Effect of hydrofluoric acid and femtosecond laser application sequence on nano-lithium silicate CAD/CAM ceramic bonding: an in vitro study

Abstract

This in vitro study investigated whether the sequence of combining hydrofluoric acid (HF) etching and femtosecond Ti: Sapphire laser irradiation influences the surface roughness (Ra) and shear bond strength (SBS) of a nano-lithium silicate CAD/CAM ceramic (Amber Mill; HASS Bio Core). Ninety rectangular block-shaped specimens (5 mm × 7 mm × 3.5 mm) were allocated to nine groups (n = 10): untreated control; HF only applied before or after crystallization; femtosecond laser only applied before or after crystallization; HF followed by laser (HF→Laser) applied before or after crystallization; and laser followed by HF (Laser→HF) applied before or after crystallization. Ra was measured with a contact profilometer (Mahr M2); surface morphology was evaluated by field emission scanning electron microscopy (GeminiSEM 300; ZEISS, Germany); and crystalline phase integrity was verified by X-ray diffraction (Rigaku DMAX-2200). Shear bond strength was determined after thermocycling (5000 cycles, 5–55 °C) using a universal testing machine (Instron Model 1082). Data were analysed with one-way ANOVA and Tukey HSD post hoc test; a supplementary two-way ANOVA (crystallization stage × treatment protocol) performed on the eight treatment groups further confirmed that application sequence remained a significant factor for both Ra and SBS (p < 0.001); the Ra–SBS correlation was assessed with Pearson’s coefficient (α = 0.05). Application sequence significantly affected both Ra and SBS (p < 0.01). Combined protocols consistently outperformed single-treatment protocols in SBS. The post-crystallization Laser→HF group achieved the highest SBS (19.295 ± 2.881 MPa), while the post-crystallization HF→Laser group produced the greatest Ra (1.538 ± 0.058 μm). A moderate positive Ra–SBS correlation was identified (r = 0.411; p < 0.01). FE-SEM and XRD analyses confirmed that neither protocol altered the fundamental crystalline phase of the material. Within the limitations of this study, application sequence is a critical determinant of bonding performance in nano-lithium silicate CAD/CAM ceramics; the post-crystallization Laser→HF combined protocol is recommended for optimal resin-cement adhesion.

Introduction

Lithium silicate-based glass ceramics are widely employed in contemporary restorative dentistry owing to their superior aesthetic properties, biocompatibility, and mechanical reliability [1, 2]. The proliferation of CAD/CAM (computer-aided design/computer-aided manufacturing) technology has further broadened their clinical utility by enabling standardised, highly reproducible fabrication of indirect restorations [3].

Amber Mill (HASS Bio Core, Republic of Korea) is a nano-lithium silicate CAD/CAM glass ceramic whose crystalline architecture differs markedly from that of conventional lithium disilicate (Li₂Si₂O₅) materials. Specifically, it contains finer and more uniformly distributed nano-scale crystals embedded within a silicate glass matrix and can be machined in either a pre-crystallized intermediate state or a fully crystallized state [4, 5]. These microstructural distinctions may render the material’s response to surface conditioning protocols substantially different from that of classical lithium disilicate ceramics, thereby warranting dedicated investigation [6].

Long-term clinical success of ceramic restorations depends critically on the quality of adhesion established between the resin cement and the ceramic bonding surface. This adhesion relies on two synergistic mechanisms: micromechanical interlocking achieved through surface roughening, and chemical coupling facilitated by silane coupling agents [7, 8]. Hydrofluoric acid (HF) etching remains the most widely adopted surface preparation method for silica-based ceramics, as selective dissolution of the glassy matrix by HF generates microporosity and substantially increases surface area available for resin infiltration [9, 10]. Despite its established efficacy, concerns regarding HF toxicity, the risk of compromising ceramic flexural strength at elevated concentrations, and the practical difficulties of intraoral application have spurred considerable interest in alternative or complementary conditioning modalities [11, 12].

Laser irradiation has emerged as a particularly promising approach for ceramic surface modification. Femtosecond lasers are especially attractive in this context because their ultra-short pulse durations confine the energy deposition zone to a sub-micrometre depth, enabling precise thermomechanical ablation without the collateral thermal damage characteristic of longer-pulse laser systems [13, 14]. Laser-assisted surface modification of dental ceramics, particularly zirconia and lithium disilicate/lithium silicate materials, has already been investigated relatively extensively, including femtosecond laser applications to zirconia surfaces and their effect on bonding [13, 15, 16], as well as combined laser-HF conditioning approaches for improving resin bonding to CAD/CAM ceramics [17–20]. More recently, direct comparisons between femtosecond laser irradiation and HF etching, and between various laser modalities and resin cement bonding to CAD/CAM ceramics, have continued to be reported [21, 22]. However, studies that have explicitly isolated application sequence (HF→Laser vs. Laser→HF) as an independent protocol variable–rather than simply comparing combined versus single-modality treatments–remain limited [23], and this gap has not, to our knowledge, been specifically examined for nano-lithium silicate CAD/CAM ceramics such as Amber Mill.

