Selective effect of atmospheric non-thermal plasma on resin cement shear bond strength to glass ceramics Authors: Emine Mustafaoğlu, Utku Kürşat Ercan, Gizem Dilara Özdemir, Bora Bagis Published: 2026 ABSTRACT AbstractStatement of problemThe effectiveness of non-thermal plasma (NTP) in improving the bond strength of different resin cements to different glass ceramics under various surface treatment protocols remains unclear. Moreover, the influence of specific plasma devices and the material-dependent response of ceramic–resin cement systems to NTP have not been sufficiently investigated.PurposeThe purpose of this in vitro study was to evaluate the effect of NTP on the shear bond strength of different resin cements to different glass ceramics by comparatively assessing multiple surface treatment protocols.Materials and methodsA total of 240 specimens of each glass ceramic group (IPS e.max CAD (E), Vitamark II (V)) were divided into 24 groups according to the combination of 6 surface treatments ((control), hydrofluoric acid (HF)+ceramic primer (Pr), Pr, HF+plasma (Pl) + Pr, Pl + Pr, and Pl) and two different resin cements (Single Bond+RelyX Ultimate (R) and ZenitCem (Z)). NTP was applied with an AC microsecond pulse power supply (5 kHz, 20 kV, 90 s). Specimens were cemented to composite cylinders. After storage in an oven, SBS was measured. The results were analyzed using three-way ANOVA, post-hoc tests, and targeted pairwise comparisons, with the significance level set at α = 0.05.The effectiveness of NTP is material- and cement-dependent and should be considered as a selective surface modification strategy rather than a universal alternative to conventional surface treatments.ResultsThe V-Pl-Z group showed significantly higher SBS values than the V-Z group, and these values were within the clinically acceptable range (p < 0.05). In the V groups treated with HF + Pr and HF + Pl+Pr, R demonstrated higher bond strength values than Z. HF surface treatment increased the SBS values for all groups (p < 0.05).ConclusionsThe effectiveness of NTP is material- and cement-dependent and should be considered as a selective surface modification strategy rather than a universal alternative to conventional surface treatmentsClinical significanceThis study demonstrates that atmospheric non-thermal plasma does not universally enhance ceramic–resin cement bonding. Its effectiveness depends on the ceramic substrate, resin cement type, and surface treatment protocol. Clinicians should therefore apply plasma selectively.Keywords: Nonthermal plasma, Plasma medicine, Resin, Ceramic, Bond strength Introduction Advances in adhesive dentistry have led to the greater use of all-ceramic restorations because of their improved aesthetic and biocompatibility properties compared with metal-supported restorations [1, 2]. Even though ceramic materials are popular in dental practice [3–8], their clinical performance is limited by their tensile strength [1]. The adhesion potential of resin cements to ceramic restorations and tooth structures is a key factor to improve the performance of dental restorations [9, 10]. An increasing number of different surface treatments have been discussed and studied to enhance dental ceramics’ clinical performance [11–12]. Surface treatments with acid etching and silane-coupling agents are commonly used applications prior to cementing ceramic restorations [13]. Hydrofluoric (HF) acid etching is a reliable protocol for the bonding of glass ceramics with resin cements [14, 15]. Some authors, however, state that surface treatment with HF acid negatively affects the mechanical properties of etched dental ceramics because of its potential for creating surface defects [16–19]. This indicates the necessity of innovative strategies to develop bonding procedures [20–22]. Plasma is the fourth state of matter [23, 24]. It is composed of reactive species, ions, electrons, free radicals, and neutrals [24, 25]. Such composition leads to modifications to plasma-treated surfaces, which may result in a reactive surface that increase the surface hydrophilicity and cement wetting capacity, without significantly affecting the bulk properties of dental ceramics [9, 26–33]. Nonthermal plasma (NTP) treatment has been suggested as an alternative to HF acid etching that can improve the adhesion effectiveness and durability between ceramic surfaces and resin cements [6, 26, 34]. In dentistry, NTP treatment is a new technique that is used for inducing the surface modification of titanium implants [35, 36], improving the mechanical properties of fiber-reinforced acrylic resins [37], and improving the adhesion effectiveness and durability between fiber post surfaces and core resins [38] and between resin cements and ceramic surfaces [39]. Previous studies have shown that the effects of non-thermal plasma (NTP) on dental silicate ceramics are influenced by the plasma device, treatment parameters, and surface conditioning protocol [40]. However, studies evaluating a specific atmospheric non-thermal plasma device operating with an AC microsecond-pulsed power supply, while simultaneously comparing multiple surface treatment protocols, different glass ceramics, and different resin cements, are limited. Furthermore, the material- and cement-dependent effectiveness of plasma treatment in comparison with conventional surface treatments has not been comprehensively investigated. Therefore, the aim of this in vitro study was to evaluate the effect of atmospheric non-thermal plasma surface treatment on the shear bond strength of two different resin cements