The spinal cord, considered to be the most important path of the human body, when injured induces severe motor dysfunction. Therefore, patients affected by lesions on the spinal cord, are most of the time unable to walk, stand or perform motor activities that are trivial for healthy people. To provide a better quality of life for these patients, extensive research and effort have been put by both neuroscientists and engineers to provide clinical therapies for pain relief and locomotion restoration together with dedicated platforms that could deliver these therapies. Currently, for these purposes, epidural spinal cord stimulation is widely used. Apart from being used as a method to reduce pain, it has also been proven to promote locomotion recovery. Apart from clinical trials, it is of great importance to understand the mechanisms that occur while delivering specific therapies. To this end, more exploratory research is mostly conducted in rodents. However, the availability of tailored neurotechnologies, for experiments conducted in small animals, is limited mostly due to size constraints. Moreover, when developing implantable devices that would target the spinal cord, careful selection of the materials used is equally important. However, understanding the underlying mechanism leading to a specific behaviour or motor outputs requires exploring and quantifying new methods of stimulation. For instance, optogenetics has been gaining a lot of popularity in the field of neural stimulation as it is a more specific technique that could help neuroscientists map the neuronal circuitry within the human body. Thus, apart from developing spinal cord implants that resemble best the anatomy of the body, while inducing as little stress as possible on the spinal cord, for exploratory reasons the developed implants must provide optogenetic compatibility. Therefore, this thesis reports the development as well as the characterization of both passive and active spinal cord implants with optogenetic compatibility.
To achieve the desired goal of having a fully implantable, flexible spinal cord implant with optogenetic compatibility, a scalable and reproducible microfabrication process has been developed. Materials such as graphene, for transparency, flexibility and conductivity were used to develop the microelectrode arrays. Moreover, soft, polymeric encapsulation was employed to sustain the high flexibility and transparency of the implant. The end result of the microfabrication process would lead to a device consisting of a multi-layered graphene structure between two polymeric-based encapsulation layers and metal test pads for interconnection to the outside world. However, towards achieving this final structure, several challenges were encountered. Suspension of the implants after developing them on a rigid substrate, yet ensuring high quality for the graphene layer leads to several iterations of the fabrication process. Despite the challenges encountered, several prototypes were successfully developed. However, having prototypes that can only validate the process flow would not suffice. Therefore, extensive evaluation of the devices has been conducted and reported. Methods such as Raman spectroscopy and optical transmittance to evaluate the graphene layer or cyclic voltammetry and electrochemical impedance spectroscopy to characterize the performance of the fabricated devices were employed. The degree of transparency obtained using the reported microfabrication process was ~78 %, leading to the conclusion that the number of graphene layers for the final device was 10. It has been proven that graphene does not deteriorate over time when soaked in saline solution for several consecutive days and apart from that, the graphene-based implants showed no performance deterioration when bent over rods down to 3 mm in diameter. Moreover, the graphene electrodes provided impedance values of ~8 kΩ at 1 kHz frequencies, values comparable to what literature has previously reported. Apart from developing a passive graphene-based spinal cord implant, the focus of this thesis was also to fabricate and characterize an active implant. However, embedding active components with a flexible, graphene-based array of electrodes is not trivial. Therefore, system integration of small test chips was investigated and after several iterations of flip-chip bonding processes, a complete, active, graphene-based prototype was obtained. The measurements performed after the bonding process have proven that both bonding on graphene-only as well as on graphene and metal substrates is possible and the four-point measurement results indicated resistance values ranging from 10 mΩ up to 16 Ω for individual connections, depending on the substrate used.
Therefore, with this research project, not only the first fully transparent, graphene-based spinal cord implants have been developed but also the results obtained from their characterization illustrate that the process is stable and the performance of the devices is promising.