The primary aim of this study was therefore to determine whether the sequence in which HF etching and femtosecond laser irradiation are combined differentially influences the Ra and SBS of a nano-lithium silicate CAD/CAM ceramic, and to compare combined protocols against their respective single-treatment counterparts. Crystallization stage (pre- vs. post-crystallization) was included as a secondary variable to examine whether it modulates the effects of application sequence.

The null hypotheses tested were:

• H₀₁: Surface treatment protocol and application sequence will not significantly affect the surface roughness (Ra) of the ceramic.

• H₀₂: Surface treatment protocol and application sequence will not significantly affect the shear bond strength (SBS) of the ceramic.

• H₀₃: Crystallization stage will not modulate the influence of application sequence on Ra or SBS.

Materials and methods

The experimental design of this study is depicted in Fig. 1.

Fig. 1.

Experimental design

Materials

Nano-lithium silicate CAD/CAM ceramic blocks (Amber Mill; HASS Bio Core, Gangwon-do, Republic of Korea), shade A2, Low Translucency (LT), were used throughout this study. The chemical composition and manufacturer details of the material are summarised in Table 1. No batch/lot number was provided by the manufacturer on the blocks used in this study.

Table 1.

Description and composition of the CAD/CAM glass-ceramic material used in this study

Material	Type/Definition	Manufacturer (Country)	Composition (wt%)
Amber Mill	Nano-lithium silicate glass-ceramic, shade A2, LT	HASS Bio Core, Gangwon-do, Republic of Korea	SiO₂, Li₂O, K₂O, MgO, Al₂O₃, P₂O₅ and other oxides

Specimen preparation

A total of 90 specimens measuring 5 mm × 7 mm × 3.5 mm were sectioned from Amber Mill CAD/CAM ceramic blocks using a precision saw (Minitom; Struers A/S, Ballerup, Denmark). All specimens were wet-polished with 600-grit SiC abrasive paper, ultrasonically cleaned in distilled water for 5 min (Eurosonic Energy; Euronda, Marano Vicentino, Italy), and air-dried before being randomly allocated to nine experimental groups (n = 10; Table 2).

Table 2.

Experimental groups and surface treatment protocols (n = 10 per group)

Group	Cryst. Stage	Treatment	Protocol Sequence	n
1	—	Control	No treatment → Silane → Cement	10
2	Pre	HF only	HF → Crystallization → Silane → Cement	10
3	Post	HF only	Crystallization → HF → Silane → Cement	10
4	Pre	Laser only	Laser → Crystallization → Silane → Cement	10
5	Post	Laser only	Crystallization → Laser → Silane → Cement	10
6	Pre	HF→Laser	HF + Laser → Crystallization → Silane → Cement	10
7	Post	HF→Laser	Crystallization → HF + Laser → Silane → Cement	10
8	Pre	Laser→HF	Laser + HF → Crystallization → Silane → Cement	10
9	Post	Laser→HF	Crystallization → Laser + HF → Silane → Cement	10

HF hydrofluoric acid, Pre pre-crystallization, Post post-crystallization

A CAD/CAM-compatible resin composite block (Grandio CAD/CAM; VOCO, Cuxhaven, Germany; 8 mm × 8 mm × 7 mm) served as the substrate material for all cementation procedures. The bonding surfaces of composite specimens were sandblasted with 50 μm Al₂O₃ particles (Basic Eco Sandblasting Unit; Renfert, Hilzingen, Germany; 2.5 bar; 10 mm distance; 20 s), ultrasonically cleaned, and coated with a bonding agent (Tokuyama Dental Corporation, Tokyo, Japan) that was subsequently light-cured for 20 s (O-light LED; Woodpecker, Guilin, China) prior to cementation.

Surface treatments

Specimens were randomly assigned to nine groups (n = 10) according to the surface treatment protocol applied (Table 2). In pre-crystallization groups, all surface treatments were completed before the crystallization firing cycle; in post-crystallization groups, crystallization was performed first and surface treatments were applied thereafter. This design enabled the effect of application sequence to be examined independently within each crystallization stage.

Hydrofluoric Acid (HF) etching

In HF-treated groups (Groups 2, 3, 6, 7, 8, and 9), 9% buffered hydrofluoric acid gel (Ultradent Products Inc., South Jordan, UT, USA) was applied to the ceramic bonding surface for 20 s, rinsed with copious running water for 30 s, ultrasonically cleaned in distilled water for 5 min (Eurosonic Energy; Euronda, Italy), and thoroughly air-dried.

