Cleansability of zirconia implants after exposure to different contamination media: a methodological assessment Authors: Andrina Sophia Frank, Luiza Freitas Brum Souza, Tan Fırat Eyüboğlu, Mutlu Özcan Published: 2026 ABSTRACT This study evaluated the effects of different cleansing protocols on the wettability, surface free energy, and chemical composition of zirconia surfaces contaminated with saliva or metallic particles. Zirconia specimens (N = 90) were randomly assigned to three contamination conditions (none, saliva, metallic particles) and six cleansing protocols (C-control, A-autoclave, U-ultrasonic bath, E-ethanol, C-chlorhexidine, H2O2-hydrogen peroxide). Surface wettability was assessed by contact angle measurements with water and diiodomethane, and work of adhesion (WoA) and surface free energy (SFE) were calculated using the Owens-Wendt-Rabel-Kaelble model. Elemental composition was analyzed using energy-dispersive X-ray spectroscopy (EDS). Data were analyzed with two-way ANOVA and Tukey’s post hoc test (α = 0.05). Both contamination type and cleaning method significantly influenced contact angle, WoA, and SFE (p < 0.001), including their interaction. Saliva and metallic contamination decreased surface hydrophilicity and polar SFE. Chlorhexidine was most effective, restoring low contact angles and high SFE across all conditions. Ethanol and autoclave also performed well, especially in saliva-contaminated samples. Ultrasonic cleaning effectively removed metallic debris. Hydrogen peroxide results were inconsistent, notably after metallic contamination. Contamination alters zirconia surfaces; cleaning effectiveness varies with contaminant type and protocol. Chlorhexidine is the best cleaner. Ethanol, autoclave, and ultrasonic methods are alternatives, with ultrasonic especially effective against metals because of its mechanical cleaning. Effective zirconia implant cleaning is vital to restore wettability after contamination. Findings show hydrogen peroxide alone may be inadequate, highlighting the need for more reliable cleaning methods. Introduction Ceramic materials, especially zirconia, have become popular in reconstructive and implant dentistry. Zirconia implants are a promising alternative to titanium due to their tooth-like white color, enhancing aesthetics when visible through soft tissue1. First mentioned in medicine in the late 1960s2, zirconia gained attention for orthopedic uses3. Early dental studies from the 1990 s demonstrated good stability and osseointegration in animal models4. However, zirconia implants can face complications such as bone loss, often linked to surface contamination affecting osseointegration6. Dental implants depend on primary stability, achieved mechanically during insertion, and secondary stability from new bone growth around the implant, called osseointegration7. A key factor is the implant surface’s hydrophilicity. Right after implantation, water molecules and ions interact with the surface, changing its electrostatic pattern8, which helps plasma proteins like albumin, globulin, and fibrin adhere. These proteins are released due to vascular trauma9. Protein adsorption, crucial for cell adhesion and clot stability, depends on surface hydrophilicity and protein properties. Nakanishi et al. identified soft proteins, which change conformation to adsorb, and stiff proteins, needing electrostatic attraction10. Hydrophilic surfaces enhance protein adsorption and cell attachment, aiding bone regeneration. The fibrin network in blood clots serves as a scaffold for new bone7. Lang et al. found ultra-hydrophilic titanium surfaces promote faster osseointegration than standard ones11. Before protein adsorption can occur, blood must wet the implant surface, a property called wettability, which is assessed by measuring the contact angle between a liquid (e.g., water) and the solid. A low contact angle signifies high wettability, beneficial for osseointegration11. Wettability is typically evaluated with ultrapure water and diiodomethane; the latter, being nonpolar, allows calculation of surface free energy via the Owens–Wendt–Rabel–Kaelble (OWRK) model, distinguishing polar and dispersive surface components12. Contamination of zirconia surfaces by saliva or metallic particles from surgical tools, such as drill guide sleeves, can impair wettability. Saliva adsorbs organic molecules, reducing hydrophilicity and bond strength despite cleaning13. Surgical instruments may deposit metallic debris such as titanium, chromium, or iron, altering surface chemistry and potentially causing inflammation14,15. These surface changes can compromise wettability and chemical properties, thereby impacting clinical outcomes16. The objective of this study, therefore, was to provide clinicians with guidelines for selecting effective cleansing methods for contaminated zirconia implant surfaces. Specifically, this study evaluated whether different cleaning procedures