The protocol used in this study was partially published in our previous paper9, nevertheless all necessary details to are describe herein as well. The proposed method, like any in vitro approach, has inherent limitations due to the model-based nature of the study. The modified pH cycle employed was designed to simulate acid exposure experienced by the metal components of orthodontic appliances while preventing decalcification, as demonstrated in our previous work 9. The rationale behind sample selection, including tooth size and type, has also been detailed elsewhere 26. In this study, enzymatic and bacterial activity within the oral cavity was deliberately excluded. While clinical studies account for microbiome variability and enzymatic activity in saliva, they overlook the direct chemical alterations in enamel caused by metal ion release from orthodontic appliances. These aspects necessitate simplified in vitro investigations.
As we state in the Introduction section, our aim was to elucidate the accumulation process of metal ions from a materials science perspective—specifically, the local propensity of enamel to incorporate ions. Therefore, in the experiments described below, no orthodontic appliances were mounted on the teeth. As demonstrated in our previous study9, adhesive systems provide a tight seal that prevents enamel from assimilating metal ions released from orthodontic appliances, making direct observation of such changes under the brackets impossible. Furthermore, given that brackets are typically bonded at the centre of the tooth, their attachment to either the buccal or lingual surface, as per standard practice, would preclude a direct comparative analysis of these two regions which is the main objective of this study. During orthodontic treatment, however, certain areas of the tooth surface remain uncovered by brackets and are thus exposed to metal ion release. For instance, in buccal systems, the lingual surface remains unprotected, while in all appliance types, some molars retain exposed enamel on both sides. Therefore, we consider the model employed in this study to be appropriate for the defined research objectives.
Tooth samples
A total of 7 teeth extracted for orthodontic purposes were used for the experiments. All teeth were undamaged, unbleached molars, with no signs of caries or other diseases. The samples set consisted of four upper left molars (Tooth # 1, #2, #3 and #4), two lower left molars (Tooth #5 and #6) and one upper right molar (Control). All were obtained randomly from nonsmoking men and woman aged 25–45 years. Each sample came from a different donor. To understand the variability of chemical composition among samples, they were tested for Ca, Zn, Cu, Fe, Ni, Cr, Ti before experiments as provided in section “Experiments” and section “Control Experiments”.
Sampling and further experiments’ procedure was approved by Bioethics Committee at the District Medical Chamber in Krakow, Poland (approval number: 154/KBL/OIL/2016). All methods were performed in accordance with the guidelines and regulations provided by the Committee. As agreed, teeth were granted by the Department of Dental Surgery at the University Dental Clinic in Krakow (Poland), after patients’ informed consent. After extraction, the teeth were stored in formalin (10% p. Chempur, Piekary Śląskie Poland). They were then cleaned of soft tissues with 2% papain (Merck, Darmstadt, Germany), deionized water, and fluoride toothpaste. After the samples were thoroughly rinsed, they were dried in a laboratory oven (Heratherm; Thermo Fisher Scientific, MA, USA) for 16 h at 60 °C and stored in this form in a refrigerator (4 °C) until the experiments were conducted.
Orthodontic appliances
One set of orthodontic braces was used to conduct the experiment. It consisted of the following: 20 brackets (Batch number: 901.2032. LOT#072,717; Orthoclassic OR, USA), 4 bands (size: 37; Batch number: MSDS-054; Orthodontic Design and Production, CA, USA), 1 Ni‒Ti archwire (Batch number: 61.40.580.01725. LOT#302,127; Orthoclassic OR, USA) and 1 Ni‒Cr archwire (Batch number: 61.40.520.02125. LOT#071,111; Orthoclassic OR, USA). The orthodontic archwires were replaced in the reactors every 30 cycles simulating daily changes in oral pH.
