Detection of ninhydrin-glyphosate in groundwater via the colour chart-assisted digital camera method

This study introduces a novel application of a one-time and multistandard colour chart across three instruments using colour visible spectrophotometer (colour VS), digital single-lens reflex (DSLR) cameras, and mobile camera (MC) analysis to quantify glyphosate (GP) in groundwater samples. The image captured by the digital cameras and the absorbance of the colour VS data of the ninhydrin-GP complexation reaction of the standards were detected by the colour space of the human eye model. The colour ranges were from yellow to brown (colour VS) and red (digital cameras). The results of MC provided enhanced luminance and appropriate ΔE levels, with similar relative standard deviations (RSD) to those of colour VS. The deviations for DSLR were greater and showed better theoretical detection levels. The recovery yield analysis of real-time ground water samples at 50-500 ng/mL in 75 mL was greater for the DSLR (116 %). This method was better than liquid chromatography-tandem mass spectrometry (LC-MS/MS) and MC analysis. However, the MC and LC-MS/MS methods were better in terms of accuracy (98.9 % and 84 % respectively) and precision (99.9 % for both) closeness in a 1:1 linear analytical comparison. The DSLR overestimated the concentrations due to ion interference. However, MC showed no interfering ions at 500 µg/mL due to its unique LED flash light, unlike colour VS, which was affected by phosphate and nitrate. This study was able to achieve real-time monitoring via cost-effective, user-friendly point-of-care technology, eliminating the need for expensive sensor fabrication and repeated standardisation methods. This might pave the way for possible integration of artificial intelligence, machine learning and deep learning models using colour chart data. Therefore, the our multistandard colour chart may also assist in standardizing other camera sensors through harmonization approach.