dental materials 38 (2022) 670–679 Available online at www.sciencedirect.com journal homepage: www.elsevier.com/locate/dental Tooth whitening, oxidation or reduction? Study of ]] ]]]]]] ]] physicochemical alterations in bovine enamel using Synchrotron based Micro-FTIR ⁎ Clara Babot-Marquillas a,1, Maria-Jesús Sánchez-Martín a, ,1, Jose Manuel Amigo b,c, Ibraheem Yousef d, Iris H.Valido a, Roberto Boada a, Manuel Valiente a a GTS Research Group, Department of Chemistry, Faculty of Science, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain b Ikerbasque, Basque Foundation for Science, María Díaz de Haro, 48013 Bilbao, Spain c Department of Analytical Chemistry, University of the Basque Country UPV/EHU, P.O. Box 644, 15 48080 Bilbao, Basque Country, Spain d ALBA Synchrotron Light Source, 08193, Cerdanyola del Vallès, Spain ar ti cl e i n f o ab stra ct Article history: Objectives: Tooth whitening demand formulations have not evolved significantly over Received 15 February 2021 years being hydrogen peroxide (HP) or carbamide peroxide (CP) the main active agents Received in revised form 10 January used. These oxidizing compounds need high concentrations and long application times to 2022 achieve the desired whitening results, leading to painful secondary effects. A novel treat­ Accepted 15 February 2022 ment based on a reducing agent, sodium metabisulphite (MBS) encapsulated in liposomes, demonstrated significant whitening results in just 3 minutes. The aim is to provide in­ Keywords: formation about the chemical interactions of the encapsulated MBS with the enamel Tooth whitening structure compared with CP’s and to study their penetration in the hydroxyapatite (HAP). Sodium metabisulphite Methods: Chemical imaging is performed by synchrotron-based micro Fourier transformed Liposomes infrared spectroscopy (SR-µFTIR). Continuous Stiffness Measurements (CSM) were used to Synchrotron determine the depth reached by the treatments in order to delimitate the area of study. MicroFTIR Results: Principal component analysis (PCA) showed that the main differences between PCA treatments were found in the intensity of the ν3 PO3- 4 peak related with tooth deminer­ alization produced by the acidic nature of MBS; these changes were observed in deeper regions of the enamel when MBS was encapsulated. Significance: The obtained results indicated that CP and MBS promote different changes in the HAP mineral, and that liposomes favour the diffusion of MBS into the enamel. It is the ⁎ Corresponding author. E-mail address: mariajesus.sanchez@uab.cat (M.-J. Sánchez-Martín). 1 Shared co-first authorship https://doi.org/10.1016/j.dental.2022.02.006 0109-5641/© 2022 The Author(s). Published by Elsevier Inc. on behalf of The Academy of Dental Materials. CC_BY_NC_ND_4.0 dental materials 38 (2022) 670–679 671 first time that synchrotron light is used to map chemically the bovine incisor’s enamel, and to determine the effect of a whitening treatment in terms of chemical HAP modifications, and the extent in deep of these effects. © 2022 The Author(s). Published by Elsevier Inc. on behalf of The Academy of Dental Materials. CC_BY_NC_ND_4.0 results in 3 min [13]. Indeed, the purpose of these nanos­ 1. Introduction tructures is twofold, they act as a carrier of the MBS and, on the other hand, they create a layer over the enamel surface Tooth whitening (aka tooth bleaching) [1], is a widely prac­ that improves the diffusion of the reductant towards the ticed aesthetic treatment around the world. Its effects in tooth. It was also seen that the encapsulated MBS had phy­ terms of the whitening results are widely studied using col­ sical effects over the enamel, increasing the surface rough­ ourimetric techniques [2], since the main goal is to change ness within safe limits; nevertheless, the enamel-treatment the colour, to obtain a bright, clean and white smile. How­ chemical interaction was still pending to study. ever, besides the desired final aesthetic result, the bleaching Although mammalian enamels are similar, bovine enamel treatment also has undesirable effects due to its mechanism has some differences compared with the human one. It of action. The whitening effect is based on the oxidizing contains a larger number of fibril-like interprisms, and the power of peroxides, e.g. hydrogen peroxide (HP), or carba­ average diameter of enamel crystals is bigger. It also has mide peroxide (CP), the latter is composed by urea and hy­ different decussating patterns, since in bovine enamel hy­ drogen peroxide molecules. HP decomposes in reactive droxyapatite (HAP)crystallites within a rod are not merely oxygen species (ROS) that are highly oxidative [3]. It performs parallel but twisting as fibres in a thread of wool [14]. its action penetrating through the different tooth structures Nevertheless, in the field of tooth whitening, bovine incisors of the discoloured tooth reaching the adsorbed stains in the are widely used as an in vitro model, since they are con­ enamel and dentin, and the ROS break the double bonds of sidered a valid model by ISO 28399 to test products for ex­ the staining molecules making them unable to absorb light ternal tooth whitening treatments [15], since human incisors [4]. Enamel and dentin behave as semi-permeable mem­ are quite difficult to obtain. In the present work, bovine in­ branes allowing the diffusion of HP according to Fick’s second cisors have been used as in vitro model. law [5] reaching the pulp in minutes [6]. For CP, the urea When analyzing the side effects of whitening treatments groups degrade the organic parts