influence the wettability, hydrophilicity, and elemental composition of zirconia implants compared with uncontaminated controls and identified the most effective cleaning approach for various types of contamination media. The null hypotheses tested were that (a) the presence of different contaminants would not affect the wettability of zirconia surfaces; and (b) different cleaning procedures would not alter the wettability of zirconia surfaces. Materials and methods Materials and study design The study used a factorial design with three contamination conditions (control/water, saliva, and metallic particles) and six cleaning protocols (control, autoclave, ultrasonic in water, ultrasonic in ethanol, chlorhexidine, and hydrogen peroxide), yielding 18 experimental groups. Each group comprised 27 zirconia specimens, for a total of 486 measurements. Details of the materials are presented in Table 1, and the full experimental workflow is shown in Fig. 1. Table 1General description (commercial name, manufacturer, chemical composition) of materials tested.MaterialManufacturerComposition (According to Manufacturer)E-max ZirCADIvoclar Vivadent AG, Schaan, LiechtensteinZrO2 (86.0–93.5 wt%)Y2O3 (> 6.5% – ≤ 8.0 wt%)HfO2 (≤ 5.0 wt%)Al2O3 (≤ 1.0 wt%)and other oxides (≤ 1.0 wt%)Artificial SalivaIn-lab preparation (according to Klimek)Artificial saliva, pH = 6.4T-sleeve (metal drill sleeves)Institut Straumann AG, Basel, SwitzerlandT-sleeve Ø 5 mm, H 6 mm, for guided surgery (metal)Diamond bursJota AG, Rüthi, SwitzerlandDiamond rotary instruments, Ref. 882.FG.016HydrogenPeroxideIn-lab preparationHydrogen peroxide (H2O2 5%)Chlorhexamed forteHaleon Schweiz AG, Rotkreuz, SwitzerlandChlorhexidine digluconate 0.2% (2 mg/mL)Diiodomethane (Methylene iodide)Sigma-Aldrich, St. Louis, MO, USACH2I2, ReagentPlus®, ≥ 99%, Cat. No. 15,842-9 Table 1General description (commercial name, manufacturer, chemical composition) of materials tested.MaterialManufacturerComposition (According to Manufacturer)E-max ZirCADIvoclar Vivadent AG, Schaan, LiechtensteinZrO2 (86.0–93.5 wt%)Y2O3 (> 6.5% – ≤ 8.0 wt%)HfO2 (≤ 5.0 wt%)Al2O3 (≤ 1.0 wt%)and other oxides (≤ 1.0 wt%)Artificial SalivaIn-lab preparation (according to Klimek)Artificial saliva, pH = 6.4T-sleeve (metal drill sleeves)Institut Straumann AG, Basel, SwitzerlandT-sleeve Ø 5 mm, H 6 mm, for guided surgery (metal)Diamond bursJota AG, Rüthi, SwitzerlandDiamond rotary instruments, Ref. 882.FG.016HydrogenPeroxideIn-lab preparationHydrogen peroxide (H2O2 5%)Chlorhexamed forteHaleon Schweiz AG, Rotkreuz, SwitzerlandChlorhexidine digluconate 0.2% (2 mg/mL)Diiodomethane (Methylene iodide)Sigma-Aldrich, St. Louis, MO, USACH2I2, ReagentPlus®, ≥ 99%, Cat. No. 15,842-9 General description (commercial name, manufacturer, chemical composition) of materials tested. ZrO2 (86.0–93.5 wt%) Y2O3 (> 6.5% – ≤ 8.0 wt%) HfO2 (≤ 5.0 wt%) Al2O3 (≤ 1.0 wt%) and other oxides (≤ 1.0 wt%) Hydrogen Peroxide Specimen preparation A total of 90 zirconia square specimens (Emax ZirCAD, Ivoclar Vivadent AG, Schaan, Liechtenstein; 15 × 15 × 1.5 mm³) were used in this study. All specimens were polished with a rotating polishing machine (LaboPol-21, Struers, Ballerup, Denmark) under continuous water irrigation. Each specimen was polished for 30 s on P#1000 waterproof silicon carbide paper (Struers), followed by another 30 s on P#1200 paper (Struers). After polishing, the specimens were ultrasonically cleaned (Telsonic TPC-25, Telsonic AG, Bronschhofen, Switzerland) in distilled water for 5 min at room temperature. The specimens were then air-dried and labeled according to the assigned contamination and cleaning protocols. Contamination The specimens were allocated into three groups: Control: non-contaminated; Saliva: contaminated with artificial saliva (pH = 6.40) according to Klimek et al. (1982) by immersion for 5 min, then air-dried17; Metal: metallic particles from drill guide sleeves (T-Sleeve, Institut Straumann AG, Basel, Switzerland) were manually milled into fine powder using a diamond drill head (Diamond 882.FG.016, Jota AG, Rüthi, Switzerland) also used in dentistry to mimic a clinical situation. The powder was collected and applied to the specimens in a standardized single coat using a brush. To improve reproducibility, all metallic particles were generated using the same drill sleeve and diamond bur under standardized milling conditions. The contamination procedure was performed by a single operator using a standardized single-coat brushing motion for all specimens. In addition, the consistency of metallic contamination was qualitatively verified by comparable Fe and Cr detection patterns in the EDS analyses prior to cleaning procedures. Cleansing procedures Specimens from the control and contaminated groups were allocated to five cleaning methods: C: control (remained in contaminated condition); A: specimens were sterilized in an autoclave (Device: Systec VX-150, Systec