Solutions and reagents
As part of the experiments, the solutions in the experimental reactors were changed daily. The solutions were prepared according to the modified recipe used in pH cycling experiments9. Remineralizing solution composition: 1.5 mmol CaCl2 × 2 H2O p.a. (Chempur, Piekary Śląskie Poland), 0.9 mmol KH2PO4 p.a. (Chempur, Piekary Śląskie Poland), 130 mmol KCl p.a. (Chempur, Piekary Śląskie Poland), 20 mmol HEPES bufor p.a. (4-(2-hydroxyethyl)−1-piperazine ethane sulfonic acid) (CARL ROTH, Karlsruhe, Germany),
and KOH p.a. to adjust pH = 7.0 (Chempur, Piekary Śląskie Poland). Demineralizing solution composition: 1.5 mmol CaCl2 × 2 H2O p.a. (Chempur, Piekary Śląskie Poland), 0.9 mmol KH2PO4 p.a. (Chempur, Piekary Śląskie Poland), acetic acid p.a. (50 mmol), and KOH p.a. to adjust pH = 4.3 (Chempur, Piekary Śląskie Poland).
Fluoride (0.0047 mmol NaF 99 + % (ACROS ORGANICS, Thermo Scientific Chemical, Delphi, India)), was added to the remineralization solution as it plays a crucial role in both enamel remineralization and metal corrosion41,42,43. The solutions were prepared using deionized water with a conductivity of 5 µS/cm produced by a R5Uv device (Hydrolab, Straszyn, Poland). In addition, a commercially available oral hygiene liquid (ELMEX; Colgate-Palmolive, Świdnica, Poland), recommended for oral hygiene during orthodontic treatment, was used in the daily pH cycles. The pH of the fluid, measured using a CPI-505 device with an EPS-1 electrode (Emeltron, Zabrze, Poland), was 4.50 ± 0.05.
The daily sequence of the solutions in the reactors was as follows: step 1. – demineralization (30 – 45 min); step 2. – hygiene: mouthwash (2 min.), tap water (10 s.), deionized water (10 s.); step 3 – remineralization (3 – 6 h); step 4 – demineralization (30—45 min.); step 5. – hygiene: mouthwash (2 min.), tap water (10 s.), deionized water (10 s.); step 6. – remineralization (16 h – 19.5 h). Demineralization and remineralization solutions’ volume was 100 ml, while mouthwash solution volume was 50 ml.
Experiments
After washing and cleaning the soft tissues (description in section “Tooth samples”), the lingual–buccal cross-sections of the experimental teeth (1 – 6) were exposed using sandpaper made of SiC and 1 micron Poly-Top-DUO Diamond slurry (Microdiamant, Lengwil, Switzerland) and analysed at selected sites (lateral lingual and buccal and crown lingual and buccal) using a scanning electron microscope with an energy dispersive spectrometer (SEM‒EDS) and a laser ablation inductively coupled plasma mass spectrometer (LA‒ICP‒MS) for: Ca, Zn, Cu, Ni, Ti, Cr, and Fe concentration. The results for Fe and Ca, the primary focus of this study, are presented in Figs. 1, 2, 3 (Tooth #1) and Supplementary Figs. S1–S17 (Teeth #2–#6). The analysis results for the remaining elements are available online in the Supplementary Excel file titled Analysis Before Experiments. For analytical details please see section “Control Experiments”. After the analyses, the exposed cross-section was secured with stickers (Color Coding Dots 3010; Avery Dennison Zweckform, München-Grünwald, Germany). This method effectively protects the enamel surface in experimental studies using a pH cycle and has been confirmed previously44. The stickers were replaced every 15 solution cycles.
The teeth were then placed in the reactor, along with the appliance set and underwent 360 cycles of pH changes. After termination of the pH cycles, the teeth were rinsed thoroughly with tap water (2 min) and deionized water (2 min). The exposed tooth cross-section was polished with 1 micron Poly-Top-DUO Diamond slurry (Microdiamant, Lengwil, Switzerland) to a depth of approximately 50 µm and then cleaned again with tap and deionized water and a cotton swab soaked in concentrated nitric acid (65% ultrapure; Merck, Darmstadt, Germany).
