Quantitative micro-XRF combined with X-ray imaging reveals correlations between Zn and dentin tubule porosity across entire teeth

Quantitative micro-XRF combined with X-ray imaging reveals correlations between Zn and dentin tubule porosity across entire teeth | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL View This is a preprint and has not been peer reviewed. Data may be preliminary. 9 October 2025 V1 Latest version Share on Quantitative micro-XRF combined with X-ray imaging reveals correlations between Zn and dentin tubule porosity across entire teeth Authors : Ioanna Mantouvalou 0000-0002-9696-2970 [email protected] , Leona Johanna Bauer , Vinh-Binh Truong , Yannick Wagener , Frank Förste , Oleksandra Marushchenko , Stephan Werner , Franco Lizzi , Frank Wieder , Timo Wolff , Birgit Kanngießer , and Paul Zaslansky Authors Info & Affiliations https://doi.org/10.22541/au.175999341.11496128/v1 Published VIEW Version of record Peer review timeline 308 views 156 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Bony materials are biogenic composites of protein fibers and mineral that create hierarchical structures. In the case of teeth, dentin is the main component and similar to other bones, it contains porosity at multiple length scales. It is traversed by micron-sized hollow channels known as dentinal tubules, essential for temperature and pain sensation. Tubule density and thus porosity vary throughout the macroscopic 3D structure, with porosity increasing towards the pulp. The different densities in teeth are easily revealed non-destructively in 3D by X-ray imaging using computer tomography (CT). Yet elemental composition analysis is more difficult to obtain from within the cm-sized heterogeneous bulk material. We describe an approach of merging CT measurements of young, healthy, intact bovine teeth with micro-X-ray fluorescence (micro-XRF) images of serially sectioned slices. Through the combination of multi-resolution CT measurements with elemental concentration quantification, gradients in density and element distributions such as calcium (Ca), phosphorus (P) and zinc (Zn) are revealed in 3D. While the main constituents (Ca and P) are homogeneously distributed in the structure, Zn concentration increases significantly and exponentially towards the pulp. We find an inverse association between dentin tissue density and Zn concentration localizing this element in or around tubules. Quantitative micro-XRF combined with X-ray imaging reveals correlations between Zn and dentin tubule porosity across entire teeth Ioanna Mantouvalou*, Leona Johanna Bauer, Vinh-Binh Truong, Yannick Wagener, Frank Förste, Oleksandra Marushchenko, Stephan Werner, Franco Lizzi, Frank Wieder, Timo Wolff, Birgit Kanngießer, Paul Zaslansky I. Mantouvalou, L.J. Bauer, O. Marushchenko, B. Kanngießer Department SyncLab, Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Berlin, Germany E-mail: [email protected] S. Werner Department X-ray microscopy, Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Berlin, Germany F. Wieder Division X-ray Imaging, Bundesanstalt für Materialforschung und -prüfung, 12205 Berlin, Germany T. Wolff Bruker Nano GmbH, Berlin, Germany I. Mantouvalou, L.J. Bauer, V.B. Truong, Y. Wagener, F. Förste, B. Kanngießer Institute for Physics and Astronomy, Technische Universität Berlin, Berlin, Germany F. Lizzi, P. Zaslansky Department for Operative, Preventive and Pediatric Dentistry, Charité - Universitätsmedizin Berlin, Berlin, Germany Funding: German Research foundation - project numbers 443841418 (IXdent) and 396127899 (InterDent – TP1) Keywords: dentin density, tubule porosity, Zn concentration, quantitative micro-X-ray fluorescence, micro-computer tomography Bony materials are biogenic composites of protein fibers and mineral that create hierarchical structures. In the case of teeth, dentin is the main component and similar to other bones, it contains porosity at multiple length scales. It is traversed by micron-sized hollow channels known as dentinal tubules, essential for temperature and pain sensation. Tubule density and thus porosity vary throughout the macroscopic 3D structure, with porosity increasing towards the pulp. The different densities in teeth are easily revealed non-destructively in 3D by X-ray imaging using computer tomography (CT). Yet elemental composition analysis is more difficult to obtain from within the cm-sized heterogeneous bulk material. We describe an approach of merging CT measurements of young, healthy, intact bovine teeth with micro-X-ray fluorescence (micro-XRF) images of serially sectioned slices. Through the combination of multi-resolution CT measurements with elemental concentration quantification, gradients in density and element distributions such as calcium (Ca), phosphorus (P) and zinc (Zn) are revealed in 3D. While the main constituents (Ca and P) are homogeneously distributed in the structure, Zn concentration increases significantly and exponentially towards the pulp. We find an inverse association between dentin tissue density and Zn concentration localizing this element in or around tubules. V.B. Truong: current address: Physikalisch-Technische Bundesanstalt, Berlin, Germany F. Wieder: current address: Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Berlin, Germany 1. Introduction All vertebrates (including mammals) possess an “endoskeleton” that is made of the material bone, a bio-composite of collagen fibers and apatite nanocrystals. This material has been the focus of research for centuries (see Reznikov et al. [1] ), in particular ever since the discovery of X-rays almost 150 years back. The geometry and characteristics of bone materials remain of high interest not only because of the central role of the skeleton in human health and disease, but also for engineers as bioinspiration for a robust, hierarchically organized [2] naturally grown structure that can adapt and change over time. While macroscopically the mineralized fiber composite supports mechanical tasks of the skeleton (e.g. defense), internally it contains water and biological components (enzymes, cells) within a myriad of cavities that are characteristic and a hallmark of different bony tissues. [3] Indeed, the porosity of bony materials, be it macroscopic (e.g. large marrow or tissue spaces), microscopic (blood or cell spaces), or nanoscale (sub-microscopic voids between collagen fiber bundles [1] ), is possibly almost as important for the normal function of the skeleton (mechanical properties, adaptation) as the mineralized fibers are. For example, bones serve as biochemical reservoirs. Many bones are osteocytic, i.e. they contain osteocyte cells that are interconnected by the so-called lacuna-canalicular network. This porous network of osteocytic bones enables mechano-sensation and material turnover, which is central to the adaptive response of many bones to load or damage by re/modeling. However, many bony materials do not contain osteocytes, and the design principles as well as the natural mechanisms of damage resistance and deformation response are the subject of ongoing investigations. [4–6] Dentin in teeth is one example of a bony material that is characteristic of vertebrate teeth. [7] In mammals, dentin shapes the roots that are anchored to the jawbones, while forming dome-like hollow structures that define tooth geometry and support enamel in the crown. Dentin is known for strength and toughness, [8,9] required for decades of tooth function without failure. Yet dentin lacks osteocyte cells and the capacity for tissue turnover or replacement even though it is highly porous