramp of ±5Kmin^-1 in air and N₂. The experiments showed that the temperature-resistance characteristics of sensors had stabilized after the first heating up to 1000°C in N₂. After stabilization the sensors underwent further 8 temperature cycles which correspond to over 28h of operation time between 800 - 1000°C. To reduce the loss of combustion heat, chip material around the microthruster structures was partially removed. The effects of thermal insulation were investigated with microthruster chip pairs which were clamped together mechanically. The heater power was varied up to 20W and the temperature distribution in the chip pairs with / without thermal insulation was monitored with 7 integrated thin film sensors.

Miniaturized piezoelectric structures based on the material langasite are fabricated and characterized. Single crystalline langasite is a high-temperature stable material and can be excited using the piezoelectric effect up to temperatures above 1000 °C. The scope of our research is to design and realize small piezoelectric structures to serve as resonant sensors or active electronic components at high temperatures. Wet chemical etch processes are developed which enable isotropic and anisotropic etching with rates up to 90 μm/h. Small structures such as membranes as well as cantilevers are prepared. Electrical and optical characterization provides information about the electromechanical properties of those structures. Resonance properties are determined with electrical impedance measurements in-situ at temperatures close to 1000 °C. The temperature dependent frequency shift and the resonator quality factor are determined for membranes, cantilevers and tuning forks. In addition, investigations with a laser interferometer are obtained. Thereby, the spatial distribution of the vibration displacement is observed for several piezoelectric devices.

°Stable nanostructured ultra-thin electrodes and protective passivation coatings have been developed for langasite (La₃Ga₅SiO₁₄) based surface acoustic wave (SAW) sensors that can successfully operate in harsh high temperature environments up to 1000°C. Ultrathin (<100nm) nanocomposite Pt-Rh/ZrO₂ electrode structures were fabricated by electron beam co-evaporation and characterized by X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), 4-point resistivity, and scanning electron microscopy (SEM) studies. It was found that the incorporation of ZrO₂ into the Pt-Rh electrode films retards recrystallization and de-wetting, thereby maintaining film continuity and low resistivity up to at least 1000°C. XPS results show that with heating at 800°C, the stoichiometry of the bare langasite SAW sensor surface becomes depleted of Ga in a reducing (vacuum) environment, but remains close to the bulk composition when heated in an oxidizing (air) environment. The incorporation of a thin oxynitride (SiAlON or SiZrON) coating over the entire sensor diminishes high temperature roughening and degradation of both the electrode and bare langasite surfaces. The viability and performance of these sensors was validated by experiments in which the SAW devices were tested in a controlled atmosphere laboratory furnace and also attached to rotating turbine blades within a small turbine engine operating with centripetal acceleration loads and temperatures in excess of 52,000g and 650°C, and under cyclical temperature shock conditions.

Aluminum nitride is a promising material for the use as a piezoelectric sensor material for resonance frequencies higher than 50 MHz and contains the potential for high frequency phased array application in the future. This work represents the fundamental research on piezoelectric aluminum nitride films with a thickness of up to 10 μm based on a double ring magnetron sputtering process. The deposition process of the aluminum nitride thin film layers on silicon substrates was investigated and optimized regarding their piezoelectric behavior. Therefore a specific test setup and a measuring station were created to characterize the sensors. Large single element transducers were deposited on silicon substrates with aluminum electrodes, using different parameters for the magnetron sputter process, like pressure and bias voltage. Afterwards acoustical measurements were carried out in pulse echo mode up to 500 MHz and the piezoelectric charge constants (d₃₃) were determined. As a result, two parameter sets were found for the sputtering process to obtain an excellent piezoelectric charge constant of about 7.2 pC/N maximum.

Silicon carbide (SiC) is a promising material for applications in harsh environments. Standard silicon (Si) microelectromechanical systems (MEMS) are limited in operating temperature to temperatures below 130 °C for electronic devices and below 600 °C for mechanical devices. Due to its large bandgap SiC enables MEMS with significantly higher operating temperatures. Furthermore, SiC exhibits high chemical stability and thermal conductivity. Young's modulus and residual stress are important mechanical properties for the design of sophisticated SiC-based MEMS devices. In particular, residual stresses are strongly dependent on the deposition conditions. Literature values for Young's modulus range from 100 to 400 GPa, and residual stresses range from 98 to 486 MPa. In this paper we present our work on investigating Young's modulus and residual stress of SiC films deposited on single crystal bulk silicon using bulge testing. This method is based on measurement of pressure-dependent membrane deflection. Polycrystalline as well as single crystal cubic silicon carbide samples are studied. For the samples tested, average Young's modulus and residual stress measured are 417 GPa and 89 MPa for polycrystalline samples. For single crystal samples, the according values are 388 GPa and 217 MPa. These results compare well with literature values.

The mechanical properties of two aluminium thin films manufactured by two different laboratories were investigated using in-plane uniaxial tensile stress. The geometrical parameters, in particular film thickness, were accurately measured and their influences on the mechanical properties were analyzed. The specimens provided by each supplier show significant differences of their mechanical properties. In the same way, annealing was performed on a specific set of specimens and its influence on the mechanical properties is highlighted and discussed.

This paper presents an innovative approach to develop highly sensitive 3D force sensors. In order to increase the probing sensitivity by using a material with low Young´s Modulus, a novel force sensor design based on SU-8 polymer is realized. Therefore, a low cost fabrication process is developed accompanied by design studies and an estimation of mechanical properties of the deforming elements. This paper will present the fabrication process as well as a distinguishing set of test results, analytical results, and simulations on the characterization of the presented SU-8 force sensor. The novel SU-8 design consists of an SU-8 boss-membrane with integrated piezoresistive elements. The SU-8 membranes are structured using photolithography. The SU-8 micromechanical structures are characterized to determine film stresses, bending stiffness, displacement of the stylus, and breaking points. Included in these tests are measurement of the stress behavior at different process steps and simulation of stress distribution in the membrane at different directions of loading. In addition a comparative analytical investigation of the structures is carried out particularly with regard to the displacement of the stylus. Using SU-8 as sensor material, due to low Young´s modulus, promises sensitivity much higher than what is obtained using conventional silicon force sensors.

In diaphragm-based micromachined calorimetric flow sensors, the convective heat transfer through the test fluid competes with the spurious heat shunt induced by the thin-film diaphragm where the heater and the temperature sensors are embedded. Therefore, accurate knowledge of the thermal transport properties (thermal conductivity and diffusivity) and the emissivity of the diaphragm is mandatory for design, simulation, and optimization of such devices. However, these parameters can differ considerably from those stated for bulk material and they are typically dependent on the production process. Commonly used methods for their determination require the fabrication of custom specimens. In order to overcome this serious drawback, we developed a novel technique to extract the thermal thin-film properties directly from measurements carried out on calorimetric flow sensors. Here, the heat transfer frequency response from the heater to the thermistors is measured and compared to a theoretically obtained relationship arising from an extensive two-dimensional analytical model. This model covers the heat generation by the heater, the heat conduction in the diaphragm, the radiation loss at the diaphragm's surface, and the heat sink caused by the supporting silicon frame. In this contribution, we report in detail on the measurement setup, the theoretical model for the associated parameter extraction, and the results obtained from measurements on calorimetric flow sensors featuring dielectric thin-film diaphragms made of PECVD Si₃N₄.

In this study, two approaches are compared to develop thin, multifunctional films of carbon nanotubes (CNT) which are targeted to serve as a catalyst layer in fuel cells. The first is based on the direct deposition of mixed multi- and single-wall CNTs on metalized silicon wafers, using the metallization as a sacrificial layer to subsequently detach the CNT film from the substrate. It is a less time consuming and a straight forward method compared to the alternative under investigation, the layer-by-layer technique (LbL). The LbL uses bilayers of charged nanotubes to slowly build up a film with an exactly defined thickness. The process is well controlled, but the time constants for deposition of each bilayer are rather high (i.e. about 1 h). With additional annealing steps implemented during film generation this method, however, is regarded advantageous as membranes results with improved mechanical stability and a good homogeneity.

In this paper two different photonic lab-on-a-chip devices for real-time cell screening are presented. We discuss, in parallel, screening of cell viability and cell count using a multiple total internal reflection (MIR) system, and a microbioreactor (MBR) for continuous cell culture screening. Each device takes advantage, in different ways, of the enhanced light-analyte interaction provided by the full-field approach. Their performance and how they represent an improvement to current methods, are also discussed herein. Both devices integrate photonic components and are fabricated by soft lithography (PDMS) replica molding of SU-8 masters, thus making them low-cost and easily mass produced.

In this contribution we present a sensor system to measure the CH₂-stretch ratio of suspended mammalian cells. To overcome the strong infrared absorbance of water our sensor system comprises a sample chip with three equal chambers with an inner height of only 20 μm.

Digital microfluidics utilises discrete droplets to increase the throughput of chemical and biological assays. There is a great need for sensing techniques that can help automate droplet handing. This work aims to illustrate by means of finite element modelling the factors that affect the measurement of droplet impedance in a thin-walled glass capillary. These include modelling the effect of varying capillary wall thickness, electrode spacing, electrode area, droplet size, shielding and droplet conductivity. The simulation is performed using COMSOL finite element analysis software. The model serves to aid the optimization of the measurement system for application in the measurement of the presence/absence of droplets as well as their chemical and biological content.

