SHARP: Signal-to-noise heightening via asynchronous reaction partitioning in time domain for ultrasensitive and reliable immunoassay of dibutyl phthalate

Dibutyl phthalate (DBP), a major co-contamination leaching from microplastics, is recognized as a typical endocrine disruptor. Its widespread environmental occurrence and pronounced low-dose effects pose a serious threat to public health. Consequently, an ultrasensitive detection method applicable to real samples is urgently required. Although immunoassays offer a viable solution, existing immunoassays are frequently compromised by non-specific adsorption in real samples, leading to severe signal-to-noise deterioration. To address this issue, a Signal-to-noise Heightening via Asynchronous Reaction Partitioning in time domain (SHARP) strategy is proposed. In this strategy, an optimal reaction time window is precisely identified by exploiting the kinetic difference of tyramine signal amplification, within which the specific signal is fully amplified while the non-specific background remains negligible. Based on this SHARP mechanism, a competitive ELISA (SHARP-cELISA) is constructed for DBP detection, employing tyramine-functionalized rare-earth fluorescent nanoparticle (NaYF4:Ce,Tb) as the signal label. Furthermore, Gaussian process regression (GPR) machine learning is introduced for nonparametric modeling of competitive inhibition curve, and its confidence intervals provide statistical support for the reliability of low-concentration samples. The SHARP-cELISA achieves a limit of detection of 5.4 pg/mL and spike recoveries of 94.2%-102.7% in water samples. This method offers a novel paradigm for overcoming non-specific adsorption from the perspective of time-domain partitioning.