Fluorescence ‘turn-on’ dual sensor for the selective detection of Al3+ and Zn2+ and the use of AI-based soft computing to predict machine learning outcomes

Abstract

A phenolphthalein-based fluorescence probe, N,N-(3-oxo-1,3-dihydroisobenzofuran-1,1-diyl)bis(6-hydroxy-3,1-phenylene)bis(3-methyl-1H-pyrazole-5-carbohydrazide) (PHP), was synthesized via a straightforward reaction. Intriguingly, the probe acts as a fluorescence ‘turn on’ dual chemosensor for Zn²⁺ and Al³⁺ with superb selectivity and sensitivity through selective “turn-on” fluorescence responses arising from a well-separated emission band based on its promising CHEF feature. The fluorescence intensities of the PHP–Al³⁺ and PHP–Zn²⁺ complexes at 441 and 472 nm increased in the presence of Al³⁺ and Zn²⁺, respectively, upon excitation at 370 nm. Job's plot revealed the binding stoichiometry of the probe (PHP) with both metal ions (Al³⁺ and Zn²⁺), which was determined to be 1 : 2 (PHP : Mⁿ⁺) in each case. The LOD values for Al³⁺ and Zn²⁺ were found to be 0.28 μM and 62.9 nM, respectively. The PHP–Al³⁺ complex showed a selective fluorescence ‘turn off’ response towards fluoride ions (F⁻) in solution, and this anion-responsive behaviour of the PHP–Al³⁺ complex was utilized to mimic numerous logic gates and FL functions. To avoid time-consuming, extensive experimental techniques, machine-learning soft computing tools, such as fuzzy logic, artificial neural networks (ANNs), and adaptive neuro-fuzzy inference systems (ANFIS), were used to predict the possible experimental emission intensities of the probe in the presence of Al³⁺ and F⁻.