Highly sensitive and selective non-enzymatic measurement of glucose using arraying of two separate sweat sensors at physiological pH

• Cobalt nanowires/Copper nanoparticles formed by controlled electrodeposition. • Arraying of Co/Cu sensor with a second sensor increased the selectivity. • Arraying of two non-enzymatic sensors minimized the need of alkaline medium. • Arrayed system showed a zone A deviation (clinical accuracy zone) in C-EGA method. A big challenge of non-enzymatic sensors is their lack of enough selectivity towards the analytes such as glucose, which must be addressed. In this study, a new system of arraying two different electrochemical sensors is designed to achieve glucose-selective and -sensitive non-enzymatic sweat-based sensors. Two kinds of sensors have been fabricated using (i) the controlled electrodeposition of cobalt/copper as the glucose-sensing electrode and (ii) the functionalized MWCNT/Fe 3 O 4 electrode to eliminate the interference of other existing chemicals in the sweat. The controlled electrodeposition of Co/Cu led to the nanowire-shaped growth of cobalt and nanoparticle formation of copper on the carbon working electrode, confirmed by the SEM. Also, the functionalization of MWCNT with long-chain amine groups and Fe 3 O 4 electrocatalysts has been monitored. Chronoamperometry test results of Co/Cu sensor showed a linear behavior in response to the glucose with a high sensitivity of 1909.86 µA.mM −1 .cm −2 and less than 5% error for the chemicals interferences in the sweat with an exception for the uric acid. The second sensor for minimizing the uric acid interference has been fabricated based on F-MWCNT/Fe 3 O 4 , which showed a high sensitivity of 629.75 µA.mM −1 .cm −2 towards the uric acid and no sensitivity response to the glucose. The arrayed sensors minimized the error imposed by the uric acid interference and the need for an alkaline medium for measuring the glucose in real sweat samples. The analyzed data of real sweat samples at physiological pH by Clarke's error grid method showed a “zone A” deviation (clinical accuracy zone), comparable with the blood glucose levels measured by the commercial glucometers.