due to micron-sized hollow channels known as dentinal tubules. These cavities traverse the entire dentin thickness, leading from the pulp tissue in the root/pulp chambers outwards, to the tooth periphery. As early as 1976, Garberoglio et al. [10] showed that, in human teeth, the dentinal tubules are denser and wider near the pulp. Indeed, in normal healthy mammalian dentin, tubules become thinner and more dispersed at increasing distances outward, towards the external tooth surfaces. The tubule porosity makes it possible for liquid to flow through dentin and to come into contact with odontoblasts, dentin forming cells, that reside in the pulp chamber, at the border between the pulp tissue and dentin. [11] Odontoblasts are thus not normally found inside the dentin matrix bulk. They are the cells responsible for the material creation during odontogenesis, where they create pre-dentin that mineralizes similar to osteoid in bone, produced by osteoblast cells. [12] In teeth, odontoblasts retain a tight contact to the dentin matrix that they created in the form of elongated cellular remnants, known as “odontoblast processes”, that are left within many dentinal tubules during dentin deposition. [13] Based on these structural observations, it may be understood that dentin porosity is not uniform, because it varies in different tooth regions. The porosity of dentin has been indirectly assessed by measuring the “hydraulic conductance” in cross sections of teeth, a parameter that is useful to quantify flow across dentin, for example of bacteria. [14] It is also a measure of the permeability, important for dental treatments such as bonding of biomaterials used for fillings. [15,16] In living (“vital”) teeth, the tubule-mediated porosity plays an important role in tooth pain sensation, enabling rapid liquid flow through tubules, [17] that may be converted into sensation via free nerve endings [18] or via other odontoblast responses. [19] One consequence of the varying porosity is that the density of the tissue is also not constant. In bony tissues, density is correlated to tissue stiffness [20] (though not strength [3] ). This is of significance also in teeth, where micro-computer tomography (micro-CT) measurements have demonstrated reduced absorption and lower tissue density near the pulp [21] which explains the reduced hardness in this area. [8] The signals that micro-CT measures are the product of the density of the material diluted by the presence of internal voids that may actually be smaller than the resolution of the measurement. Because of physical constraints in resolution and flux, lab measurements of dentin by micro-CT frequently cannot resolve the µm sized tubules. Under the assumption that the composition and density of the dentin matrix surrounding tubules is more-or-less constant, the tomographically determined density of dentin is to a first approximation reciprocal to the tubule porosity. Chemically, and just like all other bony materials, dentin is made of non-stochiometric carbonated apatite nanocrystals (~Ca 5 ((PO 4 ) 0.75 (CO 3 ) 0.25 ) 3 (OH)) and organic polymers, mainly collagen (~C 65 H 102 N 18 O 21 ). Thus, the elements Ca and P are the main heavy elements (with atomic numbers greater than 14) present in the matrix, though some Zn has been reported, alongside S. Light elements such as Mg as well as traces of F and Na are also found within the composite. With regards to Zn, early publications of its distribution in human dentin date back to the work of by Brudevold et al. in 1963. [22] They showed high concentrations of this element in outer enamel and they demonstrated a gradual increase of Zn in dentin towards the pulp chamber. Similar trends and increased Zn near the pulp were reported in a range of mammals [21,23] including marsupials. [24] Thus, in normal, untreated tooth dentin, an increase in Zn concentration in the bulk material is expected to be found towards the pulp of healthy teeth. Yet, at the micron level the composition of dentin in regard to Zn has not been quantified to the best of our knowledge, The increased Zn concentration and increased tubule porosity in dentin near the pulp raise the possibility that these phenomena are associated with one another, at least in natural, healthy, fully mineralized teeth. However, the presence of spatially varying porosity poses an experimental challenge when trying to quantify variations in chemical composition which is also related to the material density. Chemical composition of materials is usually measured on exposed surfaces that can easily be analyzed directly by high-resolution characterization methods such as scanning-electron microscopy coupled with energy-dispersive spectroscopy (SEM-EDX). However, such methods cannot reliably measure the bulk below the surface, as the chemical signal of EDX is limited to excitation of the upper ~1 µm of the material. X-ray methods are needed to extend this information into the full three dimensions (3D) especially because electron microscopy imaging of entire crowns and roots is usually not feasible. X-ray methods of imaging and chemical mapping are of high relevance, since X-rays can penetrate deep into bony materials, including teeth, and they can be easily adapted to measure and quantify gradients within cm-sized dimensions, such as the sizes typical of teeth. Micro-X-ray fluorescence spectroscopy (micro-XRF) is a quantitative non-destructive elemental imaging technique which relies on the detection of X-ray fluorescence following excitation with a focused X-ray beam. [25] The technique demands no specific sample preparation and can deliver elemental information originating from hundreds of micrometers deep in the bulk of composites such as dentin, due to penetration of the excited X-rays. [26] The information depth (the depth from which 1/e of the emitted fluorescence photons originate from) is density, energy- and consequently – element-specific. For example, in dentin, the signal detected for Ca arises from a depth of up to for Zn (K-radiation) emerges from depths of up to ~100 µm. [26] We tackle the challenge of quantifying elemental distributions, in particular of Zn, in the bony material dentin, taking into consideration the varying tubule density. We show the coupling of complementary X-ray imaging methods with chemical mapping of fully mineralized tooth dentin. The merging of data from CT and micro-XRF results in a detailed 3D picture of the main structural gradients of dentin, as observed in healthy intact teeth. Our approach makes it possible to quantify gradients and investigate the association between natural porosity and physiological Zn gradients within the volume of tooth material. 