Impact/Purpose:

An ideal microsensor for autonomously monitoring chemical and molecular environmental hazards in both water and air should simultaneously have a high sensitivity, a high selectivity, a large dynamic range, a low manufacturing cost, simple calibration/reset protocols, a long lifetime, field reconfigurability, and low power consumption. These requirements arise from considering the rapid deployment and autonomous operation of a microsensor network monitoring a large area. We have developed a Si-based neuron MOS transistor with a novel extended floating-gate structure that permits molecular/chemical sensing. Our sensor, called a chemoreceptive neuron MOS (CMOS) transistor, is expected to simultaneously meet all of these requirements, and can be fabricated by minor modification or simple postprocessing of conventional CMOS integrated circuits. The modular structure and fabrication of this new device permits us to use CMOS devices optimized for high sensitivity and large dynamic range and affords us complete flexibility in the design and composition of the molecular/ chemoreceptive sites. The performance of our new sensor is expected to be vastly superior to that of existing chemical microsensors, such as the ion-sensitive FET (ISFET) and the CHEMFET, in nearly every important respect resulting from the internal transistor gain and much better isolation between the electronics and microfluidics.

Description:

Within the 3 -year effort, we have established several major findings:

• Chemical sensor in fluid environment with inorganic and polymer sensing surfaces (1,5): Conventional metal oxide semiconductor field effect transistor (MOSFET)-based chemical sensing such as ISFET is subject to two main problems. The first is the isolation of the sensitive MOSFET from the sensing ambient which often contains alkali ions such as sodium and potassium as buffer solution. The earlier version of ISFET uses a very thick SiO₂ or non-CMOS compatible process with gate dielectric by Si₃N₄ or Al₂O₃ to minimize the influence of alkali ion drift inside the gate dielectric. The second problem is the need for a fluid potential to set up the bias point for circuit operations, since the gate was directly driven by the fluid potential that provides both the DC bias point and analyte perturbation. Although ISFET can use an operational amplifier (OP-AMP) stage to minimize its dependence on the fluid reference potential, difficulties still exist in electrically pinpointing the operating point due to nonideal OP-AMP characteristics. CnMOS overcomes both of these shortcomings by using an extended floating gate device structure. We have successfully demonstrated stable chemical sensing of salinity and acidity over a wide range of concentrations, and selectivity between sodium and potassium contributions.
• Chemical sensor in air environment with polymer sensing surfaces: To use ISFET-like structures for gas sensing, the device has to be operated without relying on a reference potential since the gas has very high impedance. Furthermore, the fringing field distribution is often a problem since the dielectric constant of gas varies significantly with the humidity. CnMOS overcomes these shortcomings by setting the DC operating point in the control gate, and by pinning the fringing field with additional structures on the sensing gate. We have successfully demonstrated stable sensing of CO₂, calibrated with the stripping potential electrode, whose bandwidth is much lower than that can be achieved by CnMOS.
• Neuromorphic circuits for collecting sensor responses: CnMOS has the drain signal directly as the output signal with the DC operating point and AC impedance excitation directly from the control gate. We have successfully constructed sampling circuits on the same die that can achieve high sensitivity, high bandwidth, and large dynamic range.
• Fluid actuation based on electrowetting by charge injection in CnMOS (2, 3): An integration of chemical sensors and electrowetting actuators based on CnMOS transistors has introduced a novel system-on-chip approach to the development of a microfluidic system. Preliminary experimental results have verified the design and operation of these devices in monolithic sensing and actuating schemes and have shown the potential in developing the dual-function devices with low power consumption and full compatibility with conventional CMOS technology.
• Pressure sensor with piezoelectric sensing surfaces (4): Integrated pyroelectric/piezoelectric sensors offer possibilities of low-cost sensor arrays for purposes such as laser positioning, thermal imaging, and in vivo biomedical measurements. Polyvinylidene fluoride (PVDF) offers an attractive polymeric option due to a high pyroelectric coefficient in the β-crystalline phase, high mechanical strength, low acoustic impedance, and good resistance to most corrosive chemicals. We have integrated PVDF on the sensing gate of CnMOS and demonstrated successful pressure and thermal sensing. The poled PVDF film in this study exhibits a piezoelectric coefficient of 15 pC/N and a pyroelectric coefficient of 19 μC/m²K. The film has approximately 1-μm thick Au/Pt layers on both sides. Film connections were made by conductive epoxy. The pyroelectric and piezoelectric characteristics allow enough sensitivity in the range appropriate for in vivo measurements.
• Time-resolved molecular transport in fluids (6): We have experimentally demonstrated microfluidic, electroosmotic, and electrophoretic integration with CnMOS to facilitate time-resolved sensing by taking advantage of the available high sensitivity and large bandwidth. By the CnMOS drain current, we can monitor the ion and molecular transport in various fluidic compositions. The present bandwidth limitation is from the electrical double-layer stabilization and fluidic delivery, with CnMOS entirely in the quasi-static modes producing repeatable measurements with no memory effects. With improved engineering in fluid delivery and sensing gate surfaces, our approach shows strong promise for high bandwidth CMOS-compatible lab-on-a-chip.
• Electrolyte current pulse detection with signal equalization (7): We report the first generalized empirical study of signal equalization for FET-based electrochemical sensors, which provides substantial improvement for fast electrolytic signal detection. The measurements are performed with charge-based sensors in CnMOS transistors, which have extended floating-gate structures for improved system integration of biological sensing, such as tight electrolyte isolation, reduced electrical invasiveness, and no need for a reference electrode in the analyte. With an oxide/electrolyte interface, we report the potentiometric detection of electrolytic current pulses through noisy channels filled with phosphate-buffered saline (PBS). We have achieved successful sensing with at least 1-MHz bandwidth for 100-mV, 1-μs pulses where the electrolyte current is on the order of 100 nA. Equalization is shown to improve S/N for more than 7.2 dB.

URLs/Downloads:

Record Details:

Project Information:

We have already established the preliminary process flow and testing of CMOS transistors with generic molecular receptive areas for vapor and liquid sensing (e.g., water, acetone, etc.). Our preliminary measurements have validated most of our assumptions on the performance of these devices. In the three-year proposed effort, we will fabricate prototype arrays of these novel microsensors with various molecular/chemoreceptive surface coatings and characterize their sensitivities. Surface adsorption kinetics will be studied to facilitate fast and reliable coating selection. We will start with polymer coatings that have been used in vapor and liquid sensors through volume expansion monitoring. We will gather a new table of target agents and coatings from CMOS reading to achieve selectivity. We will also develop a micropower neuromorphic electronic interface for such sensor arrays whose structure and function is based on what is known about the olfactory and gustatory sensory systems of animals. This interface, called the silicon olfactory bulb, will provide a distilled set of informative features that can be used by a recognition system to perform analysis and risk assessment.

Cost :$354,000.00

Research Component :Nanotechnology

Project IDs: