**4. Discussion**

The previously published studies carried out in controlled conditions, such as Petri dishes and hydroponics, reported that the toxicity of nanomaterials in the initial development stages of plant growth could be due to physicochemical properties, as well as particle size and shape [40,41]. In general, MeNPs show early negative consequences on the development stages of crops and this observation is confirmed in some publications [7,8,42,43].

Literature papers sugges<sup>t</sup> that *n*CeO2 generally enters plants through root uptake and may cause several effects on the early stages of plant development, such as reducing or increasing germination rates and improving, reducing or inhibiting radical growth [44,45]. When germinating seeds are exposed to nanoparticles, different effects could be verified, basically depending on the plant species and particle size and concentration [8].

It was demonstrated that *n*CeO2 having a diameter comprised in the range 50–100 nm are taken up by roots, but they hardly move towards the aerial plant fractions, while *n*CeO2 larger than 100 nm is not absorbed by roots [46,47]. We observed that the formation of particle agglomerates concerns, in particular *n*CeO2 25 nm. It is very likely that this was due to the higher specific surface than *n*CeO2 50 nm. At the same time, sp–ICP–MS analysis showed the largest number of peaks in all seedlings treated with *n*CeO2 25. Confirming previous literature findings [48], this suggests that the smaller particle size has the ability to enter into the roots more easily than 50 nm. Combining the previous evidence, we hypothesize that the *n*CeO2 25 nm agglomeration occurred inside the seedling tissues.

According to Layet et al. (2017) [49], we demonstrate that the two dicotyledons take up more *n*CeO2 than *H. lanatus*. Since seedlings of the different species growing in the same conditions, it is likely that the uptake and translocation of *n*CeO2 are influenced by species-specific physiological traits. The ability of *n*CeO2 uptake by roots and subsequent translocation to the other parts of plants was already demonstrated in crop species [50–52]. We observed a similar particle size distribution for *n*CeO2 25 nm in *L. flos-cuculi* and *D. tenuifolia*. At the same time, no relevant changes were observed for *n*CeO2 50 nm.

We recorded a negligible dissolved concentration of Ce ions in all samples, indicating that *n*CeO2 did not undergo dissolution after being absorbed by roots. On the other hand, we observed that small signals of dissolved forms of Ce correspond to the presence of bigger nanoparticles (50 nm), as previously reported [52,53]. Hence, it is likely that *n*CeO2 25 nm after being taken up by the seedling roots move through the vascular system forming aggregates. This process has been explained by the attraction between nanoparticles caused by van der Waals forces or chemical bonds [54,55]. However, since this also occurs within the plant tissues, it is still unclear whether and how species-specific factors can influence this process.

A large number of studies highlight that in some plant species, particle agglomeration happens before the passage from roots to the other parts of seedlings [42,56–58]. This statement could be justified by the plausible hypothesis that MeNPs pass through the apoplastic pathway [59] or cause the destruction of some cell walls, and in so doing, they pass through the enlarged pores [60]. However, nanoparticles could enter the vascular system where the Casparian strip is not formed [61,62] or through the lateral root junction [25,61,63,64].

Our data also indicated that the treatments with *n*CeO2 of different size influenced the size distribution of nanoparticles within the plant tissues. This could be due to the smaller size of the materials that lead to an exponential increase in surface area relative to volume in contact with roots. We can conclude that the *n*CeO2 entered inside the seedlings, and therefore, the results that will be described later are reasonably influenced by the experimental treatments. With regard to *n*CeO2 aggregation, in this study, we did not develop further observations.

With regard to the observed stimulating effect on germination of *n*CeO2, it must be said that on this point, the literature reports conflicting data. Low toxicity and reduction of seed germination were observed on *Lycopersicum esculentum* and *Zea may*s [44] and *Glycine max* [65], whereas germination of *Hordeum vulgare* was indifferent up to 2000 mg L−<sup>1</sup> *n*CeO2 [66]. As a matter of principle, a direct comparison of data from different experiments is difficult. However, if we were looking to draw a conclusion from available literature data on this point, we can say that *n*CeO2 does not cause acute toxicity in the early stages of plant development.

Leaving the nanoscale, it can be confirmed that Ce influence positively seeds germination. As for other rare earth elements, it has been suggested that Ce may have a positive effect by enhancing the effects of phytohormones on germinating seeds [67,68]. In addition, it seems that eventually, monocots species are more tolerant of Ce than dicot ones [69].

Contradictory literature evidence regards the influence of *n*CeO2 on root development in different stages of plant growth. Positive effects on root growth were observed in *Zea mays*, *Cucumis sativus* and *Lactuca sativa* [44,59]. As occurred in our species, very high tolerance to *n*CeO2 was reported for *Cucumis sativus*, *Brassica oleracea*, *Brassica napus* and *Raphanus sativus*, whose root growth resulted not affected up to *n*CeO2 2000 mg L−1. Oppositely, a slowed root development in treated *Medicago sativa* and *Lycopersicum esculentum* was reported [44].

As regards the Ce concentration (or mass per volume) in the seedlings exposed to the two dispersions, it is possible to conclude that the amount of Ce is greater for *n*CeO2 25 nm. Actually, this is not very informative in terms of the number of nanoparticles inside the plants if we consider the fact that the two dispersions had the same quantity of Ce in mass but different amounts in terms of nanoparticles. If we assume that the nanoparticles of the two dispersions are of the same shape and that the frequency of the dimension is 100% for both the dimensional class (25 and 50 nm), then we can conclude that the seedlings exposed to the *n*CeO2 25 nm were in fact exposed to 8 times the number of nanoparticles when compared to the 50 nm dispersion, since the volume goes with the cubic pattern.

Following this reasoning and assuming that the cell membrane pore size is not limiting the entrance of the nanoparticles because the mean size is greater than the sizes of the nanoparticles [70,71], it can be shown that the ratio between the number of nanoparticles taken up by the plant, calculated from the mass, is not in a ratio 1 to 8 as it would be if the two types of nanoparticles were taken up at the same level, but it is 1 to a greater number. This can lead us to speculate that the number of 50 nm nanoparticles inside the seedlings is lower than expected, and this could be due to the aggregation of the nanoparticles outside the plants that limited the entrance of the 50 nm nanoparticles more than the 25 nm nanoparticles despite their higher tendency to form clusters. On the other side, we may assume that the pores of the cell membrane act as filters and so explain why the *n*CeO2 50 nm reached a lower concentration in the seedlings. Of course, this is based on some assumptions that are not likely to occur in real conditions (the shape and size of the nanoparticle are far from being homogeneous and aggregation occurs). In addition, in more complex conditions, such as at field condition or in experiments that use real soil, the strategies that plants use to absorb nutrients from the substrate can influence, for instance, the solubility of the nanoparticles (e.g., by acidification of the rhizosphere which may change the solubility of the different chemical forms of Ce).