of the enamel, leaving on enamel, most of the studies are focused on determining empty spaces that favour the diffusion of hydrogen peroxide the changes in its mechanical properties, like surface hard­ throughout the whole enamel thickness [7]. Higher con­ ness or roughness; however, these techniques are not able to centrations and longer application times are associated with provide information about the chemical changes taking place better results [8], but also with a higher incidence of side ef­ in neither the inorganic nor the organic structures. Fourier fects like hypersensitivity (the most common secondary ef­ transformed infrared (FTIR) spectroscopy is a molecular vi­ fect of dental whitening) [9] and pulp damage [5], tooth brational technique that has been shown to be advantageous demineralization (leading to changes in roughness and to investigate mineralized tissues like teeth, providing in­ hardness) and gingival irritation [7]. Besides, the need to formation on the chemical structure at the molecular scale. apply these products for hours, either at the dentist or at The addition of a microscope to FTIR microspectroscopy home, is not convenient for the user. (µFTIR) has led to the possibility of combining biochemical Efforts have been made to find a solution to reduce HP with spatial information, i.e., chemical imaging, which has concentrations and exposure times. Novel strategies the potential to examine tissues at cellular resolution [16]. avoiding the use of peroxides are also being explored. For The use of synchrotron radiation light sources, synchrotron- instance, whitening based on piezocatalysis using BaTiO3 based FTIR microspectroscopy (SR-µFTIR), provides a much nanoparticles and vibration showed whitening results after better signal to noise ratio [17], and the possibility of using a 10 h application [10]; the use of natural compounds like smaller beam size without losing signal efficiency [16] than proteolytic enzymes (bromelain, papain and ficin) also when using traditional sources. SR-µFTIR in specular re­ showed visible results after 4 applications of 15 min each [11]. flectance mode can provide very precise chemical informa­ These solutions may solve the secondary effects derived tion in small areas, therefore it is a very suitable technique to from the peroxides, but not the inconvenience of long ap­ study the changes induced by whitening treatments and to plication times. characterize the extent of its effects with precision. Actually, In previous works, we reported how an outstanding it has been previously used to study the ablation in enamel whitening was achieved by using a reducing agent, sodium caused of pulsed CO2 lasers, allowing to detect ablations of metabisulphite (MBS), as bleaching agent when compared 9 µm [18]. with the traditional oxidizing ones[12]. In addition, we found Since whitening treatments are applied directly on the that the whitening effect increases when MBS is en­ enamel surface, the most affected parts in terms of chemical capsulated in liposomes obtaining significant whitening 672 dental materials 38 (2022) 670–679 transformations are the surface itself and the enamel im­ Table 1 – Treatment’s pH values at 25 °C. mediately below, since these parts are exposed to the highest Treatment pH concentration of the oxidant. According to this, a study re­ H2O 6.5 vealed that in vitro application of 10% CP on enamel for two DPPC lipos 5.7 weeks caused demineralization of the enamel extending to a MBS 2.9 depth of 50 µm below the enamel surface [19]. MBS lipos 2.4 Here we report a SR-µFTIR study on the chemical changes CP 8.0 induced in HAP by either oxidizing (CP) or reducing (MBS) whitening treatments in bovine incisors enamel, down to 50 µm depth from the surface. Besides, the chemical dis­ using a diamond saw and preserved in sodium azide solution tribution of inorganic and organic species in the enamel of (0.2%) until the experiment performance. Teeth were em­ bovine incisor were mapped and studied by using multi­ bedded in self-curing polyacrylic cylinders. Their surface was variate analysis (principal component analysis, PCA). To our polished to expose a window of at least 3 × 3 mm of the en­ knowledge and up to date, no study has been reported using amel surface, which permitted the Continuous Stiffness SR-µFTIR to determine the effects of MBS as a whitening Measurements (CSM). The polishing was done using a se­ treatment in terms of chemical HAP modifications, and the quence of silicon carbide paper starting at grit size P400 and extent in deep of these effects. Neither the bovine incisor sequentially increasing to P4000, under a constant flow of tap enamel has been chemically mapped using synchrotron water, and afterwards with a sequence of diamond pastes light. The aim of this work is to provide information about with a mean particle size of 3 and 1 µm followed by a slurry of the chemical interactions of the encapsulated reductant aluminium oxide with a mean particle size of 0.3 µm. After agent (metabisulfite, MBS) with the enamel structure com­ every polishing step, the samples were submerged in deio­ pared with carbamide peroxide (CP) and to study their pe­ nized water and sonicated for 30 s. The different whitening netration in the hydroxyapatite (HAP) and the changes treatments were freshly prepared and applied for 20 min on produced in the mineral and its hardness. Thus, to compare the flattened surface of the specimens. the side effects provoked by typical whitening treatments (by means of oxidation) compared to the new treatment devel­ 2.3.2. Samples