Robert J. Huber

A silicon-based, three-dimensional neural interface: manufacturing processes for an intracortical electrode array

A method is described for the manufacture of a three-dimensional electrode array geometry for chronic intracortical stimulation. This silicon based array consists of a 4.2*4.2*0.12 mm thick monocrystalline substrate, from which project 100 conductive, silicon needles sharpened to facilitate cortical penetration. Each needle is electrically isolated from the other needles, and is about 0.09 mm thick at its base and 1.5 mm long. The sharpened end of each needle is coated with platinum to facilitate charge transfer into neural tissue. The following manufacturing processes were used to create this array: thermomigration of 100 aluminum pads through an n-type silicon block, creating trails of highly conductive p/sup +/ silicon isolated from each other by opposing pn junctions; a combination of mechanical and chemical micromachining which creates individual penetrating needles of the p/sup +/ silicon trails; metal deposition to create active electrode areas and electrical contact pads; and array encapsulation with polyimide.<<ETX>>

Chemically sensitive field-effect transistors.

Sometimes unjustified, but nonetheless ever present need for more information continues to stimulate the development of new sensors and detectors. One of the more recent additions to the armory of these devices is the chemically sensitive field effect transistor (chemfet). It was born in the early seventies out of two very successful technologies: solid state integrated circuits and ion-selective electrodes (ise). It is still in its infancy but already out of the teething stage and with a very bright prospect ahead.

Resonator/Oscillator response to liquid loading.

The resonant frequency of a thickness-shear mode resonator operated in contact with a fluid was measured with a network analyzer and with an oscillator circuit. The network analyzer measures changes in the devices intrinsic resonant frequency, which varies linearly with (ρη)(1/2), where ρ and η are liquid density and viscosity, respectively. The resonator/oscillator combination, however, responds differently to liquid loading than the resonator alone. By applying the operating constraints of the oscillator to an equivalent-circuit model for the liquid-loaded resonator, the response of the resonator/oscillator pair can be determined. By properly tuning the resonator/oscillator pair, the dynamic range of the response can be extended and made more linear, closely tracking the response of the resonator alone. This allows the system to measure higher viscosity and higher density liquids with greater accuracy.

Electrostatically protected ion sensitive field effect transistors

Abstract Two type of ISFETs with electrostatic protection have been designed and tested. They both utilize an electrically conductive layer incorporated into the gate of the ISFET. This layer is connected via an on-chip MOSFET switch to the outside circuitry. In the first type the conductor is capacitively coupled to the ion-selective membrane and to the solution. In this case the device output is proportional to the time differential of the concentration change. The applicability of this device to high-speed FIA titrations has been tested. In the second device the gate electrically contacts the membrane. In this case the output is identical to that of a conventional ISFET. The signal-to-noise ratio and the electrostatic protection of this ISFET are considerably improved.

A study of insulator materials used in ISFET gates

Abstract Experiments with aqueous electrolyte-insulator-semiconductor structures showed that Si 3 N 4 is a satisfactory insulator on silicon whereas thermally grown SiO 2 is not. The results can be explained in terms of microcrack formation in SiO 2 . The breakdown voltage was found to be relatively independent of the SiO 2 thickness and crack sizes were estimated to be of the order of a few tens of angstroms. No electrically significant bulk hydration effects were found to occur in either insulator in mildly acidic solutions.

Characterization of the embedded micromechanical device approach to the monolithic integration of MEMS with CMOS

Recently, a great deal of interest has developed in manufacturing processes that allow the monolithic integration of microelectromechanical systems (MEMS) with driving, controlling, and signal processing electronics. This integration promises to improve the performance of micromechanical devices as well as lower the cost of manufacturing, packaging, and instrumenting these devices by combining the micromechanical devices with a electronic devices in the same manufacturing and packaging process. In order to maintain modularity and overcome some of the manufacturing challenges of the CMOS-first approach to integration, we have developed a MEMS-first process. This process places the micromechanical devices in a shallow trench, planarizes the wafer, and seals the micromechanical devices in the trench. Then, a high-temperature anneal is performed after the devices are embedded in the trench prior to microelectronics processing. This anneal stress-relieves the micromechanical polysilicon and ensures that the subsequent thermal processing associated with fabrication of the microelectronic processing does not aversely affect the mechanical properties of the polysilicon structures. These wafers with the completed, planarized micromechanical devices are then used as starting material for conventional CMOS processes. The circuit yield for the process has exceeded 98 percent. A description of the integration technology, the refinements to the technology, and wafer- scale parametric measurements of device characteristics is presented. Additionally, the performance of integrated sensing devices built using this technology is presented.