Femtosecond Ti: Sapphire laser irradiation

Laser surface conditioning (Groups 4, 5, 6, 7, 8, and 9) was carried out with a Ti: Sapphire femtosecond laser system (Spectra Physics, Spitfire Pro XP, TOPAS, Santa Clara, CA, USA; pulse duration: 50 fs; repetition rate: 1 kHz; central wavelength: 800 nm). The beam was focused onto the specimen surface through a biconvex lens (focal length: 10 cm; beam radius at the focal point: 18 μm, determined by the knife-edge method). Specimens were processed at 30 mW average power at a scan speed of 1 mm/s using a computer-controlled XYZ translation stage; pulse energy was modulated with optical density filters, and specimen positioning was managed via LabVIEW software (National Instruments, Austin, TX, USA). The calculated energy fluence, derived from the pulse energy and focal spot area, was approximately 2.95 J/cm². The entire 5 mm × 7 mm bonding surface of each specimen was raster-scanned in a single pass, with a scanning line overlap within the 50–90% range typically applied in femtosecond laser surface conditioning of dental ceramics. No active cooling or irrigation was applied during laser irradiation; this reflects the negligible thermal accumulation characteristic of femtosecond-pulse ablation, which was in fact the primary rationale for selecting a femtosecond laser system in the present study.

Crystallization protocol

Krystallization firing was performed in a ceramic furnace (Cameo Porcelain Furnace; Ivoclar Vivadent, Schaan, Liechtenstein) according to the manufacturer’s Low Translucency (LT) firing programme for Amber Mill (HASS Bio Core): preheating to 400 °C, heating rate of 60 °C/min to a final firing temperature of 840 °C, a holding time of 15 min at peak temperature, and controlled cooling initiated at 690 °C. Pre-crystallization groups (Groups 2, 4, 6, and 8) underwent firing after surface treatments were completed. The control group and post-crystallization groups (Groups 3, 5, 7, and 9) were fired before surface treatments were applied.

Surface characterization

All surface characterization measurements were performed prior to cementation.

Surface Roughness (Ra)

Surface roughness was measured with a contact profilometer (Mahr M2; Mahr GmbH, Göttingen, Germany). Three independent measurements were taken at different locations on each specimen, and the arithmetic mean roughness value (Ra, µm) was calculated. In pre-crystallization groups, Ra measurements were obtained after surface treatment and before the crystallization firing cycle.

Field Emission Scanning Electron Microscopy (FE-SEM)

Representative specimens from each group were examined by field emission scanning electron microscopy (GeminiSEM 300; ZEISS, Oberkochen, Germany) at ×2000 magnification. Prior to imaging, specimens were sputter-coated with gold–palladium to enhance electrical conductivity.

X-Ray Diffraction (XRD)

X-ray diffraction analysis was performed to evaluate the effect of surface treatments on the crystalline phase composition of the ceramic. Analyses were carried out with a Rigaku DMAX-2200 diffractometer (Rigaku Corporation, Tokyo, Japan) using Cu-Kα radiation (λ = 1.54 Å; 40 kV; 30 mA). Diffraction patterns were recorded over the 10–90° 2θ angular range.

Cementation and thermocycling

Following surface characterization, Monobond Plus silane coupling agent (Ivoclar Vivadent, Schaan, Liechtenstein) was applied in a thin layer to all ceramic bonding surfaces, left to react for 60 s, and then gently air-dried. Each ceramic specimen was manually centred on the bonding surface of the composite substrate under visual guidance, leaving an approximately equal composite margin surrounding the ceramic specimen on all sides; no mechanical alignment jig was used, reflecting the substantially larger dimensions of the composite substrate (8 mm × 8 mm) relative to the ceramic specimen (5 mm × 7 mm). Ceramic specimens were luted to composite substrates using a dual-cure resin cement (Panavia SA Universal; Kuraray Noritake Dental, Okayama, Japan) according to the manufacturer’s instructions; a light finger pressure was applied manually to seat the specimen and allow extrusion of excess cement, rather than a calibrated static load. Excess cement was carefully removed and specimens were light-cured for 20 s (O-light LED; Woodpecker, China). All cemented specimens were subsequently stored in distilled water at 37 °C for 24 h in a thermostatically controlled incubator (Wise Cube Incubator; Wisd Laboratory Instruments, Seoul, Republic of Korea).

Following storage, specimens were subjected to thermocycling to simulate the thermal fluctuations encountered in the oral environment (SD Mechatronic Thermocycler THE-1100; Feldkirchen-Westerham, Germany): 5000 cycles between 5 °C and 55 °C with a 30 s dwell time in each bath and a 10 s transfer time [24].

Shear bond strength test

After thermocycling, shear bond strength was measured using a universal testing machine (Model 1082; Instron, Norwood, MA, USA) at a crosshead speed of 1 mm/min. Each specimen was secured in a custom metal holding jig that immobilised the composite substrate, and a flat, chisel-shaped loading rod attached to the crosshead was positioned to contact the ceramic specimen immediately adjacent to the bonded interface, applying a shearing force approximately parallel to the interface until debonding occurred. As shear bond strength tests inherently generate non-uniform and partially tensile stress distributions at the loaded edge of the interface, the term “pure shear loading” has been avoided in favour of a more accurate description of the applied loading configuration. The maximum failure load (N) was recorded and converted to shear bond strength (MPa) by dividing by the bonding surface area (35 mm²). The bonding surface area was standardised by the dimensions of the ceramic specimen itself: the entire 5 mm × 7 mm bonding face of each ceramic specimen was luted to the larger composite substrate surface (8 mm × 8 mm), with the smaller ceramic footprint defining a constant bonding area of 35 mm² across all specimens; no additional masking or area-limiting template was used.

Failure mode analysis

Fracture surfaces of all tested specimens were examined under a stereomicroscope at ×9 magnification and classified into three categories: (1) adhesive failure—separation occurring exclusively at the ceramic–resin cement interface; (2) cohesive failure—fracture occurring entirely within the ceramic or composite substrate material; or (3) mixed failure—a combination of adhesive and cohesive fracture characteristics observed on the same surface.

Statistical analysis

Sample size was determined a priori using G*Power software (Version 3.1.9.3; Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany). Based on an effect size of Cohen’s f = 0.43 (large effect, derived from comparable studies in the literature), a significance level of α = 0.05, and a target statistical power of 0.80 for a one-way ANOVA with nine groups, a minimum of 10 specimens per group was calculated to be sufficient (achieved power = 80.4%; total N = 90).

All data were analysed using SPSS Version 22.0 (IBM Corp., Armonk, NY, USA). Data normality was assessed using the Kolmogorov–Smirnov test and by inspection of skewness and kurtosis indices. One-way ANOVA was applied for between-group comparisons, followed by Tukey’s honestly significant difference (HSD) post hoc test to identify specific pairwise differences. The Pearson correlation coefficient (r) was used to quantify the relationship between Ra and SBS values. The level of statistical significance was set at p < 0.05 for all analyses. One-way ANOVA, treating the nine experimental groups as a single categorical factor, was selected over a two-way factorial design because the experimental groups did not conform to a balanced, fully crossed factorial structure: the control group received no surface treatment and therefore lacked a corresponding crystallization-stage distinction, precluding a symmetric 2 (crystallization stage) × 4 (treatment protocol) design. Furthermore, the primary aim of the study was to identify specific, clinically interpretable differences between individual protocol combinations (e.g., pre- versus post-crystallization Laser→HF) rather than to statistically partition variance into separate main effects and an interaction term; the one-way ANOVA with Tukey HSD post hoc testing was therefore considered the most direct and interpretable approach for this group-wise comparative objective. To further verify that application sequence remained a robust factor irrespective of the analytical approach, a supplementary two-way ANOVA (crystallization stage × treatment protocol) was additionally performed on the eight treatment groups (Groups 2–9), which constitute a balanced 2 × 4 factorial structure once the untreated control group (Group 1) is excluded. This analysis tested the main effects of crystallization stage and treatment protocol, and their interaction, for both Ra and SBS. One Ra measurement in Group 3 (post-crystallization HF) could not be recovered from the original recordings; this group was therefore analysed with n = 9 for Ra, while all other groups retained n = 10 for both Ra and SBS.

Results

Surface Roughness (Ra)

The effects of different surface treatment protocols on surface roughness are summarised in Table 3. One-way ANOVA revealed statistically significant differences among groups (F = 44.5; p < 0.01; partial η² = 0.82), indicating a large effect size.

Table 3.

Mean surface roughness (Ra) and shear bond strength (SBS) values for all experimental groups

Group	Protocol	Ra (µm) Mean ± SD	SBS (MPa) Mean ± SD	n
1	Control	0.578 ± 0.015	3.571 ± 2.421	10
2	Pre-cryst. HF	1.070 ± 0.157	10.349 ± 1.519	10
3	Post-cryst. HF	0.736 ± 0.049*	14.221 ± 1.803	9*/10
4	Pre-cryst. Laser	0.830 ± 0.051	13.128 ± 3.412	10
5	Post-cryst. Laser	1.031 ± 0.307	10.484 ± 1.240	10
6	Pre-cryst. HF→Laser	1.255 ± 0.158	11.383 ± 1.734	10
7	Post-cryst. HF→Laser	1.538 ± 0.058	18.780 ± 8.289	10
8	Pre-cryst. Laser→HF	1.051 ± 0.055	15.047 ± 2.156	10
9	Post-cryst. Laser→HF	1.047 ± 0.049	19.295 ± 2.881	10

Mean arithmetic mean, SD standard deviation. Between-group comparisons were performed with Tukey’s HSD post hoc test; detailed pairwise results are presented in Table 4 (p < 0.05)

Pre-cryst. pre-crystallization, Post-cryst. post-crystallization

*One Ra measurement in Group 3 could not be recovered from the original recordings; n = 9 for this group’s Ra value only, while n = 10 was retained for its SBS value and for all other groups

The control group (0.578 ± 0.015 μm) exhibited significantly lower Ra than all surface-treated groups (p < 0.05). Among single-treatment groups, HF and laser applications produced comparable levels of surface roughness; however, neither modality alone attained the Ra values achieved by the combined protocols.