for SR-µFTIR oped by the authors (by means of reduction). Five bovine incisors (1 specimen per treatment) were used; each tooth was immersed in the corresponding freshly pre­ 2. Materials and methods pared treatment for 20 min making sure that only the enamel was in contact with the solution, avoiding the penetration of 2.1. Reagents the treatment through the root cut part where the pulp is exposed, and then brushed for 30 s with an electric tooth­ Tannic acid was purchased from Sigma-Aldrich (Missouri, brush (Oral-B Vitality Serie, BRAUN) while rinsed with dis­ USA), sodium metabisulfite from Riedel-de Haën (Seelze, tilled water. Afterwards, the specimens were embedded in an Germany), all in powder form. Carbamide peroxide was ob­ acrylic resin, and cut vertically along the central lobe to ob­ tained from Acros Organics (New Jersey, USA), in 1 g stabi­ tain a mesial view from the inside, the exposed surface was lized tablets (35% w/w H2O2). Dipalmitoylphophatidylcholine polished up to 1 µm particle size as described in Section 2.3.1, (DPPC) was purchased from Avanti Polar Lipids (Alabama, in order to obtain proper reflection properties for the mea­ USA). All chemicals were of analytical grade. Deionized water surements. was purified through a purification system from Millipore (Milford, MA, USA). 2.4. Continuous stiffness measurement 2.2. Treatments preparation The dynamic continuous stiffness method (CSM) permits to investigate the dependence of the dynamic hardness as a Deionized water was used as control. Sodium metabisulphite function of the penetration depth during a nanoindentation. solution was at a concentration of 0.47 M (MBS). Carbamide CSM was performed with a Nano Indenter® XP from MTS peroxide solution (CP) was prepared at 16% (w/v). DPPC li­ (Tenessee, USA) using a Berkovich diamond tip. Nine mea­ posomes solution (lipos DPPC) was obtained by lipid film surements per sample were performed separated 100 µm hydration [20] at a concentration of 20 mM of DPPC en­ between them. Before every measurement, the diamond in­ capsulating water. MBS liposomes (MBS lipos) solution was denter was calibrated on a standard fused silica specimen. obtained by monolayer technique at 20 mM of DPPC en­ capsulating a solution of metabisulphite at 0.47 M. The pH of 2.5. Specular reflectance SR-µFTIR measurements the different treatments is reported in Table 1. Specular reflectance spectra were acquired using the FTIR 2.3. Sample preparation microscope in the reflection mode at MIRAS beamline of ALBA synchrotron light source (Cerdanyola del Vallès, Spain) 2.3.1. Samples for continuous stiffness measurement [21], using a Hyperion 3000 microscope coupled to a Vertex 70 Ten specimens (2 specimens for treatment) of bovine incisors spectrometer (Bruker, Germany), and equipped with a Mer­ were cleaned of gross debris, and the root was sectioned cury-Cadmium-Telluride (MCT) detector. The microscope is dental materials 38 (2022) 670–679 673 utilizing a 36x Schwarzschild objective (NA = 0.52) coupled to (typically occurring in stoichiometric apatite with a relation a 36x Schwarzschild condenser to focus the synchrotron IR Ca10(PO4)6(OH)2 [27]). The band at 956 cm−1 corresponds to light on the sample. OPUS 7.5 (Bruker, Germany) was used to the symmetric stretching mode of phosphate (ν1PO43−) [25]. collect the spectra, taking a minimum of 256 co-added scans The region between 1580 and 1320 cm−1 corresponds to per spectrum in the 650–4000 cm−1 Mid-IR range and with a ν3CO32− antisymmetric stretching; in this band two maxima 4 cm−1 spectral resolution, using a 5 × 5 µm2 masking aper­ can be observed, at 1401 and 1442 cm−1, whose intensities are ture size. A gold mirror reference was used as reference to slightly different, being the peak at 1401 cm−1 a little higher; collect the background. also a peak at 1540 cm−1 (region 1535–1565 cm−1), corre­ To characterize bovine enamel physiological variations, a sponding to ν3CO32− and organic material (amide II), was mapping of the whole enamel thickness, of 500 µm long and detected [28]. Another carbonate band arises at 880–830 cm−1, 50 µm width, was performed in the mid crown (labial side) of corresponding to the out-of-plane bending mode of the C–O the control sample. To observe the effect of the enamel sur­ bond (ν2CO32− group) [28–31]. When the carbonate substitutes face in treated samples, mappings of the outer enamel re­ the hydroxyl group it is designated as type A apatite and gion, of 50 × 50 µm, were performed also in the mid crown when it substitutes the phosphate group it is designated as (labial side), starting in the enamel surface and moving to­ type B apatite, and it is described that there are equal wards the dentin direction. Spectra measurements were amounts of A and B in dental enamel [32]. According to taken in spots separated by 10 µm each, resulting in a matrix Brangule et.al. [33] peaks' deconvolution, there is a peak of 50 × 6 points in the control and 6 × 6 points in the treated doublet at 1540 and 1455 cm−1 corresponding to type A car­ samples. bonates and a doublet at 1475 and 1416 cm−1 corresponding to type B; for the ν2CO32− band, a peak at 880 cm−1 is assigned 2.6. Data treatment to type A and other at 871 cm−1 to type B. This means that the peaks in the regions 1580–1320 cm−1 and 880–830 detected in Principal component analysis (PCA), a multivariate method, the reflection mode are showing a mixture of type A and B widely used in hyperspectral images [22] aims at extracting carbonates, except for the peak at 1540 cm−1 that reflects type the major sources of variance (variability) in the map of the A as well as organic matter. sample. This variance correlates, to some extent, with the chemical compounds conforming the sample, giving an in­ 3.2. Chemical distribution in incisor bovine enamel dication of the compounds’ distribution along the surface in an unsupervised and straightforward manner. In the present The characteristic distribution of crystals in dental enamel work, PCA was applied by using HYPER-Tools version 2 [23]. confers its distinctive mechanical properties, like its excep­ In the spectral data obtained for this study, the most in­ tional elasticity and resilience [34]. It must be remarked that tense peak is the ν3 PO42-. It has been already reported that enamel is not homogeneous but an anisotropic material, when dentin is treated with acids, its intensity is in­ showing spatial variations in its chemical composition and sufficiently constant for normalization [24], causing artefacts density, as well as variations in the carbonate content and in the normalized spectra. However, the spectra were mean the degree of crystallinity [35]. centered before the PCA analysis in order to homegeize the To evaluate the chemical distribution of carbonates and variariance between the variables. A total of 625 spectra (an phosphates, an area of non-treated bovine tooth enamel was area of 25 ×25) were measured per sample. The Unscrambler fully mapped by SR-µFTIR from the surface to the dentino- X® (2016, CAMO Software AS, Norway) was used to assign enamel junction (DEJ) along a cross-section of 500 µm long peak spectra numbers and to perform maximum normal­ and 60 µm wide (Fig. 2). For the sake of clarity, the intensity of izations when needed to observe the peak shifts. The che­ the colour scale is adjusted to the intensity of the corre­ mical mapping was performed by integrating the peak area sponding highest peak. with integration mode A (bordered by the spectrum curve, Phosphate (PO43−) is the principal molecular species that the x-axis and the two wavenumber limits (700–1700 cm−1) gives rise to the HAP absorbance in the 900–1200 cm−1 region, defined, using the software OPUS 7.5 (Bruker, Germany). and its distribution can be seen from the integration of ν3 antisymmetric and ν1 symmetric stretching vibrations. For ν3 PO43−, the peak intensity is higher (white area) in the mantel 3. Results of the enamel. Moreover, it is progressively reduced as it goes deeper in the enamel (Fig. 2), and a dramatic reduction of the 3.1. Peaks assignment intensity is observed in the dentin since its phosphate con­ tent is lower and its crystallinity decreases [36]. The bands corresponding to carbonates and phosphates of For ν1 PO43−, high intensity spots were observed in several the HAP forming the enamel have been identified (Fig. 1) in a places along the enamel; Pleshko N. et al. found a direct re­ control sample. lationship between the percentage area of the v1 band (near The broad absorption bands in the 1170–965 cm−1 region 960 cm−1) and the crystal size [37], therefore those regions correspond to the apatitic phosphate groups coming from a could correspond to discrete areas with larger crystals. Car­ triply degenerated asymmetric stretching mode (ν3PO43−) bonates showed complementary intensity distribution to the [25]. Its deconvolution is described in the literature as sec­ phosphates, with the highest intensity close to the dentin ondary phase vibrations of Ca-O-P at 1103 cm−1, P-O at and lower in the outer part of the enamel. This is in ac­ 1091 cm−1, Ca-O at 1047 cm−1 and O-Ca-O at 1031 cm−1 [26] cordance with previous works reporting that the carbonate 674 dental materials 38 (2022) 670–679 Fig. 1 – Specular reflectance FTIR (carbonated hydroxyapatite) spectrum of enamel and band assignment numbers (original figure). The spectrum corresponds to the average of 6 spectrum measured in the first 10 µm of control’s outer enamel. content increases from outer to inner enamel in bovine in­ the carbonate group (antisymmetric stretching and sym­ cisors [38]. Comparing the regions 1580–1320 cm−1 and metric angular deformation), similar intensity patterns can 880–830 cm−1, corresponding to the two types of vibration of be observed (note the similar intense point at around 400 µm Fig. 2 – Chemical distribution (calculated by peak integration) of carbonates and phosphates in the middle crown of a bovine incisor control tooth (vertical section). The mappings where obtained using OPUS by performing chemical imaging, using peak integration mode A. dental materials 38 (2022) 670–679 675 Fig. 3 – Continuous stiffness measurement for the different treatments tested (2 samples per treatment): deionized water (control), Carbamide peroxide at 16% (CP 16%), empty liposomes of DPPC (Lipos DPPC), metabisulfite (MBS) and metabisulfite encapsulated in DPPC liposomes (Lipos MBS). depth) revealing its presence in the enamel. In the region 30, 40 and 50 µm depth from the enamel mantel were ob­ 1530–1550 cm−1, the highest intensity was found in the tained. dentin region at the bottom of the map, which corresponds to PCA was performed to compare the effect of these type A carbonates, but also to amide II vibrations, which are changes all over the surface analyzed. The five different found in higher proportion in the dentin. treatments were clearly identified from principal component (PC) 1, 2 and 3 (accounting for a total of 99% of the variance) (Fig. 4a). To better relate the observed PCs with the spectra, 3.3. Treatment effects Fig. 4b represents the average of the 12 spectra corresponding the first 10 µm of