Abbas Motalebizadeh

Topic

Toward Field-Deployable Microplastic Detection in Water: Microfluidic, Peptide-Mediated, and Machine-Learning-Assisted Sensing Strategies

Department of Mechanical Engineering

Date & location

• Monday, August 17, 2026
• 9:00 A.M.
• Virtual Defence

Examining Committee

Supervisory Committee

• Dr. Mina Hoorfar, Department of Mechanical Engineering, University of Victoria (Supervisor)
• Dr. Mohsen Akbari, Department of Mechanical Engineering, UVic (Member)
• Dr. Karolina Papera Valente, Department of Mechanical Engineering, UVic (Member)
• Dr. Ali Dolatabadi, Department of Mechanical Engineering, University of Toronto (Outside Member)

External Examiner

• Dr. Fariborz Taghipour, Department of Chemical and Biological Engineering, University of British Columbia

Chair of Oral Examination

• Dr. Abdul Roudsari, School of Health Information Science, UVic

Abstract

Microplastics are now pervasive in fresh and marine waters, yet routine monitoring remains difficult: established methods such as Fourier-transform infrared and Raman spectroscopy and microscopy give reliable polymer identification but require sample preparation, skilled operators, and laboratory instrumentation that limit rapid, distributed measurement. Narrowing the gap between laboratory characterization and practical on-site monitoring requires coordinating rapid sample handling, particle-sensitive detection, molecular recognition, quantitative transduction, matrix compatibility, a portable readout, and reliable interpretation. This dissertation investigated complementary microfluidic, triboelectric, peptide-mediated, electrochemical, plasmonic, and machine-learning-assisted strategies, each addressing a different layer of this problem and arranged as a staged progression from physical detection toward molecularly informed, interpretable sensing.

A droplet-based microfluidic triboelectric sensor produced electrical responses that varied systematically with microplastic particle size and concentration under controlled conditions, establishing rapid physical detection but lacking polymer selectivity. A critical review of peptide-based recognition then organized the relevant interaction mechanisms, evidentiary standards, and validation requirements, motivating two experimental platforms. Coupling a fluorescence-tagged plastic-binding peptide with electrochemical impedance spectroscopy (EIS) enabled concentration-dependent detection of polystyrene that was selective relative to the tested materials in low-ionic-strength water, with an operational detection limit near 50 ppb and performance that declined at high ionic strength. Transducing the same recognition concept optically, peptide-functionalized gold nanoparticles generated multivariable plasmonic and hydrodynamic features; a supervised machine-learning (ML) model correctly classified six of seven held-out test samples (85.7%), a preliminary result given the small test set.

Together, these studies show that field-oriented microplastic sensing must coordinate sample handling, recognition, transduction, matrix effects, calibration, and interpretation, since no single layer fully compensates for another, and they map the trade-offs among selectivity, sensitivity, simplicity, and portability. The principal contributions are the microfluidic triboelectric platform, the peptide-recognition framework, peptide-supported electrochemical quantification, a peptide nanoparticle optical platform with preliminary machine-learning interpretation, and an integrative staged framework linking these elements. Bounded by the use of model particles, limited polymer and matrix scope, and incomplete environmental validation, the dissertation provides an experimentally grounded foundation for future modular, field-oriented microplastic monitoring rather than a finished, deployed device.