Two microfluidic cartridges intended for upgrading standard laboratory instruments with automated liquid handling capability by use of centrifugal forces are presented. The first microfluidic cartridge enables purification of DNA from human whole blood and is operated in a standard laboratory centrifuge. The second microfluidic catridge enables genotyping of pathogens by geometrically multiplexed real-time PCR. It is operated in a slightly modified off-the-shelf thermal cycler. Both solutions aim at smart and cost-efficient ways to automate work flows in laboratories. The DNA purification cartridge automates all liquid handling steps starting from a lysed blood sample to PCR ready DNA. The cartridge contains two manually crushable glass ampoules with liquid reagents. The DNA yield extracted from a 32 μl blood sample is 192 ± 30 ng which corresponds to 53 ± 8% of a reference extraction. The genotyping cartridge is applied to analyse isolates of the multi-resistant Staphyloccus aureus (MRSA) by real-time PCR. The wells contain pre-stored dry reagents such as primers and probes. Evaluation of the system with 44 genotyping assays showed a 100% specificity and agreement with the reference assays in standard tubes. The lower limit of detection was well below 10 copies of DNA per reaction.

A Silicon Carbide based enhancement type field effect transistor with porous films of Iridium and Platinum as gate metallization has been investigated as a total NOx sensor operated in a temperature cycling mode. This operating mode is quite new for gas sensors based on the field effect but promising results have been reported earlier. Based on static investigations we have developed a suitable T-cycle for NO_x detection in a mixture of typical exhaust gases (CO, C₂H₄, and NH₃). Significant features describing the shape of the sensor response have been extracted allowing determination of NO_x concentrations in gas mixtures. Multivariate statistics (e.g. Linear Discriminant Analysis) have been used to evaluate the multidimensional data. With this kind of advanced signal processing the influence of sensor drift and cross sensitivity to ambient gases can effectively be reduced. Thereby, we were able to detect NO_x and furthermore determine different concentrations of NO_x even in mixtures with typical exhaust gases. It can be concluded that the performance of field effect gas sensors for NO_x determination can be enhanced considerably.

Gas sensing can be achieved by fingerprinting the ionization characteristics of distinct species. In this study, the fabrication of a miniaturized gas ionization sensor using polyimide as sacrificial layer is reported. The sensor consists of two planar metallic electrodes with a gap spacing obtained by the polyimide under-etching. This known sacrificial layer has the advantage besides a high planarization factor, to be CMOS compatible. Furthermore, its chemical resistance up to high temperatures, high resistance to radiation from both electrons and neutrons, and low outgassing are of primary importance to avoid interferences with the ionization gas sensing. A suspended micro-bridge with dimensions 20 μm width and 220 μm length has been developed and released by using etching holes in the membrane. The ionization characteristics of air at controlled temperature, humidity and pressure (21°C, 40% humidity and 1 atm) have been obtained during non-destructive electrical characterizations, with a breakdown voltage of 350 V for a 6 μm gap. The growth of metallic nanowires templated in ion track-etched polyimide on the electrode is envisioned in order to enhance the ionization field and to reduce the required measurement power of the sensor.

Non-dispersive infrared (NDIR) gas sensors make use of the specific infrared absorption of particular gas molecules in order to measure their distinctive gas concentration. The main parts of such a NDIR gas sensor are: an IR-emitter, a chamber containing the sample-gas, and an IR-detector with a filter for the characteristic absorption wavelength. The effectiveness of the IR-source for the total system is characterized by its temperature and the emissivity (i.e., the difference to blackbody radiation) of the device surface. Due to the fact that conventional metal surfaces provide a rather low emissivity, their emitting temperature must be set very high to generate sufficient IR-radiation for this kind of sensors. We developed an IR-source consisting of a stack of thin films with a much higher emissivity. Its main part is a combination of two mirrors and a dielectric layer which represent a Fabry-Perot structure. The obtained emission of the Fabry-Perot structure and the consequences for the performance of the whole NDIR gas sensor system were simulated with the enhanced transmittance matrix approach and a 3D ray tracing model. As an example, CO₂ was considered as sample gas where the major characteristic absorption occur around 4.26 μm. The theoretical results are validated by comparing them to experiments obtained with prototype devices.

The work presented here focuses on the investigations of metallo-porphyrins and their gasochromic behavior. Gasochromic materials change their color while they are exposed to a certain gas. So they offer the possibility to develop very selective chemical gas sensors. In the focus of this work is the metallo-porphyrin 5, 10, 15, 20- tetraphenylporphyrin-zinc (ZnTPP). When embedded into a polymeric matrix (PVC) the color change to the toxic gas NO₂ can be detected. During exposure to NO₂ the dye changes its color from bright purple to yellow. To develop a standalone gas sensor, the ZnTPP/PVC matrix is deposited onto a planar optical waveguide. The color change of the porphyrin dye, due to the gas exposure, can be detected in the evanescent field of the optical waveguide. Therefore the light of a high power LED is coupled into the waveguide. The color change of the porphyrin is detectable with photodiodes as variations of the decoupled light intensity. The sensor shows no cross-sensitivities to other gases like CO₂, NH₃, EtOH, CO or water vapor. NO₂ is detectable with a limit of 1 ppm.

The methods of the fabrication of MEMS platforms based on yttria stabilized zirconia (YSZ) and alumina membranes for gas sensors used in harsh environmental conditions are described. It is shown that the application of such membranes permits a decrease in MEMS power consumption at 450°C down to ~75 mW at continuous and down to ~ 1 mW at pulse heating of gas sensor.

We have developed compact devices comprising optical position sensing and driver electronics with closed loop control, capable of driving resonant 1D- and 2D-MOEMS scanner mirrors. Position encoding is realized by measuring a laser beam reflected from the backside of the mirror. In the 2D-device we use cylindrical mirrors in order to suppress the deflection of the orthogonal dimension. This reduces the problem to the control of two independent 1D-oscillations and allows accurate position sensing. The phase between the oscillations of the two orthogonal axes is actively controlled to achieve a stable Lissajous figure. In this contribution we also demonstrate that this approach is scalable for synchronization of separate MEMS mirrors.

Although capacitive and piezoresistive readouts are commonly used for microelectromechanical structures, they suffer from serious drawbacks like limited range of displacement, inherent nonlinearity, insucient sensitivity, and technological complexity. Our complementary readout approach relies on a novel hybrid optomechanical device, where the displacement range is not limited by typical constraints of capacitive or piezoresistive conversion principles. Furthermore, no electrical connections are required on the micromechanical part. Moreover, this principle enables custom linear or nolinear output characteristics at the same technological effort.

Infrared spectroscopy uses the characteristic absorption of the molecules in the mid infrared and allows the determination of the gases and their concentration. Especially by the absorption at longer wavelengths between 8 μm and 12 μm, the so called "fingerprint" region, the molecules can be measured with highest selectivity. We present an infrared optical filter photometer for the analytical determination of trace gases in the air. The challenge in developing the filter photometer was the construction of a multi-channel system using a novel filter wheel concept - which acts as a chopper too- in order to measure simultaneously four gases: carbon monoxide, carbon dioxide, methane and ammonia. The system consists of a broadband infrared emitter, a long path cell with 1.7m optical path length, a filter wheel and analogue and digital signal processing. Multi channel filter photometers normally need one filter and one detector per target gas. There are small detection units with one, two or more detectors with integrated filters available on the market. One filter is normally used as reference at a wavelength without any cross-sensitivities to possible interfering gases (e.g. at 3.95 μm is an "atmospheric window" - a small spectral band without absorbing gases in the atmosphere). The advantage of a filter-wheel set-up is that a single IR-detector can be used, which reduces the signal drift enormously. Pyroelectric and thermopile detectors are often integrated in these kinds of spectrometers. For both detector types a modulation of the light is required and can be done - without an additional chopper - with the filter wheel.

High-speed laser manipulation is a key technology in applications such as fast optical switches for coupling light into optical fibres, 2D beam steering for optical scanners, laser printers or 2D sensors. A novel design of a Polyvinylidene Fluoride (PVDF) bimorph actuator is proposed in this work as high speed switching element in the kHz range. The actuator of several cm length and 200 microns thickness takes advantage of the structural and electromechanical capabilities of the PVDF. A 0.164 g compact mirror is placed at the tip of the PVDF bimorph actuator allowing scanning of the beam in two directions. Tests made by an autocollimator have been performed in order to characterize the precision, repeatability and also the long term stability at low frequencies. To complete this characterization, a photodiode was used to measure the response of the mirror at high frequencies up to 3 kHz. Results of the FE element confirm the trends observed in the experimental setup. Experimental results indicate that the PVDF actuator responds conveniently to manipulate the laser beam at kHz frequency. In addition, due to the intrinsic properties of the PVDF, the system is compact and light-weight. This is of advantage in several domains such as e.g. aerospace or micro-engineering.

The need to perform structural health monitoring on composite primary structures in real time for their life cycle has become cardinal because of the drastic increase in composite usage on aircraft, helicopters and unmanned aerial systems. The Systems Engineering approach was followed to ensure that these efficient low weight high strength components are optimally, economically and safely utilized. Details of the phases involved in this approach are outlined. In this document the activities associated with the preliminary design phase of the Systems Engineering process will be emphasised. Glass embedded optical fibre Bragg sensors were identified as the most appropriate for the strain measurement essential for the structural health monitoring of composites. The necessary Interrogator instrumentation subsystem for data acquisition and an Algorithm analysis subsystem are outlined. Detail design aspects of only the embedded optical fibre Bragg sensor subsystem will be covered in this paper.