enamel of each treatment, since are the 3.3.1. Continuous Stiffness Measurements (CSM) most affected. Additionally, Fig. 4c presents the spectra nor­ It is known that the acid contact with teeth causes partial malized to maximum height to facilitate the appropriate loss of mineral that results in a reduction of enamel surface comparison of relative peak intensities and their shifts hardness [39]. In our work, CSM was used to probe the de­ (Fig. 4c). mineralization caused by MBS as well as the depth reached As can be observed in the loadings’ representation in by this effect, measuring the hardness variations of the en­ Fig. 4a, PC1 represents 95.14% of the variation, comprising the amel from the tooth surface to deeper parts. whole spectra, especially the phosphate peak. In the scores CSM analysis (Fig. 3) showed a hardness of about 5 GPa plot, it can be seen that PC1 (which is negative and thus along the depth analyzed (2 µm) for the control. Similar be­ correspond to the blue parts of the image) did not change for haviour was observed for CP and DPPC liposomes with a the whole depth analysed in the control sample; for MBS and hardness value at around 4.5 GPa. MBS treatment decreased CP there was a lack of this component in the surface, the hardness dramatically especially in the first nanometers, meaning that the intensity was lower for the first 10 µm deep although, a recovery of hardness is observed as the na­ when compared to control. For DPPC liposomes, it was ob­ noindenter is deepened in the enamel. MBS liposomes also served that this treatment also affected the first 10 µm of the decreased the hardness initially to almost 2 GPa, but also the enamel, and some of the inner parts were affected by a thin hardness was increased as depth was gained. Although the line (as if some liposomes have penetrated the enamel); a analysis has a depth limit set at 2 µm, the observed trend for higher reflectance intensity in some points of the sample the last two treatments is that the hardness is progressively explains the intense blue observed in some parts in the recovered. The values observed in the first nanometers of the scores map. For MBS liposomes, PC1 showed a high influence measurements correspond to the size effect, caused by geo­ in the mantel enamel (most outer part) and then in a thin line metrical dislocation of the HAP crystals when the tip is entering the inner parts; besides, it can be observed how the pressed into the sample [40]. whole space is yellow, meaning that the treatment reached the whole surface analysed. In Fig. 4b, the groups MBS, CP, 3.3.2. Treatment effects analyzed by PCA and MBS liposomes showed a reduced reflectance when According to the results obtained in Section 3.3.1, the first compared to control, and thus PC1 is related to the re­ microns were the most affected in terms of mechanical flectance peak intensity. In addition, MBS showed a decrease properties since they are the most exposed to the whitening of the 1058 cm−1 intensity, and the contribution of the treatments. Nevertheless, the chemicals diffuse into the 1091 cm−1 shoulder had almost disappeared. This points out inner structures, giving chemical structural changes in to a clear distortion of the calcium phosphate structure. deeper distances. For this reason, and in order to follow the PC2 (2.40% of the variance) loading profile presented a treatment penetration, single spot µFTIR spectra at 0, 10, 20, peak at 1114 cm−1, placed in the negative part of the loadings. 676 dental materials 38 (2022) 670–679 Fig. 4 – (a) Scores (distributed across the enamel depth, starting in the enamel surface at 0–50 µm deep), and loadings for the PCA treatments. (b) Average of the spectra corresponding to the first 10 µm for each treatment (n = 12). (c) Plot b normalized to maximum. Red dash line: MBS shift. Blue dots line, CP shift. Dash and dots grey line: characteristic shoulder of the HAP peak. It corresponds to the Ca-O-P secondary vibration of ν1 PO43- (from 1058 to 1056 cm−1), and the second related to a bath­ (reflects changes in the bandwidth) and to the whole ν2 CO32- ochromic shift of MBS, as can be observed in the specific part band. Its presence is related to the blue parts in the corre­ of the HAP peak in Fig. 4c. This shift was no observed with the sponding scores map. It can be noticed how the lack of this other whitening treatments and represents different che­ component was highly significant in DPPC liposomes, and mical interactions with the HAP structure. also, although with less intensity, in the first microns of the samples treated with MBS and MBS with liposomes. It is mainly related with a change in the relative intensity of the 4. Discussion carbonates since, as indicated in Fig. 4c with a black arrow, for DPPC liposomes it was lower than in the other treatments. The enamel mapping of the non-treated bovine incisor Also, all the surface of the CP treated tooth appears blue, showed an heterogeneous distribution of the different in­ meaning that this compound did not affect the carbonates. organic species. Physiologically, enamel is an anisotropic The phosphate peak corresponding to CP is slightly tightened material, showing spatial variations in its chemical compo­ when compared to the other treatments, that could explain sition and density [35]. HAP crystals are organized in a basic the intense blue in some points of the PC2 scores map since it structural unit called ‘rod’. Rods run continuously through all is also related with the bandwidth. the width of enamel, and close to the dentin become thinner PC3 (1.46%) loading