Miniaturized, multiplexed, graphene-based, wearable devices for non-invasive, electroosmotic, follicular glucose extraction and monitoring

The World Health Organization predicts that the worldwide incidence of diabetes will increase from 171 million in 2000 to 366 million in 2030, driving projections for the global glucose monitoring market to over USD 12 billion by 2025. This requires a technological revolution to take place: so far, no technology able to monitor glucose in the body non-invasively has been developed, leaving patients affected by diabetes to handle invasive, needle-based devices (often associated with significant user resistance, originating primarily in the pain and discomfort during usage). In this context, there is a clear need for innovative, non-invasive and pain-free technologies to enter the market. This would eliminate patient resistance towards more frequent sampling and, hence, considerably improve a diabetic's control over glycaemia; and, in time, help shift emphasis from diabetes treatment to its prevention, resulting in considerable financial savings for the healthcare system. In this thesis, a new path-selective, non-invasive, transdermal glucose monitoring system, based on a miniaturized pixel array platform, is introduced. The system samples glucose from the interstitial fluid (ISF) via electroosmotic extraction through individual, privileged, follicular pathways in the skin, accessible via the pixels of the array. A proof of principle - using mammalian skin ex vivo, as well as in-vivo tests on human volunteers - is demonstrated for specific and 'quantized' glucose extraction/detection via follicular pathways, and across the hypo- to hyper-glycaemic range in humans. Furthermore, the quantification of follicular and non-follicular glucose extraction fluxes is clearly shown, indicating that quantitative, calibration-free (i.e. needle free) glucose detection is achievable with this method. In vivo continuous monitoring of ISF-borne glucose with the pixel array was able to track blood sugar in healthy human subjects for a period of up to 6 hours, and, importantly, without inducing skin irritation. A pilot study on a small cohort (n = 10) of healthy human volunteers subjected to an oral glucose tolerance test provided the data for a preliminary Clarke Error Grid: glucose was tracked over 6 hours within acceptable limits, i.e., with all data falling within zones A and B of the grid. The progression of the work required several generations of pixel array prototypes, which evolved from using thin film technologies (which included Chemical Vapour Deposition (CVD) graphene and vacuum techniques for electrode definition), to scalable, cost-effective techniques, such as screen printing of functional (e.g. graphene and Ag/AgCl) inks. Several pixel device layouts were also investigated in view of decreasing glucose response time and increasing collection efficiency. The studies also highlighted some limitations of the current technological implementation (primarily related to electrode biofouling, and the extraction process affecting the detection) that will need to be addressed to advance the technology to commercialization. Finally, the work in this thesis also provided insight into the possibility of multiplexed detection of multiple analytes that are present and can be extracted simultaneously with glucose from the ISF.