Flexibele biosensoren voor microgolf-diëlektrische spectroscopie

This PhD thesis investigates flexible microwave sensing technologies for contactless dielectric characterization of liquids and other materials. While microwave dielectric spectroscopy is a powerful technique for material analysis, existing systems are typically rigid, laboratory-based instruments. In contrast, flexible microwave sensors reported in the literature are generally limited to narrowband detection and do not provide quantitative broadband dielectric characterization. To address these limitations, this work introduces a flexible cylindrical coplanar waveguide (CCPW) sensing platform that can be wrapped around standard laboratory vials and test tubes. The proposed approach exploits the excitation of higher-order waveguide modes within the cylindrical geometry, enabling strong electromagnetic interaction with the sample through the container walls. This phenomenon is leveraged as a sensing mechanism to enhance sensitivity while maintaining fully contactless operation. A highly sensitive one-port microwave sensor is developed and experimentally validated, achieving sensitivities approximately one order of magnitude higher than previously reported contactless microwave sensors. To support accurate measurement interpretation, a dedicated time-domain de-embedding methodology is proposed to isolate the intrinsic sensor response and extract reliable permittivity-related features. The sensing framework is further extended through temperature-assisted microwave characterization, where controlled thermal variation provides additional discriminatory information for quantitative analysis of multi-component ionic mixtures. This approach demonstrates the potential of combining microwave sensing with physics-informed feature extraction for complex mixture characterization. Building on these developments, the CCPW concept is generalized to broadband microwave dielectric spectroscopy. A coupled transmission-line model accounting for waveguide mode excitation is developed, enabling quantitative extraction of complex permittivity over a wide frequency range. The resulting flexible broadband sensor achieves permittivity extraction errors below 10% across the 2-18 GHz frequency range while preserving the advantages of contactless measurement. Finally, the thesis explores liquid-metal-based fabrication techniques for mechanically compliant microwave structures. Several implementation approaches are experimentally evaluated, providing insight into their benefits and practical limitations for future flexible microwave sensing systems. Overall, this work establishes a new framework for flexible, contactless broadband microwave dielectric spectroscopy and demonstrates its potential for portable material characterization in biomedical, chemical, and industrial applications.