Measuring vitamin B12 in food usually means bringing the sample to the laboratory. A new fluorescent sensor could help move part of that process onto something much simpler: a paper strip or cotton swab. A new study published in Biomedical Analysis describes a fluorescence-based strategy for portable vitamin B12 (VB12) detection in food samples. Researchers at Liaocheng University developed boron- and nitrogen-co-doped carbon dots (B, N-CDs) that emit bright blue fluorescence under ultraviolet light but become markedly dimmer when VB12 is present. By transferring this sensing chemistry from a liquid assay onto filter-paper strips and a specially designed cotton-swab sensor, the team created two simple formats intended to make the detection step faster, cheaper, and easier to use outside conventional instrument-heavy workflows.
Common methods for measuring VB12, including chromatography and electrochemical analysis, can require complex instruments, lengthy sample handling, or carefully controlled conditions. Many fluorescence probes also remain tied to laboratory spectrometers. To address these limitations, the researchers synthesized B, N-CDs through a one-step hydrothermal process and evaluated their response to VB12. The probe showed a linear fluorescence response from 0 to 100 μM and a detection limit of 0.02 μM. In selectivity tests, VB12 produced pronounced fluorescence quenching, while other tested B vitamins, amino acids, glucose, and common small molecules caused little comparable response.
Engineering a Fluorescent Signal for Vitamin B12
The B, N-CDs were prepared from 3-aminophenylboronic acid and p-hydroxybenzaldehyde and formed water-dispersible nanoparticles averaging about 1 nm in diameter. They showed their strongest fluorescence emission at 475 nm under 390 nm excitation, with a quantum yield of 18.15%. Mechanistic experiments indicated that VB12 associates with the carbon dots to form a non-luminescent ground-state complex, leading to static fluorescence quenching. Fluorescence-lifetime, UV-visible absorption, Raman, FT-IR, XPS, particle-size, and surface-charge measurements all supported this interaction, helping explain why the probe's blue fluorescence decreases as VB12 concentration rises.
From a Fluorescent Probe to Paper Strips and Swabs
To translate the sensing response into portable formats, the team impregnated filter paper with B, N-CDs and designed a cotton-swab device that delivers the fluorescent probe to a sensing region during use. Under 365 nm UV light, both formats became progressively darker as VB12 concentration increased. For the paper strip, image-based analysis of the green-to-blue (G/B) intensity ratio showed a strong linear relationship with VB12 from 0 to 80 μM (R² = 0.9927). The swab format was designed to combine sample collection, transfer, mixing, and signal readout in one disposable device, and the sensing step can be completed rapidly once a suitable sample is available. Both the strips and swabs retained useful fluorescence and VB12-sensing performance for at least 20 days under the storage condition evaluated.
Checking Accuracy in Real Food Samples
The researchers tested the analytical method using egg-yolk and goat-liver-powder samples and compared quantitative measurements with high-performance liquid chromatography (HPLC). In spiked recovery experiments, recoveries ranged from 97.09% to 101.22%, and the measured VB12 concentrations closely matched the HPLC results, supporting the method's accuracy in the food matrices studied.
The work shows how carbon-dot fluorescence can be adapted into visually readable portable sensors without sacrificing sensitivity or selectivity. The authors also outline practical next steps, including replacing the laboratory UV lamp with a compact battery-powered UV-LED and adding smartphone-based image analysis for automated quantification. The current real-food workflow still involves laboratory sample pretreatment before measurement, so further simplification will be important for truly field-ready testing.
DOI: 10.1016/j.bioana.2026.08.003