Application sequence exerted a decisive effect on Ra. The HF→Laser protocol generated the highest surface roughness: the post-crystallization HF→Laser group (1.538 ± 0.058 μm) exhibited significantly greater Ra than all other groups (p < 0.05). In contrast, the two Laser→HF groups displayed closely comparable Ra values (pre-crystallization: 1.051 ± 0.055 μm; post-crystallization: 1.047 ± 0.049 μm), with no statistically significant difference between them (p > 0.05). The supplementary two-way ANOVA performed on the eight treatment groups (excluding the untreated control) confirmed that treatment protocol was a significant factor for Ra (F = 50.07; df = 3; p < 0.001), while crystallization stage alone was not (F = 1.68; df = 1; p = 0.199). A significant Stage × Treatment interaction was observed (F = 18.31; df = 3; p < 0.001), consistent with the pattern already evident in the pairwise comparisons above, whereby the effect of crystallization timing on Ra varied depending on the specific treatment protocol applied.

FE-SEM analysis

FE-SEM images of representative specimens from all groups are presented in Fig. 2.

Fig. 2.

FE-SEM images (×2000 magnification; GeminiSEM 300, ZEISS, Germany) of representative specimens from each experimental group

The control group displayed a smooth, homogeneous surface morphology with an intact glassy matrix; no discernible surface roughening, micro-pitting, or porosity was observed beyond minor polishing scratches introduced during specimen preparation.

Single-treatment HF etching generated a micro-porous, irregular surface morphology resulting from selective dissolution of the glassy matrix. Single-treatment femtosecond laser irradiation produced regularly oriented micro-grooves aligned parallel to the scan direction, with overall surface integrity well preserved.

Application sequence markedly influenced surface morphology in the combined-treatment groups. In the HF→Laser protocol, subsequent laser scanning partially reorganised HF-induced micro-porosity into a more homogeneous distribution, yielding a relatively uniform roughness profile. Conversely, in the Laser→HF protocol, HF selectively dissolved the glassy matrix surrounding the laser-created grooves, deepening and broadening these channels and producing a heterogeneous micro-porous topography. This morphological divergence between the two combined protocols provides a mechanistic basis for interpreting their different SBS outcomes.

XRD analysis

XRD patterns of all groups are presented in Fig. 3.

Fig. 3.

XRD patterns of all experimental groups recorded over the 10–90° 2θ range

The X-ray diffraction patterns of all groups were largely superimposable across the 10–90° 2θ range. Characteristic diffraction peaks were concentrated at approximately 22–24° 2θ in all specimens, with no significant shifts in peak position and no emergence of new crystalline phases detected between groups. These findings confirm that neither HF etching nor femtosecond laser irradiation altered the bulk crystalline phase composition of Amber Mill; surface modifications were confined to the superficial microstructure.

Shear Bond Strength (SBS)

SBS values for all groups are presented in Table 3, and mean SBS values are illustrated in Fig. 4. One-way ANOVA revealed statistically significant between-group differences (F = 18.89; p < 0.01; partial η² = 0.65), again indicating a large effect size.

Fig. 4.

Mean shear bond strength (MPa) of all experimental groups. Error bars represent standard deviation

The control group (3.571 ± 2.421 MPa) yielded significantly lower SBS than all surface-treated groups (p < 0.05). Among single-treatment groups, the post-crystallization HF group (14.221 ± 1.803 MPa) and the pre-crystallization laser group (13.128 ± 3.412 MPa) achieved the highest SBS values, with no statistically significant difference between them (p > 0.05).

Combined protocols consistently outperformed all single-treatment groups in SBS. Critically, application sequence was the primary factor governing the magnitude of this superiority: the Laser→HF sequence produced higher SBS values than the HF→Laser sequence. The post-crystallization Laser→HF group achieved the highest SBS of all groups (19.295 ± 2.881 MPa), followed closely by the post-crystallization HF→Laser group (18.780 ± 8.289 MPa). These two groups did not differ significantly from each other (p > 0.05) but were both significantly superior to all single-treatment and pre-crystallization combined groups (p < 0.05). Detailed pairwise comparisons are provided in Table 4.

Table 4.

Tukey HSD pairwise comparisons of shear bond strength (SBS) among all experimental groups (MPa difference values)

	G1	G2	G3	G4	G5	G6	G7	G8
G2	−6.78*
G3	−10.65*	−3.87
G4	−9.56*	−2.78	1.09
G5	−6.91*	−0.14	3.74	2.64
G6	−7.81*	−1.03	2.84	1.74	−0.90
G7	−15.21*	−8.43*	−4.56*	−5.65*	−8.30*	−7.40*
G8	−11.48*	−4.70*	−0.83	−1.92	−4.56	−3.67	3.73
G9	−15.72*	−8.95*	−5.07	−6.17*	−8.81*	−7.91*	−0.52	−4.25

* p < 0.05. Negative values indicate that the row group achieved higher SBS than the column group

The supplementary two-way ANOVA performed on the eight treatment groups (excluding the untreated control) confirmed that both crystallization stage (F = 16.02; df = 1; p < 0.001) and treatment protocol (F = 9.77; df = 3; p < 0.001) were significant factors for SBS. A significant Stage × Treatment interaction was also observed (F = 6.87; df = 3; p < 0.001), corroborating the pairwise findings above and confirming that the relative advantage of the Laser→HF sequence over the HF→Laser sequence was not uniform across crystallization timing.

Ra–SBS correlation and failure mode distribution

The results of the correlation analysis between surface roughness and shear bond strength are presented in Table 5.

Table 5.

Pearson correlation between surface roughness (Ra) and shear bond strength (SBS)

Parameter	Value
Pearson r	0.411
p-value	< 0.01

Pearson correlation analysis revealed a moderate, positive, and statistically significant relationship between surface roughness and shear bond strength (r = 0.411; p < 0.01). Notably, however, the HF→Laser protocol—which generated the highest Ra values—did not coincide with the highest SBS, which was achieved by the Laser→HF protocol. This dissociation indicates that bond strength is governed not only by the magnitude of surface roughness but also by the quality of surface morphology and the application sequence.

Failure mode analysis (Fig. 5) revealed exclusively adhesive fractures in the control group. As surface treatment complexity increased, the proportion of mixed and cohesive failures rose progressively. In combined-protocol groups—particularly the post-crystallization Laser→HF and HF→Laser groups—mixed and cohesive failure modes became predominant. This shift is consistent with bond strength values approaching the cohesive strength of the ceramic–cement interface in these groups.

Fig. 5.

Failure mode distribution (adhesive, mixed, cohesive) for all experimental groups

Discussion

This study addressed a specific and, to our knowledge, previously unexamined question for this material: does the sequence in which HF etching and femtosecond laser irradiation are combined matter for bonding performance in a nano-lithium silicate CAD/CAM ceramic? Within the limitations of this in vitro model, the results demonstrate that it does—all three null hypotheses were rejected—and application sequence emerged as a primary determinant of both Ra and SBS under the tested conditions.

Application sequence—the primary finding

The most striking finding of this study is that the Laser→HF and HF→Laser protocols—despite utilising identical constituent treatments—generated markedly different SBS and Ra profiles. The post-crystallization Laser→HF group achieved the highest SBS (19.295 ± 2.881 MPa), whereas the post-crystallization HF→Laser group produced the greatest Ra (1.538 ± 0.058 μm) yet failed to generate the highest SBS. This dissociation demonstrates that maximising surface roughness does not automatically translate into maximal bond strength [17, 18].

In the Laser→HF sequence, femtosecond laser irradiation first generates regularly oriented micro-grooves through precise thermomechanical ablation with minimal heat-affected zone. Subsequent HF etching then selectively dissolves the glassy matrix surrounding these channels, deepening and widening them while simultaneously exposing crystalline surfaces that are amenable to silane-mediated chemical bonding. The net result is a topographically complex, high-surface-energy interface that optimises both micromechanical interlocking and chemical coupling. A mechanistically analogous synergy between laser pre-treatment and subsequent HF etching has been reported for CO₂ laser plus HF combinations applied to conventional lithium disilicate ceramics [19, 25]; the present findings extend this concept to femtosecond laser treatment of nano-lithium silicate materials and specifically identify sequence—rather than mere combination—as the operative variable.

In the HF→Laser sequence, acid etching first creates microporosity throughout the glassy matrix; subsequent laser scanning then partially reorganises these irregular features into a more homogeneous distribution. Although this protocol achieves the highest Ra values, the associated reduction in topographic complexity may limit the depth and geometry of resin cement infiltration, thereby constraining micromechanical interlocking despite the elevated average roughness value [20, 26]. Notably, the significant Stage × Treatment interaction identified by the supplementary two-way ANOVA (Surface Roughness (Ra) and Shear Bond Strength (SBS)) statistically corroborates this sequence-dependent behaviour, confirming that the relative performance of the Laser→HF and HF→Laser protocols is not independent of crystallization timing.

