Diabetes is a chronic disease that occurs when insulin is neither well-produced nor effectively utilized. Over 4.2 million people died from diabetes in 2019 , and this number is projected to constantly increase to about 700 million by 2040 s, according to the estimation of the International Diabetes Federation (IDF) . Precise control of blood glucose levels in daily life can not only reduce the symptom and increase the survival rates of diabetes, but also prevent or delay long-term, serious health problems, such as heart disease, vision loss, and kidney disease, so the development of glucose monitoring devices in real-time is necessary . Although commercial glucose oxidase-based glucose meters have been widely used, the sensing performance is easily interfered with by the external storage and transport conditions, which hinders their applications in glucose determination . Non-enzymatic glucose sensors (such as metal oxides), in that case, have attracted much attention for their high chemical stability and easy operation. Nevertheless, compared with glucose oxidase enzymes, the catalytic activity and specificity of non-enzymatic materials are usually lower, resulting in a poorer sensing performance [5,6,7].
CuO, which is expected to be an attractive non-enzymatic sensor for its p-type semiconductor with a narrow bandgap of 1.2 eV, has been widely explored [8, 9], especially for non-enzymatic glucose sensors [10,11,12]. Still, the detection range of most as-prepared sensors is 0-5 mM which is narrower than the glucose in human blood (4-7 mM for healthy people and ≥ 9 mM for diabetic patients) due to the weak conductivity of CuO [13, 14]. While introducing oxygen vacancies by adjusting defect structures and electronic states, such as heating the material in a reductive atmosphere, and doping nonmetals (halogen, etc.) to replace lattice oxygen [15, 16], has been suggested to improve the conductivity of CuO [17, 18]. However, these methods often require secondary heating, and the reductive atmosphere is prone to make crystal agglomerate and collapse. Thus it is urgent to develop a technique to increase the oxygen-vacancy content in CuO without secondary heating.
Metal-organic frameworks (MOFs) are composed of metal ions and organic links through coordination bonds with regular pores. These have been widely applied in biomedicine, such as sensors, bioimaging materials, and drug delivery agents [19,20,21,22]. Recently, metal oxide/C-derived MOFs have attracted much attention in the electrochemical field for designing appropriate structures with plenty of active species . For example, thermal treatment can reduce the reaction time with no post-treatment, whilst the product keeps the morphology of the initial MOF with a large surface area [24, 25]. More importantly, such treatment allows more atomically active sites to be exposed, resulting in better MOF performance. Recently, many groups found that oxygen-vacancy in MOF-derived materials can enhance electrochemical performance . Hence, exploring MOF-derived CuO with more atoms active-sites and oxygen-vacancy will be a promising way to achieve a more comprehensive linear range for glucose sensing. However, the change of oxygen-vacancy under different processing temperatures is seldom reported for the MOF-derived sensors.
Given the above-mentioned advantages and consideration, in this study, the Cu-MOF was synthesized through the coordination of copper ions and homophenolic acid, followed by the calcination in the furnace at different temperatures (350 oC, 400 oC, and 450 oC) to receive the CuO/C core-shell nanoparticles with oxygen vacancy. The thermal treatment of Cu-MOF under different temperature were further found to create active sites and increase the oxygen-vacancy for the electrocatalytic oxidation of glucose. Compared with CuO/C-350oC and CuO/C-450oC, the developed CuO/C-400 °C has the most oxygen-vacancy and the highest response toward glucose oxidation in primary media. The wide detection range of glucose was explored using CuO/C-400 °C modified glassy carbon electrode (GCE) by amperometry under an optimal applied potential at 0.5 V due to the oxygen-vacancy and adsorbed hydroxyl ions. The sensing performance was then verified in artificial serum/saliva and human blood sample in real time analysis with remarkable reproducibility.