Graphene quantum dots with tunable photoluminescence and hydrophobicity were synthesized from an abundant natural carbon source containing nitrogen, sulfur, and oxygen heteroatoms.

Sensing and detecting nitroaromatics (NAs) are essential for environmental, health, and safety reasons. Graphene quantum dots (GQDs) respond to the presence of NAs by a well-understood fluorescence quenching mechanism. However, despite the relative simplicity of fluorescence-based sensing, the limit of detection (LoD) can compare unfavorably with other methods. Here, we show that the LoD for sensors based on GQDs can be lowered by orders of magnitude using a droplet-based analyte partitioning effect. While previous efforts have attempted to improve the intrinsic GQD sensitivity via surface functionalization and size control, we show that a major improvement can be attained by changing from a bulk solution to droplet-based sensing of 2,4-dinitrotoluene and nitrobenzene. Moreover, the method is compatible with sensing from an aqueous solvent and has broader implications for many fluorescence-quenching-based sensing strategies that could benefit from partition-related enhancements.