to two different glass ceramics by comparing various surface treatment combinations. The null hypothesis of this study was that the effectiveness of non-thermal plasma treatment would not increase bond strength depending on the material, cement, and surface treatment combination. Materials and methods In the present study, a 2 × 6 × 2 factorial experimental design was used to evaluate the effects of two glass ceramic materials, six surface treatment protocols and two cement systems on shear bond strength. Accordingly, a total of 24 experimental combinations were generated. Ten specimens were allocated to each experimental combination (n = 10), resulting in a total sample size of 240 specimens. The sample size was determined based on an a priori power analysis, assuming a minimum statistical power of 80%, a Cohen’s effect size of 0.50, and a Type I error probability of α = 0.05. The materials used in the study are listed in Table 1. A flowchart of the experimental process is given in Fig. 1. Table 1Materials used in the studyMaterialCompositionProduct NameManufacturerLot. No.Feldspathic ceramicSilicon dioxide 56–64%, Aluminum oxide 20–23%, Sodium oxide 9–11%, Potassium oxide 6–8%, Calcium oxide 0.3–0.6%, Titanium dioxide 0.0-0.1%Vitablocks Mark II (V)Vita Zahnfabrik, Bad Säckingen, GermanyLithium disilicate ceramicQuartz, lithium dioxide, phosphoric oxide, alumina, potassium oxideIPS e.MAX CADIvoclar Vivadent, LiechtensteinSelf Adhesive Resin CementBarium in a bis-GMA based dental resin matrix, pigments, catalysts and additives. Filler content: 69%, weight: 51%, volume. Inorganic filler particle amount: between 0.02 and 3 MmZenitcemPresident Dental, Munich, GermanyPD793000Chemical and light curing (dual-cure) resin cementSurface modification with glass powder, 2 propinoic acid, 2 methy l-3- (trimethoxysilyl) propyl ester and phenyltrimethoxy silane; bulk material; 2-propinoic acid; 2 methyl-, 1,11- [1- (hydroxymethyl) 1,2-ethanediyl] ester; reaction products with 2-hydroxy-1,3 propanediyl dimethacrylate and phosphorus oxide; TEGDMA; silane treated silica; oxide glass chemicals(non-fibrous); sodium persulfate; TBPINRelyX Ultimate Clicker (RXU)3 M ESPE, Neuss, Germany670,438Hydrofluoric acid9.5% hydrofluoric acidPorcelain Etchant − 9.5%Buffered hydrofluoric Acid GelBisco, Schaumburg, USA1,800,004,851Porcelain Silane-PrimerGlycidoxypropyltrimethoxysilane, ethyl alcoholCeramic primerB.J.M.Laboratories Ltd., Israel4572PFXSTRAdhesiveMDP phosphate monomer, dimethylacrylate resins, HEMA, vitrebond copolymer, filler, ethanol, water, initiators, silaneSingle Bond Universal3 M, Neuss, Germany669,822Composite resinNano-sized amorphous silica and glass-filled 4-methanyloxy ethyl trimellitic acidFlowable compositeFusio, USAAcrylic resinPolymethylmethacrylateMethyl methacrylateSC Cold AcrylicImicryl Dental, Konya, Turkey Materials used in the study Polymethylmethacrylate Methyl methacrylate Flowchart of the experimental process. 6 groups according to the combination of surface treatments ; (Hydoflouric acid etching (HF); Primer application (Pr); Non-thermal Plasma application (Pl). Cements: ((Single Bond + Rely x Ultimate)(R); Zenitcem (Z). Glass ceramic specimens which were obtained from IPS e.max CAD (E) and Vitamark II (V) CAD / CAM blocks For evaluating shear bond strength (SBS), glass ceramic plates (IPS e.max CAD, Vitamark II; 2 mm in thickness) were obtained using a precision sectioning machine (IsoMet 1000; Buehler Ltd., Lake Bluff, IL, USA). Then samples were embedded in self-curing acrylic resin (SC, Imicryl Dental, Konya, Turkey) to allow the test surfaces to be exposed [6]. These glass ceramic plates were then sequentially polished with 600, 800, and 1200 grit silicon carbide abrasive papers under water cooling to create initial surface roughness standardization. Then the specimens were washed in an ultrasonic water bath to remove the impurities remaining on the specimen surface followed by air drying. Two hundred and forty composite resin cylinders were built up in a plastic mold (4 mm in inner diameter and 5 mm in height). A flowable composite resin (Fusio, Pentron, Wallingford, CT, USA) was used to fill the plastic mold and then it was light polymerized according to the manufacturer’s recommendation. The experimental design was structured to enable a comprehensive evaluation of non-thermal plasma in comparison with conventional and combined surface treatment protocols across different ceramics fabricated using CAD/CAM systems and different resin cement systems. Glass ceramic specimens obtained from IPS e.max CAD (E) and Vitamark II (V) CAD/CAM blocks were randomized into 6 groups according to the combination of surface treatments: hydrofluoric acid etching (HF), primer application (Pr), and nonthermal plasma application (Pl): Group 1: Untreated (C) Group 1: Untreated (C) Group 2: HF + Pr Group 2: HF + Pr Group 3: Pr Group 3: Pr Group 4: Pl Group 4: Pl Group 5: HF + Pl+Pr Group 5: HF + Pl+Pr Group 6: Pl + Pr Group 6: Pl + Pr Cements: ((Single Bond + RelyX Ultimate) (R), ZenitCem (Z) Cements: ((Single Bond + RelyX Ultimate) (R), ZenitCem (Z) For the NTP application, an AC microsecond pulse power supply was used. Glass ceramic surfaces were treated with NTP for 90 s. The power supply was operated at 5 kHz frequency and 20 kV voltage by maintaining the discharge gap at 2 mm for the plasma treatment of samples. Porcelain Etchant Buffered (9.5% hydrofluoric acid) was applied to the IPS e.max CAD (E) substrate surface for 60 s. Porcelain Etchant Buffered (9.5% hydrofluoric acid) was applied to the Vitamark II (V) substrate