Usage of fiber-optic Bragg gratings (FBG) for strain measurement is well-known technique in structural health monitoring (SHM). However, this technique based on shift of spectral peak, suffers from different spurious signals, particularly caused by thermal effect. We present here a method for impact damage detection of composite materials based on FBG without thermal disturbance. This method is based on the broadening of the spectral peak in dependence on the FBG separation from the impact damage. We used a sensing configuration where an Optical Spectrum Analyzer was interrogated with an array of 10 FBGs with central wavelength between 812 and 817nm that were bonded on a CFRP composite plate with 8 plies. We performed two groups of impact experiments: by impact energy of about 10J and 20J. We found there was no any significant shift of the spectral peak after the impact. Contrary, we confirmed in the last experiments the spectral peak broadening caused by impact load. The spectrum broadening primarily depends on the strain gradient generated into the FBGs, i.e. into the damaged area. We made a correlation between the peak width and FBG-impact damage separation. There are three characteristic regions; near to impact, abrupt region and almost even region.

The non-ferroelectric polar wurtzite aluminium nitride (AlN) material has been shown to have potential for various sensor applications both utilizing the piezoelectric effect directly for pressure sensors or indirectly for acoustic sensing of various physical, chemical and biochemical sensor applications. Especially, sputter deposited AlN thin films have played a central role for successful development of the thin film electro-acoustic technology. The development has been primarily driven by one device - the thin film bulk acoustic resonator (FBAR or TFBAR), with its primary use for high frequency filter applications for the telecom industry. AlN has been the dominating choice for commercial application due to compatibility with the integrated circuit technology, low acoustic and dielectric losses, high acoustic velocity in combination with comparably high (but still for some applications limited) electromechanical coupling. Recently, increased piezoelectric properties (and also electromechanical coupling) in the AlN through the alloying with scandium nitride (ScN) have been identified both experimentally and theoretically. Inhere, the utilization of piezoelectricity in electro-acoustic sensing will be discussed together with expectation on acoustic FBAR sensor performance with variation in piezoelectric material properties in the parameter space around AlN due to alloying, in view of the Sc_xAl_1-xN (0<x<0.5) solid solution findings. Original calculations are presented to support the discussion.

We develop shunt capacitive RF MEMS switches in III-V technology making use of materials which can be alternative to the ones commonly used, in order to overcome some technological constraints concerning the RF MEMS reliability. Specifically, we evaluate the potential of tantalum nitride (TaN) and tantalum pentoxide (Ta₂O₅) to be used for the switches actuation pads and dielectric layers, respectively. To this scope, a compositional, structural and electrical characterization of TaN and Ta₂O₅ films as a function of the deposition parameters, such as the substrate temperature and the sputtering mixture composition, is performed. The realized switches show good actuation voltages, in the range 15- 20 V, an insertion loss better than -0.8 dB up to 30 GHz, and an isolation of ~ -40 dB at the resonant frequency. A comparison between the measured S-parameter values and the results of a circuit simulation is also presented and discussed, providing useful information on the operation of the fabricated switches.

A phenomenological approach is herein followed to describe several superimposed effects occurring within the structure of microswitch for radio frequency application (RF-MEMS). This device is operated via a nonlinear electromechanical action imposed by applied voltage. Unfortunately, it is usually affected by residual stress, after microfabrication, therefore axial and flexural behaviors are coupled. This coupling increases actuation voltage to achieve the so-called "pull-in" condition. Moreover, temperature may strongly affect strain and stress distributions, respectively. Environmental temperature, internal dissipation of material, thermo-elastic and Joule effects play different roles on the microswitch flexural displacement. Sometimes buckling phenomenon evenly occurs. Finally, if stress concentration on microbeam cross section is sufficiently large, plastic behavior may arise. All those aspects make difficult an effective computation of pull-in voltage, understanding actual behavior of microsystem, particularly when several loading cycles are applied, and predicting its life. Analysis, experiments and numerical methods are herein applied to test case suggested by industry to investigate step by step the microswitch operation. Effects on pull-in and switch contact are investigated. Multiple electro-thermo-mechanical coupling is finally modeled to have a preliminary and comprehensive description of microswitch behavior and of its structural reliability.

In this work we present a reconfigurable impedance matching network for RF (Radio Frequency) applications, entirely manufactured in MEMS technology (RF-MEMS). The network features four impedance sections. The way they load the RF line (i.e. in series or shunt configuration) as well as the type of impedance they realize (purely capacitive, inductive, or both in parallel) are reconfigured by means of RF-MEMS cantilever-type ohmic switches. A few specimen of the network have been fabricated using the RF-MEMS technology platform available at FBK and experimentally characterized. In particular, the electromechanical characteristic of the RF-MEMS switches is observed, and the upward bending of the switches contact tips made the characterization of the RF behavior impossible. The non-planarity is due to the accumulation of residual stress within the suspended Gold layer during the release of suspended structures, and is currently being mitigated by performing a low-temperature release step. Electromagnetic simulations (S-parameters) of the RF-MEMS network are also reported in this paper, showing the wide range of impedance transformations enabled by such a complex device based on MEMS technology.

Three different multi-energy domain coupled system-level models are used to simulate the closing and opening transients of a respective ohmic contact type RF-MEMS switch. The comparison of simulated and measured data shows that, due to the presence of multiple structural modes, none of the system-level models is able to capture exactly the initial closing and contact phase whilst dynamic pull-in. The system-level model, that uses a mechanical submodel based on modal superposition, produces the result closest to the real situation. Notably, the effective residual air gap, assumed whilst contact between the membrane with high surface roughness and the contact pads of the switch, is the most influential parameter in the simulation of the closing transient, as this parameter strongly affects the air damping on the device during pull-in. This finding demonstrates that a reliable model of air damping is a vital prerequisite for the predictive simulation of pull-in and pull-out transients.

Modeling of gold microbeams for characterizing MEMS packaging solutions in terms of strains induced to the MEMS devices as well as hermetic sealing capability is presented. The proposed test structures are meant to be manufactured by the surface micromachining front-end technology available at FBK. They are based on arrays of rectangular-shaped cantilever beams as well as clamped-clamped bridges, with a width of 20 μm and a length ranging from 100 to 400 μm, to be realized by a 2 μm thick film of electroplated gold. The resonant frequency of the microbeams is modeled by FEM simulations as a function of substrate deformations, which could be induced by the package. Clamped-clamped bridges show a linear change with respect to the square of the resonant frequency up to 1800 ppm/μstrain in case of in-plane deformations. The impact of temperature excursions is also simulated, in order to use these structures for assessing thermally induced deformations. Cantilever beams are modeled as variable capacitors to detect out-of-plane deformations. Finally, both an analytical model and FEM simulations are used to study cantilever beams as resonators for detecting pressure changes, showing an impact on the quality factor in a range from 1-2 bar down to 10^-3-10^-2 mbar.

In this paper, an electrostatic actuator linearization will be introduced, which is based on an existing hardware-efficient iterative square root algorithm. The algorithm is solely based on add and shift operations while just needing n/2 iterations for an n bit wide input signal. As a practical example, the nonlinear input transformation will be utilized for the design of the primary mode controller of a capacitive MEMS gyroscope and an implementation of the algorithm in the Verilog hardware description language will be instantiated. Finally, measurement results will validate the feasibility of the presented control concept and its hardware implementation.

This paper deals with an optimization study of a vibration energy harvester. This harvester can be used as autonomous source of electrical energy for remote or wireless applications, which are placed in environment excited by ambient mechanical vibrations. The ambient energy of vibrations is usually on very low level but the harvester can be used as alternative source of energy for electronic devices with an expected low level of power consumption of several mW. The optimized design of the vibration energy harvester was based on previous development and the sensitivity of harvester design was improved for effective harvesting from mechanical vibrations in aeronautic applications. The vibration energy harvester is a mechatronic system which generates electrical energy from ambient vibrations due to precision tuning up generator parameters. The optimization study for maximization of harvested power or minimization of volume and weight are the main goals of our development. The optimization study of such complex device is complicated therefore artificial intelligence methods can be used for tuning up optimal harvester parameters.

Piezoelectric energy microgenerators are devices that continuously generate electricity when they are subjected to varying mechanical strain due to vibrations. They can generate electrical power up to 100 μW which can be used to drive various sensing and actuating MEMS devices. Today, piezoelectric energy harvesters are considered autonomous and reliable energy sources to actuate low power microdevices such as wireless sensor networks, indoor-outdoor monitoring, facility management and biomedical applications. The advantages of piezoelectric energy harvesters including high power density, moderate output power and CMOS compatible fabrication in particular with aluminum nitride (AlN) have fuelled and motivated researchers to develop MEMS based energy harvesters. Recently, the use of AlN as a piezoelectric material has increased fabrication compatibility, enabling the realization of smart integrated systems on chip which include sensors, actuators and energy storage. Piezoelectric MEMS energy microgenerator is used to capture and transform the available ambient mechanical vibrations into usable electric energy via resonant coupling in the thin film piezoelectric material. Analysis and modeling of piezoelectric energy generators are very important aspects for improved performance. Aluminum nitride as the piezoelectric material is sandwiched between two electrodes. The device design includes a silicon cantilever on which the AlN film is deposited and which features a seismic mass at the end of the cantilever. Beam theory and lumped modeling with circuit elements are applied for modeling and analysis of the device operation at various acceleration values. The model shows good agreement with the experimental findings, thus giving confidence in the model.

This work reports the analysis, modeling and characterization of an electrostatic vibration-to-electric energy scavenger. The scavenger was manufactured by the FBK in-house foundry by using a MEMS (Micro-Electro- Mechanical System) surface micro-machining technology. Extensive finite element analyses (FEA) were performed for the compact modeling of the device and presented here. The scavenger was characterized by means of dynamical measurements as a further validation of the FE model. A random vibration analysis is also utilized for the emulation of the real device operating conditions.