GmbH, Linden, Germany) at 121 °C for 30 min; U: specimens were cleaned in an ultrasonic bath (Telsonic TPC-25, Telsonic AG, Bronschhofen, Switzerland) with distilled water for 5 min; E: specimens were cleaned in an ultrasonic bath (Telsonic TPC-25, Telsonic AG, Bronschhofen, Switzerland) with 70% ethanol for 5 min; CHX: specimens were rinsed with Chlorhexamed forte 0,2% (2 mg/ml Chlorhexidini digluconas Haleon Schweiz AG, Rotkreuz, Switzerland) for 30 s; H2O2: specimens were rinsed with 5% hydrogen peroxide (H2O2) for 30 s. Energy-dispersive X-ray spectroscopy Before contamination, after contamination, and after each cleaning method, the elemental chemical composition of the specimens was measured using Energy Dispersive Spectroscopy (EDS) (TM4000Plus, Hitachi, Tokyo, Japan). EDS analyses were performed under standardized acquisition conditions, using consistent central surface regions for all specimens to minimize measurement variability. Surface morphology was assessed in secondary electron (SE) mode at an accelerating voltage of 15 kV and a working distance of approximately 10 mm. Elemental chemical composition was evaluated by energy-dispersive X-ray spectroscopy (EDS) using the integrated EDS system equipped with a silicon drift detector (SDD). EDS analyses were performed in mapping mode with an acquisition time of 10 min to qualitatively assess the spatial distribution of elements on the specimen surfaces. Analyses were conducted without sputter coating. Wettability and surface free energy For the second test method, surface wettability was evaluated by measuring the static contact angle (θ) with a contact angle goniometer (Krüss Easy Drop, Krüss GmbH, Hamburg, Germany) and Drop Shape Analysis software. Each specimen was first analyzed with distilled water and then with 99% diiodomethane, both dispensed as 1 µL droplets. These test liquids were selected from the software’s internal Liquid Database, which automatically applies the corresponding physical parameters. For each liquid, the software calculated the contact angle (θ) and the work of adhesion (WoA) using the following equation:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{W}\mathrm{o}\mathrm{A}={\gamma}_{\mathrm{L}}(1+\mathrm{cos}{\uptheta\:})$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\gamma _{{\mathrm{L}}}$$\end{document} is the liquid’s surface tension, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\uptheta\:}$$\end{document} is the contact angle between the droplet and the solid surface. By performing the measurements with two liquids of known surface tension components (one polar, one non-polar), a system of two equations is obtained and solved to determine the surface energy components for each specimen. The surface free energy of the solid (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma}_{\mathrm{S}}$$\end{document}) and its components were calculated using the Owens-Wendt-Rabel-Kaelble (OWRK) model, which separates the total surface energy into a dispersive component (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{S}}^{\mathrm{d}}$$\end{document}) and a polar component (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{S}}^{\mathrm{p}}$$\end{document}). The model assumes that:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma}_{\mathrm{S}}={{\upgamma\:}}_{\mathrm{S}}^{\mathrm{d}}+{{\upgamma\:}}_{\mathrm{S}}^{\mathrm{p}}$$\end{document} and is based on the equation:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma}_{\mathrm{L}}\left(1+\mathrm{cos}{\uptheta\:}\right)=2\left(\sqrt{{{\upgamma\:}}_{\mathrm{S}}^{\mathrm{d}}{{\upgamma\:}}_{\mathrm{L}}^{\mathrm{d}}}+\sqrt{{{\upgamma\:}}_{\mathrm{S}}^{\mathrm{p}}{{\upgamma\:}}_{\mathrm{L}}^{\mathrm{p}}}\right)$$\end{document} where: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{L}}^{\mathrm{d}}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{L}}^{\mathrm{p}}$$\end{document} are the dispersive and polar components of the liquid’s surface tension, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{S}}^{\mathrm{d}}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{S}}^{\mathrm{p}}$$\end{document} are the unknown dispersive and polar components of the solid’s surface energy. The test liquids used and their surface tension properties were: Distilled water: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma}_{\mathrm{L}}$$\end{document}= 72.8mN/m, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{L}}^{\mathrm{d}}$$\end{document} = 26.4 mN/m, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{L}}^{\mathrm{p}}$$\end{document} = 46.4 mN/m; Diiodomethane (99%): \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma}_{\mathrm{L}}$$\end{document}= 50.8 mN/m, with \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{L}}^{\mathrm{d}}$$\end{document} = 50.8 mN/m and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upgamma\:}}_{\mathrm{L}}^{\mathrm{p}}$$\end{document} = 0 mN/m. All measurements