2. Results Figure 1: a) and b) High magnification BSE-SEM images collected from a polished bovine tooth slice, imaged both near the pulp and 1 mm away, respectively. Tubules appear as dark hollow voids. In b) many tubules are surrounded by bright rings of high-density peritubular dentin (PTD). c) Reconstructed virtual slice of a 1 µm thick dentin lamella, milled within a normal tooth by means of a focused ion-beam (FIB), imaged tomographically by X-ray transmission microscopy. d) A higher magnification image of intertubular dentin (ITD) in the area marked in blue in c) shows traces of the mineralized collagen fibers of ITD, observed at nm resolution between tubules. Figure 1 shows typical example backscattered scanning electron microscopy (BSE-SEM) images of cut and polished bovine dentin, revealing exposed tubules observed near the pulp (Figure 1a) as well as approximately 1 mm outwards (Figure 1b). Such images in a healthy juvenile untreated bovine molar demonstrate the increase in tubule porosity observed near the pulp. Due to differences in the electron backscatter signal, tubules appear as dark hollow voids within the mineralized dentin matrix (“intertubular dentin” - ITD). In crown dentin, about half-way between the pulp and enamel, many tubules become surrounded with a hypermineralized peritubular sheath [27,28] (“peritubular dentin” - PTD) that leads to a small increase in the mean tissue density. [29] The same tissue imaged by transmission X-ray microscopy (TXM) tomography reveals, at nm resolution and in 3D, the fibrous nature of ITD. Figure 1c shows a reconstructed slice within a 1 µm thick lamella (sized 14 µm x 17 µm in lateral dimensions) extracted ~90 µm away from the pulp (see SI, section 1, for details). At such proximity to the pulp, no PTD is present, and the more-or-less uniform density of the ITD is revealed, as demonstrated by a close-up image of the matrix between tubules (Figure 1d), in which traces of fibers are visible. While electron and transmission X-ray microscopy provide insightful high-resolution results, they are limited in revealing gradients in dentin density across large mm-cm sized samples. For such larger length scales, commercial lab micro-CT instruments provide 3D data with reasonable high spatial resolutions. Figure 2a depicts a 3D rendering of typical micro-CT data of one of several bovine teeth (T4) with a second rendering showing the same data with a virtual cutout that makes it possible to examine the internal pulp chamber surrounded by dentin. Slices along this tooth (e.g. panel c) reveal characteristics of such samples: The uppermost image shows density variations in a slice in the micro-CT data (gray value units). The virtual cut reveals high absorption in enamel (the diagonal upper right-side feature) and a moderate gradient in density (brightness) in dentin, observed when exploring the image information from the pulp in the center, outwards toward the external tooth surfaces (periphery). The three other figure panels show the results of micro-XRF mapping of a physical slice in this tooth, revealing the K-line intensities of the elements Ca, P and Zn. Whereas the micro-CT data represents the density in a 24 µm thick (virtual) slice in the tomography data, the micro-XRF information originates from different depths beneath the physically cut and polished tooth slice surface. The XRF signal for P and Ca (emerging from a thickness of up to 30 µm) is emitted from the mineral component of the dentin composite, probing a similar volume as the micro-CT data. This is because the excited XRF radiation for these elements is only detectable from rather shallow dentin layers. The Zn K fluorescence originates from greater depths beneath the slice surface (up to ~100 µm). The Ca and P K fluorescence intensities are high, corresponding to a strong signal from the high-prevalence dentin mineral, and both elements appear to be relatively homogeneous within dentin (and with both higher intensities also in enamel). Notably, the Ca to P ratio does not change in different dentin regions, suggesting a uniform composition of the apatite crystal stoichiometry. In contrast, the Zn signal is strongly enhanced toward the pulp. To that end, and even though mineral (Ca, P) is the main cause for absorption, it is curious to observe how none of the chemical maps matches the absorption distribution seen in the micro-CT (panel a) data. Close examination shows in fact that the Zn signal appears to be inversely proportional to the absorption revealed by the micro-CT data. Figure 2: Comparisons of micro-XRF and micro-CT data; micro-CT rendering of a full tooth (T4) (a) with virtual cutout exposing variable densities and internal structures (b). c) Micro-CT slice of the same volume showing the absorption in gray value units. Beneath it, the P, Ca, and Zn K-line fluorescence intensity distributions of a physical cut slice are shown. Fluorescence intensity values are given in counts per second (CPS), higher counts to a first approximation correlate with higher concentrations of the heavier elements. Micro-XRF measurements of multiple slices from root and crown regions of several teeth (T1-T4), show similar results. The Zn maps reveal a recurring pattern of an increase in Zn signal in dentin near the pulp. We used point-by-point XRF measurements to generate mean spectra (see Figure S2 in the SI) of all areas with similar Zn K-line intensities in layers with increasing distances from the pulp (Figure 3a). These spectra yield information from which it is possible to quantify the mass fractions of the main elements, Ca and P as well as the trace Zn. During quantification, the “dark matrix”, i.e. all elements, which are not detected in the XRF spectrum is taken into consideration. It typically includes light elements such as C, N and O where fluorescence detection is difficult due to extensive self-absorption and low fluorescence yield. The results of the quantification can be plotted on a topological scale (from the pulp region to the outermost periphery) as shown in Figure 3b demonstrating the high similarity observed between different teeth. Quantitative mass fraction evaluations reveal very similar results for all investigated spectra within ca. 10 % uncertainty, as described in detail in the methods section. Plotting the data in this manner makes it easy to compare the Ca and P distributions between different crown and root slices: it becomes apparent that the crowns and roots of different young healthy teeth are visibly very similar. Specifically, the major apatite mineral elements Ca and P mass fraction values only mildly differ between different distances to the pulp. This is surprising, because the micro-CT data clearly show that the near-pulp regions have a lower density. To understand this, it is necessary to recall that the information depth for XRF signals emitted by Ca and P is restricted, due to self-absorption by the tooth matter, to arise mostly from near-surface areas (~30 µm and ~10 µm beneath the tooth slice surface, [26] respectively). Within such thin layers there can only be a couple of tubules, so that variations in tubule density have negligible effects on the detected Ca and P signals. In contrast, the Zn mass fractions (with the XRF signal arising from an information depth of ~100 µm [26] ) shows a marked gradient across the tooth thickness. Our quantification reveals a tenfold increase in concentration from about 50 ppm in peripheral dentin regions, increasing up to 500 ppm near the pulp. This trend is independent of the tooth or the location of the slice in the tooth, and recurred in all our analyzed slices. Figure 3: a) Topological maps of the Zn K intensity used for the generation of mean spectra. Note that the tooth slices are not represented with a comparable size scale, for better clarity here. A correct size scale of these slices is provided in Figures 2 and 4. b) Mass fraction values for Ca, P and Zn as a function of the topological distance defined in a) derived through a quantification of the mean spectra. The quantitative XRF results yield a relationship between Zn K intensity and Zn mass fraction, which to a first approximation is linear. Thereby, the measured micro-XRF intensity images of the slices can be converted into mass fraction maps, yielding fully quantitative images of chemical composition of the different slices as presented in the bottom images of Figure 4a. The absolute tissue density values for these tooth slices were obtained by measuring