Machine learning-assisted smartphone platforms for carbon quantum dot interfaces in optical chemical sensing

Recent developments in carbon quantum dot (CQD) nanomaterials and smartphone-based optical sensing have enabled new directions in portable chemical analysis. In this review, the relationships between CQD structural characteristics, surface chemistry, and photophysical behavior are examined in the context of fluorescence-based sensing. Emphasis is placed on how synthetic strategies, heteroatom doping, and surface functionalization influence emission properties, analyte recognition, and signal stability through interfacial interactions. Key photophysical mechanisms, including electron transfer, energy transfer, and fluorescence quenching, are discussed to clarify their roles in signal transduction within CQD-based sensing systems. The complexity of optical signal acquisition using smartphone cameras is also analyzed, highlighting the effects of device-dependent imaging pipelines, illumination variability, and environmental conditions on analytical performance. Machine learning approaches are further considered for extracting multidimensional features from imaging data, enabling improved calibration, pattern recognition, and robustness in non-linear sensing environments. Reported applications, including bacterial identification, latent fingerprint visualization, epigenetic biomarker detection, and trace ion monitoring, are evaluated to illustrate the analytical capabilities of CQD–smartphone sensing platforms.