Vibration energy harvesting devices based on piezoelectric bimorphs have attracted widespread attention. In this work experimental set-ups are developed to assess the performances of commercially available piezoelectric energy harvesters. Bending tests allow determining the equivalent bending stiffness of the scavengers. On the other hand, dynamic tests allow obtaining frequency response functions in terms of produced voltage and power outputs vs. base acceleration around the fundamental resonance frequency. The results allow determining the influence of the voltage feedback on the dynamic response of the devices, the dependence of output voltages and powers on the applied resistive loads, the values of the loads and frequencies for which the output power is maximized, as well as the comparison of the experimental data with those obtained by using the recently developed coupled electromechanical modal model. All of this creates the preconditions for the development of optimized vibration energy harvesting devices.

The modern drive towards mobility and wireless devices is motivating intense research in energy harvesting (EH) technologies. In an effort to reduce the battery burden of people, we are investigating a novel piezoelectric wearable energy harvester. As piezoelectric EH is significantly more effective at high frequencies, in opposition to the characteristically low-frequency human activities, we propose the use of an up-conversion strategy analogous to the pizzicato musical technique. In order to guide the design of such harvester, we have modelled with Finite Elements (FE) the response and power generation of a piezoelectric bimorph while it is "plucked", i.e. deflected, then released and permitted to vibrate freely. An experimental rig has been devised and set up to reproduce the action of the bimorph in the harvester. Measurements of the voltage output and the energy dissipated across a series resistor are reported and compared with the FE predictions. As the novel harvester will feature a number of bimorphs, each plucked tens of times per step, we predict a total power output of several mW, with imperceptible effect on the wearer's gait.

In this paper a novel MEMS design of a piezoelectric generator for fluid-actuated energy conversion is presented. An exemplary energy scavenging scenarios is introduced. Moreover, the electro-mechanical properties of different piezoelectric thin films (PZT, AlN, ZnO) are introduced and compared to each other. Depending on the specific load scenario optimum piezoelectric cantilever shapes are investigated. A point load at the free-end of a cantilever requires a triangular cantilever shape. Compared to a classical rectangular shape the electrical area energy density is three times larger. A maximum area energy density value of 35 nJ/mm² is measured for the triangular cantilever shapes. A uniform load as occurring for fluid-actuated energy harvesting calls for a triangular-curved cantilever shape and is also superior to classical geometries.

An in-wheel wireless and battery-less piezo-powered tire pressure sensor is developed. Where conventional battery powered Tire Pressure Monitoring Systems (TPMS) are marred by the limited battery life, TPMS based on power harvesting modules provide virtually unlimited sensor life. Furthermore, the elimination of a permanent energy reservoir simplifies the overall sensor design through the exclusion of extra circuitry required to sense vehicle motion and conserve precious battery capacity during vehicle idling periods. In this paper, two design solutions are presented, 1) with very low cost highly flexible piezoceramic (PZT) bender elements bonded directly to the tire to generate power required to run the sensor and, 2) a novel rim mounted PZT harvesting unit that can be used to power pressure sensors incorporated into the valve stem requiring minimal change to the presently used sensors. While both the designs eliminate the use of environmentally unfriendly battery from the TPMS design, they offer advantages of being very low cost, service free and easily replaceable during tire repair and replacement.

We propose and demonstrate a design methodology for embedded systems satisfying low power requirements suitable for self-powered sensor and actuator nodes. This design methodology focuses on 1. smart energy management at runtime and 2. application-specific System-On- Chip (SoC) design at design time, contributing to low-power systems on both algorithmic and technology level. Smart energy management is performed spatially at runtime by a behaviour-based or state-action-driven selection from a set of different (implemented) algorithms classified by their demand of computation power, and temporally by varying data processing rates. It can be shown that power/energy consumption of an application-specific SoC design depends strongly on computation complexity. Signal and control processing is modelled on abstract level using signal flow diagrams. These signal flow graphs are mapped to Petri Nets to enable direct high-level synthesis of digital SoC circuits using a multi-process architecture with the Communicating-Sequential-Process model on execution level. Power analysis using simulation techniques on gatelevel provides input for the algorithmic selection during runtime of the system, leading to a closed-loop design flow. Additionally, the signal-flow approach enables power management by varying the signal flow and data processing rates depending on actual energy consumption, estimated energy deposit, and required Quality-of-Service.

Resonant microcantilevers are widely spread for different applications, such as atomic force microscopy or biosensor devices. The use of piezoelectric films on top of the cantilevers is a very promising technology due to the possibility of implementing an all electrical actuation/detection scheme. Among the different piezoelectric materials, polycrystalline aluminium nitride (AlN) is a very attractive material as, in contrast to typical alternatives such as lead zirconate titanate or zinc oxide, it is fully compatible with standard silicon technology, what facilitates its integration in silicon-based devices. In this study, our objective is to use sputter-deposited polycrystalline AlN as the actuation material for cantilever-based biosensors and to study the performance of high frequency modes for bio-detection in liquid media. Our results demonstrate a limit of mass detection of 0.07 ng for a 300x200 μm² cantilever in liquid media, which represents a value below the mass of a monolayer of IgG on a 300x200 μm² area.

The paper presents two example approaches to energy harvesting. Mechanical energy harvesting system is based on vibrational minigenerator. Basic relations of its analytical model are given in order to obtain an idea about the operating conditions. Electromagnetic harvesting system is based on tuned resonant nano-structure. Its concepts allows impedance matching in order to operate in given frequency range. The matching properties are verified by means of numerical finite element analysis. For power management of vibration energy harvesting system several circuit design concepts are presented together with simulation results and basic properties comparison.

In RF-MEMS switches many reliability issues are related to the metal contacts in the switching area. The characteristics of this contact influence not only contact resistance and insertion loss, but also the most relevant switch failure mechanisms that are wear of ohmic contact, adhesion and stiction. Gold is widely used for this purpose because of its good conductivity and chemical inertness, but is a soft metal, and the development of hard contact materials with low resistivity is of great interest for RF-MEMS switch reliability. It is possible to increase the contact hardness preserving the convenient gold properties alternating gold layers with thin layers of different metals. The material becomes harder not only by simple alloying but also by the presence of interfaces which act as barriers for mechanical dislocation migration. A detailed study of mechanical, electrical and morphological properties of gold-chromium, gold-platinum and gold-palladium multilayers is presented and discussed. It is found that the annealing treatments are important for tuning hardness values, and a careful choice of the alloying metal is essential when the material is inserted in a real switch fabrication cycle, because hardness improvements can vanish during oxygen plasma treatments usually involved in RF-switches fabrication. Platinum is the only metal tested that is unaffected by oxidation, and also modifies the chromium adhesion layer diffusion on the contact surface.

This work aims for utilizing human ocular motion for the self-sufficient power supply of a minimally invasive implantable monitoring system for intraocular pressure (IOP). With a proven piezoelectric functionality (d₃₃>5 pm/V), nanocrystalline thin films of aluminum nitride (AlN) provide a good capability for micromechanical energy harvesting (EH) in medical applications. Many d₃₁-mode microcantilever architectures are poorly suited for human-induced EH: Resonant mass-spring-damper systems are tested under high, narrow-band excitation frequencies. However, human motions, e.g. vibrations of eyeballs are marked by their low frequency, unpredictable, mainly aperiodic and time-varying signature. Different vibration types and directions are 3-dimensionally superimposed. Saccadic eye movements are favorable for inertial microgenerators because of their high dynamic loading (ω≤1000°/s). Our generator concept (symmetric active/active-parallel-bimorph cantilever) enables a high structural compliance by maximizing the piezoactive volume at very low cantilever thicknesses (<1 μm). An increased length and seismic mass enable an effective excitation by low-level aperiodic vibrations such as saccadic acceleration impulses. Analytic calculations and FEA-simulations investigate the potential distribution and transient response of different bimorph structures (length 200- 1000 μm, width 20-200 μm) on broadband vibrations. First released monomorph and bimorph structures show very low resonant frequencies and an adequate robustness.

This article presents a new approach for airborne particle generation for optical tweezers. The used element is a 500 nm thin aluminum nitride membrane with an integrated heating element. Thus the membrane works as thermo-mechanical actor. The membrane device is characterized concerning their mechanical and thermal behavior. Successful airborne particle generation is demonstrated with 10 μm silicon dioxide spheres. They are lifted up some 10th of μm from the membrane surface. The development and test of this device serves as starting point for experiments with optical tweezers in air.

The characterization of the first in-plane mode of AlN-actuated piezoelectric microcantilevers was done using different techniques. The top electrode of the cantilever was designed to allow for an efficient electrical actuation of these in-plane modes. The detection of the electrically induced in-plane movement was performed optically with the help of a stroboscopic microscope and electrically by means of an impedance analyzer. The quality factor and the resonant frequencies of the in-plane modes were estimated from the above techniques. Our results show quality factor values as high as 3000 for the first in-plane mode in air.