surface for 2 min. Following acid etching, the ceramic surfaces were thoroughly rinsed under running water for 60 s to remove residual acid and reaction by-products, and subsequently dried with compressed air. The ceramic primer was uniformly applied to the ceramic surfaces with a disposable brush applicator and subsequently air-dried for 5 s. After the surface treatments, composite resin cylinders were cemented to the ceramic samples with ZenitCem or RelyX Ultimate resin cement. For the RelyX Ultimate resin cement groups, Single Bond Universal adhesive was actively applied to the ceramic bonding surface using a disposable microbrush, in accordance with the manufacturer’s instructions. The adhesive layer was gently air-thinned to obtain a uniform distribution. For specimen preparation, all composite resin cylinders were prepared using a plastic mold with a fixed internal diameter of 4 mm and a height of 5 mm, and transparent tapes with a central perforation were placed on the ceramic surfaces to standardize and control the bonding area. Flowable composite resin was carefully placed into the mold to prevent air bubble entrapment, and polymerization was performed using an LED light-curing unit. During cementation, each composite cylinder was centrally positioned on the treated ceramic surface, and the circular interface defined by the perforated tape and mold was considered the standardized bonding area. To prevent uncontrolled enlargement of the bonding interface, excess resin cement around the cylinder margins was carefully removed before polymerization. Light curing was performed for 20 s using an LED curing unit, according to the manufacturer’s recommendations. Before SBS testing, the bonded specimens were inspected to verify proper cylinder positioning, absence of obvious excess cement, and preservation of the standardized bonding diameter. During the same inspection procedure, all ceramic specimens were visually examined under magnification using dental loupes and an LED light source to detect visible defects, including surface cracks, edge chipping, irregularities, or voids within the composite build-up, that could potentially affect bond strength measurements. This procedure was performed to reduce bias related to pre-existing ceramic or composite defects, prevent variability in the bonding interface, and improve specimen standardization among the experimental groups. Specimens showing visible defects, structural irregularities, incorrect cylinder positioning, obvious excess cement, or loss of the standardized bonding diameter were excluded and replaced with new specimens. Then these samples were stored in an oven (37 °C, 100% humidity, 24 h) before the SBS test. SBS was measured using a universal testing machine (Shimadzu, AGS-X 5kN, Shimadzu Corporation, Kyoto, Japan). In the SBS test, a knife edge-shaped apparatus was used at a crosshead speed of 1 mm/min until failure occurred. Subsequently, SBS (MPa) was determined by dividing the ultimate failure load by the bonding interface area. After SBS testing, the fractured surfaces of all specimens were systematically evaluated to determine the failure mode. Initial examination was performed using a stereomicroscope (Stemi 508; Carl Zeiss Jena GmbH, Jena, Germany) to identify the predominant failure location and the amount and distribution of residual resin cement on the ceramic surface. Failure modes were classified according to predefined criteria as adhesive, cohesive, or mixed failures. Type I failure was defined as adhesive failure, corresponding to separation at the resin cement–ceramic interface, and was recorded when the ceramic surface appeared clean or nearly clean. Type II failure was defined as cohesive failure within the material and was subclassified as cohesive failure within the resin cement, composite resin cylinder, or ceramic substrate. Cohesive failure was recorded when fracture occurred predominantly within one of these materials without complete separation at the bonding interface. Type III failure was defined as mixed failure, in which both adhesive interfacial separation and cohesive fracture areas were observed within the same specimen. The same classification criteria were applied to all specimens to improve consistency. Scanning electron microscopy (SEM; Zeiss, Germany) was used at 100× magnification to further evaluate the failure modes and fracture surface characteristics in greater detail. The normality of data distribution was assessed using the Shapiro–Wilk test, and the homogeneity of variances was evaluated using Levene’s test. A three-way ANOVA with Type III sums of squares was performed to analyze the main effects and all interaction effects. For statistically significant main effects, Tukey’s HSD post-hoc test was applied, whereas simple main effects analysis was used for the evaluation of interaction effects. To assess the targeted effect of plasma treatment, pairwise comparisons were performed using either the independent-samples t-test or the Mann–Whitney U test, as appropriate, and Cohen’s d effect size was calculated. The clinically relevant threshold was defined as 8–9 MPa, which is widely accepted in the literature. The level of statistical significance was set at α = 0.05, and all analyses were performed using IBM SPSS Statistics version 30. Results Table 2 shows the mean and standard deviation values for the SBS of composite cylinders with glass ceramic specimens. Table 2Descriptive Statistics of SBS Values by GroupCeramicSurface TreatmentCement n Mean ± SD (MPa)95% CIMed.Min–MaxVitamark IIPrimerRelyX+Bond1010,22 ± 3,997,37–13,0710,254,19–15,88Vitamark IIPrimerZenitCem108,80 ± 2,546,98–10,618,555,56–13,59Vitamark IIHF+PrimerRelyX+Bond1018,56 ± 2,4916,78–20,3319,1914,96–21,61Vitamark IIHF+PrimerZenitCem1013,77 ± 3,6711,14–16,4013,576,93–19,41Vitamark IIPlasma+PrimerRelyX+Bond1010,35 ± 3,138,11–12,589,764,22–14,86Vitamark IIPlasma+PrimerZenitCem1011,90 ± 2,0710,42–13,3811,368,78–16,22Vitamark IIHF+Plasma+PrimerRelyX+Bond1019,93 ± 2,7317,97–21,8819,8417,13–26,46Vitamark IIHF+Plasma+PrimerZenitCem1016,01 ± 3,8113,29–18,7416,498,30–20,88Vitamark IIUntreatedRelyX+Bond108,00 ± 1,506,92–9,078,185,40–9,74Vitamark IIUntreatedZenitCem107,12 ± 2,595,27–8,976,524,41–13,06Vitamark IIPlasmaRelyX+Bond1010,77 ± 3,658,15–13,3811,265,09–16,51Vitamark IIPlasmaZenitCem1010,77 ± 2,339,10–12,4410,487,22–14,90IPS e.max CADPrimerRelyX+Bond108,90 ± 2,986,77–11,038,735,08–13,98IPS e.max CADPrimerZenitCem105,39 ± 2,443,64–7,144,702,54–8,62IPS e.max CADHF+PrimerRelyX+Bond1019,59 ± 6,3315,06–24,1220,309,01–28,78IPS e.max CADHF+PrimerZenitCem1015,60 ± 5,4811,68–19,5215,054,99–22,90IPS e.max CADPlasma+PrimerRelyX+Bond1011,45 ± 3,389,03–13,8711,256,75–16,48IPS e.max CADPlasma+PrimerZenitCem108,11 ± 2,566,28–9,958,473,79–12,36IPS e.max CADHF+Plasma+PrimerRelyX+Bond1021,56 ± 3,1019,34–23,7821,2417,30–26,60IPS e.max CADHF+Plasma+PrimerZenitCem1016,71 ± 5,4812,78–20,6318,419,69–26,66IPS e.max CADUntreatedRelyX+Bond105,33 ± 2,493,55–7,114,422,56–9,59IPS e.max CADUntreatedZenitCem102,98 ± 0,432,67–3,292,992,25–3,61IPS e.max CADPlasmaRelyX+Bond105,99 ± 2,614,12–7,865,692,63–11,04IPS e.max CADPlasmaZenitCem104,23 ± 1,223,36–5,114,272,42–6,60Mean ± SD mean ± standard deviation, CI confidence interval, Med. median, HF hydrofluoric acid, MPa megapascal. Each group included n = 10 specimens. Bold values indicate mean SBS values < 8 MPa, below the clinical threshold Descriptive Statistics of SBS Values by Group Mean ± SD mean ± standard deviation, CI confidence interval, Med. median, HF hydrofluoric acid, MPa megapascal. Each group included n = 10 specimens. Bold values indicate mean SBS values < 8 MPa, below the clinical threshold According to the three-way ANOVA results, the main effect of surface treatment was statistically significant and showed the highest effect size among the tested factors (F = 22.561, p < 0.001, η²p = 0.343). In addition, the ceramic × surface treatment interaction (p = 0.013, η²p = 0.064) and the surface treatment × cement system interaction (p = 0.026, η²p = 0.057) were statistically significant. Figure 2; Table 3. Ceramic Material × Surface Treatment Interaction Plot (Mean ± SEM). The red dashed lines indicate the 8–9 MPa clinical threshold. Left: RelyX+Bond; Right: ZenitCem Table 3Three-Way ANOVA Results (Type III Sum of Squares)EffectSSdfF p η²pCeramic (C)8,73810,7970,3730,004Surface Treatment (S)1236,286522,561< 0,0010,343Primer/Cement System (P)10,15310,9260,3370,004C × S Interaction162,29052,9620,0130,064C × P Interaction10,85810,9910,3210,005S × P Interaction143,06152,6110,0260,057C × S × P Interaction42,94950,7840,5620,018Error2367,220216– – –SS sum of squares, df degrees of freedom, η²p partial eta squared. η²p ≤ 0.06 indicates a small effect, 0.06–0.14 indicates a medium effect, and ≥ 0.14 indicates a large effect. p-values are presented in italics; statistically significant p-values (p < 0.05) are presented in bold italics Three-Way ANOVA Results (Type III Sum of Squares) SS sum of squares, df degrees of freedom, η²p partial eta squared. η²p ≤ 0.06 indicates a small effect, 0.06–0.14 indicates a medium effect, and ≥ 0.14 indicates a large effect. p-values are presented in italics; statistically significant p-values (p < 0.05) are presented in bold italics In the Tukey’s HSD post-hoc analysis of the main effect of surface treatment, significant differences were observed among the surface treatment groups. The HF + Pl+Pr and HF + Pr groups showed higher SBS values than most non-HF groups. The Pl + Pr group had significantly higher SBS values than the Pl group (+ 2.51 MPa, p = 0.045) and the untreated group (+ 4.596 MPa, p < 0.001). For the V ceramic, the highest mean SBS values were recorded in the HF-containing surface treatment groups for both cement systems. In the R group, the mean SBS values were 19.93 ± 2.73 MPa for HF + Pl+Pr and 18.56 ± 2.49 MPa for HF + Pr. In the Z group, the corresponding values were 16.01 ± 3.81 MPa and 13.77 ± 3.67 MPa, respectively. Among the V combinations, the mean SBS value of the untreated Z group was below the clinical threshold of 8 MPa (7.12 ± 2.59 MPa). For the E ceramic, the highest mean SBS values were also recorded in the HF-containing surface treatment groups. In the R group, the mean SBS values were 21.56 ± 3.10 MPa for HF + Pl+Pr and 19.59 ± 6.33 MPa for HF + Pr. In the Z group, the corresponding values were 16.71 ± 5.48 MPa and 15.60 ± 5.48 MPa, respectively. In the E–R combinations, the mean SBS values of the untreated group (5.33 ± 2.49 MPa) and the Pl group (5.99 ± 2.61 MPa) were below the clinical threshold of 8 MPa. In the E–Z combinations, the mean SBS values of the untreated group (2.98 ± 0.43 MPa), Pl group (4.23 ± 1.22 MPa), and Pr group (5.39 ± 2.44 MPa) were below the clinical threshold of 8 MPa Fig. 3. SBS Values According to Surface Treatment and Cement System (Mean ± SD). The red dashed line indicates the 8 MPa clinical threshold. Numbers indicate the mean SBS values. Left: VITAMARK II; Right: E-MAX In the Mann–Whitney U comparison, plasma-treated groups showed higher mean SBS values than non-plasma-treated groups. The difference was significant in the Z groups (+ 2.35 