were conducted under controlled ambient conditions. Results were reported as the mean of nine measurements per group, and the total surface energy \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{(\gamma}}_{\mathrm{S}})$$\end{document}, along with its polar and dispersive components, was reported in mN/m. Statistical analysis Because the data for contact angle, work of adhesion, and surface free energy met the assumptions of normality and homogeneity of variance, as verified by the Shapiro-Wilk and Levene’s tests (p > 0.05), a 2-way analysis of variance (ANOVA) was performed (α = 0.05) using SPSS version 21 (IBM Analytics, New York, USA). This analysis evaluated the effects of contaminants and cleaning methods, with pairwise comparisons conducted using Tukey’s post-hoc test. Elemental composition data were analyzed descriptively and qualitatively. Results The results were significantly influenced by contamination type, cleansing protocol, and their interaction, as shown by all evaluated parameters. For water contact angle, two-way ANOVA revealed statistically significant effects of contamination (F = 112.75, p < 0.001, η²=0.58), cleaning method (F = 148.26, p < 0.001, η²=0.63), and their interaction (F = 13.51, p < 0.001), indicating that both contamination type and cleaning method accounted for a substantial proportion of the observed variance (Table 2). Table 2Means and standard deviations for Contact angle (degrees) and Work of Adhesion (mN/m) measurements.ContaminantCleanerWaterDiiodomethaneContact Angle (θ)Mean ± SDWork of Adhesion (WoA)Mean ± SDContact Angle (θ)Mean ± SDWork of Adhesion (WoA)Mean ± SDControlC71.66 ±0.69 BC95.70 ±2.48 FGH44.29 ±0.81 CDE87.13 ±0.80 BCDEA54.73 ±2.56 DE114.51 ±8.07 CD37.32 ±0.86 FGH91.16 ±0.22 ABU65.96 ±5.83 BCD100.72 ±15.80 EFG46.64 ±0.89 ABCD85.64 ±0.69 DEE65.32 ±2.78 BCD102.92 ±9.66 DEFG40.30 ±0.61 EFG89.53 ±0.50 ABCDCHX19.34 ±2.49 G140.97 ±3.18 A36.52 ±1.18 FGH91.57 ±0.96 ABH2O276.46 ±1.33 AB89.81 ±4.92 GHI46.31 ±0.59 ABCD85.88 ±0.45 CDESalivaC14.53 ±0.51 G143.25 ±0.53 A40.37 ±0.68 EF89.50 ±0.18 ABCDA37.32 ±3.23 F129.97 ±7.59 AB39.99 ±1.08 EFG89.67 ±0.74 ABCDU61.32 ±2.45 CDE107.48 ±8.20 CDEF44.91 ±0.84 BCDE86.75 ±0.46 BCDEE48.94 ±2.41 EF120.27 ±6.87 BC39.92 ±0.38 EFG89.75 ±0.18 ABCDCHX15.78 ±2.11 G142.48 ±2.25 A34.50 ±2.62 H92.32 ±2.32 AH2O265.90 ±1.97 BCD102.39 ±6.81 DEFG43.94 ±1.02 CDE87.45 ±1.19 ABCDEMetallicparticlesC65.96 ±5.83 BCD100.72 ±15.80 EFG34.97 ±1.31 GH92.34 ±0.76 AA57.70 ±4.12 CDE110.93 ±13.38 CDE41.61 ±1.19 DEF88.73 ±0.99 ABCDU78.89 ±0.83 AB86.82 ±3.11 HI50.91 ±0.53 A82.82 ±0.48 EE76.49 ±2.64 AB89.66 ±9.69 GHI48.30 ±1.33 ABC85.23 ±1.73 DECHX21.78 ±1.79 G140.14 ±2.39 A38.44 ±0.84 FGH90.55 ±0.61 ABCH2O286.70 ±0.90 A76.99 ±3.41 I49.69 ±0.85 AB83.67 ±0.37 E*Uppercase letters indicate statistical differences between different contaminants and cleaners by Two-way ANOVA and Tukey HSD post-hoc tests (α = 0.05). C: Control; A: Autoclave; U: Ultrasonic bath; E: Ethanol; CHX: Chlorhexidine; H2O2: Hydrogen peroxide. Table 2Means and standard deviations for Contact angle (degrees) and Work of Adhesion (mN/m) measurements.ContaminantCleanerWaterDiiodomethaneContact Angle (θ)Mean ± SDWork of Adhesion (WoA)Mean ± SDContact Angle (θ)Mean ± SDWork of Adhesion (WoA)Mean ± SDControlC71.66 ±0.69 BC95.70 ±2.48 FGH44.29 ±0.81 CDE87.13 ±0.80 BCDEA54.73 ±2.56 DE114.51 ±8.07 CD37.32 ±0.86 FGH91.16 ±0.22 ABU65.96 ±5.83 BCD100.72 ±15.80 EFG46.64 ±0.89 ABCD85.64 ±0.69 DEE65.32 ±2.78 BCD102.92 ±9.66 DEFG40.30 ±0.61 EFG89.53 ±0.50 ABCDCHX19.34 ±2.49 G140.97 ±3.18 A36.52 ±1.18 FGH91.57 ±0.96 ABH2O276.46 ±1.33 AB89.81 ±4.92 GHI46.31 ±0.59 ABCD85.88 ±0.45 CDESalivaC14.53 ±0.51 G143.25 ±0.53 A40.37 ±0.68 EF89.50 ±0.18 ABCDA37.32 ±3.23 F129.97 ±7.59 AB39.99 ±1.08 EFG89.67 ±0.74 ABCDU61.32 ±2.45 CDE107.48 ±8.20 CDEF44.91 ±0.84 BCDE86.75 ±0.46 BCDEE48.94 ±2.41 EF120.27 ±6.87 BC39.92 ±0.38 EFG89.75 ±0.18 ABCDCHX15.78 ±2.11 G142.48 ±2.25 A34.50 ±2.62 H92.32 ±2.32 AH2O265.90 ±1.97 BCD102.39 ±6.81 DEFG43.94 ±1.02 CDE87.45 ±1.19 ABCDEMetallicparticlesC65.96 ±5.83 BCD100.72 ±15.80 EFG34.97 ±1.31 GH92.34 ±0.76 AA57.70 ±4.12 CDE110.93 ±13.38 CDE41.61 ±1.19 DEF88.73 ±0.99 ABCDU78.89 ±0.83 AB86.82 ±3.11 HI50.91 ±0.53 A82.82 ±0.48 EE76.49 ±2.64 AB89.66 ±9.69 GHI48.30 ±1.33 ABC85.23 ±1.73 DECHX21.78 ±1.79 G140.14 ±2.39 A38.44 ±0.84 FGH90.55 ±0.61 ABCH2O286.70 ±0.90 A76.99 ±3.41 I49.69 ±0.85 AB83.67 ±0.37 E*Uppercase letters indicate statistical differences between different contaminants and cleaners by Two-way ANOVA and Tukey HSD post-hoc tests (α = 0.05). C: Control; A: Autoclave; U: Ultrasonic bath; E: Ethanol; CHX: Chlorhexidine; H2O2: Hydrogen peroxide. Means and standard deviations for Contact angle (degrees) and Work of Adhesion (mN/m) measurements. Metallic particles *Uppercase letters indicate statistical differences between different contaminants and cleaners by Two-way ANOVA and Tukey HSD post-hoc tests (α = 0.05). C: Control; A: Autoclave; U: Ultrasonic bath; E: Ethanol; CHX: Chlorhexidine; H2O2: Hydrogen peroxide. In general, CHX produced the highest surface wettability across all conditions, yielding the lowest contact angles, whereas H₂O₂ tended to produce higher contact angles, especially after metallic particle contamination. The effectiveness of cleansing methods depended on the type of prior contamination, with U and E treatments showing intermediate behavior (Fig. 2; Table 2). The work of adhesion (WoA) derived from water contact angles was also significantly affected by contamination (F = 64.95, p < 0.001, η²0.44), cleansing protocol (F = 36.42, p < 0.001, η²=0.49), and their interaction (F = 8.20, p < 0.001), indicating that both main factors explained a substantial proportion of variance in WoA outcomes (Fig. 3; Table 2). Similarly, the diiodomethane contact angle showed significant differences based on contamination (F = 14.80, p < 0.001, η²=0.31), cleaning method (F = 47.15, p < 0.001, η²=0.52), and their interaction (F = 9.02, p < 0.001), indicating a substantial contribution of both factors to dispersive surface interactions (Table 2). Corresponding WoA values from diiodomethane contact angles followed a similar pattern, with significant effects of all factors (contamination: F = 7.27, p = 0.002, η²=0.24; cleaner: F = 22.22, p < 0.001, η²=0.41; interaction: F = 3.98, p = 0.001), indicating that the observed differences were driven by meaningful proportions of explained variance (Table 2). The surface free energy (SFE) was significantly influenced by contamination, cleansing method, and their interaction. This effect was consistent across all SFE components. The dispersive component (σᵈ) was significantly affected by contamination (F = 17.65, p < 0.001, η²=0.29), cleaning protocol (F = 51.77, p < 0.001, η²=0.54), and their interaction (F = 9.92, p < 0.001), indicating a substantial influence of cleaning-related variance on dispersive surface energy. In the control condition, CHX and autoclave treatment (A) produced the highest dispersive surface energy values, whereas ultrasonic cleaning (U) and H₂O₂ showed the lowest σᵈ values. After saliva contamination, CHX maintained the highest dispersive component, while U and H₂O₂ resulted in reduced σᵈ values. Following metallic particle contamination, the control and CHX groups exhibited the highest dispersive components, whereas ethanol (E) and H₂O₂ showed the lowest values, indicating that some cleaning protocols reduced non-polar surface interactions after metallic contamination (Table 3). Table 3Means and standard deviations of surface energy parameters in mN/m.ContaminantCleanerσd (DispersiveComponent)σp (PolarComponent)σ (Total Surface Energy)ControlC37.38 ±0.43 CDE0.34 ±0.01 FGHI37.71 ±0.43 FGA40.91 ±0.42 AB0.69 ±0.06 DE41.59 ±0.42 ABCDU36.11 ±0.48 EFG0.50 ±0.14 EFGH36.61 ±0.50 GHE39.45 ±0.31 ABC0.46 ±0.06 EFGH39.91 ±0.34 CDEFCHX41.27 ±0.57 AB1.48 ±0.05 AB42.75 ±0.54 AH2O236.30 ±0.32 DEFG0.26 ±0.02 HI36.55 ±0.33 GHISalivaC39.41 ±0.34 ABC1.61 ±0.01 A41.02 ±0.33 ABCDA39.58 ±0.55 ABC1.15 ±0.07 BC40.73 ±0.58 BCDU37.02 ±0.45 CDEF0.59 ±0.07 DEFG37.63 ±0.42 FGE39.65 ±0.19 ABC0.86 ±0.06 CD40.50 ±0.22 BCDECHX41.99 ±1.11 A1.51 ±0.04 A43.50 ±1.08 AH2O237.55 ±0.54 CDE0.47 ±0.05 EFGH38.02 ±0.53 EFGMetallicparticlesC41.98 ±0.60 A0.41 ±0.14 EFGHI42.40 ±0.60 ABCA38.75 ±0.61 BCD0.66 ±0.10 DEF39.42 ±0.59 DEFU33.76 ± 0.30 G0.24 ±0.01 HI34.00 ±0.30 IE35.18 ±0.72 EFG0.28 ±0.05 GHI35.47 ±0.73 GHICHX40.36 ±0.41 AB1.47 ±0.03 AB41.83 ±0.41 ABCDH2O234.44 ±0.47 FG0.11 ±0.01 I34.55 ±0.47 HI*Uppercase letters indicate statistical differences between different contaminants and cleaners according to two-way ANOVA and Tukey`s HSD post-hoc tests (α = 0.05). Table 3Means and standard deviations of surface energy parameters in mN/m.ContaminantCleanerσd (DispersiveComponent)σp (PolarComponent)σ (Total Surface Energy)ControlC37.38 ±0.43 CDE0.34 ±0.01 FGHI37.71 ±0.43 FGA40.91 ±0.42 AB0.69 ±0.06 DE41.59 ±0.42 ABCDU36.11 ±0.48 EFG0.50 ±0.14 EFGH36.61 ±0.50 GHE39.45 ±0.31 ABC0.46 ±0.06 EFGH39.91 ±0.34 CDEFCHX41.27 ±0.57 AB1.48 ±0.05 AB42.75 ±0.54 AH2O236.30 ±0.32 DEFG0.26 ±0.02 HI36.55 ±0.33 GHISalivaC39.41 ±0.34 ABC1.61 ±0.01 A41.02 ±0.33 ABCDA39.58 ±0.55 ABC1.15 ±0.07 BC40.73 ±0.58 BCDU37.02 ±0.45 CDEF0.59 ±0.07 DEFG37.63 ±0.42 FGE39.65 ±0.19 ABC0.86 ±0.06 CD40.50 ±0.22 BCDECHX41.99 ±1.11 A1.51 ±0.04 A43.50 ±1.08 AH2O237.55 ±0.54 CDE0.47 ±0.05 EFGH38.02 ±0.53 EFGMetallicparticlesC41.98 ±0.60 A0.41 ±0.14 EFGHI42.40 ±0.60 ABCA38.75 ±0.61 BCD0.66 ±0.10 DEF39.42 ±0.59 DEFU33.76 ± 0.30 G0.24 ±0.01 HI34.00 ±0.30 IE35.18 ±0.72 EFG0.28 ±0.05 GHI35.47 ±0.73 GHICHX40.36 ±0.41 AB1.47 ±0.03 AB41.83 ±0.41 ABCDH2O234.44 ±0.47 FG0.11 ±0.01 I34.55 ±0.47 HI*Uppercase letters indicate statistical differences between different contaminants and cleaners according to two-way ANOVA and Tukey`s HSD post-hoc tests (α = 0.05). Means and standard deviations of surface energy parameters in mN/m. Metallic particles *Uppercase letters indicate statistical differences between different contaminants and cleaners according to two-way ANOVA and Tukey`s HSD post-hoc tests (α = 0.05). Likewise, the polar component (σᵖ) showed highly significant sensitivity to all factors (contamination: F = 94.61, p < 0.001, η²=0.56; cleaner: F = 120.08, p < 0.001, η²=0.67; interaction: F = 13.51, p < 0.001), indicating that a large proportion of variance in polar surface energy was attributable to these effects. CHX consistently produced the highest σᵖ values across all contamination conditions, particularly after saliva contamination, indicating increased polar surface interactions and enhanced surface wettability. In contrast, H₂O₂ and ultrasonic cleaning generally resulted in the lowest σᵖ values, especially after metallic particle contamination, suggesting reduced polar contribution to the surface free energy. Saliva contamination increased σᵖ values in the untreated control specimens, whereas metallic contamination markedly reduced the polar component in most groups except CHX, which maintained elevated polar surface energy (Table 3). The total surface energy (σ) also showed significant differences by contamination (F = 28.74, p < 0.001, η²=0.38), cleaner (F = 73.18, p < 0.001, η²=0.59), and their interaction (F = 10.12, p < 0.001), indicating that both factors accounted for substantial variability in total surface energy (Fig. 4). Elemental analysis (EDS) showed that all specimens retained the characteristic composition of yttria-stabilized zirconia (Zr, O, Y, Hf), regardless of contamination or cleaning protocol. In the control group (no contamination), only minor variations in elemental percentages were observed across cleaning methods, and no extraneous elements were detected, confirming that none of the protocols altered the zirconia chemistry in the absence of contaminants (Fig. 5). Following saliva contamination, residual elements including Cl, Na, K, P, and Al were present in the control (C), autoclave (A), and ultrasonic (U) groups. In contrast, ethanol (E), chlorhexidine (CHX), and hydrogen peroxide (H2O2) effectively eliminated these residues, restoring a clean elemental profile (Fig. 6). Following metallic particle contamination, iron (Fe) and chromium (Cr) were detected on the surfaces of the control (C) and autoclave (A) groups. However, ultrasonic (U), ethanol (E), and chlorhexidine (CHX) completely removed these metallic residues. H2O2, although effective in other conditions, still left traces of iron, suggesting limited efficacy in this context (Fig. 7). Discussion This study evaluated the effectiveness of various cleaning protocols on contaminated zirconia surfaces by analyzing wettability, surface free energy (SFE), and chemical composition. Two null hypotheses were tested: (1) that the presence of different contaminants would not affect the wettability of zirconia surfaces, and (2) that different cleansing procedures would not alter zirconia wettability. Based on the findings, both hypotheses were rejected. Regarding the first hypothesis, contamination type significantly influenced all analyzed surface parameters. Both saliva and metallic particles increased contact angles, reduced the work of adhesion (WoA), and decreased surface energy, particularly the polar component. The large effect sizes observed across wettability, WoA, and SFE indicate that contamination accounted for a substantial proportion of surface variability. These findings confirm that contamination compromises physicochemical characteristics essential for early osseointegration. Organic macromolecules from saliva13 and inorganic debris from surgical instruments14,15 can hinder protein adsorption and alter surface electrostatic profiles8,9, thereby reducing biological reactivity. Therefore, the presence of contaminants refutes the first null hypothesis. The second hypothesis was also rejected because different cleansing protocols produced significantly distinct outcomes across all contamination conditions. Wettability, WoA, and both polar and dispersive components of surface energy were strongly dependent on the cleaning method. Consistently large effect sizes demonstrate that the cleaning protocol was a dominant contributor to variance in surface properties. These differences were confirmed by two-way ANOVA results (p < 0.001), indicating that zirconia surface properties can be restored or compromised depending on the cleaning strategy. Among the tested protocols, chlorhexidine (CHX) consistently improved zirconia surface properties, showing the lowest water contact angles and highest SFE values, with recovery of the polar component crucial for protein adhesion and clot formation11. The very large effect sizes associated with CHX indicate a substantial shift in surface behavior rather than a marginal improvement. These findings align with previous reports that CHX removes organic contamination and enhances