In Vitro Effect of Sequential Compressive Loading and Thermocycling on Marginal Microleakage of Digitally Fabricated Overlay Restorations Made from Five Materials

Featured Application

This study highlights that selecting restorative materials with superior marginal integrity, such as zirconia and graphene-reinforced polymers, can enhance the clinical performance and longevity of CAD/CAM overlay restorations. The findings support the use of these materials for high-stress posterior restorations where long-term marginal sealing is critical to prevent secondary caries and restoration failure.

Abstract

Marginal microleakage compromises the longevity and biological seal of indirect restorations. Despite the growing adoption of computer-aided design and manufacturing (CAD/CAM) and three-dimensional (3D) printing technologies, limited evidence compares the marginal integrity of these materials under combined mechanical and thermal stresses. This study evaluated and compared the marginal microleakage of overlay restorations fabricated from five contemporary restorative materials, IPS e.max^® ZirCAD Prime, BioHPP^®, G-CAM, VarseoSmile Crown^Plus, and IPS e.max^® CAD, after sequential compressive loading and thermocycling. A total of 125 extracted human molars were prepared for standardized 1.5 mm-thick CAD/CAM overlay restorations and assigned to three experimental conditions: control, sequential compressive loading (3 × 500 N), and thermocycling (6000 cycles between 5 °C and 55 °C) followed by loading. Microleakage was assessed using 2% methylene blue dye and stereomicroscopy. Data were analyzed using Fisher’s exact test and Fleiss’ Kappa (α = 0.05). G-CAM and IPS e.max^® ZirCAD Prime exhibited the lowest microleakage across all testing conditions, while BioHPP^® showed the highest values. Both sequential compressive loadings and thermocycling significantly increased microleakage in all materials (p < 0.001). The results indicate that material type significantly influences marginal sealing, with G-CAM and IPS e.max^® ZirCAD Prime maintaining superior marginal integrity compared with other materials tested.

1. Introduction

The clinical longevity of indirect overlay restorations largely depends on the integrity of the tooth–restoration interface. Marginal microleakage allows bacterial penetration and fluid diffusion, which can lead to dentine demineralisation, secondary caries, and eventual restoration failure [1,2,3,4,5,6,7]. Therefore, maintaining a durable marginal seal is essential for the long-term success of adhesive restorations [1,8,9,10,11,12,13,14].

Recent advances in dental biomaterials have introduced multiple restorative options fabricated through computer-aided design and computer-aided manufacturing (CAD/CAM) or three-dimensional (3D) printing technologies. Among these, next-generation monolithic zirconia, lithium-disilicate ceramics, high-performance polymers, and graphene-reinforced resin-matrix ceramics have gained increasing clinical popularity [15,16,17,18,19,20,21]. These materials differ markedly in microstructure, elastic modulus, and bonding behaviour, all of which may influence their response to intraoral mechanical and thermal challenges [22,23,24,25,26,27,28,29].

Although several studies have evaluated the marginal integrity of individual restorative materials under specific laboratory conditions, limited comparative evidence exists among contemporary CAD/CAM and 3D-printed materials when subjected to sequential loading and thermal ageing, which more closely simulate combined oral stresses. Understanding how these materials behave at the marginal interface is critical for clinicians selecting indirect restorative options for posterior teeth exposed to high occlusal forces and temperature fluctuations [1,14,18,30,31,32,33,34,35,36,37,38].

The five materials tested in this study present distinctly different microstructural configurations. IPS e.max^® ZirCAD Prime is a gradient monolithic zirconia combining a highly translucent cubic-rich superficial layer with a tetragonal core that provides transformation toughening and high flexural strength. IPS e.max^® CAD is a partially crystallised lithium-disilicate glass-ceramic composed of elongated Li₂Si₂O₅ crystals embedded in a glassy matrix, which becomes more densely packed following crystallisation.

BioHPP^® is a high-performance PEEK-based polymer containing 20% ceramic fillers, characterised by a semi-crystalline polymeric backbone with dispersed inorganic particles that increase stiffness while preserving resilience. G-CAM is a graphene-reinforced resin-matrix biopolymer in which nanometric graphene platelets are incorporated into an organic polymer network, enhancing elastic behaviour and potentially improving crack-deflection mechanisms. VarseoSmile CrownPlus is a 3D-printed hybrid resin composed of a cross-linked polymer matrix with ceramic microfillers, produced through photopolymerisation; its mechanical behaviour is highly dependent on polymerisation kinetics and the layer-by-layer manufacturing process.

These structural differences may influence stress distribution at the tooth–restoration interface, susceptibility to thermal fatigue, and ultimately marginal sealing.

The purpose of this in vitro study was to compare the marginal microleakage of overlay restorations fabricated from five modern restorative materials, IPS e.max^® ZirCAD Prime, BioHPP^®, G-CAM, VarseoSmile Crown^Plus, and IPS e.max^® CAD, after compressive loading and thermocycling.

The null hypothesis stated that there would be no significant differences in marginal microleakage among the tested materials or experimental conditions.

2. Materials and Methods

2.1. Study Design and Ethical Approval

This in vitro experimental study used 125 extracted human molars obtained for This in vitro experimental study used 125 extracted human molars obtained for periodontal or orthodontic reasons. A preliminary pilot study was conducted to estimate the appropriate sample size for the microleakage analysis. Based on the variability observed in this pilot phase and on sample sizes commonly reported in comparable in vitro studies, a group size of 25 teeth per material (10 per experimental condition) was determined to provide adequate statistical robustness while maintaining methodological feasibility. Written informed consent was obtained from all donors. The protocol was reviewed and approved by the Ethics Committee of the Universitat Internacional de Catalunya (approval code PRT-ELM-2020-01). All procedures complied with institutional and international guidelines for the use of human teeth in research.

2.2. Sample Selection and Storage

Molars free from caries, cracks, or restorations were selected. Crown dimensions were standardized according to anatomical references described by Nelson and Ash and verified using a digital caliper (Mahr) [39]. Each tooth was mechanically cleaned with an ultrasonic scaler (SP Neutron Power; Satelec/Acteon Group, Mérignac, France) and hand curettes (Hu-Friedy). After disinfection in 0.5% chloramine-T for 24 h, specimens were stored in artificial saliva (80% artificial saliva + 20% of 0.1% chloramine-T solution) at 4 °C and used within 3 months to preserve structural integrity.

Sample selection was performed by one calibrated operator following the inclusion and exclusion criteria. Random allocation to the experimental groups was conducted by an external investigator not involved in the study procedures. After coding all specimens with alphanumeric identifiers, random assignment to the five material groups and the subsequent experimental conditions was performed using a computer-generated randomization sequence.

2.3. Periodontal Ligament Simulation and Mounting

Three landmarks were marked on each tooth: (1) cementoenamel junction (CEJ), (2) a line 2 mm coronal to the CEJ for restoration margins, and (3) a line 1 mm apical to the CEJ representing bone level. The root surface was coated with a thin layer of molten wax (Red Modelling Special Wax; Cera Reus, Reus, Spain) at ≈90 °C to create a 0.1–0.3 mm space for periodontal-ligament simulation. Each wax-coated tooth was fixed in an acrylic resin mould Paladur (Kulzer GmbH, Hanau, Germany) using a paralleling device (Candulor AG, Glattpark, Switzerland). After polymerisation, the wax was removed, and the space was filled with polyether impression material Impregum (3M Oral Care, St. Paul, MN, USA) (Figure 1).

2.4. Cavity Preparation and Scanning

An independent researcher randomly assigned 25 teeth to each material group, labeling every specimen with a unique alphanumeric code. All cavity preparations were performed by the same calibrated operator to ensure uniformity and minimize methodological variability.

Preparations followed a standardized overlay design, consisting of 1.5 mm occlusal reduction, 1 mm axial reduction, a 1 mm chamfer finish line, and a 6° axial taper with rounded internal line angles. A silicone putty index Zetalabor (Zhermack SpA, Badia Polesine, Italy) was used for dimensional control. Interproximal boxes were excluded to simulate vital teeth (Figure 2).

Each sample was scanned with an intraoral scanner Trios Move 3 (3Shape A/S, Copenhagen, Denmark). The digital impressions were saved in STL format and imported into CAD software (Exocad v3.1; Exocad) for restoration design. The metal stylus used for the compressive load testing was also scanned to define the antagonist geometry (Figure 3).

2.5. Restoration Fabrication

Overlays were fabricated by milling or 3D printing according to each manufacturer’s instructions for the following materials: IPS e.max^® ZirCAD Prime (Ivoclar Vivadent AG, Schaan, Liechtenstein), BioHPP^® (Bredent GmbH & Co. KG, Senden, Germany), G-CAM (Graphenano Dental S.L., Yecla, Spain), VarseoSmile Crown^Plus (BEGO GmbH & Co. KG, Bremen, Germany), and IPS e.max^® CAD (Ivoclar Vivadent AG, Schaan, Liechtenstein).

BioHPP^® overlays were milled from industrially pre-pressed PEEK-based high-performance polymer discs (Bredent). A 5-axis milling unit was used under manufacturer-specified feed rates and bur diameters suitable for semi-crystalline polymers, ensuring accurate marginal reproduction. After milling, restorations were finished using a sequential polishing protocol involving tungsten-carbide cutters, polymer-specific abrasive rubbers, and high-gloss polishing pastes to standardize surface smoothness and marginal quality.

G-CAM overlays were produced by milling graphene-reinforced biopolymer discs using a 5-axis milling strategy with controlled machining parameters to reduce the risk of chipping the polymer–graphene matrix. Margins were refined following the manufacturer-recommended polishing workflow, which included abrasive points and high-shine discs designed for graphene-reinforced polymers, ensuring homogeneous marginal contours.

IPS e.max^® ZirCAD Prime restorations were milled in the pre-sintered state using zirconia-specific milling burs and a 5-axis subtractive protocol. After milling, all restorations underwent a full sintering cycle in a high-temperature furnace according to Ivoclar specifications to achieve final density and the material’s characteristic gradient-translucency structure. A standardized mechanical polishing procedure using fine-grit diamond instruments was performed to minimize surface irregularities and maintain consistent marginal finishing (Figure 4).

IPS e.max^® CAD overlays were manufactured by milling partially crystallized lithium-disilicate blocks (“blue stage”) using CAD/CAM protocols specific to glass-ceramic substrates. Following milling, each restoration underwent the crystallization firing cycle recommended by the manufacturer, enabling full Li₂Si₂O₅ crystal development and the final mechanical properties. A controlled finishing and polishing sequence was performed after crystallization to ensure uniform marginal integrity across all samples.

VarseoSmile Crown^Plus restorations were produced by additive manufacturing using a stereolithography-based 3D-printing system. The material was polymerized layer by layer according to BEGO parameters for layer thickness, exposure time, and light intensity. After printing, restorations were thoroughly rinsed in isopropanol to remove uncured resin, followed by post-curing under controlled light intensity and temperature. Final surface finishing was carried out using sequential polishing steps compatible with light-cured hybrid resins.

All manufacturing processes followed the respective manufacturer’s instructions and were standardized to ensure consistency in surface quality, marginal finishing, and reproducibility across all material groups.

2.6. Adhesive Cementation

Each restoration was luted using its corresponding resin cement and adhesive protocol (

Table 1. Surface pretreatment protocols for each restorative (overlay) material, including abrasion, chemical conditioning, and adhesive application.

Material	Abrasion Parameters	Chemical Pretreatment	Adhesive & Curing Protocol
BioHPP®	110 µm Al2O3, 2 bar, 10 s, 10 mm distance	—	Visio.link (Bredent); light-cured 90 s @ 1000 mW/cm2
G-CAM	50 µm Al2O3, 2 bar, 15 s, 10 mm distance	—	Scotchbond (3M); light-cured 20 s @ 1000 mW/cm2
IPS e.max® ZirCAD Prime	30 µm Al2O3, 1 bar, 15 s, 10 mm distance	Ivoclean (60 s), rinse; Monobond® Plus (60 s)	—
VarseoSmile CrownPlus	50 µm glass beads, 2 bar, 15 s, 10 mm distance	5% HF (20 s), rinse 2 min, dry with oil-free air	Scotchbond (3M); light-cured 10 s @ 1000 mW/cm2
IPS e.max® CAD	—	5% HF (20 s), rinse 2 min; Monobond® Plus (60 s)	Scotchbond (3M); light-cured 10 s @ 1000 mW/cm2

and

Table 2. Resin cementation and polishing systems used for each material, including standardized pressure, light-curing, and finishing protocols.

Material	Resin Cement Used	Cementation Protocol	Polishing System Used
BioHPP®	Variolink® Esthetic DC	Preheated cement; 1 kg load × 10 min; light-cured 20 s/surface	Polish & Go (Bredent)
G-CAM	RelyX™ Unicem 2	1 kg load × 10 min; light-cured 20 s/surface	KemicPolitur (Aidite)
IPS e.max® ZirCAD Prime	SpeedCEM® 100	1 kg load × 10 min; light-cured 20 s/surface	OptraGloss® (Ivoclar Vivadent)
VarseoSmile CrownPlus	Variolink® Esthetic DC	1 kg load × 10 min; light-cured 20 s/surface	VarseoSmile Polishing Set (BEGO)
IPS e.max® CAD	Variolink® Esthetic DC	1 kg load × 10 min; light-cured 20 s/surface	OptraGloss® (Ivoclar Vivadent)

). Cementation was performed under a constant 1 kg load for 10 min using a custom parallelising device to ensure uniform pressure (Figure 5). All cementation procedures were performed by the same calibrated operator to ensure consistency across all restorations.

