Elemental analysis by EDX
EDX can determine the elements present, but not the bonding conditions. We detected Ca, Cl, Cu, Fe, K, Mg, P+Pt, S, Si and Zn in the teeth (see Figure 5). The content of each individual element, except Fe, showed highly significant differences between the inner structure and the surface (results from Wilcoxon-test: p<.0001*, for p-values see Supplementary Table 1).
In the inner tooth structure, the following proportions were found (sorted from high to low mean content): P+Pt (mean ± standard deviation: 0.56 ± 0.26 atomic %), Fe (0.53 ± 0.35), Ca (0.47 ± 0.23), S (0.34 ± 0.12), Mg (0.26 ± 0.09), Zn (0.06 ± 0.03), Cu (0.05 ± 0.04), Si (0.04 ± 0.13), Cl (0.03 ± 0.03) and K (0.03 ± 0.02). All lateral teeth showed a rather small proportion of these elements in the inner tooth structure, compared to the surface (see below) (see Supplementary Table 2 for elemental content in the inner structure of the different teeth). Between the tooth types, we could not detect significant differences by pairwise comparison for the individual elemental content (see Supplementary Table 3 for p-values).
In the surface layer, the following proportions were detected (sorted from high to low mean content): Si (mean ± standard deviation: 5.79 ± 6.40 atomic %), Ca (2.28 ± 2.28), P+Pt (0.85 ± 0.60), S (0.67 ± 0.30), Fe (0.60 ± 0.72), Mg (0.34 ± 0.20), Cu (0.14 ± 0.11), Zn (0.12 ± 0.08), Cl (0.08 ± 0.08) and K (0.05 ± 0.03). In each tooth, especially Si and Ca were present in larger proportions in comparison to the other elements (see Supplementary Table 2 for elemental content of the surfaces).
When the results from the outer tooth surface were sorted according to the tooth region (see Figure 2 for nomenclature and 6 for results), we found that the cusps contained highest content of all elements (Ae), followed by the styli (terminal), the styli (bases), the sides, the bases, and finally the bulges with the lowest content of Ae (see Supplementary Table 4). This gradient in Ae is primarily due to the content and distribution of Si (at cusps, mean ± standard deviation: 16.24 ± 2.76 atomic %; at bulges: 0.27 ± 0.14) and Ca (at cusps: 5.56 ± 2.01; at bulges: 0.05 ± 0.05). But the Fe content, which was present in lower proportions, probably also contributes to the gradient, (at cusps: 1.26 ± 0.75; at bulges: 0.09 ± 0.07) (see Supplementary Table 4 for elemental content). The decrease of Si, Ca and Fe from tip to base of the teeth were found in each tooth type (see Supplementary Table 5 for elemental content of the different teeth and their regions) and differences when comparing the various regions of each tooth were highly significant (see Supplementary Table 6 for p-values). For the distribution of Cl, Cu, K, Mg, P+Pt, S and Zn no clear gradient could be detected, even though most regions differed highly significantly as determined by pairwise comparison (see Supplementary Table 6 for p-values). These elements were present in rather small proportions in each tooth (mean is <1 atomic %; see Supplementary Tables 2 and 4). When comparing the different teeth, we could not determine clear differences between them (see Supplementary Figure 2). In most cases we detected highly significant differences between the individual regions of the surface when comparing the different teeth by pairwise comparison for Si, Ca and Fe (see Supplementary Tables 3 for p-values).