Dual-emission CdTe/AgInS2 photoluminescence probe coupled to neural network data processing for the simultaneous determination of folic acid and iron (II)

• Type: Article in International Scientific Journal
• Date: 2020

Authors

Castro, RC (Author) Other; Ribeiro, DSM (Author) Other; Pascoa, RNMJ (Author) Other; Jose X Soares (Author) FFUP; Mazivila, SJ (Author) Other; Joao L M Santos (Author) FFUP

Journal

Title: Analytica Chimica Acta

Vol. 1114

Pages: 29-41

ISSN: 0003-2670

Publisher: Elsevier

Indexing

ISI Web of Knowledge - 0 Citations

Scopus - 3 Citations

Other information

Authenticus ID: P-00S-09P

DOI: 10.1016/j.aca.2020.04.007

Abstract (EN): This work focused on the combination of CdTe and AgInS2 quantum dots in a dual-emission nanoprobe for the simultaneous determination of folic acid and Fe(II) in pharmaceutical formulations. The surface chemistry of the used QDs was amended with suitable capping ligands to obtain appropriate reactivity in terms of selectivity and sensitivity towards the target analytes. The implementation of PL-based sensing schemes combining multiple QDs of different nature, excited at the same wavelength and emitting at different ones, allowed to obtain a specific analyte-response profile. The first-order fluorescence data obtained from the whole emission spectra of the CdTe/AgInS2 combined nanoprobe upon interaction with folic acid and Fe(II) were processed by using chemometric tools, namely partial least-squares (PLS) and artificial neural network (ANN). This enabled to circumvent the selectivity issues commonly associated with the use of QDs prone to indiscriminate interaction with multiple species, which impair reliable and accurate quantification in complex matrices samples. ANN demonstrated to be the most efficient chemometric model for the simultaneous determination of both analytes in binary mixtures and pharmaceutical formulations due to the non-linear relationship between analyte concentration and fluorescence data that it could handle. The R-p(2) and SEP% obtained for both analytes quantification in pharmaceutical formulations through ANN modelling ranged from 0.92 to 0.99 and 5.7-9.1%, respectively. The obtained results revealed that the developed approach is able to quantify, with high reliability and accuracy, more than one analyte in complex mixtures and real samples with pharmaceutical interest.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 13