MPa, p = 0.012), but not in the R groups (+ 1.57 MPa, p = 0.110). Significant differences were also observed for V + Z (+ 3.80 MPa, p = 0.001) and E + Z (+ 5.09 MPa, p < 0.001). After HF etching, the additional effect of plasma treatment was not statistically significant Table 4. Table 4Overall Comparison Between Plasma-Treated and Non-Plasma-Treated Groups (Mann–Whitney U Test)GroupPlasma-treated Mean ± SD (MPa)Non-plasma-treated Mean ± SD (MPa)Difference (MPa)Test p All groups12,31 ± 5,9210,35 ± 6,11+ 1,96MWU0,004*VİTAMARK II13,29 ± 4,5911,08 ± 4,87+ 2,21MWU0,006*E-MAX11,34 ± 6,929,63 ± 7,11+ 1,71MWU0,062nsRelyX+Bond13,34 ± 6,3311,77 ± 6,45+ 1,57MWU0,110nsZenitCem11,29 ± 5,348,94 ± 5,46+ 2,35MWU0,012*VİTAMARK II+ ZenitCem12,15 ± 3,478,35 ± 4,25+ 3,80MWU0,001**E-MAX + ZenitCem9,68 ± 6,554,59 ± 4,26+ 5,09MWU< 0,001***Plasma-treated groups: Plasma+Primer, HF+Plasma+Primer, and Plasma. Non-plasma-treated groups: Primer, HF+Primer, and Untreated. Additional rows were included for the VITAMARK+ZenitCem and EMAX+ZenitCem combinations. p < 0.05; ** p < 0.01; *** p < 0.001; ns: not significant. p-values are presented in italics; statistically significant p-values (p < 0.05) are presented in bold italics Overall Comparison Between Plasma-Treated and Non-Plasma-Treated Groups (Mann–Whitney U Test) Plasma-treated groups: Plasma+Primer, HF+Plasma+Primer, and Plasma. Non-plasma-treated groups: Primer, HF+Primer, and Untreated. Additional rows were included for the VITAMARK+ZenitCem and EMAX+ZenitCem combinations. p < 0.05; ** p < 0.01; *** p < 0.001; ns: not significant. p-values are presented in italics; statistically significant p-values (p < 0.05) are presented in bold italics In the targeted comparisons of plasma effect, at level A (untreated vs. Pl), statistically significant differences were observed in the V + R group (+ 2.77 MPa, p = 0.040) and in both Z ceramic groups. In the E + Z group, the Plasma group showed a statistically significant increase compared with the untreated group (p < 0.01); however, the absolute mean SBS value remained below the clinical threshold. At level B (Pr vs. Pl + Pr), statistically significant differences were observed in the Z groups. The mean SBS value was higher in the Pl + Pr group than in the Pr group for both V + Z (+ 3.10 MPa, p = 0.008) and E + Z (+ 2.72 MPa, p = 0.026). In the R groups, the difference between Pr and Pl + Pr was not statistically significant. At level C (HF + Pr vs. HF + Pl+Pr), no statistically significant differences were observed in any ceramic–cement combination Fig. 4. Targeted Comparisons of Plasma Effect. A Untreated vs. Plasma; B Primer vs. Plasma+Primer; C HF+Primer vs. HF+Plasma+Primer). Numbers indicate the mean SBS values. The red line indicates the 8 MPa clinical threshold. p < 0.05; ** p < 0.01; *** p < 0.001 The SBS values of the two ceramic groups were compared within the same surface treatments and cement types. For Z cement type in the Pl, control, Pr, and Pl + Pr surface treatments groups, V samples showed significantly higher SBS values when compared with the E samples (p < 0.05). When the SBS values of different cement types were compared within the same surface treatment and glass ceramic groups, significant differences were determined (p < 0.05). For V ceramic groups receiving the HF + Pr and HF + Pl+Pr surface treatments, R showed significantly higher SBS values when compared with Z (p < 0.05). Morphological changes on the CAD/CAM block surfaces resulting from different surface treatments were evaluated by SEM (AxioCam ERc 5s, Zeiss, Germany). The SEM images were obtained at 100× and 500× magnification and then micrographs were compared with each other. Considering the SEM micrographs (Fig. 5), alterations and micro-retentive areas were more clearly visible on the surfaces of the HF samples compared with the untreated samples. HF acid etching surface treatment increased the surface area on all the CAD/CAM block (E, V) surfaces. However, small changes, mostly the same surface characteristics, were observed between HF and HF + Pl samples. There were no large morphological differences between HF and HF + Pl samples for all materials. FE-SEM micrographs of various CAD/CAM materials after different surface treatments. ( at 500 x magnification, scale bar indicates 20 μm ) The SBS values of HF + Pl samples and HF samples were similar as observed in the SEM micrographs for those samples. The untreated samples exhibited a significantly lower bond strength value when compared with HF + Pl and HF samples (p < 0.05). Figure 6 shows the failure modes of all the glass ceramic groups. In the specimens that received Pl surface treatment, mix/cohesive failure was observed rather than adhesive failure. Representative micrographs of different failure modes were obtained by SEM (Zeiss, Germany) (Fig. 7). Failure mode distribution for all groups after the SBS test. 6 groups according to the combination of surface treatments ; (Hydoflouric acid etching (HF); Primer application (Pr); Non-thermal Plasma application (Pl). Cements: ((Single Bond + Rely x Ultimate)(R); b, Zenitcem (Z). Glass ceramic specimens which were obtained from IPS e.max CAD (E) and Vitamark II (V) CAD / CAM blocks FE-SEM micrographs of debonded specimens. a adhesive failure. b cohesive failure. c mixed failure. (100x magnification) Discussion In the present study, atmospheric-pressure dielectric barrier discharge (DBD) plasma was applied to the surfaces of feldspathic and lithium disilicate CAD/CAM ceramics in ambient air without the use of an additional process gas. The null hypothesis of this