Sacco, A; Mandrile, L; Tay, L-L; Itoh, N; Raj, A; Moure, A; Del Campo, A; Fernandez, J F; Paton, K R; Wood, S; Kwan, H; Adel, T; Hight Walker, A R; Ferreira, E R M; Theissmann, R; Koch, T; Giovannozzi, A M; Portesi, C; Rossi, A M (2023) Quantification of titanium dioxide (TiO2) anatase and rutile polymorphs in binary mixtures by Raman spectroscopy: an interlaboratory comparison. Metrologia, 60 (5). 055011

Preview

Text eid9923.pdf - Published Version Available under License Creative Commons Attribution. Download (800kB) | Preview

Abstract

Titanium dioxide is used as white pigment in varous application. Besides plastics, coatings and paper also food (E171), pharma and cosmetics, even photocatalytic applications like sunscreen or other photocatalytic active coatings use TiO2. In nature as well as in its industrial application, TiO2 occurs mainly in two polymorphs: anatase and rutile; these forms differ in physical and chemical properties, industrial applications, stability, and market price. Some standard methods are currently employed internationally for the determination of TiO­2 content in samples (ISO 591−1, ASTM D3720−90), but they require extensive preparation, and do not discern between the two forms or are accurate only for small fractions of either polymorph. Raman spectroscopy is a well-suited chemical characterization technique for measuring TiO2 in a fast, non-invasive way because of its ease-of-use and the large cross section of the material, while requiring no particular reagent or sample preparation.

In this work, an interlaboratory comparison (ILC) on TiO2 polymorph quantification of commercial anatase/rutile binary mixtures with 11 international participants, conducted in the framework of Versailles Project on Advanced Materials and Standards (VAMAS), is presented. Diverse Raman spectroscopy setups are used by the participating laboratories. Samples were prepared by the lead participant (LP) laboratory and sent to all participants. The LP analyzed the collected data by means of partial least squares (PLS) regression after spectral preprocessing. The resulting models all show a bias better than 2 % in the quantitative analysis over the concentration range of 5 % to 95 % mixture ratios, with many datasets suggesting substantial improvement margins on this figure. The results of this ILC provide validation of Raman spectroscopy as a reliable method for TiO2 phases quantification with important industrial and academic applications worldwide.

Item Type:	Article
Keywords:	interlaboratory comparison, titanium dioxide, Raman spectroscopy, partial least squares, machine learning
Subjects:	Advanced Materials > Functional Materials
Divisions:	Electromagnetic & Electrochemical Technologies
Identification number/DOI:	10.1088/1681-7575/acf76d
Last Modified:	15 Feb 2024 10:52
URI:	http://eprintspublications.npl.co.uk/id/eprint/9923

Actions (login required)

View Item