calibration samples together with high resolution micro-CT measurements of the entire cut slices. For this purpose calibrated (density-determined) reference samples of dentin and enamel were prepared from a fifth tooth. A linear calibration curve utilizing dentin, enamel and the known density of supporting glass slides facilitated the generation of quantitative images of tissue density (top images of Figure 4a). Both calibration curves (for Zn and density) are shown in Figure S3 in the SI. Figure 4 compares slices across 3 teeth (crown and root) and shows averages of 5 virtual slices of cross-sections in the calibrated micro-CT data, displayed along with the quantitative Zn distributions obtained in the corresponding physical slices. Whereas the dentin tissue density values span 1.75 g cm -3 to 2 g cm -3 the Zn mass fractions occupy a range between 50 ppm and 500 ppm. Both density and mass fraction values display predominantly radial patterns with increasing density and decreasing Zn values at increasing distances from the pulp toward the tooth periphery. Interestingly, in some of the dentin regions in particular near the pulp, circumferential variations in density can be discerned (e.g. T3 crown), possibly due to tubule blockage typically attributed to dentin sclerosis. [30] Such patterning is not visible in the Zn mass fraction images. Figure 4: a) Quantitative Zn mass fraction and tissue density images of 6 selected slices in teeth T1-T3. Zn mass fraction and tissue density appear to be inversed. From each slice, 2 line profiles (roughly vertical (v) and horizontal (h)) were used to plot profiles as shown in panel (b) (300 µm thick line average). b) Example line profiles from the crown slice of T1 show tissue density (top) and Zn mass fraction (bottom) as a function of distance to the pulp (marked in grey). Enamel (in the crowns) and cementum (in root slices) are excluded from the fitting, to focus on the trends within dentin. The Zn mass fraction profiles and tissue density profiles are fitted with exponential functions (bold lines). (Nomenclature: T1, v1: Tooth 1, vertical/horizontal profile 1/2, values from the first/second half of each profile.) In each of the slices, 4 line profiles were extracted (positions marked with yellow lines, one roughly horizontal and one vertical, Figure 4a). Figure 4b shows the density (top) and Zn mass fraction (bottom) of the four corresponding profiles for the T1 crown slice, plotted against distance to the pulp. Both the tissue density curves as well as the Zn mass fractions can be described with an exponential dependency as a function of the distance to the pulp (fit = area * exp( - exponent * distance) + offset). Figure 5: a) Box plot of all exponents extracted from the tissue density line profiles of all slices in Figure 4a. The decrease of tissue density is slightly different between root and crown, with a wide spread in curve shapes. (b) Box plot of the exponential fits of the Zn mass fraction as a function of distance to the pulp. A statistically significant difference between crown and root can be seen. c) - d) Zn mass fractions are plotted against the tissue density for crown and root data, separately, and fitted with exponential functions. (Nomenclature: T1, v1: Tooth x, vertical/horizontal profile 1/2, values from the first/second half of each profile). The boxplots in Figure 5a and 5b summarize the exponential fit parameters from all the cross-sectional slices of T1, T2 and T3. An exponential decrease of density is observed towards the pulp but this varies considerably. There is a slightly higher mean exponent for the crown data. Additionally, the crown data show a higher variance which could be explained by more PTD or tubule blockage (see Figure S5 in the SI). For the Zn mass fractions, the exponential behavior is distinctive with a faster increase in the crowns compared to the root cross sections, as seen by the lower exponents observed in the root fits. In Figure 5c and 5d the tissue density is plotted against the Zn mass fractions separately for the crown and root slices. A base (minimum) level of about 50 ppm of Zn is found for mature dentin (across multiple teeth), in the same area where we consistently observe a density of about 2 g/cm³. Wherever the tissue density decreases, Zn mass fractions increase, with the lower density dentin regions near the pulp exceeding Zn values of 300 ppm. The graphs suggest an exponential behavior for the dependency between Zn mass fraction and tissue density. Figure 6: 3D representation: chemical picture in 3D of two teeth, one cut using cross sections (T1, a) and one longitudinally (T4, b). The micro-CT renderings (left) show the intact teeth and the Zn data (false color images, right) is virtually cut open (blue rectangles) to demonstrate the full information in 3D. The 2D maps of Zn, (n = 32 cross sections and n = 14 longitudinal slices of teeth T1 and T4, respectively) were correlated in 3D with the CT data obtained initially of the intact teeth. The matched data sets allow us to co-align and visualize the distribution of all measured elements in 3D, see also the video of T4 in the SI. The results are shown in Figure 6 depicting the 3D Zn distribution in false colors with a virtual cutout to demonstrate the internal distribution of the elements. 3. Discussion Our work reveals opposite trends in the Zn concentration and in the local dentin tissue density, by merging complementary X-ray measurements from tomography with X-ray fluorescence elemental mapping at the micrometer length scale. The result is a quantitative description of the normal, physiological Zn distributions in the crowns and roots of teeth with healthy dentin. With increasing proximity to the pulp, we observe a moderate, exponential decrease in the tissue density while at the same time, we reveal a pronounced exponential increase in the Zn mass fraction. We find this to be the case in all the healthy, fully-formed (~4 years old) teeth tested. Importantly, this appears to be characteristic of mammalian dentin, and therefore also in bovine teeth, a model system frequently used in dental research. [31] Therefore this Zn enrichment near the pulp is the normal chemical composition. There appears to be an association with increased tubule porosity (inversed to dentin tissue density) and therefore Zn might be preferentially localized to tubules around the pulp of mammalian teeth. To understand our results, we consider a simple model that integrates known measurements of the dentin composite structure and porosity. The model considers a surface inside dentin, oriented perpendicular to the tubules. In a representative cross-sectional area A i - tubules become larger at increasing proximity to the pulp. In any such cross section, the area occupied by tubules A T increases as a function of the number of tubules n(x) as well as the tubule diameter D(x) , (see Figure 7a). The ITD (corresponding only to the dentin nanocomposite material) area A D can therefore be described by \(A_{D}\left(x\right)=A_{i}-A_{T}=A_{i}-n\left(x\right)\frac{\pi}{4}{D(x)}^{2}\). (1) Considering the values published by Garberoglio et al. [10] the void and tubule cross-sectional areas can be estimated (see Figure 7b). We find that as the ITD area decreases, the void area increases. Figure 7: a) Simplified schematic representation of changes in number and area of tubules in dentin, b) dentin and void area calculated with the values from Garberoglio et al. [10] and eq. 1. c) Averaged line profile values of the tissue density and the Zn mass fraction determined in our work for the crown and