study was that the effectiveness of non-thermal plasma treatment would not increase bond strength depending on the material, cement, and surface treatment combination. Based on the obtained results, the null hypothesis was rejected. An important finding of the present study was the material-, cement-, and surface treatment–dependent effect of atmospheric non-thermal plasma application. The effect of plasma treatment appeared to be more evident in the ZenitCem cement system (p = 0.012), particularly on surfaces that did not receive HF etching. The V + Z combination was among the groups in which plasma treatment showed a more pronounced and statistically significant effect. In the E + Z group, plasma treatment resulted in a statistically significant increase in SBS compared with the untreated condition; however, the absolute SBS value remained below the clinical threshold. In contrast, hydrofluoric acid–based surface treatments consistently yielded higher bond strength values regardless of the ceramic type. These findings indicate that non-thermal plasma does not provide a universal bonding improvement, but may be more effective in specific ceramic–resin cement combinations and under certain surface conditioning protocols. Considering the limited number of studies evaluating this specific plasma device and the comparative assessment of multiple surface treatments, the present study provides novel and clinically relevant insights into the selective effectiveness of atmospheric non-thermal plasma. According to Borges and colleagues, cementation plays an important role in the clinical success of ceramics [2]. Some authors suggest that strong resin bonding depends on the composition of the ceramic material and the surface treatment [41]. Airborne particle abrasion is a well-known technique for increasing the surface roughness of ceramics. However, it is not recommended for cementing silica-based ceramics as it may lead to cracks and a high loss of porcelain material [42]. Together with micromechanical interlocking, chemical bonding helps to increase bond strength. For glass ceramics, surface roughening and chemical bonding are achieved via HF acid etching and silane treatment [43]. Although HF acid etching is a generally accepted application for strong resin bonding with glass ceramics, some researchers state that HF may affect ceramic strength negatively since it may lead to cracks [44]. Furthermore, numerous authors still recommend eliminating HF from the etching procedure [21, 22]. The effects of NTP surface treatments on different materials have been investigated in dentistry [13–16]. According to some studies, NTP surface treatment appears to be a promising alternative for improving the surface wettability of ceramics [45–47]. HF etching mainly enhances bonding by dissolving the glassy phase of the ceramic matrix, increasing surface roughness, and promoting micromechanical interlocking between the ceramic and resin cement [48]. In contrast, atmospheric plasma primarily acts through surface activation rather than pronounced micromechanical roughening. Plasma treatment may generate polar functional groups, such as carboxyl groups, on the outermost surface layer, thereby increasing wettability and potentially improving chemical interaction with the adhesive [49–51]. NTP has been used to chemically modify the ceramic surface without negatively affecting ceramic strength [9, 47]. According to Han and colleagues, nonthermal atmospheric pressure plasma surface treatment improved the wettability of feldspathic ceramic surfaces. Moreover, plasma surface treatment increased ceramic bonding by producing carboxyl groups [46]. In a study by Lanza et al. investigating the effects of novel surface treatments on the shear bond strength of a self-adhesive resin cement to lithium disilicate glass ceramic, the combined application of hydrofluoric acid and non-thermal plasma (NTP) was reported to yield the highest overall bond strength among the evaluated surface pretreatment protocols [52]. Liao et al. investigated the effects of atmospheric non-thermal argon plasma (NTP) treatment on enhancing the micro-shear bond strength (µSBS) of lithium disilicate (LD) to dental resin cements and concluded that NTP surface treatment represents a promising technology for improving the short- and long-term micro-shear bond strength of LD bonded to dental resin cements [53]. In a study evaluating the effects of surface modification of feldspathic porcelain by corona discharge on bonding to resin cement, it was reported that the shear bond strength (SBS) values of corona-discharge-treated porcelain were comparable to those obtained with conventionally hydrofluoric acid (HF)–etched porcelain [54]. Non-thermal plasma protocols reported in the literature show substantial variability in gas type, exposure time, treatment power, operating pressure, and source-to-specimen distance. In most dental ceramic studies, plasma treatment times have generally ranged from 10 to 90 s, whereas longer activation periods, including 5–15 min, 30 min, and 60 min, have been used only in limited studies [40]. Treatment power and working distance also vary considerably among protocols, indicating that there is no standardized NTP application protocol [40]. Previous studies have further suggested that longer treatment times or higher NTP power may result in more pronounced changes in ceramic surface roughness and wettability [46, 55]. In the present study, a 90-s exposure time and a 2-mm discharge gap were selected based on previous evidence identifying this duration as optimal