In this study, a total of 80 ti-bases (Ti-Base NB RS 4.3 L, Sirona Dental Systems GmbH, Bensheim, Germany) and 80 Vita ENAMIC® PICN crowns (VITA ENAMIC® for CEREC®/inLab, VITA Zahnfabrik, Bad Säckingen, Germany) were divided into 8 groups of 10 samples each. A similar test method was used in previous studies [19]. The materials for the bonding protocol and the conventional pretreatment protocol were used according to the manufacturer’s instructions from VITA Zahnfabrik. The crown was designed using CEREC SW® 5.0 software (Sirona Dental Systems GmbH, Bensheim, Germany). The milling blocks used consisted of a ceramic network (86 vol%) and a polymer network (14 vol%). The crown design was digitally positioned in the block so that part of the design was outside the block and the later milled crown had a flat basal structure (red marking) (Fig. 1). This design provided an even and flat support surface for the pull-off device (Zwick 1425, ZwickRoell GmbH & Co. KG, Ulm, Germany). The abutment in the PICN blocks was determined by the manufacturer, guaranteeing an identical shape and cement gap in all samples. To avoid contamination of the surfaces, gloves were worn during all procedures and changed between the different pretreatments.

The design of the PICN crown incorporates design components that extend beyond the block (red marking)

In all groups, the ti-bases were pretreated in the same way as follows: Group (A)—(DP): sandblasting vertically to the ti-base at 10 mm distance using a dental sandblaster (P-G 400, Harnisch + Rieth GmbH & Co. KG, Winterbach, Germany) with 50 μm Al2O3 (Plurakorund, Pluradent AG & Co. KG, Offenbach, Germany), at 1.0 bar for approximately 10 s and application of the bonding agent (VITA ADIVA M-Prime, Metal/Alloy primer, Harvard Dental International GmbH, Hoppegarten, Germany) for 10 s. Different pretreatment protocols have been applied in the PICN crowns (Table 1).

The PICN crowns in groups B, C, BP and CP underwent HF treatment (VITA Ceramics Etch 3 ml syringe, hydrofluoric acid 5% etch gel, VITA Zahnfabrik, Bad Säckingen, Germany). Specifically, the inner surfaces of the crowns in these groups were pretreated with HF for a duration of 60 s. At the end of the 60 s, the HF was completely removed with running water and then with a steam jet device (Triton SLA, Bremer Goldschlägerei Wilh. Herbst GmbH & Co. KG, Bremen, Germany). Afterward, the surfaces were dried with an oil-free, clean air flow. In group C, HF pretreatment was followed by conditioning with a silane bonding agent (VITA ADIVA C-Prime, Harvard Dental International GmbH, Hoppegarten, Germany) for 10 s. Overall, the inner crown surfaces of groups C, D, CP and DP were pretreated with the silane agent. First, a drop of the silane was placed in a sterile tray. The inner surface of the crowns was then evenly and thinly moistened with the bonding agent using an application brush. The silane was applied to the crown surface for 10 s. Afterward, the surface was gently treated with an oil-free, clean air flow until complete drying was achieved. In the groups AP, BP, CP and DP, plasma conditioning (CAP) (piezobrush PZ3, relyon Plasma GmbH, Regensburg, Germany) was performed with the piezobrush PZ3 plasma device (prior to silane application, Group CP, DP). The piezobrush PZ3 employs advanced piezoelectric direct discharge (PDD) technology to produce CAP [20]. The plasma unit was operated at 100% power (18 W, 240 V, < 50° C) for a period of 30 s throughout the pretreatment process. The needle nozzle for non-conductive materials was selected as the device for conditioning the inner surface of the PICN crowns to ensure uniform treatment of the inner surface of the crown.

