ERC for Wireless Integrated MicroSystems (WIMS) an NSF-University-Industry Collaboration

Cochlear implants for the deaf are the most successful of all neural prostheses; however, pitch perception remains relatively poor due to wire limitations. Commercial wire-bundle cochlear arrays are limited to about twenty wires (sites) by the size of the scala tympani into which the arrays must be inserted.

Low-power, small-size analog-to-digital converters (ADC) have numerous applications in areas ranging from power-aware wireless sensing nodes for environmental monitoring to biomedical monitoring devices in point-of-care (PoC) instruments. Despite continued improvements to ACDs in recent years, we still strive to make them smaller and more power efficient.

In response to the need for ever-better methods of screening passengers and luggage for concealed explosives at transportation terminals, we are developing microsystems that can rapidly differentiate markers of nitro-aromatic explosives (e.g., TNT) from background interferences at trace concentrations in ambient air. The INTREPID prototype (upper left panel) is a gas chromatographic microsystem (?GC) designed to rapidly capture, preconcentrate, inject, separate, and selectively detect targeted marker compounds.

Carbon nanotubes (CNTs) are seamless hollow cylinders that have unique and exceptional mechanical, thermal, electrical, and chemical properties. Our ability to fabricate highly organized CNT assemblies, having critical dimensions ranging from micron to milli-meter scales, offers opportunity to harness the properties of CNTs in MEMS devices.

The SAR analog-to-digital conversion architecture is popular because of its energy efficiency; however, the SAR architecture is less used for resolutions above 10b. A two-comparator architecture, incorporating deliberate comparator offset and pre-amplifier power management, reduces comparator metastability and comparator power consumption in a 12b 11MS/s SAR ADC.

Advances in neuroscience depend on the availability of supporting technology for the high-density stimulation and recording of neural activity. An important area of research focuses on the cochlear nucleus (CN) of the auditory pathway, where details of the connections between the anterior ventral cochlear nucleus (AVCN) and the dorsal cochlear nucleus (DCN) are yet to be fully understood.

High-temperature pressure sensors have uses in numerous industrial sectors including gas turbine engines, coal boilers, internal combustion engines, and oil/gas exploration machinery. These environments require durable sensors, making sensors without moving parts and without intermediate transduction steps advantageous. To meet these challenges, microdischarge-based pressure sensors have been developed.

A new process has been developed for fabricating silicon-glass microsystems such as the implantable intraocular pressure sensor now being pursued by the WIMS ERC. The process allows the realization of glass die of arbitrary two-dimensional shape containing silicon (or metal) feedthroughs as small as 20?m in diameter. Shaping glass and producing feedthroughs in it are both long-standing problems in MEMS.

This study was designed to investigate whether a non-penetrating microwire device can serve as a brain machine interface (BMI) for a motor neural prosthesis. These devices offer higher spatial and temporal resolution than standard electroencephalography or electrocorticography electrode arrays without penetrating the neocortex.

Brain machine interfaces have been recognized as a powerful tool in helping patients with neural disorders. Wireless transmission of potentially hundreds of signals to extracranial processing units must address three major limitations: bandwidth, implant area, and power consumption. For example, without compression, a 32-channel system with a sampling rate of 25KHz per channel and 10-bits of data precision generates data at 8Mbps.

Micromachined Joule-Thomson (J-T) coolers have applications ranging from cryosurgery to cooling infrared detectors in space and portable applications. The counter-flow recuperative heat exchangers in the J-T coolers must maintain good stream-to-stream heat conductance while restricting stream-wise conduction in order to achieve a high effectiveness and allow a large enthalpy difference between the two streams.

The average power consumption of sensor platforms, which are often low duty cycle, can be greatly reduced by applying strong power gating while idling. A key component in power gating is the timekeeping device while the system is in the idle mode. Since the timekeeping device, or timer, is always active, it is often the dominant source for energy loss and must oscillate at a low frequency (e.g., from sub-Hz to 10Hz). Crystal oscillators are too expensive and large for tiny sensor platforms, and typical watchdog timers consume uW of power, which is too high.

One of the key components in neural prosthetic systems is the microprobe, which is responsible for interfacing with neurons. This project aims to design and fabricate a novel all-diamond neural probe. Polycrystalline diamond is used as the material for the probe shank, interconnects/leads, and electrodes. Diamond is chosen because of its unique properties. It has one of the largest Young’s moduli (~1011Pa) of all known materials.