Tooth surfaces were cleaned with an oil-free water spray and selectively etched with 37% orthophosphoric acid (15 s for dentine and 30 s for enamel), rinsed for 1 min, and gently air-dried. A universal adhesive Scotchbond (3M Oral Care, St. Paul, MN, USA)was applied and either light-cured or left uncured according to the manufacturer’s instructions. Finishing and polishing procedures were standardized across all materials to eliminate surface irregularities and ensure uniform marginal quality.

2.7. Experimental Groups and Loading Protocols

Specimens were randomly assigned to three experimental groups (n = 25 per material) using computer-generated random numbers by an independent investigator.

Table 3. Experimental group allocation showing aging procedures, mechanical loading, and specimen distribution per material and group.

Group	Aging Procedure	Mechanical Loading	Specimens Per Material	Total Specimens per Group
Group 1 (Control)	None	None	5	25
Group 2 (Loading)	None	3 × 500 N compressive loads	10	50
Group 3 (Aging + Loading)	6000 thermocycles (5–55 °C)	3 × 500 N compressive loads	10	50

summarizes the experimental conditions.

Compressive loading was applied using a universal testing machine ZwickRoell Z005 (ZwickRoell GmbH & Co. KG, Ulm, Germany) with a bullet-shaped metal stylus (6 mm diameter) at a crosshead speed of 0.5 mm/min. Three consecutive compressive loads of 500 N were applied at a 0° angle, each with a holding time of 15 s. A compressive load of 500 N was selected to reproduce a high-stress parafunctional scenario in the posterior region, exceeding normal masticatory forces (100–200 N). Using an upper-limit physiological load provided a demanding mechanical challenge capable of detecting material-dependent differences. The three sequential loading events were intended to mimic repeated high-stress contacts without full cyclic fatigue.

Thermal aging consisted of alternating immersion in water baths at 5 °C and 55 °C, with 2-min cycles (30 s per bath + 1 min transition), totaling 6000 cycles, equivalent to approximately 5 years of clinical service.

2.8. Microleakage Evaluation

After testing, specimens were coated with two layers of nail varnish (Beter), leaving a 1 mm band below the cementation line uncoated. Samples were immersed upside down in 2% methylene blue dye (Xalabarder Farma S.L., Barcelona, Spain) for 24 h at 37 °C, rinsed for 10 min, and sectioned mesiodistally through the central fossa using a precision saw Isomet 1000 (Buehler Ltd., Lake Bluff, IL, USA) at 375 rpm under water irrigation.

Sections were examined under a stereomicroscope Stereo Discovery V8 (Carl Zeiss Microscopy GmbH, Jena, Germany) at 40× magnification equipped with a digital camera AxioCam ERc 5s (Carl Zeiss Microscopy GmbH, Jena, Germany). Microleakage was scored at four points per sample (two mesial and two distal) by three calibrated examiners who were blinded to group allocation.

Leakage was rated on a five-point ordinal scale (0–4) according to dye penetration depth (

Table 4. Scoring criteria for the qualitative assessment of microleakage based on the depth and extent of dye penetration along the restoration–tooth interface.

Score	Description
0	No dye penetration observed
1	Dye penetration up to half the chamfer margin (≤0.5 mm)
2	Dye penetration across the entire marginal interface (≥1 mm)
3	Dye penetration along the margin and into the ascending wall of the cavity
4	Dye penetration along the margin and throughout the entire restoration–cavity interface

).

2.9. Statistical Analysis

A total of 1500 scores were obtained across groups. Inter-examiner agreement was assessed with Fleiss’ Kappa (κ), where κ < 0.4 = poor, 0.4–0.7 = Good, >0.7 = very good. Group and material effects were analysed using Fisher’s exact test for multiple comparisons (α = 0.05). Analyses were performed in R software v4.4.3 (R Foundation for Statistical Computing) using packages openxlsx, ggplot2 and ggpubr.

3. Results

3.1. Inter-Examiner Agreement

Marginal microleakage was evaluated in all 125 restorations. Three independent calibrated examiners recorded scores at four reference points per specimen. Fleiss’ κ indicated good inter-examiner agreement (κ ≈ 0.60, p < 0.001) (

Table 5. Fleiss’ kappa index was used to assess inter-examiner agreement for the analyzed variables.