for achieving the highest wettability under similar conditions, as well as on preliminary observations showing comparable surface responses. The 5 kHz frequency and 20 kV voltage were used as device-specific parameters that provided stable plasma discharge, and all plasma settings were kept constant throughout the experiment to ensure reproducibility. In our previous study the effects of NTP surface treatment on the wettability and surface roughness of different CAD/CAM blocks as well as the bond strength of dual-cure composite resin cements were investigated. In accordance with the results of that previous study, while 30-s NTP treatment caused an increase in the wettability of the feldspathic ceramic-based CAD/CAM materials, extension of NTP treatment time to 90 s caused an increase in surface roughness. The SBS of feldspathic ceramic-based CAD/CAM materials and the resin cement was also increased after 90 s of NTP treatment. Similarly, in the present study, plasma surface treatment appeared to increase the bond strength values of both Z and R in V specimens when compared with the untreated group. Plasma treatment may modify material surfaces through various surface-related effects, including cleaning, activation, etching, and coating [56]. Plasma surface treatment is thought to improve bonding mainly through surface activation, increased surface energy, and enhanced wettability [57, 58]. During plasma exposure, reactive species or functional groups such as OH, O₃, H₂O₂, and NO may be generated on the treated surface without substantial structural alteration [59]. This effect may also be accompanied by surface cleaning [60]. Accordingly, these surface modifications may promote adhesive spreading and interfacial interaction, contributing to higher bond strength. The clinically acceptable minimum bond strength reported in the literature varies widely, ranging from 8 to 30 MPa [61–64]. Kalra et al. suggested that, considering the masticatory forces acting on anterior and posterior teeth and the relationship in which strength is directly proportional to force and inversely proportional to bonding area (Strength = F/A), a minimum bond strength of approximately 8–9 MPa may be regarded as clinically acceptable [61]. Based on this threshold, the present findings indicate that the statistically significant effect of plasma treatment may have clinical relevance only in selected material–cement combinations. Specifically, among the plasma-treated and non-plasma-treated surface protocols, the V-R and V-Z groups showed bond strength values within the clinically acceptable range, whereas the E-Z group remained below the clinical threshold. For the Pl + Pr surface treatment protocol, clinically acceptable values were also obtained in the Z group. The SBS values of the HF + Pl+Pr samples were not significantly different from those of the HF + Pr samples. Another result of our study was that regardless of the cement type, HF surface treatment was a more effective pretreatment than Pl. Previous studies have shown that bond strength values obtained after cementation following plasma treatments are generally lower than those achieved with conventional hydrofluoric acid (HF) etching [40]. Similar findings were also observed in the present study, which supports the existing literature. Furthermore, similar results were obtained from the comparison of surface characterization of selected samples. SEM evaluations confirmed that when comparing the HF and HF + Pl treated samples, large morphological surface changes were not observed between these samples. This finding is clinically relevant, as it suggests that the strong effect of hydrofluoric acid etching may overshadow the potential benefits of plasma treatment, emphasizing the importance of selecting appropriate surface treatment strategies rather than applying plasma indiscriminately. According to Graiff and colleagues [65], the application of HF etching and silane surface treatments to the ceramic surface improved bond strength. Similarly, the SBS of the HF + Pl+Pr treatment was significantly greater than that of the Pl + Pr treatment for all groups. Regardless of ceramic and cement type, HF + Pr exhibited significantly higher values than silane alone. This result is consistent with those of previous studies, as HF acid etching was shown to improve the SBS of ceramics with resin cements when compared with other surface treatments [66–68]. In addition, all surface treatment protocols involving HF etching provided bond strength values exceeding the clinically acceptable threshold for both ceramic materials and both resin cement systems. The significantly higher SBS values observed for Vitamark II compared with IPS e.max CAD in the control, plasma, primer, and plasma+primer groups when using the self-adhesive resin cement indicate that the bonding performance is strongly influenced by the ceramic substrate. The higher glass phase content and more homogeneous amorphous structure of feldspathic ceramics may facilitate improved surface wettability and interaction with self-adhesive resin cements following surface energy–modifying treatments. In contrast, the higher crystalline content of lithium disilicate glass ceramics may limit the effectiveness of plasma and primer applications alone, resulting in lower bond strength values in the absence of hydrofluoric acid etching. The selective response observed after plasma treatment indicates that its effect was not uniform across all experimental groups, but was dependent on the material type and resin cement system. In particular, the effect