Parallel-plate electrostatic dual-mass resonator

A surface-micromachined two-degree-of-freedom system that was driven by parallel-plate actuation at antiresonance was demonstrated. The system consisted of an absorbing mass connected by folded springs to a drive mass. The system demonstrated substantial motion amplification at antiresonance. The absorber mass amplitudes were 0.8 - 0.85 micrometer at atmospheric pressure while the drive mass amplitudes were below 0.1 micrometer. Larger absorber mass amplitudes were not possible because of spring softening in the drive mass springs. Simple theory of the dual-mass oscillator has indicated that the absorber mass may be insensitive to limited variations in strain and damping. This needs experimental verification. Resonant and antiresonant frequencies were measured and compared to the designed values. Resonant frequency measurements were difficult to compare to the design calculations because of time-varying spring softening terms that were caused by the drive configuration. Antiresonant frequency measurements were close to the design value of 5.1 kHz. The antiresonant frequency was not dependent on spring softening. The measured absorber mass displacement at antiresonance was compared to computer simulated results. The measured value was significantly greater, possibly due to neglecting fringe fields in the force expression used in the simulation.

Electro-thermal modeling of a microbridge gas sensor

Fully CMOS-compatible, surface-micromachined polysilicon microbridges have ben designed, fabricated, and tested for use in catalytic, calorimetric gas sensing. To improve sensor behavior, extensive electro-thermal modeling efforts were undertaken using SPICE. The validity of the SPICE model was verified by comparing its simulated behavior with experimental results. The temperature distribution of an electrically-heated microbridge was measured using an IR microscope. Comparisons among the measured distribution, the SPICE simulation, and distributions obtained by analytical methods show that heating at the ends of a microbridge has important implications for device response. Additional comparisons between measured and simulated current-voltage characteristics, as well as transient response characteristics, further support the accuracy of the model. A major benefit of electro-thermal modeling with SPICE is the ability to simultaneously simulate the behavior of a device and its control/sensing electronics. Results for the combination of a unique constant-resistance control circuit and microbridge gas sensor ar given. Models of in situ techniques for monitoring catalyst deposition are shown to be in agreement with experiment. Finally, simulated chemical response of the detector is compared with the data, and methods of improving response through modifications in bridge geometry are predicted.

Out with the old in with the new

Abstract The materials issues facing the microelectromechanical systems (MEMS) community can be understood best in terms of the historical context. The field began almost as an afterthought among those engaged in integrated circuit production. These researchers recognized early on that the same processes used in the production of circuits could be re-ordered to make very small mechanical devices. The huge investment made by the electronics community in silicon technologies — and the relative ease with which these techniques could be adapted to device production — made them an obvious resource for early MEMS designers. It is no accident that polycrystalline silicon is the most commonly used structural material. But with the continued expansion of MEMS devices into new areas of application, the limitations of silicon (Si) usefulness became clear. The need to combine electronics and MEMS on the same chip ( iMEMS ), improve the wear characteristics of moving parts, and achieve a greater mass of moving parts in MEMS inertial sensors have led researchers away from a ‘one-material-fits-all’ approach. Instead the search is on for materials that more directly serve specific ends. The need for biocompatibility in the emerging field of bio-MEMS has added urgency to the quest for new materials, since Si-based materials cannot meet every bio-MEMS need.

Fabrication and performance of ion implanted I 2 L devices

Integrated injection logic devices based on vertical NPN transistors and lateral PNP injectors, and fabricated using ion-implant doping exclusively are described. Two basic processes are outlined. One uses an implanted low-energy boron pre-deposited layer, thermally driven in, to form the NPN base and p+injectors. The other process uses high-energy implanted boron to form the active base region of the NPN device below the silicon surface. Both techniques use ion-implanted phosphorus for the n+collectors and collars. Results of speed-power measurements for a 7-stage ring oscillator and a 50-stage inverter string are described. The high-energy implanted boron process can be used together with Schottky-barrier collectors to further reduce the power-delay product.