root profiles (Figure 4 and 5), the standard deviation uncertainty is plotted as shaded regions. Measured independently by our XRF methods, we find an exponential increase in the Zn mass fraction in dentin tissue nearing the pulp (Figure 4). Specifically, there is a 5-10-fold increase in Zn concentration within 1 mm distance to the pulp chamber, where a mass fraction of up to 500 ppm is found at the transition between dentin and the pulp tissue. On the opposite, outer rims of the teeth sections, Zn mass fraction is the lowest, ca. 50 ppm observed far from the pulp. In such peripheral areas, tubule density and diameters are the lowest. When comparing the exponential increase of the Zn mass fraction to the tubule porosity in Figure 7b a tight association is observed. This observation makes it possible to speculate on where the Zn elements are localized in the tissue. There have been proposals that Zn is incorporated in hydroxyapatite, substituting for Ca in the crystal structure. [22,32] While the increase of Zn that we observe toward the pulp is substantial, the absolute amount in comparison to the dominant Ca concentration is very small (compare 50 ppm - 500 ppm for Zn to 27 % for Ca). Therefore, neither the Ca to P ratio nor the c-lattice parameter is expected to change significantly with the increase in Zn. Indeed, microbeam X-ray diffraction (micro-XRD) line profiles in 200 µm thick samples of bovine root sections (SI Figure S4), annealed to release any collagen compression effects on the nanocrystals, reveal no significant change in the crystal lattice in the regions showing strong Zn concentration increase. Line profiles from the external cementum all the way to the pulp reveal some crystal lattice decrease only up to about 1 mm from the pulp (see details in the SI). In other words, the regions in dentin where density decreases and Zn concentration increases near the pulp are the same areas where the c-lattice parameter is rather constant, suggesting that the Zn concentration in dentin is not changing the apatite lattice parameter in dentin. Other crystal inclusions such as CO 3 -2 substitution [33–35] are more likely to be the cause of lattice deformations in dentin. Therefore, we propose that Zn near the pulp is not significantly incorporated into the apatite nanocrystals of native, healthy mammalian dentin. Our results suggest that it is more likely that Zn resides in a different habit in teeth, taking into account that several literature studies [21,36] report that Zn is embedded within the organic matrix as a key component of enzymes such as matrix metalloprotease (MMP) or alkaline phosphatase (ALP). This observation matches findings by other authors, for example Stock et al. [21,37] who used nm-sized XRF beams to show Zn enhancement in bovine teeth around tubules. Those authors proposed that Zn enhancement was related to PTD, though they did not show this. It is well known in fact that PTD is mainly found in mid-crown regions [38] and not near the pulp. Our data show that the increase in Zn concentration near the pulp is seen for both crown and root dentin. This suggests that it is organic material in or around the tubules that harbors Zn not the PTD. Indeed, the increase in Zn in the near-tubular dentin may be a remnant or indirect trace of the mineralizing front for both PTD or ITD. [21] Since Zn has been associated with mineralization fronts, possibly concentrated in ALP enzymes, [36,39] the higher concentration of Zn raises the possibility that some enzymes may become entrapped in the ITD surrounding tubules, as the tissue mineralizes. We find that the average dentin tissue density in all our quantified tooth slices decreases at increased proximity to the pulp, with average dentin tissue values typically spanning 1.8 g/cm³ to 2 g/cm³. One would expect tissue density to follow the dentin area as depicted in Figure 7b. Deviations from the exponential dependency may hint to the presence of PTD or to tubule blockage, as we occasionally detect in selected areas (e.g. due to dentin sclerosis). Indeed, though we investigate young, healthy teeth, we observe in multiple regions (e.g. Figure S5 in the SI, sclerosis marked with black *) tubule blockage with an unexpectedly higher mineral density. This explains the patches of higher mineral density observed in some of the slices depicting density distributions circumferentially around the pulp (Figure 4a). Our 3D microXRF data, collected from healthy, non-treated relatively young bovine teeth, provide the first quantitative distributions of natural Zn in healthy dentin collected in a series of slices that we used to generate a complete 3D volumetric quantification. The 3D distribution of Zn is shown in Figure 6 based on different teeth where both longitudinal and cross-sectional slices were measured. The Zn signal in dentin is always strongly (~10-fold) enhanced towards the pulp. Though we are not able to show similar data in young, mature, untreated human teeth (due to ethical restrictions), the findings in bovine teeth highly match and help explain previous reports in human teeth. It is therefore likely that the Zn is bound to (organic) molecules and is not free to diffuse, hence it is entrapped in the dentin, but Zn might become mobile and possibly chemically reactive, if demineralization takes place (e.g. due to acidic processes such as caries, but perhaps also acid-etching/filling placement). This is different from diffusion of clinically used zinc containing pastes and biomaterials (such as ZnO) placed e.g. during root canal treatment, that have known bactericidal and anti-pathogenic activity and where the ionic stability is different, totally detached from the biological molecules and Zn that comprise dentin as a biological nanocomposite. [37] The relatively high natural concentration of Zn found near the pulp does not normally prevent caries: on the contrary, higher concentrations of Zn have been found in carious regions [40,41] that are a significant concern when dentists treat deep cavities, where the fillings are placed near the pulp (i.e. in areas that have high natural Zn content). Here one may speculate that natural Zn has a possibly negative dental-care effect, as Zn within the natural MMP enzymes may become active to participate in dentin degradation (e.g. collagenases) associated with caries progression. This may be related in part also to previous work that identifies correlation between dentin permeability (mediated by to porosity) and adhesion of fillings. [16] Our results suggest that the naturally elevated Zn concentrations near the pulp, observed in both the crown and root, may contribute to enzymatic degradation of dental restorative bonding to deep-placed fillings. With this in mind, and in particular with an increased population of ageing individuals with root exposure and possible deep caries, treatment beyond prevention might require alternative approaches that do not make use of dentin etching, demineralization and release of entrapped Zn. 5. Conclusion The combination of micro-CT and micro-XRF made it possible to generate a quantitative correlation of Zn concentration inside the bulk of healthy dentin tissue of varying density to reveal the 3D gradients that are typical for the macro-structure of young healthy bovine teeth. Dentin porosity in bovine teeth changes significantly at different distances from the pulp amounting to differences on the order of 10-20 % within regions of tissue that encase the pulp tissue. The composition of the dentin intertubular matrix is more or less homogeneous, the only significant chemical change observed in the natural tissue is the Zn distribution. We observe an inverse relationship between Zn concentration and tissue density following a seemingly exponential dependency that results in a 5-10-fold increase in Zn near the pulp. This substantial increase is associated with an increase in tubule diameter. If dentin ITD composition is more or less homogeneous in all tooth regions, our findings hint to a possibility that a major source of Zn is from the tissue within tubules, and not the ITD. Possible sources of such Zn might be odontoblast membrane enzymes or various matrix proteases that could localize preferentially in or around tubules, which suggests that very little if any Zn is localized within mineral nanocrystals, as we recently demonstrated for fish-bones. [5] While this study shows data from bovine teeth, Zn enhancement towards the pulp has been repeatedly reported also in human teeth, [21–24] showcasing the usefulness of the abundantly available bovine tooth material as a model that reproducibly simulates human healthy untreated dentin material. 