The samples prepared in this way were reanalysed using LA‒ICP‒MS at the same sites of the cross-section as before the experiments and then imaged using SEM‒EDS. The results for Fe and Ca, the primary focus of this study, are presented in Figs. 1, 2, 3 (Tooth #1) and Supplementary Figs. S1–S17 (Teeth #2–#6). The analysis results for the remaining elements are available online in the Supplementary Excel file titled Analysis After Experiments. For analytical details please see section “Analytical methods” and Supplementary Table S1.
Control experiments
The core objective of the experiment was to examine changes in the enamel’s chemical composition, specifically the accumulation and spatial distribution of Fe ions released from orthodontic appliances over time. To ensure rigorous experimental control, the initial chemical composition of each tooth was analysed before the experiment and re-examined at the same locations post-experiment, accounting for the laser penetration depth within the margin of error.
Control for the potential impact of pH cycling on enamel composition was based on a previously published experimental set conducted on 54 teeth divided into 107 samples9. These studies demonstrated no statistically significant differences in enamel composition between the control groups subjected either to pH cycles without appliances or to no exposure at all. However, to confirm additionally that Fe local distribution observed after the experiments in present experimental set primarily resulted from the presence of the orthodontic appliance component in solution, one tooth from the control group of the previous study9 was analysed using LA-ICP-MS profiling. The analysis was conducted on both the labial and crown lingual and buccal sites.
The selected tooth, originating from the control group (of previous study) titled Experiment #2, was bisected along the lingual‒buccal plane9. One half of the tooth underwent 360 pH cycles, while the other remained unexposed, serving as a reference for pre-experimental conditions. The procedures for profile preservation, analysis, and pH cycling were identical to those applied to the experimental samples9. Results of the analysis are presented in Supplementary Figs. S6 and S7 and in Supplementary Excel files titled Analysis Before Experiments and Analysis After Experiments.
Analytical methods
Imaging of the dental samples and analysis of the alloy composition of the parts of the orthodontic appliances were performed using a variable pressure field emission scanning electron microscope coupled with an FEI Quanta 200 FEG energy dispersive spectrometer (Thermo Fisher Scientific, OR, USA) at 20 kV (SEM‒EDS).
The analysis of the spatial distribution of Fe in the enamel was performed using LA‒ICP‒MS. For this purpose, an ICP‒MS NexION 300 spectrometer (Perkin Elmer, MA, USA) coupled with an LSX-213 laser ablation system (CETAC, NE, USA) was used. The operating parameters of the ICP‒MS measurement system and laser ablation conditions along with detection limits for Fe are listed in Supplementary Table S1. NIST SRM 1400 and NIST SRM 610 (Sigma–Aldrich, Merck, Steinheim am Albuch, Germany) were used as reference materials45. The method of making maps of element distribution has been described in previous publications46,47,48. Supplementary Table S1 presents detection limits for analysed elements.
Statistics
TIBICO Statistica version 13.3. and 14.0.0.15 (TIBICO, Palo Alto, CA, USA) was used for all statistical analyses in this study. The normal distribution was verified with a use of Kolmogorov–Smirnov (K–S) test with Lilliefors correction (K–S–L) and Shapiro–Wolf test (S–W). The Levene’s test was used to check the homogeneity of variances. The normal distribution data with homogeneous variances were subjected to Student’s t-test; a nonparametric U Manna-Whitney test with discontinuity correction was applied, otherwise. The statistical parameters including significance p values are presented in Tables 1 and 2. The correlation between metal concentrations within the first 80 laser measurement points from the enamel surface (approximately the first 600 µm in depth) was assessed using the Pearson test. Results are presented in Tables 3 and 4. Statistical analysis was conducted on the results obtained for the labial-lingual and buccal regions. Data collected from the crown regions were not included in the interpretation as they were beyond objectives of this study.