was negative at 1076 cm−1, i.e. blue in [41]; this creates a mineral gradient, in which the HAP is more the PC3 scores map, and positive at 1006 cm−1 (red in the packed in the outer enamel, leaving less space for organic score map). The first is related to a hypsochromic shift of CP matter. The frequencies and the intensity ratio between the dental materials 38 (2022) 670–679 677 HAP spectrum bands depend on the local environment, and the c-axis increases with sulphur substitution because of the they also vary depending on the modifications on the crys­ length of the atomic bond, as well as an hypsochromic shift tallinity degree [42]. Accordingly, phosphates showed higher happens in the phosphate peak [45]. While, in PC1, MBS af­ reflectance in the outer enamel, where their concentration is fected only the outer layers of the tooth, MBS liposomes af­ higher and the crystal framework is neater, while the re­ fected the whole analysed surface, especially in the middle flectance decreased close to the dentin, where the crystals part; we relate this effect to the increased diffusion promoted become thinner. Instead, in the inner parts of the enamel by the liposomes [13]. Instead, the bathochromic shift af­ carbonates showed higher and the intensity of the HAP peak fected only the tooth treated with MBS. was reduced. In fact, carbonates modify the lattice para­ Vasluianu et al. [42], studying the changes induced in HAP meters of the structure as a consequence of the different size by HP whitening treatments, observed a decrease in the re­ of the substituting ions; the carbonate content increases flectance and an hypsochromic shift for the main FITR peaks across outer to inner enamel, which is associated with de­ of HAP; they related it, again, to the shortening of P-O, even if creased crystallinity (thus decreased reflectance) and in­ the pH of the treatments was 7.4. In our experiment, CP pH creased solubility [43]. was 8 as in the commercial treatments, since its effectiveness CSM results showed that both, MBS and MBS with lipo­ is increased at this pH [46]. Accordingly, a decrease in the somes, treatments decreased the enamel hardness beyond reflectance (PC1) and an hypsochromic shift (PC3) were ob­ 2 µm depth, although the trend revealed that the hardening served. At basic pH, HAP dissolution should not occur, but was partially recovered at larger depth. We related it to ero­ protein degradation in the tooth organic matrix might be sive demineralization, a process that takes place when the happening. Indeed, the urea in CP along with the ammonium tooth is in contact with a solution that has lower concentra­ ions (NH4+) formed in contact with water act on the hydrogen tion of calcium and phosphate than physiological fluid bonds of the proteins weakening their structure [7]. Since the (plaque fluid), and lower pH than HAP critical pH (5.5). Such components of the organic matter form a support structure damage is characterized by an initial softening of the enamel that stabilizes the crystalline enamel layer [47], any change in surface, which extent is reported to be between 0.2 and 3 µm, the organic matter could cause an effect in the reflectance in depending on the immersion time and the acids under study the first 30 µm, as observed in PC1. According to Pleshcko [39]. If the acid contact is maintained, this process is followed et al. [37], a narrowing of the phosphate peak is related with by continuous layer-by-layer dissolution of the enamel crys­ increased crystallinity; this effect was observed for CP treat­ tals, leading to a permanent loss of tooth volume. But, if the ment, probably because the crystallinity was better detected contact ceases, several salivary protective mechanisms come after the loss of organic matter, because of its better reflective into play, such as dilution and clearance of the erosive agent properties. from the mouth, neutralization and buffering of acids (within Regarding the carbonates, MBS treatments promoted the 2–5 min after the exposure), and formation of the acquired loose of this component at the same extent as it affected the pellicle. When the acid is removed, remineralization takes PO43- peak (PC1), while CP did not affect the content at all, place in the partly demineralized surface enamel. according to the results observed in the scores map of PC2. Considering the sample thickness (around 0.5 cm thick), This confirms the previous observation, that the mechanism and the area of the analyzed spots (5 ×5 µm), the presence of by which the compounds interact with the enamel is dif­ the treatment components should not be significant over the ferent: the first produces an acid attack to the mineral con­ signal emitted by the enamel itself when measuring in re­ tent, the second by promoting the organic matrix flection mode. Therefore, no contribution of liposomes, CP or decomposition. MBS vibration peaks were expected to be observed, but only the induced modifications in the HAP structure. The treatments applied produced different effects on the 5. Conclusions FTIR HAP reflectance spectra. The most important change observed was the decrease in the v3 PO43-, which can be re­ From the reported results, the following conclusions can be lated to a loss of mineral content in the tooth surface [44] extracted: (erosion). Kim et al. [24] also observed a decrease in the ab­ MBS treatments soften the first 10 µm of enamel, as hap­ sorbance, but also an hypsochromic shift of v3 PO43- peak in pens in the initial states of tooth decay. But, since saliva the acid-eroded enamel. They related