Publications

Mobile applications based on gas sensing present new opportunities for low-cost air quality monitoring, safety, and healthcare. Metal oxide semiconductor (MOX) gas sensors represent the most prominent technology for integration into portable devices, such as smartphones and wearables. Traditionally, MOX sensors have been continuously powered to increase the stability of the sensing layer. However, continuous power is not feasible in many battery-operated applications due to power consumption limitations or the intended intermittent device operation. This work benchmarks two low-power, duty-cycling, and on-demand modes against the continuous power one. The duty-cycling mode periodically turns the sensors on and off and represents a trade-off between power consumption and stability. On-demand operation achieves the lowest power consumption by powering the sensors only while taking a measurement. Twelve thermally modulated SB-500-12 (FIS Inc. Jacksonville, FL, USA) sensors were exposed to low concentrations of carbon monoxide (0–9 ppm) with environmental conditions, such as ambient humidity (15–75% relative humidity) and temperature (21–27 ◦C), varying within the indicated ranges. Partial Least Squares (PLS) models were built using calibration data, and the prediction error in external validation samples was evaluated during the two weeks following calibration. We found that on-demand operation produced a deformation of the sensor conductance patterns, which led to an increase in the prediction error by almost a factor of 5 as compared to continuous operation (2.2 versus 0.45 ppm). Applying a 10% duty-cycling operation of 10-min periods reduced this prediction error to a factor of 2 (0.9 versus 0.45 ppm). The proposed duty-cycling powering scheme saved up to 90% energy as compared to the continuous operating mode. This low-power mode may be advantageous for applications that do not require continuous and periodic measurements, and which can tolerate slightly higher prediction errors.

JTD Keywords: Smartphone, Metal-oxide semiconductor, Gas sensor, Low power, Temperature-modulation, Interferences

Biogenic magnetite (Fe3O4) has been identified in human brain tissue. However, abnormal concentration of magnetite nanoparticles in the brain has been observed in different neurodegenerative pathologies. In the case of Alzheimer's disease (AD), these magnetic nanoparticles have been identified attached to the characteristic brain plaques, which are mainly formed by fibrils of amyloid β peptide (Aβ). However, few clues about the formation of the magnetite-Aβ complex have been reported. We have investigated the interaction between these important players in the AD with superconducting quantum interference, scanning electron microscope, surface plasmon resonance, and magnetic force microscopy. The results support the notion that the magnetite-Aβ complex is created before the synthesis of the magnetic nanoparticles, bringing a highly stable interaction of this couple.

JTD Keywords: Alzheimer's disease, Biogenic magnetite, Amyloid β peptide (Aβ), Superconducting quantum interference, Scanning electron microscope, Surface plasmon resonance, Magnetic force microscopy