An anomalous enhancement of fluorescence of carbon dots (CDs) was observed via lanthanum (La) doping. La-doped CDs (La-CDs) were prepared through microwave pyrolysis within 4 min. With La³⁺ doping, the emission band shifted from blue to green although La³⁺ is non-fluorescent. The quantum yield and fluorescence lifetime improved by about 20% and 35%, respectively. All experiment results indicate that La³⁺ doping is an effective way to tune fluorescence and improve the performance of CDs. Another unique attribute of La-CDs is high sensitivity to Fe³⁺. The La-CD-based fluorescence probe was established and used for sensitive and selective detection of Fe³⁺ with a limit of detection of 91 nmol/L. The proposed fluorescence probe also was successfully employed to visualize intracellular Fe³⁺ in live HeLa cells through cell imaging. It was also shown that yttrium exhibited the same fluorescence enhancement effect as La. The results may provide a new route for preparing CDs with special properties.

Analyte oriented selection of the preparation procedure is essential for improving the sensing performance. Herein, we report a one-step approach to synthesize highly fluorescent polymeric nanoparticles through a solvothermal method and their application for sensitive detection of TNT in soil samples by a fluorescence spectrometric method. The polymeric nanoparticles with bright cyan fluorescence were obtained through polycondensation between melamine and terephthalaldehyde at 180 °C for 3 h. The nanoparticles were fully characterized and the effect of synthesis methods was compared based on the fluorescence intensity and quantum yield (QY). The results indicated that trace amounts of TNT could effectively quench the fluorescence of the polymeric nanoparticles. The effect of the possible co-existing substances was systematically investigated and it was proved that nanoparticles could be used for selective and robust detection of TNT. A limit of detection of 1.8 nmol L⁻¹ was achieved. The standard addition recoveries for different soil samples were in the range of 95.3% and 106.4% with relative standard deviations below 5.2%.