A key challenge in realizing implantable multi-electrode microsystems for neural interface applications is the integration of the front-end electrodes with their interface circuitry. A low profile is important to allow the dura to be replaced over the implant, decoupling it from the skull and allowing it to move with the brain. A new approach to array formation uses a parylene overlay cable to connect a 3-D microelectrode array to a custom-designed signal conditioning chip and thence to the external world.

The use of microsystems for many emerging wireless sensing applications has increased the demand for on-site, small-volume, and replacement-free energy sources as opposed to conventional batteries. On-site energy scavenging from various environmental sources including ambient heat, solar energy, and vibration has been introduced as efficient and promising approaches. This work focuses on the use of body heat generated by beetles as an energy source. The goal of this project is to develop a micro thermoelectric generator (TEG) with the area of approximately 1cm2 that is capable of generating 20–50µW/cm2/°C from a beetle’s body heat.

Applying nano-scaled materials and technologies to microsystems has been a primary theme of the research conducted in the WIMS Center since its inception. Nowhere is this more clearly illustrated than in the chemiresistor (CR) arrays we have developed for the WIMS µGC detector. In the latest advancement, we have employed electron-beam lithography to create interdigital electrodes (IDE) with individual Au/Cr fingers and spaces measuring just 300nm in width (~8µm in length).

The Narada wireless sensor platform has been under development at the WIMS Center since 2005. Narada is designed for field deployment in large-scale infrastructure systems where long communication range, power efficiency, and reliability are all necessities. This past summer, the platform was field tested on the Yeondae Bridge, located in Icheon, Korea.

An integrated platform for sensor applications, dubbed the “Phoenix Processor,” has been designed in a 0.18?m process with a square area of 915?m, making on-die battery integration feasible. The Phoenix Processor uses a comprehensive sleep strategy with a unique power gating approach, an event-driven CPU with compact ISA, a custom low-leakage memory cell, as well as data memory compression, and adaptive leakage management.

Parkinson’s disease is a progressive disorder of the central nervous system, affecting more than three million people in the United States. Deep-Brain Stimulation (DBS) is an emerging therapeutic technology for hypokinetic neurological disorders, such as Parkinson’s disease. In order to achieve the most effective treatment results, the stimulation parameters should be adjusted based on the individual patient’s condition, which can be achieved by a neuro-physiological feedback algorithm.

Several of the most critical factors affecting the performance of the WIMS µGC relate to the DRIE-Si/glass channels used as gas chromatographic (GC) separation microcolumns, specifically, the consistency and uniformity of stationary phase deposition, the deactivation of surface-adsorption sites on the microcolumn walls, and the stability of the stationary phase following repeated thermal cycling.

Most anticipated applications of microanalytical systems to environmental monitoring require detection of target compounds in the parts-per-billion or parts-per-trillion concentration range. Because detector technologies are not sensitive enough to achieve limits of detection in this range, a preconcentration step is necessary prior to separation and detection.

A 64-channel, single-unit, neural recording microsystem has been developed and tested for the first time. The penny-size device is capable of wirelessly recording from different regions of the brain, detecting action potentials (spike events) above a user-programmed threshold on all the channels simultaneously, or digitizing any one of the channels with 8-bit resolution.

Past work has produced 3-D electrode arrays using microassembled planar 2-D probes, but the arrays have been tedious to assemble using micromachined spacers, limiting their availability for practical use. Also, they have had relatively high rise above the cortical surface and have consumed a wide footprint there due to the use of lead transfer wings on the probes.

In the vast majority of studies of microfabricated sensor arrays for analysis of volatile organic compounds (VOC), the sensors employed in the arrays operate on the same transduction principle. With most of these single-transducer (ST) arrays, a thin interfacial film of a sorptive polymer serves to reversibly concentrate vapors near the surface of each sensor.

Electronic and micromechanical devices/circuits, such as resonant sensors, low-noise amplifiers, and micromechanical resonators, exhibit superior performance when operated at low temperatures. To take advantage of these potential performance gains, low-power microscale cooling systems must be developed that can be integrated easily with a wide variety of microfabricated devices. The goal of this project is to develop microthermoelectric coolers capable of cooling MEMS and electronic devices below 200K, while dissipating less than 100mW of power, all in a chip no larger than a few millimeters on a side.