6. Experimental Section/Methods Bovine tooth samples Several (n > 5) untreated bovine teeth (3-4 years old animals slaughtered for human consumption) were provided by Fleischhandelsgesellschaft Henke mbH, Germany, ensuring similar dietary conditions. After extraction and mechanical cleaning, the teeth were stored in 0.5 % Chloramine T following standard tooth storage procedures routinely performed at the dental clinic of the Charité Berlin. Scanning electron microscopy Several (n = 4) slices were cut across multiple hydrated bovine teeth, after embedded in self-curing dental acrylic (Technovit), see below. Each slice was ground and polished with a series of SiC grinding papers polished up to the finest polishing paper with a grit of 4000. The samples were allowed to slowly dry in air, before imaging in an electron microscope (Phenom-XL, Thermofisher, Eindhoven, NL) under low vacuum (60 Pa) imaging conditions. Transmission X-ray microscopy From one of the slices a 1 µm thin lamella was prepared, see description in the SI. Transmission X-ray microscopy measurements were performed at the U41 TXM Beamline of the synchrotron radiation facility BESSY II operated by the Helmholtz-Zentrum Berlin. At a set energy of 1 keV, 118 radiographs at angles between -60° and +57°, were collected with a step size of 1°and 4 s exposure time at a magnification of 1449. Subsequently, 10 flat field images were measured. For reconstruction, Fiji/ImageJ with different plugins was used. A flat-field normalization was performed using the minimum standard deviation method. With the TomoJ plugin, [42] the stack of radiographs was first aligned. Then, phase retrieval was performed with the ANKAphase plugin [43] (δ/β = 6.45, distance = 0.0000002667, pixel size = 0.02 µm) and, finally, reconstruction was performed again with the TomoJ plugin (iterative / art / 25 iterations / 0.2 relaxation coefficient / fista optimisation is on). Computer tomography For imaging of intact teeth, 3D data was collected prior to drying. T1-T3 were measured with a medical cone beam computer tomography (CBCT) setup (Veraviewepocs 3D R100, J. MORITA EUROPE) which utilizes a rotating X-ray source and detector system for patient treatment. The X-ray source was set to 75 kV and 10 mA and a 0.2 mm Cu filter was used. With an exposure time of 9.3 s the scans resulted in CT images with 125 µm 3 isotropic voxels. Standard clinical settings were used as defined by the manufacturer, and the data was reconstructed using the Morita inbuilt control and reconstruction software. For tooth T4, an industrial micro-CT setup (Xradia 620 Versa, ZEISS at BAM) was used. The X-ray cone beam was set to 90 kV and 0.1 mA. To reduce effects of beam hardening, a filter was used to absorb the softer X-rays, situated directly after the source beam window. Data (2400 projections) yielded a tomographic dataset with a voxel size of 24 µm³, reconstructed with the standard Zeiss reconstruction pipeline using the filtered-back-projection method. Mechanical sample preparation (T1-T4 + references) The teeth were dried in a sequence of increasing ratios of alcohol-water solutions (70 % alc., 80 % alc., 90 % alc., 100 % alc.), for gentle dehydration. [44] Subsequently, the teeth were embedded in Technovit 7200 VLC and light-cured (EXAKT 520, EXAKT Advanced Technologies GmbH), placed in an oven at 40 °C to complete the hardening. The embedded teeth were sliced using a water-cooled diamond band saw (EXAKT 310) to produce slices of a thickness of approximately 200 µm. Three of the bovine teeth were cut horizontally (T1-T3), resulting in 33, 36, 39 slices, respectively, while the fourth bovine tooth (T4) was cut longitudinally, resulting in 18 slices. Each slice was mounted on a plexiglass slide (Walter Messner GmbH). The thickness and density of these slides were measured before the sample preparation. The tooth slices were subsequently polished using a microgrinder (EXAKT 400) with polishing paper (grain size P2500). The thicknesses of all slices were documented to span 140 µm to 470 µm. For calibration purposes, a fifth dried and embedded bovine tooth was used to extract two pieces, one made of pure enamel and the other of pure dentin. The origin of the dentin piece was selected to be cut-out of material halfway between the pulp and the periphery, taken from a region just below the crown. The density of these samples was measured by weighing the pieces and by determining the volume using micro-CT (see next section), measured with voxel sizes of 10 µm 3 . The determined values were: ρ D = (2.025 +/- 0.075) g/cm³ for dentin and ρ E = (2.75 +/- 0.05) g/cm³ for enamel. Slice imaging and density calibration From the large number of slices produced, 7 characteristic slices (crown and root cross sections from T1-T3 + one central slice from T4) were tomographically imaged using an industrial EasyTom 150 S micro-CT scanner with 15 µm³ voxel size (100 kV, 0.3 mm Cu filter, 2 avg, 1400 projections, CMOS framerate of 2 FPS). The two density-measured homogeneous samples of pure dentin and enamel were also measured along with the same parameters to serve as density calibration samples. The reconstructed micro-CT grey value data were converted into density values using these calibration samples ( ρ D and ρ E ) and considering the ρ S density value of the acrylic glass slides on which the tooth sections were mounted. A linear fit through the values derived from the main peaks visible in CT histograms yielded a calibration curve of\(\rho=0.75\frac{g}{cm^{3}}+3.8*10^{-5}\frac{g}{cm^{3}}*(grey\ value)\). Micro-XRF measurements and quantification of sum spectra Elemental mapping by micro-XRF was performed with a modified commercial spectrometer (M4 Tornado Plus, Bruker Nano GmbH) which utilizes a rhodium microfocus X-ray tube, a polycapillary lens and a silicon drift detector (SDD) with a carbon-based light element window. [45] The lateral resolution (FWHM) derived by a knife-edge scan at the Zn K fluorescence line energy (8.64 keV) is 22.5 µm. The measurements were acquired in air under ambient conditions and with a tube acceleration voltage of 50 kV and current of 1 mA. All bovine tooth slices were imaged in-plane with a lateral step size of 60 µm and a measurement time of 60 s per point. All maps were deconvolved using the software SpecFit. [46] The resulting elemental maps were visualized by the software ImageJ (v. 1.53q). For quantification of elemental compositions, n = 3 slices from the three different crowns of teeth T1-T3 and n = 3 slices from the