ZnO nanostructures, especially in the form of dense arrays of nanorods or belts have the ability to efficiently convert mechanical energy to electrical energy. One of the drawbacks though for the exploitation of nanorod technology for commercial devices is the ability to make the electrical contacts to these nanostructured piezoelectric converting elements. Although technologies have been developed that provide solutions for electrical contact issues, metal contact on uniform thin films are much simpler, and can readily be implemented to commercial mass-produced applications. At the same time it is known that high piezoelectric coefficients ZnO uniform films with columnar grains having their c-axis perpendicular to the substrate are required. In this work, we investigate the growth of uniform ZnO films, using a low temperature, low cost hydrothermal process typically used for the fabrication of ZnO nanorods. Under appropriate conditions coalescence of the nanorods occur resulting in uniform films with a columnar structure. The study focuses on understanding the role of the growth factors in order to be able to fully control the proposed process. Moreover, the hydrothermal method is further exploited for the fabrication of uniform ZnO nanostructures on patterned substrates with Au interdigitated electrodes (IDE) using standard lithography as a proof-of-concept of the applicability of the method to standard microfabrication techniques. The piezoelectric films with the IDEs are electrically characterized in order to assess the electrical properties of the grown films. From this analysis, process conditions have been identified for the growth of uniform nanostructured ZnO films, suitable for piezoelectric microgenerators.

In this work, we present a detailed study of the first vibration modes of a micro-tuning fork electrically actuated by means of a differential electrode lying on a piezoelectric AlN layer. The actuation through a differential electrode allows the optimal excitation of the modes of interest, including in-plane modes. Lateral and perpendicular displacements have been measured for a complete characterization of the resonator. Promising results have been obtained, including high values of the quality factor, which reach values above 4300 in air for the anti-phase in-plane modes.

The load-deflection (LD) method is a common and convenient procedure to extract the Young's modulus and the internal tensile stress of thin-film diaphragms from measurements of the maximum transverse deflection to a uniformly distributed load. This technique allows simultaneous determination of both parameters by fitting a theoretical to an experimental LD characteristic. Consequently, a proper knowledge of such a theoretical relationship is of utmost importance to obtain accurate values. We deduced a novel LD formula covering all relevant elastomechanical bending and stretching effects. It enables an easy but still accurate extraction of the Young's modulus and the internal tensile stress from LD measurements on circularly-shaped diaphragms. This LD relationship was derived from an adaptation of Timoshenko's membrane bending theory, where the in-plane and the out-of-plane deflections were approximated by a series expansion and a polynomial, respectively. Utilizing the minimum total potential energy principle yielded an infinite-dimensional system of equations which was solved analytically resulting in a compact closed-form solution. The flexibility of our approach is demonstrated by extracting the Young's modulus and the internal tensile stress of three disparate diaphragm materials made of either sputtered AlN, PECVD Si_xN_y, or microfiltered carbon nanotubes (bucky paper).

Copper (Cu) is commonly used as metallization for a wide range of microelectronic devices. Typically, organic circuit boards as well as ceramic and glass-ceramic substrates use galvanic deposited Cu films for this purpose. However, due to a thickness of several microns the lateral resolution in the μm-region being required e.g. for novel high frequency applications can not be guaranteed when applying this technology. Hence, sputter deposition is envisaged for the realization of Cu thin films on glass, LTCC (low temperature co-fired ceramics) and alumina substrates. The reliability of 300 nm thick Cu thin films is investigated under accelerated aging conditions, utilizing a test structure which consists of 20 parallel lines stressed with current densities up to 1•10⁺⁶ A•cm^-2 at temperatures between T= 100°C and 200°C. To detect the degradation via the temporal characteristics of the current signal a constant voltage is applied according to the overall resistance of the test structure. Knowing the mean time to failure (MTF) and the activation energy at elevated temperatures conclusions on the migration mechanism can be drawn. Whereas on LTCC substrates the activation energy of E_a~ 0.75 eV is similar to other face centered cubic metals such as silver, the higher activation energies of about E_a~ 1 eV on glass and alumina indicate a suppression of back diffusion especially at enhanced temperature levels. Therefore, the overall electromigration resistance is lower compared to Ag. This effect is predominantly caused by a stable oxide layer being formed at high temperatures acting as passivation layer.

This paper describes ongoing work to develop a low cost, passive wireless chemical sensor using a microfabricated inductor with interdigitated capacitors (IDC). A self-resonant-structure (SRS) is designed by incorporating IDC electrodes in the inter-winding space of the inductor. The distributed IDC capacitance is affected by dielectric constant and conductivity of its environment or material under test (MUT). This serves as a capacitive transducer changing the resonant frequency of the SRS. The SRS is interrogated using a non-contact inductively coupled reader coil. The shift in resonance frequency of the SRS is used to detect material properties of the environment/MUT. The dielectric constant (ε) and conductivity (σ) can be used to provide information about the surrounding environment. The ε and the σ are determined by fitting and extraction from circuit models of the IDC. Relationship between sensor layout and coupling factor between sensor and reader is investigated. Optimizations of the coupling factor based on this relationship are discussed. IDC design trade-offs between the sensor's sensitivity and coupling factor are investigated. The sensor's response to variety of liquid MUTs with a wide range of dielectric constant and conductivity is presented.

In this contribution we present an imaging platform that operates in standard incubators, allowing the investigation of multiple cell cultures at the same time. Also, the sensor platform operates with disposable multi-well plates. The system consists of four image sensors (Charge-Coupled Devices) built into a custom made holder, in which a multi-well plate can be positioned without the need of further alignment. The focal points of the image sensors are fixed at the bottom of the wells of interest. Above the multi-well plate, white light emitting LEDs with an aperture are mounted to provide orthogonal illumination over the sensors. For the observation and understanding of tumour progression, cell proliferation and cell motility studies are of great importance. For such studies, a sensor system that monitors cell motility in four wells simultaneously (of a 24-well plate) has been designed and realized. In the present work we focus on measuring the motility of adherently grown mammalian epithelial cells. Simultaneous, real-time observation of stimulated (with hepatocyte growth factor) and control samples of MDCK (Madin-Darby Canine Kidney) cells has been carried out. The presented imaging platform is an attractive and versatile alternative to conventional time-lapse microscopy to monitor the motility of individual cells. Furthermore, the high-throughput feature makes its use advantageous for the simultaneous tracking of biological samples under different conditions.

A MEMS capacitive sensor is basically an electrostatic transducer and an analytical approach to evaluate the pull-in voltage associated with a clamped circular plate or a circular membrane due to a bias voltage is presented. The approach is based on a linearized uniform approximation of the nonlinear electrostatic force due to a bias voltage and the use of a 2D load-deflection model for the clamped plate or membrane. In particular, the large deflection of the clamped thin, circular and isotropic plate is investigated when a transverse loading and an initial in-plane tension load are applied. The transition from plate behavior to membrane behavior is analyzed. The edge zone region is explored and properties of this region are given. The resulting electrostatic pressure on the diaphragm of the MEMS capacitive sensor and the pull-in voltage are studied.

The human senses are of extraordinary value but we cannot change them even if this proves to be a disadvantage in modern times. However, we can assist, enhance and expand these senses via MEMS. Current MEMS cover the range of the human sensory system, and additionally provide data about signals that are too weak for the human sensory system (in terms of signal strength) and signal types that are not covered by the human sensory system. Biomimetics deals with knowledge transfer from biology to technology. In our interdisciplinary approach existing MEMS sensor designs shall be modified and adapted (to keep costs at bay), via biomimetic knowledge transfer of outstanding sensory perception in 'best practice' organisms (e.g. thermoreception, UV sensing, electromagnetic sense). The MEMS shall then be linked to the human body (mainly ex corpore to avoid ethics conflicts), to assist, enhance and expand human sensory perception. This paper gives an overview of senses in humans and animals, respective MEMS sensors that are already on the market and gives a list of possible applications of such devices including sensors that vibrate when a blind person approaches a kerb stone edge and devices that allow divers better orientation under water (echolocation, ultrasound).

This paper reports the design of a piezoelectric energy harvesting module for a tire based wireless sensor node. System considerations comprise the generator design, material impact and the generator interface circuitry. A design procedure is presented, which allows identifying a geometry design space for the piezoelectric microgenerator consistent with given specifications. For the addressed application a large dynamic force range occurs for a given mass. The acceleration is in the range of some ten up to some thousand units of gravitational acceleration. Therefore, a conventional generator cantilever design with a mass in the gram-range is critical. For our design we use a piezoelectric MEMS generator approach without seismic mass. The intrinsic mass of the cantilever is in the microgram region and the resulting acceleration forces are very small. For the energy transfer from the environment to the generator we suggest a non-resonant excitation scheme. Tire related forces during the period of tread shuffle passage are to be used for a pulsed excitation of the generator. Based on analytical modeling the system parameters are calculated for a given generator geometrical design. The results are utilized to identify a design space consistent with given requirements and operation conditions.

In this paper the design and fabrication of an integrated micro energy harvester capable of harvesting electrical energy from low amplitude mechanical vibrations is presented. A specific feature of the presented energy harvester is its capability to harvest vibrational energy from different directions (3D). This is done through an innovative approach for electrets placed on vertical sidewalls and thereby allowing for miniaturization of 3D capacitive energy harvester on monolithic CMOS substrates. A new simple electret charging method using ionic hair-dryers/hair ionizers is reported and shown that it can be effectively used for electrets-based micro energy harvesters.

The modified micromachined thermopile consists of linearly arranged thermopairs instead of the usual loops. This device can act as a series of antennas and senses the millimeter wave radiation [1]. The antenna-like operation was demonstrated by the strong dependence on the polarization. The working principle is similar to the bolometers in that respect that the absorbed radiation heats up the semiconductor strips, but the temperature increment is sensed by the Seebeck effect instead of the resistance increment. Therefore there is no read-out current and the voltage output starts from zero. In the present work we are going to search absorption spectra of the device. It is shown that the resonance depends mainly on the dimensions and shape of the chip; the linear thermopiles are rather E-field probes. The fabrication of the device will be also outlined as well as the experimental results.