Combined versus single-treatment protocols

Both combined protocols significantly exceeded all single-treatment groups in SBS, confirming additive or synergistic effects of combining the two conditioning modalities [27, 28]. Single HF and single laser treatments achieved intermediate SBS values broadly consistent with previously published data [13, 14, 29]. Souza et al. systematically evaluated sequential laser and chemical surface treatments for lithium disilicate ceramics and reported consistent superiority of combined approaches over their single-treatment counterparts [19], corroborating the present findings. The underlying mechanism appears to involve complementary contributions of micromechanical interlocking—predominantly driven by HF-induced microporosity—and controlled surface topography from femtosecond laser ablation, which together create a more favourable bonding substrate than either treatment alone can achieve.

Ra–SBS relationship

The moderate Ra–SBS correlation identified in this study (r = 0.411; p < 0.01) confirms that surface roughening contributes meaningfully to bonding but is neither the sole nor dominant governing factor. As has been emphasised in the literature, adhesive performance also depends critically on the geometric quality of the surface topography, surface free energy, and the capacity of the resin cement to infiltrate the conditioned substrate [30, 31]. Porto et al. demonstrated that HF concentration and etching duration jointly influence both profilometric parameters and SBS, underscoring the insufficiency of Ra alone as a predictor of bond strength [32]. Almiro et al. similarly showed that acid concentration, etching time, and the use of a two-component silane system interactively optimise bonding of resin composites to lithium disilicate ceramics, a finding that aligns with the Ra–SBS pattern observed in the present study [33]. Regarding the clinical adequacy of the bond strengths observed, no universal minimum SBS threshold for durable ceramic-resin bonding has been established in the literature; reported acceptable ranges vary considerably depending on restoration type, loading conditions, and testing methodology, with general dental cement performance criteria citing shear bond strengths in the range of approximately 5–20 MPa as clinically relevant. The HF-only group in the present study (14.221 ± 1.803 MPa) falls within this range, suggesting that conventional HF etching alone may already provide clinically adequate bonding for many indications. Nevertheless, the additional bond strength gained through the combined Laser→HF protocol (up to 19.295 ± 2.881 MPa) is supported by the concurrent shift toward predominantly cohesive failure modes (Sect. 3.5), indicating that these higher values approach or exceed the cohesive strength limit of the ceramic-cement interface–a pattern generally regarded as indicative of optimal, rather than merely statistically superior, adhesion.

FE-SEM and XRD findings

The FE-SEM images revealed group-specific surface morphologies fully consistent with the Ra ranking [34]. In HF→Laser groups, laser scanning partially homogenised the HF-induced porosity, producing the highest Ra but a more uniform topographic distribution. In Laser→HF groups, HF deepened and broadened the laser-created grooves into a heterogeneous micro-porous morphology, generating a more complex interface with greater potential for three-dimensional resin infiltration—a finding that mechanistically explains the superior SBS of these groups despite their lower Ra compared with HF→Laser groups. The preservation of characteristic XRD peaks at approximately 22–24° 2θ across all groups confirms that neither HF etching nor femtosecond laser irradiation altered the fundamental bulk crystalline phase of Amber Mill [35, 36]. This finding has direct clinical relevance, as it implies that the aggressive conditioning protocols investigated here do not compromise the structural integrity of the ceramic material.

Failure mode analysis

The progressive shift from exclusively adhesive failure in the control group to predominantly mixed and cohesive failures in combined-protocol groups is fully consistent with published data on thermocycled lithium disilicate specimens [37]. The high proportion of cohesive failures in post-crystallization combined groups—particularly Laser→HF—indicates that bond strength in these groups approached or exceeded the cohesive strength limit of the ceramic–cement interface, which is widely regarded as a hallmark of optimal adhesion. Detogni et al. similarly reported an increase in cohesive and mixed failure rates following thermocycling of lithium disilicate ceramics [37], supporting the interpretation that thermocycling-induced maturation of the resin-ceramic interface can enhance apparent bond performance under certain conditioning conditions.

Amber mill-specific considerations

Demiroğlu and Ongun reported that Amber Mill generally displays higher baseline Ra values than conventional lithium disilicate ceramics under comparable polishing conditions [2], an observation consistent with the control group Ra measured in the present study. Alhomuod et al. emphasised the critical role of pre-cementation silane application in maximising SBS for Amber Mill [6], reinforcing the importance of the complete bonding protocol adopted in the present study. Munoz et al. characterised the relationship between firing protocol, translucency, and microstructural features of Amber Mill, noting that the material’s distinct crystalline architecture may produce surface treatment responses that differ from those of classical lithium disilicate ceramics [5]. The present study’s contribution lies in the systematic evaluation of HF and femtosecond laser application sequence specifically on Amber Mill; while laser-assisted and combined laser-HF conditioning of dental ceramics has been investigated in other materials [13, 15–20], the present findings indicate that application sequence should be considered, alongside the choice of surface treatment modality, when optimising bonding protocols for nano-lithium silicate ceramics.