appeared to be more pronounced in the V material. Vitamark II and IPS e.max CAD ceramics differ in their crystalline phase content, silica content, and aluminum oxide composition [69], which may contribute to their different responses to surface modification. Accordingly, the effect of plasma surface treatment on different glass-ceramic surfaces may not be identical and may be influenced not only by the chemical composition of the ceramic but also by its microstructural organization and molecular arrangement. The SEM findings, showing different surface modification patterns on glass-ceramic surfaces with distinct structural characteristics, further support this explanation. Etch-and-rinse adhesive resin systems are generally considered to result in higher bond strengths when compared with self-adhesive and self-etch resin cements on glass ceramic surfaces [70]. In the present study, within the V ceramic groups, the HF + Pr and HF + Pl+Pr surface treatments resulted in significantly higher SBS values with R than with Z. It is possible that R, which has an adhesive agent, reacts more to the HF-treated glass ceramic surface than Z, which is a self-adhesive resin cement and used single step and alone [35]. This result was supported by previous studies in which the self-adhesive resin cement bond strength exhibited a significantly lower bond strength value when compared with etch-and-rinse procedure resin cements [70–72]. Chen et al. have reported that the weaker physical properties of self-adhesive cements than traditional resin cements could be considered a reason for this behavior [71, 72]. Taken together, the bonding performance of plasma-treated ceramics appears to be governed by both surface treatment–cement and surface treatment–material interactions. The response to atmospheric nonthermal plasma may therefore depend on the microstructural characteristics of the ceramic substrate, the physicochemical properties of the resin cement system, and the specific surface treatment protocol used. Accordingly, plasma treatment should not be considered merely as a surface roughening method, but rather as a multifactorial interfacial phenomenon involving surface activation, wettability enhancement, and material- and cement-dependent bonding mechanisms. In light of these selective effects, atmospheric nonthermal plasma should not be regarded as a direct substitute for HF etching; instead, it may be evaluated as a potential alternative for selected material–cement–surface treatment combinations. The present study has several limitations. First, although the shear bond strength (SBS) test is one of the most commonly used methods for evaluating bond strength in resin-bonded restorations [73], it has inherent methodological limitations. The complex stress distribution generated during loading may result in cohesive bulk fractures within the material rather than failure occurring exclusively at the bonded interface [73]. Second, thermocycling was not performed; therefore, the durability of the adhesive interface after artificial aging and the influence of long-term factors, such as material aging, on bond strength could not be assessed. This limits the clinical interpretation of the findings and restricts the ability to make direct clinical recommendations. Third, the effectiveness of NTP treatment may vary considerably depending on the plasma system, gas type, exposure time, power, source-to-specimen distance, restorative material, surface treatment combination, resin cement system, and subsequent adhesive protocol. Accordingly, since the effect of plasma treatment cannot be generalized based on a single application protocol, further studies evaluating different plasma parameters in combination with various adhesive systems and ceramic substrates are needed. Finally, because of the in vitro design of this study, intraoral conditions could not be fully reproduced. Moreover, future studies are required to evaluate clinically applicable intraoral versions of plasma treatment and to determine their safety, feasibility, and effectiveness under real oral conditions. Further experimental and clinical studies are therefore needed to validate the routine applicability of NTP treatment as a surface pretreatment method in clinical practice. Conclusions Within the limitations of this in vitro study, NTP surface treatment selectively improved the SBS of self-adhesive resin cement to feldspathic glass ceramic, reaching a clinically acceptable and statistically significant level in the Vitamark II–ZenitCem combination without HF etching. However, HF etching remained the most effective surface treatment and produced clinically acceptable bond strength values in all HF-containing groups. The bonding effectiveness of NTP was dependent on the ceramic substrate, resin cement type, and surface treatment protocol. Therefore, NTP should be considered a selective surface modification strategy rather than a universal alternative to conventional HF-based surface treatments. Acknowledgements Acknowledgements : The authors gratefully acknowledge the financial support provided by the Scientific Research Projects Unit of Izmir Katip Celebi University (Project No: 2018-TDU-DİŞF-0052). Authors’ contributions Emine Mustafaoglu wrote the main manuscript text and prepared figures. Bora Bağış, Gizem Dilara Özdemir, Utku Kürşat Ercan, Emine Mustafaoğlu reviewed the manuscript. Funding This study was supported by Izmir Kâtip Çelebi University Scientific Research Projects with project number 2018-TDU-DİŞF-0052. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.