Machine learning-aided Cu3(HHTP)2/COOH-MWCNTs electrochemical sensing platform for simultaneous detection of dopamine and uric acid in human serum.

The simultaneous determination of dopamine and uric acid in complicated biological matrices remains challenging due to overlapping electrochemical signals and interference from coexisting species. Herein, a machine learning-assisted electrochemical sensor based on a Cu3(HHTP)2/COOH-MWCNTs hybrid is developed for the simultaneous determination of these analytes. Benefiting from the unique π-π stacking and π-d conjugation characteristics of Cu3(HHTP)2, as well as the excellent coordination effect between COOH-MWCNTs and Cu2+ and the good dispersibility of carbon nanotubes themselves, the composite material exhibits more prominent surface activity, significantly boosting the exposure of active sites and accelerating the electron transfer kinetics at the electrode-electrolyte interface. On this basis, we constructed an electrochemical sensing platform for the simultaneous detection of DA and UA in human serum samples. The sensor show excellent performance with dual linear ranges of 0.5-15 and 15-80 μM for DA, and 1-200 and 200-350 μM for UA, achieving low LODs of 0.039 μM and 0.083 μM, respectively (S/N = 3). Furthermore, machine learning (ML) algorithms were employed to optimize electrochemical signal processing, enabling more accurate and robust analysis of complex data. The proposed approach achieved successful application in the simultaneous determination of DA and UA in human serum matrix samples.