Using the extract of pomegranate peel as a natural indicator for colorimetric detection and simultaneous determination of Fe3+ and Fe2+ by partial least squares–artificial neural network

• Published in: Journal of chemometrics 2023-01, Vol.37 (1), p.n/a
• Main Authors: Khani, Shokoofeh, Mohajer, Fatemeh, Mohammadi Ziarani, Ghodsi, Badiei, Alireza, Ghasemi, Jahan B.
• Format: Article
• Language: English
• Subjects: Absorption spectra; artificial neural network; Artificial neural networks; Biocompatibility; Calibration; Cations; chemometrics; Colorimetry; Complex formation; Detection limits; Drinking water; Fe3+ and Fe2+ sensor; Ferric ions; Ferrous ions; Functional groups; Least squares; Mixtures; Multivariate analysis; Neural networks; Organic constituents; partial least squares; Plant extracts; pomegranate; Speciation; Spectrophotometry; Waterfalls
• ISSN: 0886-9383; 1099-128X
• DOI: 10.1002/cem.3390

A simple, novel, biocompatible, and sensitive spectrophotometric method was developed for colorimetric detection and speciation of Fe3+ and Fe2+ ions. The method is based on the complex formation of Fe3+ and Fe2+ with organic constituents containing functional groups in pomegranate peel extract (PG). Developing of the color is specifically established in the presence of Fe3+ and Fe2+. Accordingly, in addition to direct detection of both cations, we carried out a quantitative multivariate model for their accurate determination in real matrix aqueous samples. Due to spectral interference, the simultaneous determination of Fe3+ and Fe2+ mixtures using spectrophotometry is a problematic issue. A combined multivariate as partial least squares (PLS)–artificial neural network (ANN) was used to create and adjust a model and predict said cations in test and real sample sets. In this context, the calibration model is based on absorption spectra in the 400–900 nm range for 98 different mixtures of Fe3+ and Fe2+. Calibration matrices contained 1–12 and 1–10 μg ml−1 of Fe3+ and Fe2+, respectively. The detection limits for Fe3+ and Fe2+ were 0.53 and 0.14 μg ml−1, respectively. The root mean square error of prediction (RMSEP) for Fe3+ and Fe2+ with PLS‐ANN was 0.78, 1.65 and 1.062, 0.894 in Kalugan waterfall and Tehran tap water. This strategy can simultaneously determine Fe3+ and Fe2+ in real matrix samples, and the reliability of the determination is acceptable. We proposed a new and green natural indicator for colorimetric detection and speciation of iron ions. The PLS‐ANN was applied for deconvolution of UV–vis spectra. The Fe3+ and Fe2+ were quantitatively determined in real samples. The introduced technique is biocompatible, inexpensive, simple, and fast.