On-site semi-quantitative analysis for ammonium nitrate detection using digital image colourimetry

Digital image colourimetry was successfully applied in the semi-quantitative analysis of ammonium nitrate using Griess’s test with zinc reduction. A custom-built detection box was developed to enable reproducible lighting of samples, and was used with the built-in webcams of a netbook and an ultrabook for on-site detection. The webcams were used for colour imaging of chemical reaction products in the samples, while the netbook was used for on-site colour analysis. The analytical performance was compared to a commercial external webcam and a digital single-lens reflex (DSLR) camera. The relationship between Red-Green-Blue intensities and ammonium nitrate concentration was investigated. The green channel intensity (I_G) was the most sensitive for the pink-violet products from ammonium nitrate that revealed a spectrometric absorption peak at 546 nm. A wide linear range (5 to 250mgL^− 1) with a high sensitivity was obtained with the built-in webcam of the ultrabook. A considerably lower detection limit (1.34 ± 0.05mgL^− 1) was also obtained using the ultrabook, in comparison with the netbook (2.6 ± 0.2mgL^− 1), the external web cam (3.4 ± 0.1mgL^− 1) and the DSLR (8.0 ± 0.5mgL^− 1). The best inter-day precision (over 3 days) was obtained with the external webcam (0.40 to 1.34%RSD), while the netbook and the ultrabook had 0.52 to 3.62% and 1.25 to 4.99% RSDs, respectively. The relative errors were + 3.6, + 5.6 and − 7.1%, on analysing standard ammonium nitrate solutions of known concentration using I_G, for the ultrabook, the external webcam, and the netbook, respectively, while the DSLR gave − 4.4% relative error. However, the I_G of the pink-violet reaction product suffers from interference by soil, so that blank subtraction (|I_G-I_Gblank| or |A_G-A_Gblank|) is recommended for soil sample analysis. This method also gave very good accuracies of –0.11 to − 5.61% for spiked soil samples and the results presented for five seized samples showed good correlations between the various imaging devices and spectrophotometer used to determine ammonium nitrate concentrations. Five post-blast soil samples were also analysed and pink-violet product were observed using Griess’s test without zinc reduction indicating the absence of ammonium nitrate. This demonstrates significant potential for practical and accurate on-site semi-quantitative determinations of ammonium nitrate concentration.