Atomic force microscopy (AFM) enables resolving features in the nanometer regime and allows in contrast to scanning tunneling microscopy (STM) measuring in liquids and a true atomic resolution also on isolating samples. But as all other scanning probe techniques the AFM requires serial data acquisition and suffers therefore from a low temporal resolution. Fast dynamical processes like most chemical reaction can not be observed in real time and the application as a tool for quality controls, e.g. in semicunductor indutries, fails often due to an unprofitable high time consumption. To enhance the scanning speed of an AFM we have developed, realized and tested different novel AFM-probes - one key component for high speed AFM- with small cantilevers and integrated sharp tips. Although such cantilevers have been presented and demonstrated on specific applications by different groups, our focus concentrates on widely dispersed applications and on aspects of mass fabrication like reproducibility and yield.

Human-machine interfaces can convey information via visual, audio and/or haptic cues during a navigation task. The visual and audio technologies are mature, whereas research has to be focused on haptic technologies for mobile devices. In this work, a tactile refreshable screen is proposed which allows its user the exploration of maps and navigational tasks in an egocentric perspective. The proposed device consists of an array of actuators which can display various patterns. The actuation technology is based on a magneto-rheological fluid which is injected in a chamber with an elastomeric membrane using a micro pump. The fluid pressure deforms the membrane in order to display a pattern. The fluid properties are used to form a valve in each cell. A permanent magnet, a ferromagnetic core, and a coil form a closed magnetic circuit with a gap where the magneto-rheological fluid can flow; the magnetic field interacts with the fluid and prevents the filling or draining of the chamber. Applying a current to the coil counteracts the magnetic field generated by the magnet and the fluid can circulate freely in order to inflate or deflate the membrane. The design, fabrication and integration of the device in addition to the results of finite element simulations and experimental measurements are reported.

In this work, the latest results of the design, fabrication and characterization of a new MEMS piezoresistive pressure sensor are presented. It is made of silicon using a boron diffusion process to create piezoresistors. Significant changes in the layout as well as in the micro-fabrication process have been made, e.g. anodic bonding of a Pyrex cover on the backside. These lead to a very precise pressure sensor, which is tailor made for high dynamic measurements in fluids with a total pressure up to 4 bar. This new piezoresistive pressure sensor has been developed in order to meet the special requirements of measurements in fluid mechanics, particularly with regard to the non-intrusive nature of the sensor. The sensor development, starting with the simulation of mechanical stresses within the diaphragm is described. These calculations have lead to an optimized placement of the piezoresistors in order to achieve a maximum sensitivity. The result of this work is a sensor which has well known properties. Important parameters including sensitivity, resonance frequency and maximum load are described precisely. These are necessary to enable new measurements in the boundary layer of fluids. The experiments and the initial results, e.g. its linearity and its dynamic capability are demonstrated in several figures.

The integration of RFID tags into packages offers the opportunity to combine logistic advantages of the technology with monitoring different parameters from inside the package at the same time. An essential demand for enhanced product safety especially in pharmacy or food industry is the monitoring of the time-temperature-integral. Thus, completely passive time-temperature-integrators (TTI) requiring no battery, microprocessor nor data logging devices are developed. TTI representing the sterilization process inside an autoclave system is a demanding challenge: a temperature of at least 120 °C have to be maintained over 45 minutes to assure that no unwanted organism remains. Due to increased temperature, the viscosity of a fluid changes and thus the speed of the fluid inside the channel increases. The filled length of the channel represents the time temperature integral affecting the system. Measurements as well as simulations allow drawing conclusions about the influence of the geometrical parameters of the system and provide the possibility of adaptation. Thus a completely passive sensor element for monitoring an integral parameter with waiving of external electrical power supply and data processing technology is demonstrated. Furthermore, it is shown how to adjust the specific TTI parameters of the sensor to different applications and needs by modifying the geometrical parameters of the system.

Low temperature co-fired ceramics (LTCC) has established as a widespread platform for advanced functional ceramic devices in different applications, such as in the space and aviation sector, for micro machined sensors as well as in micro fluidics. This is due to high reliability, excellent physical properties, especially in the high frequency range, and the possibility to integrate passive components in the monolithic LTCC body, offering the potential for a high degree of miniaturisation. However, for further improvement of this technology and for an ongoing increase of the integration level, the realization of miniaturized structures is of utmost importance. Therefore, novel techniques for micro-machining are required providing channel structures and cavities inside the glass-ceramic body, enabling for further application scenarios. Those techniques are punching, laser cutting and embossing. One of the most limitations of LTCC is the poor thermal conductivity. Hence, the possibility to integrate channels enables innovative active cooling approaches using fluidic media for heat critical devices. Doing so, a by far better cooling effect can be achieved than by passive devices as heat spreaders or heat sinks. Furthermore, the realization of mechanic devices as integrated pressure sensors for operation under harsh environmental conditions can be realized by integrating the membrane directly into the ceramic body. Finally, for high power devices substantial improvement can be provided by filling those channel structures with electrical conductive material, so that the resistivity can be decreased drastically without affecting the topography of the ceramics.

The direct on-disk wireless temperature measurement system [1,2] presented at μTAS 2010 was further improved in its robustness. We apply it to an IR thermocycler as part of a centrifugal microfluidic analyzer for polymerase chain reactions (PCR). This IR thermocycler allows the very efficient direct heating of aqueous liquids in microfluidic cavities by an IR radiation source. The efficiency factor of this IR heating system depends on several parameters. First there is the efficiency of the IR radiator considering the transformation of electrical energy into radiation energy. This radiation energy needs to be focused by a reflector to the center of the cavity. Both, the reflectors shape and the quality of the reflecting layer affect the efficiency. On the way to the center of the cavity the radiation energy will be diminished by absorption in the surrounding air/humidity and especially in the cavity lid of the microfluidic disk. The transmission spectrum of the lid material and its thickness is of significant impact. We chose a COC polymer film with a thickness of 150 μm. At a peak frequency of the IR radiator of ~2 μm approximately 85 % of the incoming radiation energy passes the lid and is absorbed within the first 1.5 mm depth of liquid in the cavity. As we perform the thermocycling for a PCR, after heating to the denaturation temperature of ~ 92 °C we need to cool down rapidly to the primer annealing temperature of ~ 55 °C. Cooling is realized by 3 ventilators venting air of room temperature into the disk chamber. Due to the air flow itself and an additional rotation of the centrifugal microfluidic disk the PCR reagents in the cavities are cooled by forced air convection. Simulation studies based upon analogous electrical models enable to optimize the disk geometry and the optical path. Both the IR heater and the ventilators are controlled by the digital PID controller HAPRO 0135 [3]. The sampling frequency is set to 2 Hz. It could be further increased up to a maximum value being permitted by the wireless temperature data transmission system. As we are controlling a significantly non-linear process the controller parameters need to be optimized for all temperatures relevant for the PCR thermocycling process. Such we get a dynamic system for both, the heating and the cooling process. Heating rates up to 5 K/s with our IR heater (100 W electrical power) could be achieved. Cooling rates of instantly 1.3 K/s at 20 Hz rotation frequency could be even further increased by higher rotation frequencies, faster air circulation, optimization of the controller parameters or an active air cooling unit.

We present the development and characterization of a fiber-optic colorimetric gas sensor combined with the electronic circuitry for measurement control and RFID communication. The gas sensor detects ammonia using a 300 μm polyolefin fiber coated with a gas-sensitive polymer film. The spectral and time-dependent sensitivity of various polymer films was tested in transmission measurements. Light from a standard LED at λ = 590 nm was coupled into the polyolefin fiber through the front face. A prototype of the gas sensor with the direct coupling method was tested under realistic measurement conditions, i.e. battery-driven and in a completely autonomous mode. The sensor system showed good sensitivity to the ammonia concentrations and response times in the order of minutes. The achievable power consumption was below 100μW.The films contained the pH-sensitive dyes bromocresol purple or bromophenol blue embedded in either ethyl cellulose or polyvinyl butyral, and optionally tributyl phosphate as plasticizer. The bromophenol blue based films showed a strong reaction to ammonia, with saturation concentrations around 1000 ppm and response times of about 15 seconds to 100ppm. The colorimetric reaction was simulated using a simple kinetic model which was in good agreement with the experimental results.

Miniaturized sensors for fluid viscosity often utilize shear vibrations and thus measure a thin film of fluid. To probe the bulk of a sample, pressure waves can be utilized instead. Then, however, the so-called longitudinal viscosity is determined, which can be equally useful for condition monitoring applications. Moreover, this parameter has not yet been investigated in detail such that material data are scarce. In this paper, we report on a prototype setup utilizing standing acoustic pressure waves in a small sample chamber. The impact of these resonances on the impedance of a PZT transducer is modeled and investigated experimentally. It is demonstrated that with this setup sound velocity and the longitudinal viscosity of liquid samples can be investigated.

We present a novel electronic circuit for a capacitive non-contact volume sensor to be used for the determination of dispensed liquid droplets in the nanoliter range. Beside the ability of online droplet volume measurement the presented sensor features also sensitivity to certain droplet parameters like droplet velocity, liquid type and even small droplet deformations during the droplet's flight. The sensor principle is based on the capacitive change caused by a droplet while it passes an open plate capacitor. In contrast to the already published results [1, 2], the presented circuit amplifies the change of capacitance by a measurement bridge that feeds a differential amplifier. The required bridge regulation is realised by an adjustable analogue voltage divider combined with an all-pass filter to modulate the reference signal in phase and amplitude. The implementation of the analogue adjustment is crucial to enable the required high sensitivity. The circuit is able to amplify changes in capacitance in the range from ▵C ≈ {0.6 to 2.7 fF} which are caused by single droplets of volumes of V ≈ {20 to 70 nl}). The output voltage signals can reach up to U_max= 2.3V. The sensor sensitivity to the droplet volume reaches S_i = 82 mV/nl with an accuracy of ▵V = ± 4 nl. With this sensitivity even droplet deformation occurring within the measurement capacitor can be observed as multiple signal peaks. These are caused from lateral (towards the electrodes) and longitudinal extension of the droplet's shape that influence the effective capacitance by ▵C_max≈ 0.2 fF.

