Titanium dioxide (TiO₂) is a transition metal oxide widely used as a white pigment in various applications, including food. Due to the classification of TiO₂ nanoparticles by the International Agency for Research on Cancer as potentially harmful for humans by inhalation, the presence of nanoparticles in food products needed to be confirmed by a set of independent studies. Seven samples of food-grade TiO₂ (E171) were extensively characterised for their size distribution, crystallinity and surface properties by the currently recommended methods. All investigated E171 samples contained a fraction of nanoparticles, however, below the threshold defining the labelling of nanomaterial. On the basis of these results and a statistical analysis, E171 food-grade TiO₂ totally differs from the reference material P25, confirming the few published data on this kind of particle. Therefore, the reference material P25 does not appear to be the most suitable model to study the fate of food-grade TiO₂ in the gastrointestinal tract. The criteria currently to obtain a representative food-grade sample of TiO₂ are the following: (1) crystalline-phase anatase, (2) a powder with an isoelectric point very close to 4.1, (3) a fraction of nanoparticles comprised between 15% and 45%, and (4) a low specific surface area around 10 m² g^–1. GRAPHICAL ABSTRACT

Titanium dioxide is a metal oxide used as a white pigment in many food categories, including confectionery. Due to differences in the mass fraction of nanoparticles contained in TiO₂, the estimated intakes of TiO₂ nanoparticles differ by a factor of 10 in the literature. To resolve this problem, a better estimation of the mass of nanoparticles present in food products is needed. In this study, we focused our efforts on chewing gum, which is one of the food products contributing most to the intake of TiO₂. The coatings of four kinds of chewing gum, where the presence of TiO₂ was confirmed by Raman spectroscopy, were extracted in aqueous phases. The extracts were analysed by transmission electron microscopy (TEM), X-ray diffraction, Fourier Transform Raman spectroscopy, and inductively coupled plasma atomic emission spectroscopy (ICP-AES) to establish their chemical composition, crystallinity and size distribution. The coatings of the four chewing gums differ chemically from each other, and more specifically the amount of TiO₂ varies from one coating to another. TiO₂ particles constitute the entire coating of some chewing gums, whereas for others, TiO₂ particles are embedded in an organic matrix and/or mixed with minerals like calcium carbonate, talc, or magnesium silicate. We found 1.1 ± 0.3 to 17.3 ± 0.9 mg TiO₂ particles per piece of chewing gum, with a mean diameter of 135 ± 42 nm. TiO₂ nanoparticles account for 19 ± 4% of all particles, which represents a mass fraction of 4.2 ± 0.1% on average. The intake of nanoparticles is thus highly dependent on the kind of chewing gum, with an estimated range extending from 0.04 ± 0.01 to 0.81 ± 0.04 mg of nano-TiO₂ per piece of chewing gum. These data should serve to refine the exposure scenario.