Limitations

Several limitations of this study merit acknowledgement. The use of a standardised CAD/CAM resin composite block as the substrate material—rather than natural tooth tissue—was a deliberate methodological choice to isolate ceramic surface variables and minimise biological variability, but it inevitably limits direct extrapolation to clinical cementation scenarios. Only one nano-lithium silicate ceramic brand was tested; findings may not be fully generalisable to other lithium silicate or lithium disilicate materials with different compositional profiles. The femtosecond laser parameters employed represent a single point within a broad optimisation parameter space, and systematic parameter studies are warranted to identify the conditions that maximise bonding performance. The absence of surface energy or water contact angle measurements limits the mechanistic interpretation of chemical bonding contributions. Future studies incorporating wettability assessment, X-ray photoelectron spectroscopy (XPS) surface characterisation, and long-term fatigue testing under simulated clinical loading conditions would substantially complement the present findings [3, 15, 16].

Additionally, specimen alignment and cementation seating pressure were performed manually under visual guidance rather than with a standardised mechanical jig or calibrated load, which may have introduced a degree of procedural variability across specimens.

Finally, the clinical applicability of the femtosecond laser-based protocols warrants critical consideration. While the post-crystallization Laser→HF protocol achieved the highest bond strength, femtosecond laser systems are costly, require specialised equipment and technical expertise, and are not routinely available in most dental clinics or laboratories–in contrast to HF etching, which is simple, inexpensive, and widely accessible. Consequently, the recommendation of a combined laser-HF protocol for routine clinical practice cannot be justified by the observed bond strength gain alone; its adoption would need to be weighed against practicality, accessibility, and cost-effectiveness in everyday clinical settings, particularly given that the HF-only protocol already achieved bond strength values within clinically relevant ranges (Sect.  4.3).

Conclusions

Within the limitations of this in vitro study, the following conclusions were drawn:

1. Application sequence is a critical determinant of bonding performance: the Laser→HF sequence yielded the highest shear bond strength (19.295 ± 2.881 MPa), while the HF→Laser sequence produced the greatest surface roughness (1.538 ± 0.058 μm). Both H₀₁ and H₀₂ were rejected.

2. Combined surface treatment protocols (Laser→HF and HF→Laser) consistently outperformed single-treatment protocols (HF alone or laser alone) in shear bond strength, confirming a synergistic interaction between femtosecond laser conditioning and HF etching.

3. A moderate positive Ra–SBS correlation was identified (r = 0.411; p < 0.01); however, elevated Ra did not universally predict higher SBS, indicating that surface morphology quality and application sequence are equally important determinants of adhesion alongside average roughness magnitude.

4. XRD and FE-SEM analyses confirmed that neither HF etching nor femtosecond laser irradiation altered the fundamental crystalline phase composition of the nano-lithium silicate ceramic, demonstrating that both conditioning modalities are safe with respect to material integrity.

5. Within the conditions of this in vitro study, the post-crystallization Laser→HF combined protocol achieved the highest bond strength among the tested groups for this nano-lithium silicate CAD/CAM ceramic. These findings suggest that application sequence warrants consideration alongside the selection of surface treatment modality; however, given the single-material, single-cement, single laser-parameter set, and limited-aging design of this study, further validation under long-term fatigue loading, water ageing, and other clinically relevant conditions is required before this protocol can be recommended for routine clinical use.

Abbreviations

CAD/CAM

Computer-Aided Design / Computer-Aided Manufacturing

FE-SEM

Field Emission Scanning Electron Microscopy

HF

Hydrofluoric Acid

HSD

Honestly Significant Difference

LED

Light-Emitting Diode

MPa

Megapascal

Ra

Arithmetic Mean Surface Roughness

r

Pearson Correlation Coefficient

SBS

Shear Bond Strength

SD

Standard Deviation

SPSS

Statistical Package for the Social Sciences

XPS

X-ray Photoelectron Spectroscopy

XRD

X-ray Diffraction

ANOVA

Analysis of Variance

µm

Micrometer

°C

Degree Celsius

%

Percent

<

Less than

>

Greater than

α

Significance Level (Alpha)

η²

Eta-Squared (Effect Size)

Authors’ contributions

RF designed the study, performed the literature review, conducted the experimental procedures, analyzed the data, and drafted the manuscript. BDY supervised the study process, contributed to methodological planning, reviewed the manuscript critically, and approved the final version of the manuscript. All authors read and approved the final manuscript.

Funding

This research was supported by the Dicle University Scientific Research Projects Coordination Unit as project number DİŞ.25.015. and DİŞ.26.015.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Local Ethics Committee of the Faculty of Dentistry, Dicle University (approval number: 2025-73). All procedures were performed in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki Declaration and its later amendments.

This study did not involve human participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.