This paper reports on the potential of microelectro discharge machining (μEDM) as an innovative method for the fabrication of 3D microdevices. To demonstrate the wide capabilities of μEDM two different high-potential 3D microsystems - a microfluidic device for the dispersion of nanoparticles and a star probe for microcoordinate metrology - are presented. For the fabrication of these microdevices a μEDM-milling machine with integrated microwire electro discharge grinding (μWEDG) module is utilized. To gain optimized process conditions as well as a high surface quality an adequate adaption of the single erosion parameters such as energy, pulse frequency and spark gap has to be carried out and are discussed below. The dispersion micromodule is used for pharmaceutical screening applications in a high pressure range up to 2000 bar. At the channel bottom a surface roughness of R_a = 80 nm is achieved. In case of the star probe it is possible to produce shaft and sphere out of one piece. The fabricated stylus elements have sphere diameters of 40-200 μm. For both applications μEDM offers a flexible, precise, effective and cost-efficient fabrication method for the machining of hard and resistant materials.

Wafer bonding became during past decade an important technology for MEMS manufacturing and wafer-level 3D integration applications. The increased complexity of the MEMS devices brings new challenges to the processing techniques. In MEMS manufacturing wafer bonding can be used for integration of the electronic components (e.g. CMOS circuitries) with the mechanical (e.g. resonators) or optical components (e.g. waveguides, mirrors) in a single, wafer-level process step. However, wafer bonding with CMOS wafers brings additional challenges due to very strict requirements in terms of process temperature and contamination. These challenges were identified and wafer bonding process solutions will be presented illustrated with examples.

A silicon cantilever with slender geometry based Micro Electro Mechanical System (MEMS) for nanoparticles mass detection is presented in this work. The cantilever is actuated using a piezoactuator at the bottom end of the cantilever supporting frame. The oscillation of the microcantilever is detected by a self-sensing method utilizing an integrated full Wheatstone bridge as a piezoresistive strain gauge for signal read out. Fabricated piezoresistive cantilevers of 1.5 mm long, 30 μm wide and 25 μm thick have been employed. This self-sensing cantilever is used due to its simplicity, portability, high-sensitivity and low-cost batch microfabrication. In order to investigate air pollution sampling, a nanoparticles collection test of the piezoresistive cantilever sensor is performed in a sealed glass chamber with a stable carbon aerosol inside. The function principle of cantilever sensor is based on detecting the resonance frequency shift that is directly induced by an additional carbon nanoparticles mass deposited on it. The deposition of particles is enhanced by an electrostatic field. The frequency measurement is performed off-line under normal atmospheric conditions, before and after carbon nanoparticles sampling. The calculated equivalent mass-induced resonance frequency shift of the experiment is measured to be 11.78 ± 0.01 ng and a mass sensitivity of 8.33 Hz/ng is obtained. The proposed sensor exhibits an effective mass of 2.63 μg, a resonance frequency of 43.92 kHz, and a quality factor of 1230.68 ± 78.67. These results and analysis indicate that the proposed self-sensing piezoresistive silicon cantilever can offer the necessary potential for a mobile nanoparticles monitor.

An often reported problem during production and operation of silicon MEMS is stiction. It describes the sticking of movable MEMS parts to surrounding structures. The probability of the occurrence of stiction is linked to the surface energy of the MEMS. Self assembling monolayers can be used to reduce the surface energy and therefore the probability of stiction. These monolayers have to resist high temperatures up to 400°C to be compatible with various micro-production processes, e.g., eutectic bonding. Several groups tried to coat such monolayers with different success and results. One problem is the instability of the coating method due to water contaminations of the coating solution. To circumvent this error source, an experimental setup was designed and built up to minimize the water content of the monolayer solvent and ensures reproducible conditions during the coating process. The required set of liquids is piped through a system of valves and tubes to rinse a trench with a silicon die. To avoid contamination of the liquids with water, the setup is partly placed in a box flushed with nitrogen. With this experimental setup, the surface energy γ_s of the MEMS structures had been reduced from 18.1 mJ/m² to 33.1 μJ/m² and 36.6 μJ/m² for FDTS and DDMS, respectively.

Digital microfluidics combines the advantages of low consumption of reagents with a high flexibility of processing fluid samples automatically. For applications in life sciences not only the processing but also the characterization of fluid analytes is crucial. In this contribution a microfluidic platform combining the actuation principle of electro wetting on dielectrics for droplet manipulations and the sensor principle of impedance spectroscopy for the characterization of fluid composition and condition is presented. The fabrication process of the microfluidic platform comprises physical vapor deposition and structuring of the metal electrodes onto a substrate, the deposition of a dielectric isolator and a hydrophobic top coating. The key advantage of this microfluidic chip is the common electric nature of the sensor and the actuation principle, so no additional sensor integration is necessary. Multiple measurements on fluids of different composition (including rigid particles and biologic cells) and of different conditions (temperature, sedimentation) were performed as well as online monitoring of in process parameters changing over time. These sample applications demonstrate the versatile applications of this combined technology.

A novel MEMS based scanning head comprising an array of comb drive actuators is developed for parallel SPM (Scanning Probe Microscope) imaging to enhance the performance of the currently available nano-measuring machines and effectively reduce the measurement time for large specimen. The scanning-head consists of an row of seven scan heads, which are realized on one chip by MEMS technology. The actuation of the scan-head is realized by an electrostatic comb-drive, in which the displacement can be measured by detecting the capacitance change of the actuator. The main shaft of the scan-head protrudes out of the MEMS chip to sense the surface topography of a specimen under test. To further improve the lateral resolution of the micro-SPM head, an AFM tip can be mounted onto the end of the actuator's main shaft. Design, simulation and fabrication of the scan-head array will be presented. The displacement of the actuator was calibrated using a laser interferometer. The actuator has a good linearity. Mechanical performance such as stiffness and eigenfrequency were also investigated. Preliminary experiments proved the feasibility of the scanninghead array for large-scale topography measurements.

Micromachined thermal infrared emitters using heavily boron doped silicon as active material have been developed. The proposed fabrication process allows the integration of infrared emitters with arrays of thermopile infrared detectors to achieve integrated non dispersive infrared (NDIR) microspectrometers. A set of emitters with a common radiating silicon slab size (1100x300x8μm³) has been successfully fabricated and characterized. The working temperature of Joule heated radiating elements has been controlled by means of DC or pulsed electric signals, up to temperatures exceeding 800°C. Measured thermal time constants, in the order of 50 ms, enable direct electrical modulation of emitted radiation up to a frequency of 5Hz with full modulation depth. The temperature distribution in the radiating element has been analyzed with infrared thermal imaging.

Promising results on flexible and large area pressure sensors for human-neuroprostheses and humanneurobotic interface assessment are presented. Array sensors of 4x4 and 8x8 based on polymer and plastic electronics have been fabricated and characterized with a resolution of 1 and 0.25 cm². The working pressure range is between 0 and 1 Kg/cm² and the offset measured is around 0.03 Kg/cm². The response time is around 1 ms.

In this contribution a feasibility study on resonating sensors for rheologic properties such as e.g., viscosity facilitating measurements at tunable frequencies is presented. For the concepts presented in this work, sample liquids are subjected to time harmonic shear stresses induced by a resonating wire and a suspended resonating platelet, respectively. From the resulting frequency response the liquid's rheological properties can be deduced by fitting the parameters of an appropriate closed-form model representing the physical behavior of the sensors. To allow large penetration depths of the shear waves being imposed by the resonating mechanism into the test liquid, it is desired to have oscillators with resonance frequencies in the low kilohertz range. Large penetration depths become important when examining complex liquids such as multi-phase systems as, e.g., emulsions. For the investigation of liquids showing shear thinning (or thickening) or viscoelastic behavior, it is necessary to record the liquid's characteristics not only at one single frequency but in a range of different frequencies, which in the best case should cover several decades of resonance frequencies. For this purpose, especially in the case of resonating microsensors, it is desired to have devices, which can be operated at tunable frequencies without changing their geometries. For the two concepts presented in this work, the ability of tuning the sensor's resonance frequency is based on varying the normal stresses within tungsten wires. The use of appropriate materials and different micro-fabrication techniques are discussed and the applicability of the devices for rheological measurements are outlined. The models are compared to measurement results and the capability of the particular resonator for accurate and reliable sensing is discussed.

Energy-autarkic sensor systems use energy from their environment for operation and data transfer. This needs both an efficient energy harvesting strategy and an appropriate energy storage management but also ultra-low-power technologies for sensors, signal processing, and wireless signal transfer. Binary Zero-Power Sensors (BIZEPS) are threshold switches which take the switching energy directly from the quantity to be measured. Hence, they deliver the switching states "ON" and "OFF" without any additional electrical energy.