these observations to buffers acid and promotes remineralization once acid contact changes in the local structure; in apatite, each P atom is has been removed, this first microns will be naturally re­ linked to four Ca atoms via a shared oxygen atom (P–O–Ca mineralized again. atomic bridges), and the exposure to acid breaks the Ca–O CP and MBS promote different changes in the HAP mi­ bonds and consequently reduces the P–O bond length, due to neral; CP shortens the P-O bond while MBS seems to elongate the redistribution of the electron density in the vicinity of the it. Moreover, MBS promotes the loss of carbonates while CP bridging oxygen. According to that, under the acidic condi­ doesn't, which is probably related to the solution’s pH. tions of MBS treatments, a decrease in the reflectance due to When comparing MBS and MBS Liposomes, it is observed the acidic nature of MBS (PC1) was observed, but a bath­ how liposomes favour the diffusion of MBS to inner layers, ochromic shift, i.e., an elongation took place instead of a since the effects of MBS can be observed in deeper enamel. shortening of the P-O bond. This could be explained by a SO42- Thus, the encapsulated MBS whitening effect is highly im­ substitution of the PO43- group; the lattice parameter along proved in terms of time when compared to MBS alone or CP. 678 dental materials 38 (2022) 670–679 [14] Wang C, Li Y, Wang X, Zhang L, Tiantang FuB. The enamel Acknowledgements microstructures of bovine mandibular incisors. Anat Rec 2012;295:1698–706. https://doi.org/10.1002/ar.22543 This work was supported by the Spanish Project CTM 2015- [15] Standardization IO for. ISO 28399:2020 Dentistry — External 65414-C2-1-R; and by the Universitat Autònoma de Barcelona tooth bleaching products. 2020. with the "Personal d′Investigació en Formació" scholarship. The [16] Dumas P, Sockalingum GD, Sulé-Suso J. Adding synchrotron SR-µFTIR experiment was performed at MIRAS beamline at radiation to infrared microspectroscopy: what’s new in biomedical applications? Trends Biotechnol 2007;25:40–4. ALBA Synchrotron (Cerdanyola del Vallès, Spain), with the https://doi.org/10.1016/j.tibtech.2006.11.002 collaboration of ALBA staff (Dr. Ibraheem Yousef ). This work [17] Miller LM, Dumas P. Chemical imaging of biological tissue has been funded by ALBA Synchrotron through granted pro­ with synchrotron infrared light. Biochim Biophys Acta posals (grant references: 2017092361and 2017021987). Biomembr 2006;1758:846–57. https://doi.org/10.1016/j. bbamem.2006.04.010 [18] Fan K, Bell P, Fried D. Rapid and conservative ablation and references modification of enamel, dentin, and alveolar bone using a high repetition rate transverse excited atmospheric pressure CO[sub 2] laser operating at λ=9.3 μm. J Biomed Opt 2006;11:064008https://doi.org/10.1117/1.2401151 [1] Li Y, Greenwall L. Safety issues of tooth whitening using [19] Efeoglu N, Wood D, Efeoglu C. Microcomputerised peroxide-based materials. Br Dent J 2013;215:29–34. https:// tomography evaluation of 10% carbamide peroxide applied doi.org/10.1038/sj.bdj.2013.629 to enamel. J Dent 2005;33:561–7. https://doi.org/10.1016/j. [2] Joiner A, Luo W. Tooth colour and whiteness: a review. J Dent jdent.2004.12.001 2017;67:S3–10. https://doi.org/10.1016/j.jdent.2017.09.006 [20] Ryman BE, Tyrrell DA. Liposomes - methodology and [3] Torres C, Crastechini E, Feitosa F, Pucci C, Borges A. applications. Front Biol 1979;48:549–74. Influence of pH on the effectiveness of hydrogen peroxide [21] Nicolas J, RibÃ3 L, Colldelram C, Crisol A, Dumas P, Ellis G, whitening. Oper Dent 2014;39:E261–8. https://doi.org/10. et al. Mechanical design of MIRAS, infrared 2341/13-214-L microspectroscopy beam line at ALBA synchrotron. JACoW [4] Kwon SR, Wertz PW. Review of the mechanism of tooth Geneva, Switz 2017:403–8. https://doi.org/10.18429/jacow- whitening. J Esthet Restor Dent 2015;27:240–57. https://doi. medsi2016-fraa03 org/10.1111/jerd.12152 [22] Amigo JM, Babamoradi H, Elcoroaristizabal S. Hyperspectral [5] Llena C, Martínez-Galdón O, Forner L, Gimeno-Mallench L, image analysis. A tutorial. Anal Chim Acta 2015;896:34–51. Rodríguez-Lozano FJ, Gambini J. Hydrogen peroxide diffusion https://doi.org/10.1016/j.aca.2015.09.030 through enamel and dentin. Materials 2018;11:1–10. https:// [23] Mobaraki N, Amigo JM. HYPER-Tools. A graphical user- doi.org/10.3390/ma11091694 friendly interface for hyperspectral image analysis. Chemom [6] Cooper JS, Bokmeyer TJ, Bowles WH. Penetration of the pulp Intell Lab Syst 2018;172:174–87. https://doi.org/10.1016/j. chamber by carbamide peroxide bleaching agents. J Endod chemolab.2017.11.003 1992;18:315–7. [24] Kim IH, Son JS, Min BK, Kim YK, Kim KH, Kwon TY. A simple, [7] Goldberg M, Grootveld M, Lynch E. Undesirable and adverse sensitive and non-destructive technique for characterizing effects of tooth-whitening products: a review. Clin Oral bovine dental enamel erosion: attenuated total reflection Investig 2010;14:1–10. https://doi.org/10.1007/s00784-009- Fourier transform infrared spectroscopy. Int J Oral Sci 0302-4 2016;8:54–60. https://doi.org/10.1038/ijos.2015.58 [8] Borges AB, Zanatta RF, Barros ACSM, Silva LC, Pucci CR, [25] Bachmann L, Diebolder R, Hibst R, Zezell DM. Infrared Torres CRG. Effect of hydrogen peroxide concentration on absorption bands of enamel and dentin tissues from human enamel color and microhardness. Oper Dent 2015;40:96–101. and bovine teeth. Appl Spectrosc Rev 2003;38:1–14. https:// https://doi.org/10.2341/13-371-L doi.org/10.1081/ASR-120017479 [9] Soares DG, Basso FG, Pontes ECV, Garcia LDFR, Hebling J, De [26] Arboleda A, Franco M, Caicedo J, Tirado L, Goyes C. Synthesis Souza Costa CA. Effective tooth-bleaching protocols capable and chemical and structural characterization of of reducing H2O2 diffusion through enamel and dentine. J hydroxyapatite obtained from eggshell and tricalcium Dent 2014;42:351–8. https://doi.org/10.1016/j.jdent.2013.09. phosphate. Ing y Compet 2016;18:69. https://doi.org/10. 