root regions of those teeth were used, with one additional longitudinal slice taken from T4. The Zn K fluorescence intensity maps were treated as topological images in which isolines were traced manually, separating six areas of similar Zn intensity with respect to the pulp, as shown in Figure 4a. Entire spectra were added from points collected in these areas, resulting in six sum spectra for each investigated slice (a total of n = 42 for 7 slices). These sum spectra were converted into mass fractions using the proprietary evaluation package FPQ-Tools, developed by Bruker. This software uses a fundamental parameter approach to forward calculate full spectra - and takes into account effects such as continuum scattering and the detector response (including escape, shelf and pile-up peaks). Spectra of samples of all thicknesses as well as layered systems with the relevant self-absorption effects can be simulated. Here, the background was modeled with a stripping background up to 12 keV (15 cycles) and calculated analytically for the high energy regions of each spectrum. Calibration of the spectrometer parameters (including solid angle of detection, detector efficiency and optic parameters) was performed by measuring single-element reference samples as well as multi-element glass reference samples (Breitländer GmbH). The uncertainty of the quantification due to this calibration procedure is in the range of 5 %. The limited thickness of the tooth slices (140 µm – 470 µm) results in the possibility of excitation or fluorescence radiation being transmitted to and from the underlying acrylic glass slides. Therefore, as sample model, a layered system was assumed with the dentin slices situated above plexiglass. The plexiglass slides, of known composition, were modelled as C 5 O 2 H 8 (8 % H, 60 % C, 32 % O) and had a thickness of 2 mm and a density of ρ S = 1.18 g/cm³. For the dentin slices (section 2.3), the thickness, the density and the dark matrix (which comprises elements not detectable in our micro-XRF spectra, due to limitations in excitation and detection) are necessary input parameters. The thickness of the individual slices was measured, see above. For the assignment of the components of the dark matrix, values from a previous publication were used, [47] because the bovine tooth samples originated from the same source. As explained elsewhere (see Bauer et al. [47] ), the dark matrix (light elements C, N, H and O) has an impact on quantification of all elements in the micro-XRF measurements. Therefore, to reach quantitative concentrations of Ca, P and Zn in the present work, dark matrix values of 2.2 % for H, 5 % for N, 35 % for O, 16 % for C, 1.2 % for Na and 1.2 % for Mg were inserted in the quantification procedure. When changing the fixed ratios of the dark matrix elements in the quantification, maximum uncertainties can be estimated. When changing the concentration values of the dark matrix within their uncertainties [26] , variations of 4 % for P, 6 % for Ca and 9 % for Zn were obtained for an example sum spectrum on T4, see SI, section 6. Therefore, 10 % uncertainty can be considered the upper limit for all quantification values presented here. We also evaluated the impact of density variations on the quantification results by examining and multi-fitting the sum spectra of the longitudinal slice of T4. Quantification of the 6 spectra summed-up in that slice was conducted both with a fixed mean density value for dentin ( ρ D = (2.025 +/- 0.075) g/cm³) and also with variable density values derived from line profiles across the calibrated micro-CT data ( ρ D,i = 2.1 g/cm³, 2.1 g/cm³, 2 g/cm³, 1.92 g/cm³, 1.88 g/cm³, 1.82 g/cm³ for spectrum 1 to 6, see Figure 3). When comparing the difference between the use of a mean density versus gradually varying density for determining mass fractions of Zn, Ca and P, the difference observed was <1 %, implying that differences of ~ 10 % density have negligible impact on the quantification of concentrations of the main dentin elements, see SI. Therefore, for the purpose of all further micro-XRF quantification of Ca, P and Zn, the fixed mean density ρ D was used. Alignment of micro-XRF and CT data Image registration was used to align and match the micro-XRF with the micro-CT 3D data. Some micro-XRF slices at the edges of the teeth were omitted due to enhanced geometry effects of the curved outer surface. The following workflow was applied to the rest: Initially, all CT images (125 µm/px or 24 µm/px) were interpolated/binned to match the pixel sizes of the micro-XRF measurements (60 µm/px). Both, the micro-CT as well as the Ca K distributions of the micro-XRF data were segmented into binary masks to separate the tooth slice from background. For the segmentation, different methods were used depending on the image quality: global threshold, Otsu threshold and Gaussian mixture model (for details see SI, section 7). As a result, for each tooth a 3D CT data set of binary masked data and several 2D micro-XRF data sets of binary masked data were available. Thereafter, the CT data was rotated in 3D to match the micro-XRF slices plane assuming that the cutting process delivers parallel slices. The resulting rotated CT stack was used for alignment and further processing explained briefly in the following. For registration of the micro-XRF data to micro-CT, the first micro-XRF slice was compared to all micro-CT slices using a rigid template-matching custom written python code. The algorithm searches for the best match in (x, y, z, θ) with θ defined as the rotation of the micro-XRF slices in the xy-plane. Optimal matching is found when the error value R(x, y, z, θ) is minimized. This is repeated for all other micro-XRF slices with the boundary condition that an iteration through the micro-CT model can only start from a CT slice that is above the last matched CT location. The output was the best match of each micro-XRF measurement to the 3D CT data cube. Figures and Visualization All graphs were prepared with OriginPro 2021b and all pixel images with Fiji/Image J 1.53q. Microsoft Powerpoint 2013 was used for figure composition and schematic drawings. Acknowledgements This work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – project numbers 443841418 (IXdent) and 396127899 (InterDent – TP1). The authors would like to thank Fabian Nitsche and Dimitrijs Docenko from the FPQ team of Bruker AXS and Hamza Elfarraj for the micro-CT measurement of the reference samples for dentin and enamel. We thank the Helmholtz-Zentrum Berlin für Materialien und Energie for the allocation of synchrotron radiation beamtime. We thank Ivo Zizak for his help with micro-XRD data collection at the mySpot beamline of BESSY II and Holger Kropf for assistance with the FIB-SEM instrument. Conflict of Interest The authors declare no conflict of interest. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Supporting Information Supporting Information is available from the Wiley Online Library or from the author. Received: ((will be filled in by the editorial staff)) Revised: ((will be filled in by the editorial staff)) Published online: ((will be filled in by the editorial staff)) References [1] N. Reznikov, R. Shahar, S. Weiner, Acta Biomater 2014 , 10 , 3815. [2] P. Fratzl, R. Weinkamer, Progress in Materials Science 2007 , 52 , 1263. [3] J. D. Currey, R. Shahar, Journal of Structural Biology 2013 , 183 , 107. [4] K. Sauer, A. Silveira, V. Schoeppler, A. Rack, I. Zizak, A. Pacureanu, N. Nassif, I. Mantouvalou, W. de Nolf, C. Fleck, R. Shahar, P. Zaslansky, Acta Biomaterialia 2024 , 179 , 164. [5] A. Silveira, A. Davydok, C. Krywka, M. Scheel, T. Weitkamp, C. Fleck, R. Shahar, P. Zaslansky, Advanced Science 2025 , 12 , 2410617. [6] S. Weiner, R. Shahar, Acta Biomater 2024 , 179 , 1. [7] P. W. Lucas, Dental Functional Morphology: How Teeth Work , Cambridge University Press, Cambridge, 2004 . [8] R. Z. Wang, S. Weiner, J Biomech 1998 , 31 , 135. [9] J. H. Kinney, S. J. Marshall, G. W. Marshall, Critical Reviews in Oral Biology & Medicine 2003 , 14 , 13. [10] R. Garberoglio, M. Brännström, Archives of Oral Biology 1976 , 21 , 355. [11] A. Nanci, Ed., Ten Cate’s Oral Histology: Development, Structure, and Function , Elsevier, 2024 . [12] M. Robin, C. Djediat, A. Bardouil, N. Baccile, C. Chareyron, I. Zizak, P. Fratzl, M. Selmane, B. Haye, I. Genois, J.