Printed circuit boards (PCBs) are used to mechanically support and electrically connect electronic components in various commercial electronic devices.
Conventional rigid PCBs typically consist of an organic resin like epoxy or polyimide, reinforced with around 50-70 % of woven glass fiber. Cu sheets etched through photoresist masks usually define the conductor patterns. [125], [205], [221] In this thesis, commercially manufactured PCB-based resonance sensors (Publication I) were first tested as a proof-of-concept for wirelessly monitoring the degradation of bioresorbable polymers. The multilayer structure included an inductor coil, interdigitated and parallel-plate capacitors, as well as via through holes. This sensor architecture gives an idea about the structural features that can be fabricated in large quantities using conventional materials. It can thus be used as a comparison reference for bioresorbable devices.
The first step towards fabricating an LC circuit using bioresorbable materials was a conductor wire (Publication II), which was used to form planar coils. In fact, the coils themselves can be considered as resonators whose capacitance consists of parasitic capacitances between the coil turns. This kind of coils can be used for example as sensors [21], [222] or as wirelessly powered microheaters if a meander resistor is connected to the coil [146]. More sensing options can be achieved by pairing two coils with a dielectric layer in between [223]–[225]. PDTEC was chosen as the core fiber material due to its thermal properties (Tg = 99 °C) that enable melting the PCL (Tm = 57 °C) coating onto the core fiber without severely damaging it. It was shown that the evaporated Mg layer on the core fiber remained almost as conductive after the PCL melting process than before, which was a crucial advancement towards realizing a bioresorbable wireless pressure sensor.
The wireless pressure sensors (Publications III and IV) consisted of two similar metallized PDTEC substrates that were attached onto a holed spacer using PCL films as an adhesive. The substrates were melt processed to enable better control over the substrate thickness and surface roughness, as compared to solvent casting which has been commonly used for preparing bioresorbable substrates for electronic
devices [73], [199], [226], [227]. The patterning of the e-beam evaporated Mg (7.0-7.5 µm) or magnetron sputtered Zn (~4 µm) conductors was based on an earlier work by Salpavaara et al. [104]. In the type 2 pressure sensors, laser-cut metal masks replaced the 3D printed plastic masks to prevent sputtered Zn from permeating under the mask. This kind of blurring has been described earlier as a result of non-directional sputtering deposition and may lead to short-circuits if the shadow masks are not in good contact with the substrate [228].
The operation of the pressure sensors was based on bending of the substrates towards each other in the holes of the spacers as pressure was applied. This brought the metallized capacitor plates closer to each other, which changed the capacitance of the LC circuit and thereby the resonance frequency of the device. We showed that with one exception, the sensor sensitivities and initial resonance frequencies between different sensors (n=8) were consistent, even with a manual fabrication process.
The sensor architecture was designed to avoid through hole vias or other galvanic contacts between the conductor layers. This was the main advantage of the presented pressure sensors, along with the possibility to construct the sensors from two identical metallized substrates. In the literature, Luo et al. avoided the challenge of fabricating through hole vias by folding a single substrate that contained an electrical interconnect between the inductor coil and the upper capacitor plate [73], [74].
Boutry et al. used microstructured PGS pyramids between the pressure-sensitive capacitor plates to shorten the response time of their sensor [76]. Furthermore, their inductor coils were laminated on a dielectric PLLA film (50 µm) to separate them from the region of pressure-sensitive capacitors.
The main goal in all the reported sensor architectures was to keep their complexity and the required processing steps as low as possible. During this thesis, another pressure sensor architecture was also tested, where Mg was deposited directly onto a dielectric PCL layer. The pressure-responsive capacitor was created by hot embossing a round cavity onto the upper PDTEC substrate (or lid) and evaporating Mg onto the bottom of this cavity [37]. These endeavors were abandoned due to more complex fabrication procedures and poorer sensor performance compared to the presented sensors.
To summarize the position of our bioresorbable pressure sensors with respect to existing literature, it needs to be first emphasized that the used materials, fabrication methods and sensor architectures vary significantly between different studies. Direct comparison is therefore difficult. In this study, the simple Mg pressure sensor architecture with two similar metallized substrates required only two different, easy-to-manufacture components (holed multilayer spacer and metallized polymer
substrates). The resulting simple assembly process is the main asset of our pressure sensors compared to other reported bioresorbable wireless pressure sensors. In this study, the multilayer spacer was manually assembled from compression molded polymer sheets, but on a commercial scale it could be extruded as a continuous multilayer sheet and then cut into shape. Other bioresorbable pressure sensors in the literature include transfer printing of electrodeposited conductors and folding a single substrate to provide galvanic connections between the two layers [73], [74], or stacking numerous layers of different materials [76], [229].
Bioactive glasses are under-represented as substrate materials for biodegradable electronics and contain a lot of potential for photolithographic methods and other post-processing methods that are being used with conventional non-degradable wafers. Previous studies have evaluated borate-based glasses as sensor substrates, demonstrating their fast dissolution profile with mass loss measurements and wired electrical measurements of prototype devices [115], [116]. In this study, S53P4 (Bonalive®) bioactive glass discs were introduced as substrates that have a better chemical and thermal resistance compared to typical biodegradable polymers. S53P4 is a silicate based bioactive glass composition, which has been commercially used as a synthetic bone graft material in orthopedic surgery [60]. The discs used in this study enabled spin coating an insulating PDTEC layer onto a metallized Mg layer, which was used to demonstrate its chemical resistance. Another option would be to sputter deposit or evaporate dielectric layers (e.g. SiO2) onto the glass, as done by Unda et al. [115]. This kind of inorganic dielectric layers would be thermally more resistant compared to their polymeric counterparts, which should enable depositing an inorganic water barrier layer onto the final resonator structure. In this study, the highly water-sensitive Mg conductors on the insulating PDTEC layer were protected with non-degradable Parylene to enable studying the sensors under immersion. As a conclusion, the robust glass substrates allowed fabricating the resonator structure step by step without damaging the underlying layers.
The compression sensors (Publication IV) were prepared onto PLDLA 96:4 screws using Parylene-insulated Mo wire to demonstrate a scheme for fabricating solenoidal resonator structures. Although the Parylene coating was not biodegradable, similar chemical vapor deposition methods have been recently developed for biodegradable polymers as well [198]. Extrusion coating the Mo wire could provide a further option. Karipott et al. have earlier reported an approach, where a non-degradable temperature sensor has been mounted into a hollow orthopedic screw [39]. This kind of modular approach could be beneficial in terms
of device assembly, if the sensor part could be fabricated separately and then installed into e.g. an injection molded screw.