For adhesive bonding a dual curing composite resin was used (VITA ADIVA IA-CEM; Harvard Dental International GmbH, Hoppegarten, Germany). The abutments (TiBase NB RS 4.3, Sirona Dental Systems GmbH, Bensheim, Germany) were screwed into an implant analogue (NobelParallel Conical Connection RP 4.3 × 11.5 mm, Nobel Biocare Services AG, Zürich, Switzerland) and fixed in a plastic clamp (Mediplast AB, Malmö, Sweden). The components of the adhesive were mixed using automix cannulas (3 M Deutschland GmbH, Neuss, Germany), disposing of the first mixed portion for every single bonding procedure. The screw channel was sealed with cotton wool. The pretreated surfaces of the titanium adhesive abutment and of the PICN crown were then completely covered with a thin layer of adhesive cement. The PICN crown was carefully and accurately placed onto the ti-base until it was seated in the final position. Excess adhesive was removed from the screw channel with a microbrush. The bonded parts were clamped in a special hybrid abutment bonding aid (HPdent GmbH, Gottmadingen, Germany) to guarantee standardized pressure in all specimens. The crowns were then cured from all sides for 3–5 s at 1200 mW/cm2 with a light curing lamp (Elipar S10, 3 M Deutschland GmbH, Neuss, Germany). Excess cement was removed with a LeCron spatula (Henry Schein Dental GmbH, Langen, Germany). The bonded hybrid abutments were kept in the hybrid abutment bonding aid for 10 min. To avoid an oxygen inhibition layer, glycerine gel (Liquid Strip, Ivoclar Vivadent AG, Schaan, Liechtenstein) was applied to the adhesive gap, followed by light polymerisation for additional 30 s. The time was recorded with a stopwatch (GEFU GmbH, Eslohe, Germany). After 10 min, the bonded PICN crowns were removed from the fixation device and remained at room temperature for at least 24 h. Subsequently, the adhesive gap was polished with ceramic polishers operating at a maximum speed of 5000 rpm.

All bonded PICN crowns underwent thermocycling to simulate oral aging. The specimens were thermocycled using a device (Thermocykler Willytec, SD Mechatronik GmbH, Feldkirchen-Westerham, Germany) alternating between cold (+ 5.0 °C) and warm (+ 55.0 °C) distilled water. Each cycle consisted of 30-s exposures to these temperatures, followed by 5 s of dripping and transfer periods between water basins. This process was repeated for a total of 5000 cycles, each lasting 80 s. Following thermocycling, the specimens were stored in distilled water at a temperature of 23 °C until further processing.

A universal testing machine (Zwick 1425, ZwickRoell GmbH & Co. KG, Ulm, Germany) was employed to perform pull-off tests with the objective of determining the maximum bond strength required to remove the crown. To prevent misinterpretation, the fracture cut-off threshold was set at 50 N. An implant (NobelParallel Conical Connection RP 4.3 × 11.5 mm, Nobel Biocare Services AG, Zürich, Switzerland) was secured using a wedge grip attached to the lower part of the machine, while the PICN crowns were placed in the upper part using a custom-made specimen holder designed specifically for this purpose (Fig. 2). The specimens were then screwed onto the implant with a torque of 35 Ncm. The holder ensured a flat contact surface by means of a polished steel disc with a central recess for the specimen. The gripping system uses a ball joint to ensure that the bond strength is applied straight and vertically, ensuring that the measuring axis corresponds to the test axis. The tests were conducted at a velocity of 1 mm/min, and were recorded on video for further analyses. All pull-off tests continued until the crowns fully detached from the adhesive bases.

Custom-made holder (yellow) and wedge grip (red)

Following the completion of the pull-off tests, each surface of both components from the respective test groups underwent a detailed visual analysis using an optical microscope (VHX-1000, Keyence Deutschland GmbH, Neu-Isenburg, Germany) at 30× magnification, with images recorded for documentation. During the inspection, three observers identified any remaining adhesive residues and categorised fractures as either adhesive or cohesive. Adhesive fractures typically occur when the adhesive's maximum strength is exceeded. In cases of pure adhesive fracture, all adhesive remains on one substrate without any residue on the other. This can be observed in cases where the adhesive adheres entirely to the titanium bonding base or PICN crown. Furthermore, an adhesion fracture may occur on both components to be joined, whereby the adhesive residues on the two components to be joined fit together like a jigsaw puzzle.

The statistical analysis of the data was carried out using the programmes IBM SPSS Statistics (SPSS Statistics Version 27, IBM Corp., Armonk, New York, USA) and SAS (SAS 9.4., SAS Visual Statistics, 100 SAS Campus Drive, Cary, NC, USA). With regard to the central questions the following null hypotheses were formulated and tested as part of a statistical analysis:

A three-factor ANOVA was performed to compare the mean value of the BS with regard to the various pretreatment measures and the interactions between the pretreatment components, in which the effect size was determined. A Shapiro–Wilk test and a Levene test were carried out beforehand to check whether the model assumptions of the ANOVA were significantly violated. The significance level was set at p = 0.05.

For the comparison of the mean BS based on the different failure modes, a Fisher’s Exact Test was carried out. The significance level was set at p = 0.05.