A new concept for a tube-like dielectric elastomer actuator (DEA) utilizes rigid micro-electrodes to stabilize the tube structure in azimuthal direction. A number of those electrodes are electrically and mechanically interconnected and fully embedded into a silicone layer. Those electromechanical networks represent individual actuator groups. An axial arrangement of a number of those electrode groups forms a single actuator filament. Applying an electrical voltage induces mechanical tension into those electrode groups by the effect of Maxwell-stress. The interaction of individual electrodes causes a change of the total length of selected actuator filaments. A circular arrangement of a number of actuator filaments allows bending of the tube in any direction. The electrode thickness appears as passive non compliant parameter within the axial length of an actuator filament. The desired tube actuator is focused on thin walled structures with an outer diameter less than 6 mm and an available width of the wall of less than 0.4 mm. The challenge for the fabrication of this electrode structure is mainly based on the elastic properties of the 2K-silicone substrate, unfavourable adhesion characteristics of the silicone, a desired minimal electrode thickness of less than 5 microns and a required homogeneon electrode body. The present work introduces the tube-actuator concept and details fabrication concepts for the embedded electrode structure considering actuator-related electrical and mechanical requirements.

With respect to varying operation conditions, only sensors directly installed in the engine can detect the current oil condition hence enabling to get the right time for the oil change. Usually, only one parameter is not sufficient to obtain reliable information about the current oil condition. For this reason, appropriate sensor principles were evaluated for the design of sensor arrays for the measurement of critical lubricant parameters. In this contribution, we report on the development of a sensor array for engine oils using laboratory analyses of used engine oils for the correlation with sensor signals. The sensor array comprises the measurement of conductivity, permittivity, viscosity and temperature as well as oil corrosiveness as a consequence of acidification of the lubricant. As a key method, rapid evaluation of the sensors was done by short term simulation of entire oil change intervals based on artificial oil alteration. Thereby, the compatibility of the sensor array to the lubricant and the oil deterioration during the artificial alteration process was observed by the sensors and confirmed by additional laboratory analyses of oil samples take.

We present a resonant MEMS sensor for viscosity and mass density measurements of liquids. The device is based on Lorentz-force excitation and has an integrated piezoresistive readout. The sensing element is a rectangular vibrating plate suspended by four beam springs. Due to in-plane vibrations, the plate contribution to the overall damping is low. Additionally, the plate increases the moving mass of the sensor and therefore the quality factor of the resonant system. Two of the plate-carrying springs comprise piezoresistors. With two additional resistors on the silicon rim, they form a half Wheatstone-bridge. Through the conductive layer of the beam springs a sinusoidal excitation current is driven. In the field of a permanent magnet, the Lorentz force excites plate vibrations resulting in a bridge unbalance. The sensor was experimentally tested with glycerol-water mixtures at constant temperature. We recorded both the frequency and the phase of the bridge output. By evaluating the properties of the resonant system, it is possible to extract the viscosity and the mass density of the mixtures simultaneously.

The aim of this work is to design and fabricate a flow sensor using an artificial hair cell (AHC) inspired by biological hair cells of fish. The sensor consists of a single cilium structure with high aspect ratio and a mechanoreceptor using force sensitive resistor (FSR). The cilium structure is designed for capturing a drag force with direction due to flow field around the sensor and the mechanoreceptor is designed for sensing the drag force with direction from the cilium structure and converting it into an electric signal. The mechanoreceptor has a symmetric four electrodes to sense the drag force and its direction. To fabricate the single cilium structure with high aspect ratio, we have proposed a new design concept using a separated micro mold system (SMS) fabricated by the LIGA process. For a successful replication of the cilium structure, we used the hot embossing process with the help of a double-sided mold system. We used a composite of multiwall carbon nanotube and polydimethylsiloxane (MWCNT-PDMS). The performance of the mechanoreceptors was measured by a computer-controlled nanoindenter. We carried out several experiments with the sensor in the different flow rate and direction using the experimental test apparatus. To calibrate the sensor and calculate the velocity with direction based the signal from the sensor, we analyzed the coupled phenomena between flow field and the cilium structure to calculate the deflection of the cilium structure and the drag force applying to the cilium structure due to the flow field around sensor.

In recent years, the piezoresistive properties of different semiconductor thin films, with chemical and mechanical stability, have been studied in order to use them as base material in the fabrication of strain sensors for high temperature applications. In this context, this work compares the performance of strain sensors based on sputter-deposited semiconductor thin films such as titanium dioxide (TiO₂), silicon carbide (SiC) and diamond-like carbon (DLC) operated at temperatures up to 250°C. The structure of each sensor consists of four thin-film resistors, configured in Wheatstone bridge, with Ti/Au electrical contacts. These strain sensors reported here differ from our previous works in the types of materials used and in the quantity/configuration of the thin-film resistors. The strain sensors were fabricated by photolithography techiques in conjunction with lift-off processes. The beam-bending experiments were performed to characterize the sensors.

Present work proposed design, finite element tools simulation and prototype fabrication of a low cost energy autonomous, maintenance free, flexible and wearable micro thermoelectric generator (μTEG), finalized to power very low consumption electronics Ambient Assisted Living (AAL) applications. The prototype, integrating an array of 100 thin films thermocouples of Sb₂Te₃ and Bi₂Te₃, generates, at 40 °C, an open circuit output voltage of 430 mV and an electrical output power up to 32 nW with matched load. In real operation conditions of prototype, which are believed to be very close to a thermal gradient of 15°C, the device generates an open circuit output voltage of about 160 mV, with an electrical output power up to 4.18 nW. In this work we proposed design, thermal simulation and fabrication of a preliminary flexible and wearable micro thermoelectric generator (μTEG), finalized to power very low consumption electronics for Ambient Assisted Living (AAL) applications. Presented simulations show the performances of different fabrication solution for the PDMS/Kapton packages, considering flat and sloped walls approach for thermal gradient enhancement.

Thermoelectric modules convert thermal energy into electrical energy and vice versa. At present bismuth telluride is the most widely commercial used material for thermoelectric energy conversion. There are many applications where bismuth telluride modules are installed, mainly for refrigeration. However, bismuth telluride as material for energy generation in large scale has some disadvantages. Its availability is limited, it is hot stable at higher temperatures (>250°C) and manufacturing cost is relatively high. An alternative material for energy conversion in the future could be silicon. The technological processing of silicon is well advanced due to the rapid development of microelectronics in recent years. Silicon is largely available and environmentally friendly. The operating temperature of silicon thermoelectric generators can be much higher than of bismuth telluride. Today silicon is rarely used as a thermoelectric material because of its high thermal conductivity. In order to use silicon as an efficient thermoelectric material, it is necessary to reduce its thermal conductivity, while maintaining high electrical conductivity and high Seebeck coefficient. This can be done by nanostructuring into arrays of pillars. Fabrication of silicon pillars using ICP-cryogenic dry etching (Inductive Coupled Plasma) will be described. Their uniform height of the pillars allows simultaneous connecting of all pillars of an array. The pillars have diameters down to 180 nm and their height was selected between 1 micron and 10 microns. Measurement of electrical resistance of single silicon pillars will be presented which is done in a scanning electron microscope (SEM) equipped with nanomanipulators. Furthermore, measurement of thermal conductivity of single pillars with different diameters using the 3ω method will be shown.

This paper reports a method on the manufacturing of through wafer via holes in silicon with tapered walls by Deep Reactive Ion Etching (DRIE) using the opportunity to change the isotropy in the DRIE equipments during processing. By using consecutively anisotropic and isotropic etching steps it is possible to enlarge the dimension of via holes on one side of the wafer, while on the other side dimension is set by the initial etching window. The method was used for two etching windows sizes (100μm and 20μm respectively) on 200μm and 300μm thick wafers. The aim was to manufacture tapered walls via having a good control over the walls angle. Different Bosch process recipes providing different walls roughness were used. Via holes with tapered walls (2° to 22°) were manufactured using this method. An angle deviation smaller than 10% of the manufactured via holes along the wafers was observed.

This paper describes the development of a new technological process for manufacturing MEMS with out-of-plane motion in order to develop voltage references with high stability (deviation from the nominal value less than some 10^-6per year) and working in alternating current (AC). In this application, the AC voltage reference is defined at the pull-in point of the MEMS variable capacitor made of two electrodes, one fixed and the other movable driven by an AC current having a frequency much higher than the mechanical resonant frequency. This process has been developed to get MEMS devices exhibiting high nominal capacitances, which makes the development of the read-out electronics easier, and metallized electrodes to avoid charge effects on the stability of the voltage standard. Moreover, this process allows to realize stoppers, in order to get a good operational reliability.

Tactile sensors are basically arrays of force sensors. Most of these force sensors are made of polymers or conductive rubbers to lower the cost, especially in the case of large area low-medium resolution tactile sensors. The price to pay for such decrease in the cost and complexity is a worse performance. Hysteresis and drift are the two main sources of error. This paper presents a method to reduce the error caused by hysteresis. This method is based on the generalized Prandtl- Ishlinskii model that has been applied to characterize hysteresis and saturation nonlinearities in smart actuators. The classical Prandtl-Ishlinskii model is not suitable because the lack of symmetry the output curves from the sensor show. Other alternatives like the Preisach model are too complex to implement, especially taking into account that a tactile sensor provides many data to process. The approximation error depends on several parameters as well as on the envelope functions that are chosen. Different alternatives are explored in the paper. Moreover, the model can also be inverted. This inverted model allows obtaining the force values from the tactile sensor output while reducing the errors caused by hysteresis. Since implementations of tactile sensors usually have the electronics close to the raw sensor, and this hardware is also commonly based on a microcontroller or even on a FPGA, it is possible to add the algorithms presented in this paper to the set of compensation and calibration procedures to run in the smart sensor.