Bimodal neural interfaces: Design, fabrication and characterisation considerations for the seamless operation across electrical and acoustic domains

Neurological disorders are the leading cause of disability, affecting millions around the world, and often resulting in severe motor, sensory, or cognitive impairments. Conventional pharmaceutical treatments can alleviate symptoms, but rarely restore full neurological function, and their lack of specificity often leads to significant long-term side effects. Miniaturized neural interfaces, on the other hand, have the potential of providing critical insights into the nervous system, as well as, enabling more precise therapies for patients. Although electrical stimulation has primarily been the method of interaction with the nervous system, ultrasound is emerging as a minimally invasive and highly specific alternative for neuromodulation, but also communication and wireless power transfer. Bimodal neural interfaces, combining electrical and acoustic modalities, within a single, fully integrated platform, are of great importance for directly comparing their effects, yet to date, no such tools are available. Developing bimodal neural interfaces is a complex task that demands a deep understanding of the interactions between different modalities, materials, and components.

To this end, several research questions have been addressed. The first, relates to one of the final steps in the development of bimodal neural interfaces, the encapsulation. Given that miniaturized micromachined ultrasonic devices are a fundamental building block of the proposed systems, how can they be best encapsulated in polymers that are mechanically compliant with the soft biological tissue, while still preserving their acoustic performance? The second research question focuses on the development and integration of the constituent components of a bimodal neural interface. In particular, it addresses how can micromachined ultrasonic transducers and microelectrodes coexist and not interfere with each other during operation. Furthermore, the requirements differ depending on the intended application. In the case of an in-vitro bimodal neural interface, flat and rigid micromachined devices are required, whereas for implantable applications, thin, flexible, curved transducers and electrodes are required, substantially increasing the design and fabrication complexity of the devices.

This thesis aims to address current technological limitations by designing, fabricating, and validating two different types of miniaturized bimodal neural interfaces for both in-vitro and implantable applications. To achieve this, the interaction between micromachined ultrasonic transducers (MUTs) and common polymer-based coatings used in neural interface packaging was first evaluated through simulations and measurements on the individual materials as well as encapsulated transducers. All tested materials exhibited high acoustic transparency (>94 %), but performance was also influenced by mechanical factors such as stiffness, residual stress, and flexural rigidity. Building on these insights, two bimodal neural interfaces were then developed. For the first, an in-vitro microelectrode-microtransducer array (MEMTA), a less conventional, reduced-step microfabrication process was adopted, employing maskless photolithography and sputtering through shadow masks. Furthermore, characterization through a multi-domain framework, treating the device as an integrated system rather than as a collection of independent components, was also employed. Next, a cuff-shaped implantable bimodal neural interface comprising thin-film Au electrodes and flexible MUTs embedded in thermoplastic polyurethane, was prototyped to assess how encapsulation affects membrane displacement and acoustic output in curved MUTs.

This work outlines design, fabrication, and analysis approaches that overcome current technological limitations and advance the development of bimodal neural interfaces. Nevertheless, further studies are needed to evaluate the long-term stability of the devices in wet environments, including the effects of microvibrations on the encapsulated MUTs and the impact of bending on encapsulation integrity and device performance. In addition, the electroacoustic effects that may arise in the proposed bimodal neural interfaces when operating in wet ionic environments require careful investigation. Furthermore, developing ultrasonically powered bimodal neural interfaces could fully unlock the potential of next-generation, miniaturized, and versatile neural interfaces.