Electronic Military & Defense magazine was developed for engineers, program managers, project managers, and those involved in the design and development of electronic and electro-optic systems for military, defense, and aerospace applications.
Issue link: https://electronicsmilitarydefense.epubxp.com/i/350588
device, despite being sensitive to added mass, may not be of significant use in a real-world application, owing to the fact that the efficiency of a single NEMS device is extremely low. Consider the case of a single functionalized nanodevice oper- ating in a very dilute aqueous or gaseous solution as illus- trated in Figure 1. The sensor may easily have the sensitivity to detect a single analyte molecule, but the effective cross- sectional area of a nanodevice is miniscule. If it is assumed that the analyte particles go through Brownian motion, then the time that it will take for the analyte to come into contact with the sensor can be estimated as: Δ t ~ 1 ( 1 ) 2 (1.2) with D defined as the diffusion constant, A as the active NEMS area, and n A as the number of analyte molecules per unit volume. One could potentially detect a signal due to the presence of tens to hundreds of molecules using a single sensor. However, in a typical rare ( n A ~small) mixture, it would take an unacceptably long time (Δ t ~ large) for an analyte molecule to come in contact with a single sensor — thus rendering it ineffective. Along the same lines, performing a complicated mechanical computing or signal processing task by a single device would also be difficult. The above scenario can be overcome by employing an array of nanodevices distributed over space operating in parallel. If an array of N sensors is working in parallel, the corresponding detection time is roughly scaled down by 1/N, retaining the sensitivity of a single device and thereby positively detecting analyte molecules in a dilute solution in a reasonable amount of time. In addition, it also leads to direct improvement in sensitivity through signal averaging. Nanosensor arrays, therefore, could provide powerful new approaches to biothreat detection — with sensitivity at the single molecule level. The discussion so far suggests that it is desirable to oper- ate many sensors in parallel. In resonant sensor operation, the actuator excites the sensor harmonically around its fundamental resonance, and the sensor converts the subse- quent mechanical response into electrical signals — while the resonator interacts with the environment. This immedi- ately imposes the fact that to independently operate N NEMS devices, one needs ~2N connections and ~2N detection circuits/electronics. As N gets larger, this quickly becomes impractical. To reduce the number of connections to the array, the actuation-detection technique for the NEMS devices should be multiplexed. Multiplexing is a method of providing a sin- gle channel to actuate the NEMS motion and a single channel to detect its motion. In addition, the necessary information — such as resonance frequency — of each resonator in the array should be detectable by a single set of electronics. To get the individual device information, it is clear that each device should be individually addressable in some fashion. This, however, imposes an arduous limit on the minimum amount of time for collecting the information from all of the array elements. These basic concepts in parallel operation of NEMS are illustrated in Figure 2. In Figure 2(a), N devices are operating in parallel with 2N connections and 2N detection circuits. In principle, the minimum time for the measurement of all the information from the array is the integration time of the slow- est electronics. In contrast, in Figure 2(b), the actuation and detection channels are multiplexed as well as the electronics. The measurement time increases depending upon how the data is collected from individual sensors. Proposed Sensor Array Design And Fabrication We propose an optical multiplexing approach incorporating an adaptive photorefractive holography technique to actuate and detect sensor array motion. This approach is indepen- dent of the number of elements in the array and allows for broadband parallel detection with independent addressabil- ity using a charge-coupled device (CCD) with high spatial resolution. Doubly clamped nanomechanical beams were fabricated using a top-down approach commonly employed in the microelectronics industry. The process begins on a commer- cial silicon-on-insulator (SOI) wafer. These heterogeneous substrates contain a buried layer of silicon dioxide (SiO 2 ) acting as a sacrificial insulator between the device layer and the bulk silicon handle. The thickness of the top silicon layer determines the thickness of the resonator, which ranges from 150 nm to 300 nm. The sacrificial layer thickness varies from 150 nm to 400 nm based on the wafer used. The fabrication sequence begins with optical lithography to define the micro- scopic features, which include contact (finger) pads that serve as electrical connections between the sensors and measure- ment electronics. Photolithography enables the transfer of a previously designed contact pad pattern from an optical mask onto a thin film of polymer. To create the beam features that are in sub-micron range, an electron beam lithography (EBL) process was adopted. To suspend the nanomechanical reso- nators, both dry and wet etching techniques were used. The ultimate goal of this research is to develop a system for the excitation and detection of vibrational characteristics of large-scale NEMS arrays. Optical excitation of vibrations in nanoresonators is employed where an amplitude modulated laser source is used to heat the beams photothermally, and actuation is achieved through the thermoelastic effect. The principle behind exciting NEMS resonators by photothermal Technology 28 Electronic Military & Defense Annual Resource, 4th Edition D An A Figure 2 (a and b)