hydrophilicity11. In addition, the amphiphilic cationic nature of chlorhexidine digluconate may itself contribute to reduced contact angles and increased surface free energy through adsorption onto the zirconia surface and modification of interfacial interactions11,19. However, the present study evaluated only physicochemical surface characteristics, and the biological implications of these interactions were not investigated. Autoclaving yielded favorable outcomes, particularly in saliva-contaminated specimens. Although less effective than CHX, autoclaving increased polar and total SFE and was associated with clean elemental profiles. Moderate-to-large effect sizes suggest a meaningful but less pronounced contribution to surface restoration. Thermal sterilization may therefore promote partial surface reactivation18, although effects on surface microtopography were not evaluated. Ethanol performed reliably in both saliva- and metal-contaminated scenarios. Its effectiveness was confirmed by EDS analysis, and it maintained intermediate SFE values. Moderate effect sizes indicate a consistent but limited impact on surface reactivation, supporting ethanol as a clinically accessible cleaner for general decontamination19. On the other hand, ultrasonic cleaning showed a mixed profile. It effectively removed metallic residues but was less effective in restoring hydrophilicity and polar surface energy. Observed moderate effect sizes indicate that its impact was primarily mechanical rather than chemical. Cavitation from high-frequency vibrations dislodges inorganic particles without relying on chemical interactions20. Only one exposure protocol was tested, and alternative parameters may yield different outcomes. Hydrogen peroxide (H₂O₂) performed least favorably. Although capable of removing saliva-related contaminants, it failed to eliminate metallic residues and consistently yielded lower wettability and SFE, particularly the polar component. Smaller effect sizes indicate a limited contribution to surface restoration, which may explain its inconsistent performance. These findings may relate to the limited reactivity of H₂O₂ with inorganic contaminants and interference with the zirconia oxide layer21. Overall, CHX was the most effective and broadly applicable cleaning agent, followed by autoclaving and ethanol, while ultrasonic cleaning was best suited for metallic debris. Differences in effect size magnitude support the hierarchical performance observed among protocols. However, EDS analysis was qualitative, and quantitative elemental comparison was beyond the study scope. Elemental analysis corroborated these findings, showing that while all specimens retained characteristic zirconia elements (Zr, O, Y, Hf), contaminants such as Cl, Na, K, and Fe persisted after inadequate cleaning. CHX, ethanol, and ultrasonic cleaning were most effective in removing these residues, particularly after metallic contamination. Effect size patterns emphasize that both chemical and mechanical decontamination contributed meaningfully to surface composition differences. Taken together, these findings reinforce the importance of preserving or restoring zirconia surface hydrophilicity and chemical purity. High polar surface energy and a clean chemical composition are essential for protein adsorption, fibrin formation, and bone healing7,8,11. Large effect sizes across key parameters indicate robust alterations in surface behavior, although results should be interpreted as physicochemical indicators rather than direct predictors of biological outcomes. The in vitro design and absence of biological testing limit clinical extrapolation. Flat zirconia specimens were used instead of implant surfaces, and contamination models did not fully replicate surgical conditions. Only single concentrations and exposure times were tested. Surface roughness, long-term stability, and quantitative elemental analysis were not assessed. Future studies should include clinically relevant implant surfaces, biological outcomes, and aging protocols to enhance translational relevance. Conclusions From this study, the following could be concluded:Contamination of zirconia surfaces with saliva or metallic particles can significantly change their wettability and surface energy, especially by decreasing the polar component crucial for osseointegration.The effectiveness of cleaning methods varies depending on the protocol used, with chlorhexidine (CHX) offering the most reliable restoration of surface hydrophilicity and contaminant removal.Autoclaving and ethanol also performed well, while ultrasonic cleaning was mainly effective against metallic debris.Hydrogen peroxide (H₂O₂), on the other hand, showed limited and inconsistent results. Overall, these differences highlight notable changes in the surface characteristics of zirconia. Contamination of zirconia surfaces with saliva or metallic particles can significantly change their wettability