001 25100/iyc.v18i1.2178 [10] Wang Y, Wen X, Jia Y, Huang M, Wang F, Zhang X, et al. [27] Gadaleta SJ, Paschalis EP, Betts F, Mendelsohn R, Boskey AL. Piezo-catalysis for nondestructive tooth whitening. Nat Fourier transform infrared spectroscopy of the solution- Commun 2020;11:1328. https://doi.org/10.1038/s41467-020- mediated conversion of amorphous calcium phosphate to 15015-3 hydroxyapatite: new correlations between X-ray diffraction [11] Ribeiro JS, Barboza A, da S, Cuevas-Suárez CE, da Silva AF, and infrared data. Calcif Tissue Int 1996;58:9–16. https://doi. Piva E, et al. Novel in-office peroxide-free tooth-whitening org/10.1007/BF02509540 gels: bleaching effectiveness, enamel surface alterations, [28] Lopes CCA, Limirio PHJO, Novais VR, Dechichi P. Fourier and cell viability. Sci Rep 2020;10:1–8. https://doi.org/10.1038/ transform infrared spectroscopy (FTIR) application chemical s41598-020-66733-z characterization of enamel, dentin and bone. Appl Spectrosc [12] Marquillas CB, Procaccini R, Malmagro MV, Sánchez Martín Rev 2018;53:747–69. https://doi.org/10.1080/05704928.2018. MJ. Breaking the rules: tooth whitening by means of a 1431923 reducing agent. Clin Oral Invest 2019:1–7. https://doi.org/10. [29] Gross K, Yang TC, Kareiva A. Curr Biol CB 2014;24:503–6. 1007/s00784-019-03140-3 https://doi.org/10.1039/c4ce00119b [13] Babot-Marquillas C, Sánchez-Martín MJ, Rodríguez-Martínez [30] Rokaya D, Srimaneepong V, Sapkota J, Qin J, Siraleartmukul J, Estelrich J, Busquets MA, Valiente M. Flash tooth K, Siriwongrungson V. Polymeric materials and films in whitening: a friendly formulation based on a dentistry: an overview. J Adv Res 2018;14:25–34. https://doi. nanoencapsulated reductant. Colloids Surf B Biointerfaces org/10.1016/j.jare.2018.05.001 2020;195:111241https://doi.org/10.1016/j.colsurfb.2020. 111241 dental materials 38 (2022) 670–679 679 [31] Chen KH, Cheng WT, Li MJ, Yang DM, Lin SY. Calcification of [40] Voyiadjis GZ, Yaghoobi M. Review of nanoindentation size senile cataractous lens determined by Fourier transform effect: experiments and atomistic simulation. Crystals infrared (FTIR) and Raman microspectroscopies. J Microsc 2017;7:8–10. https://doi.org/10.3390/cryst7100321 2005;219:36–41. https://doi.org/10.1111/j.1365-2818.2005. [41] Shahmoradi M, Bertassoni LE, M. Elfallah H, Swain M. 01491.x Fundamental structure and properties of enamel, dentin and [32] Fleet M. Biological Apatites. Carbonated Hydroxyapatite cementum Mahdi. DAdv Calcium Phosphate Biomater Mater. Synth. Appl. Pan Standford Publishing; 2015. p. 2014;vol. 2:1–39. https://doi.org/10.1007/978-3-642-53980-0 199–234. [42] Vasluianu R, Forna DA, Zaltariov M, Murariu A. In vitro study [33] Brangule A, Gross KA. Importance of FTIR spectra using ATR-FTIR method for analyze the effects of the deconvolution for the analysis of amorphous calcium carbamide peroxide on the dental structure. Rev Chim phosphates. IOP Conf Ser Mater Sci Eng 2015:77. https://doi. 2016;67:2475–8. org/10.1088/1757-899X/77/1/012027 [43] Xu C, Reed R, Gorski JP, Wang Y, Walker MP. The distribution [34] Beniash E, Stifler CA, Sun CY, Jung GS, Qin Z, Buehler MJ, of carbonate in enamel and its correlation with structure et al. The hidden structure of human enamel. Nat Commun and mechanical properties. J Mater Sci 2012;47:8035–43. 2019;10:1–13. https://doi.org/10.1038/s41467-019-12185-7 https://doi.org/10.1007/s10853-012-6693-7 [35] Sakae T. Variations in dental enamel crystallites and micro- [44] González-López S, Torres-Rodríguez C, Bolaños-Carmona V, structure. J Oral Biosci 2006;48:85–93. https://doi.org/10.2330/ Sanchez-Sanchez P, Rodríguez-Navarro A, Álvarez-Lloret P, joralbiosci.48.85 et al. Effect of 30% hydrogen peroxide on mineral chemical [36] Verdelis K, Crenshaw MA, Paschalis EP, Doty S, Atti E, Boskey composition and surface morphology of bovine enamel. AL. Spectroscopic imaging of mineral maturation in bovine Odontology 2016;104:44–52. https://doi.org/10.1007/s10266- dentin. J Dent Res 2003;82:697–702. https://doi.org/10.1177/ 014-0189-7 154405910308200908 [45] Tran LK, Stepien KR, Bollmeyer MM, Yoder CH. Substitution [37] Pleshko N, Boskey A, Mendelsohn R. Novel infrared of sulfate in apatite. Am Miner 2017;102:1971–6. https://doi. spectroscopic method for the determination of crystallinity org/10.2138/am-2017-6088 of hydroxyapatite minerals. Biophys J 1991;60:786–93. [46] Young N, Fairley P, Mohan V, Jumeaux C. A study of https://doi.org/10.1016/S0006-3495(91)82113-0 hydrogen peroxide chemistry and photochemistry in tea [38] Sydney-Zax M, Mayer I, Deutsch D. Carbonate content in stain solution with relevance to clinical tooth whitening. J developing human and bovine enamel. J Dent Res Dent 2012;40:e11–6. https://doi.org/10.1016/j.jdent.2012.07. 1991;70:913–6. https://doi.org/10.1177/00220345910700051001 016 [39] Lussi A, Schlueter N, Rakhmatullina E, Ganss C. Dental [47] Mcguire JD, Walker MP, Dusevich V, Wang Y, Gorski JP. erosion - an overview with emphasis on chemical and Dentin enamel junction. Connect Tissue Res 2014;55:33–7. histopathological aspects. Caries Res 2011;45:2–12. https:// https://doi.org/10.3109/03008207.2014.923883 doi.org/10.1159/000325915