Development of an inductive sensor for superparamagnetic nanoparticles detection and quantification

Details

Supervisors

Raskin, Jean-Pierre

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil électricien, à finalité spécialisée

Abstract

This research falls in line with the development of a point-of-care system for water quality monitoring by means of a paper-based sensor using the Later Flow Assay (LFA) technique. Specifically, this LFA employs superparamagnetic nanoparticles as label to detect bacteria in water. The main objective is to create a low-cost, portable system suitable for regions with significant water quality issues, but limited resources and equipment for monitoring. The main focus of this work is the design of a magnetic sensor to detect and quantify nanoparticles at the LFA test line, with the constraint of using only off-the-shelf components to ensure affordability and portability. Two inductor-based solutions are explored. The first solution is based on a design proposed in an article, that uses two pairs of primary-secondary coils. The primary inductors induce currents in their respective secondary coils, and the presence of nanoparticles is detected by measuring variations in the induced current. However, this method proved unsuitable in this context of this work, due to both sensitivity issues and complexity of implementation. The second solution measures the frequency variation in the oscillation frequency of an LC-oscillator. The presence of nanoparticles in the inductor of the oscillator causes a shift in its inductance value, thereby altering the frequency at which the circuit oscillates. This approach yielded promising results, detecting down to 15 µg of nanoparticles on a sample, and displaying a linear relationship between the mass of nanoparticles and the amplitude of the frequency variation. Moreover, both the oscillator and the entire frequency-conditioning chain can be implemented with off-the-shelf components and a Teensy microcontroller. Several improvements were identified that could further lower the limit of detection reached in this initial iteration.