-M. Krafft, G. Costentin, T. Azaïs, F. Artzner, M.-M. Giraud-Guille, P. Zaslansky, N. Nassif, Advanced Science 2024 , 11 , 2304454. [13] N. Shuhaibar, A. R. Hand, M. Terasaki, Anat Rec (Hoboken) 2021 , 304 , 1820. [14] O. W. Reeder, R. E. Walton, M. J. Livingston, D. H. Pashley, J Dent Res 1978 , 57 , 187. [15] K.-W. Lee, H.-H. Son, M. Yoshiyama, F. R. Tay, R. M. Carvalho, D. H. Pashley, Am J Dent 2003 , 16 Spec No , 68A. [16] D. H. Pashley, R. M. Carvalho, Journal of Dentistry 1997 , 25 , 355. [17] M. Lin, Z. Y. Luo, B. F. Bai, F. Xu, T. J. Lu, PLOS ONE 2011 , 6 , e18068. [18] A. Aminoshariae, J. C. Kulild, J Endod 2021 , 47 , 1696. [19] M. Khatibi Shahidi, J. Krivanek, N. Kaukua, P. Ernfors, L. Hladik, V. Kostal, S. Masich, A. Hampl, V. Chubanov, T. Gudermann, R. A. Romanov, T. Harkany, I. Adameyko, K. Fried, J Dent Res 2015 , 94 , 945. [20] J. D. Currey, Journal of Biomechanics 1969 , 2 , 477. [21] A. C. Deymier-Black, A. Veis, Z. Cai, S. R. Stock, Scanning 2014 , 36 , 231. [22] F. Brudevold, L. T. Steadman, M. A. Spinelli, B. H. Amdur, P. Grøn, Archives of Oral Biology 1963 , 8 , 135. [23] R. M. Frank, M. L. Sargentini-Maier, J. C. Turlot, M. J. F. Leroy, Archives of Oral Biology 1989 , 34 , 593. [24] W. M. G. Parker, J. W. Adams, D. P. Hocking, E. M. G. Fitzgerald, G. Shaw, M. B. Renfree, A. R. Evans, Biol Trace Elem Res 2025 , DOI 10.1007/s12011-024-04502-z. [25] B. Beckhoff, B. Kanngießer, N. Langhoff, R. Wedell, H. Wolff, Eds., Handbook of Practical X-Ray Fluorescence Analysis , Springer Science & Business Media, 2007 . [26] L. J. Bauer, H. A. Mustafa, P. Zaslansky, I. Mantouvalou, Acta Biomaterialia 2020 , 109 , 142. [27] S. Amarie, P. Zaslansky, Y. Kajihara, E. Griesshaber, W. W. Schmahl, F. Keilmann, Beilstein J. Nanotechnol. 2012 , 3 , 312. [28] B. K. B. Berkovitz, G. R. Holland, B. J. Moxham, Oral Anatomy, Histology and Embryology , Elsevier, 2017 . [29] I. Zanette, B. Enders, M. Dierolf, P. Thibault, R. Gradl, A. Diaz, M. Guizar-Sicairos, A. Menzel, F. Pfeiffer, P. Zaslansky, Sci Rep 2015 , 5 , 9210. [30] F. Kabartai, T. Hoffmann, C. Hannig, Medical Hypotheses 2015 , 85 , 887. [31] G. H. Yassen, J. A. Platt, A. T. Hara, Journal of Oral Science 2011 , 53 , 273. [32] M. T. Rahman, A. Hossain, C. H. Pin, N. A. Yahya, Biol Trace Elem Res 2019 , 187 , 51. [33] S. Wong, A. Eaton, C. Krywka, A. Nair, C. Drouet, A. Deymier, Journal of the Mechanical Behavior of Biomedical Materials 2025 , 166 , 106962. [34] B. Wingender, M. Azuma, C. Krywka, P. Zaslansky, J. Boyle, A. Deymier, Acta Biomaterialia 2021 , 122 , 377. [35] J.-B. Forien, C. Fleck, C. Krywka, E. Zolotoyabko, P. Zaslansky, Journal of the Mechanical Behavior of Biomedical Materials 2015 , 50 , 171. [36] M. C. Dean, J. Garrevoet, S. J. M. Van Malderen, F. Santos, M. Mirazón Lahr, R. Foley, A. Le Cabec, Biology 2023 , 12 , 1455. [37] S. R. Stock, A. Veis, A. Telser, Z. Cai, Journal of Structural Biology 2011 , 176 , 203. [38] J. R. Dorvee, A. Deymier-Black, L. Gerkowicz, A. Veis, Connect Tissue Res 2014 , 55 Suppl 1 , 9. [39] A. Togari, S. Arakawa, M. Arai, S. Matsumoto, General Pharmacology: The Vascular System 1993 , 24 , 1133. [40] H. H. Harris, S. Vogt, H. Eastgate, P. A. Lay, J Biol Inorg Chem 2008 , 13 , 303. [41] K. Moriwaki, Y. Shimpuku, S. Furuyama, Radioisotopes 1999 , 48 , 172. [42] C. O. S. Sorzano, F. de Isidro-Gómez, E. Fernández-Giménez, D. Herreros, S. Marco, J. M. Carazo, C. Messaoudi, Journal of Structural Biology: X 2020 , 4 , 100037. [43] T. Weitkamp, D. Haas, D. Wegrzynek, A. Rack, Journal of Synchrotron Radiation 2011 , 18 , 617. [44] H. Shemesh, T. Lindtner, C. A. Portoles, P. Zaslansky, J Endod 2018 , 44 , 120. [45] T. Lachmann, G. van der Snickt, M. Haschke, I. Mantouvalou, J. Anal. At. Spectrom. 2016 , 31 , 1989. [46] F. Förste, L. Bauer, K. Heimler, B. Hansel, C. Vogt, B. Kanngießer, I. Mantouvalou, J. Anal. At. Spectrom. 2022 , 37 , 1687. [47] L. J. Bauer, F. Wieder, V. Truong, F. Förste, Y. Wagener, A. Jonas, S. Praetz, C. Schlesiger, A. Kupsch, B. R. Müller, B. Kanngießer, P. Zaslansky, I. Mantouvalou, Anal. Chem. 2024 , 96 , 8441. Table of contents Quantitative micro-XRF combined with X-ray imaging reveals correlations between Zn and dentin tubule porosity across entire teeth The combination of calibrated micro-CT and quantitative micro-XRF measurement facilitates the correlation of density and composition in the 3D macro-structure of teeth. Zn concentration in dentin increases by a factor of 10 towards the pulp. Incidentally, the overall density decreases, showing an association between Zn and the dentin tissue porosity. Information & Authors Information Version history V1 Version 1 09 October 2025 Peer review timeline Published VIEW Version of Record 19 Jan 2026 Published Copyright This work is licensed under a Non Exclusive No Reuse License. Collection View Keywords dentin density micro-computer tomography quantitative micro-x-ray fluorescence tubule porosity zn concentration Authors Affiliations Ioanna Mantouvalou 0000-0002-9696-2970 [email protected] Helmholtz-Zentrum Berlin für Materialien und Energie GmbH View all articles by this author Leona Johanna Bauer Helmholtz-Zentrum Berlin für Materialien und Energie GmbH View all articles by this author Vinh-Binh Truong Technische Universität Berlin Fakultät II Mathematik und Naturwissenschaften View all articles by this author Yannick Wagener Technische Universität Berlin Fakultät II Mathematik und Naturwissenschaften View all articles by this author Frank Förste Technische Universität Berlin Fakultät II Mathematik und Naturwissenschaften View all articles by this author Oleksandra Marushchenko Helmholtz-Zentrum Berlin für Materialien und Energie GmbH View all articles by this author Stephan Werner Helmholtz-Zentrum Berlin für Materialien und Energie GmbH View all articles by this author Franco Lizzi Charite - Universitatsmedizin Berlin View all articles by this author Frank Wieder Bundesanstalt für Materialforschung und -prüfung View all articles by this author Timo Wolff Bruker Nano GmbH View all articles by this author Birgit Kanngießer Technische Universität Berlin Fakultät II Mathematik und Naturwissenschaften View all articles by this author Paul Zaslansky Charite - Universitatsmedizin Berlin View all articles by this author Metrics & Citations Metrics Article Usage 308 views 156 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Ioanna Mantouvalou, Leona Johanna Bauer, Vinh-Binh Truong, et al. Quantitative micro-XRF combined with X-ray imaging reveals correlations between Zn and dentin tubule porosity across entire teeth. Authorea . 09 October 2025. DOI: https://doi.org/10.22541/au.175999341.11496128/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175999341.11496128/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ffe2786da4f4807',t:'MTc3OTQ3NzM3Nw=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();