and surface energy, especially by decreasing the polar component crucial for osseointegration. Contamination of zirconia surfaces with saliva or metallic particles can significantly change their wettability and surface energy, especially by decreasing the polar component crucial for osseointegration. The effectiveness of cleaning methods varies depending on the protocol used, with chlorhexidine (CHX) offering the most reliable restoration of surface hydrophilicity and contaminant removal. The effectiveness of cleaning methods varies depending on the protocol used, with chlorhexidine (CHX) offering the most reliable restoration of surface hydrophilicity and contaminant removal. Autoclaving and ethanol also performed well, while ultrasonic cleaning was mainly effective against metallic debris. Autoclaving and ethanol also performed well, while ultrasonic cleaning was mainly effective against metallic debris. Hydrogen peroxide (H₂O₂), on the other hand, showed limited and inconsistent results. Overall, these differences highlight notable changes in the surface characteristics of zirconia. Hydrogen peroxide (H₂O₂), on the other hand, showed limited and inconsistent results. Overall, these differences highlight notable changes in the surface characteristics of zirconia. Fig. 1Flowchart of experimental design. Flowchart of experimental design. Fig. 2Means and standard deviations for water Contact angle (degrees). Uppercase letters indicate statistical differences between cleaners considering each contaminant: (a) Control; (b) Saliva; and (c) Metallic particles (1-way ANOVA and Tukey’s HSD post-hoc tests (α = 0.05). Means and standard deviations for water Contact angle (degrees). Uppercase letters indicate statistical differences between cleaners considering each contaminant: (a) Control; (b) Saliva; and (c) Metallic particles (1-way ANOVA and Tukey’s HSD post-hoc tests (α = 0.05). Fig. 3Means and standard deviations for water Work of Adhesion (mN/m). Uppercase letters indicate statistical differences between cleaners considering each contaminant: (a) Control; (b) Saliva; and (c) Metallic particles (1-way ANOVA and Tukey’s HSD post-hoc tests (α = 0.05). Means and standard deviations for water Work of Adhesion (mN/m). Uppercase letters indicate statistical differences between cleaners considering each contaminant: (a) Control; (b) Saliva; and (c) Metallic particles (1-way ANOVA and Tukey’s HSD post-hoc tests (α = 0.05). Fig. 4Means and standard deviations for Total Surface Energy (mN/m). Uppercase letters indicate statistical differences between cleaners considering each contaminant: (a) Control; (b) Saliva; and (c) Metallic particles (1-way ANOVA and Tukey`s HSD post-hoc tests (α = 0.05). Means and standard deviations for Total Surface Energy (mN/m). Uppercase letters indicate statistical differences between cleaners considering each contaminant: (a) Control; (b) Saliva; and (c) Metallic particles (1-way ANOVA and Tukey`s HSD post-hoc tests (α = 0.05). Fig. 5Representative EDS spectra, and elemental composition (wt%) of zirconia specimens after different cleaning protocols without prior contamination. All surfaces presented only expected elements (Zr, O, Y, Hf) with no evidence of external residues. Representative EDS spectra, and elemental composition (wt%) of zirconia specimens after different cleaning protocols without prior contamination. All surfaces presented only expected elements (Zr, O, Y, Hf) with no evidence of external residues. Fig. 6Representative EDS spectra, and elemental composition (wt%) of zirconia specimens after contamination with saliva and subsequent cleaning. Residual elements such as Cl, Na, K, P, and Al were detected in the control, autoclave, and ultrasonic groups. In contrast, ethanol (E), CHX, and H2O2 showed only the characteristic zirconia elements. Representative EDS spectra, and elemental composition (wt%) of zirconia specimens after contamination with saliva and subsequent cleaning. Residual elements such as Cl, Na, K, P, and Al were detected in the control, autoclave, and ultrasonic groups. In contrast, ethanol (E), CHX, and H2O2 showed only the characteristic zirconia elements. Fig. 7Representative EDS spectra, and elemental composition (wt%) of zirconia specimens after contamination with metallic particles and cleaning. Iron (Fe) and chromium (Cr) residues were observed in the control and autoclave groups. Ethanol (E), CHX, and ultrasonic (U) cleaning effectively eliminated these contaminants, while H2O2 showed residual iron, indicating partial decontamination. Representative EDS spectra, and elemental composition (wt%) of zirconia specimens after contamination with metallic particles and cleaning. Iron (Fe) and chromium (Cr) residues were observed in the control and autoclave groups. Ethanol (E), CHX, and ultrasonic (U) cleaning effectively eliminated these contaminants, while H2O2 showed residual iron, indicating partial decontamination.