Variable	Evaluation Type	Kappa	Z-Value
Microleakage M1	GLOBAL	0.59	18.74
Microleakage D1	GLOBAL	0.58	19.25
Microleakage M2	GLOBAL	0.61	20.61
Microleakage D2	GLOBAL	0.61	20.15

Kappa values below 0.4 indicate poor agreement, 0.6 good agreement, whereas values above 0.7 indicate very good agreement. MM1 = Mesial evaluation 1; MD1 = Distal evaluation 1; MM2 = Mesial evaluation 2; MD2 = Distal evaluation 2.

). It should be noted that intra-examiner agreement was not assessed for sample selection or tooth preparation, as these procedures were performed entirely by a single calibrated operator. Agreement analysis was conducted exclusively for the microleakage scoring process. Fisher’s exact test showed significant differences among restorative materials and between experimental conditions involving mechanical loading and thermal ageing.

3.2. Effect of Loading and Ageing on Marginal Microleakage

All materials exhibited increased microleakage after sequential compressive loading (Group 2) compared with the control (Group 1) (Figure 6). Microleakage scores were expressed as percentage distributions across categories 0–4.

Significant differences in marginal leakage were detected among materials. IPS e.max^® ZirCAD Prime presented the lowest leakage at both mesial and distal sites, followed by G-CAM. BioHPP^® and IPS e.max^® CAD displayed the highest leakage values, particularly in severe categories (3 and 4). Mild leakage occurred least often in ZirCAD Prime and most in VarseoSmile Crown^Plus.

Thermal ageing followed by loading (Group 3) further increased microleakage in all materials except in G-CAM, which maintained superior sealing. In contrast, BioHPP^® and IPS e.max^® CAD demonstrated markedly higher leakage. VarseoSmile Crown^Plus showed intermediate behaviour with frequent but less severe scores.

Thermal ageing aggravated leakage, particularly in IPS e.max^® CAD and BioHPP^®, whereas ZirCAD Prime and G-CAM exhibited greater dimensional stability (Figure 7).

3.3. Multiple Comparisons and Representative Images

Table 6. Multiple comparison analysis of marginal microleakage for all measurement sites (MM1, MM2, DM1, DM2) across the five restorative materials.

Variable	Microleakage Degree	G-CAM	IPS e.max® CAD	IPS e.max® ZirCAD Prime	BioHPP®	VarseoSmile CrownPlus	p-Value	SR
MM 1	0	41.7%	4.48%	38.81%	7.4%	7.46%	<0.001	*
MM 1	1	14.29%	14.29%	14.29%	0%	57.14%
MM 1	2	0%	42.86%	0%	14.2%	42.86%
MM 1	3	0%	58.33%	0%	0%	41.67%
MM 1	4	0%	25%	5.56%	61.1%	8.33%
DM 1	0	45.45%	0%	43.64%	9.1%	1.82%	<0.001	*
DM 1	1	16%	32%	0%	8%	44%
DM 1	2	0%	36.36%	22.73%	9.1%	31.82%
DM 1	3	6.67%	60%	0%	0%	33.33%
DM 1	4	0%	15.15%	3.03%	63.6%	18.18%
MM 2	0	46%	4%	32%	4%	14%	<0.001	*
MM 2	1	9.09%	22.73%	22.73%	9.1%	36.36%
MM 2	2	12.5%	31.25%	12.5%	12.5%	31.25%
MM 2	3	11.54%	42.31%	23.08%	11.5%	11.54%
MM 2	4	0%	19.44%	2.78%	58.33%	19.44%
DM 2	0	39.34%	4.92%	22.95%	13.1%	19.67%	<0.001	*
DM 2	1	18.75%	12.5%	12.5%	0%	56.25%
DM 2	2	0%	45.45%	22.73%	9.1%	22.73%
DM 2	3	17.65%	41.18%	41.18%	0%	0%
DM 2	4	0%	23.53%	5.88%	58.8%	11.76%

MM1 = Mesial microleakage evaluation 1; MM2 = Mesial microleakage evaluation 2; DM 1= Distal microleakage evaluation 1; DM2 = Distal microleakage evaluation 2. 0–4: Qualitative scale previously defined. (SR) = Significant result. Asterisks indicate statistically significant differences across all marginal microleakage sites evaluated (MM1, MM2, DM1, DM2).

summarises the multiple comparisons for Group 3 according to the restorative material tested. Data are expressed as mean ± standard deviation. p values were adjusted using the FDR method, and significant results (SR) are indicated with asterisks (*) (Table 6). Statistically significant differences correspond to all evaluated marginal sites (MM1, MM2, DM1, DM2), as indicated by the asterisks.

In Figure 8, we can observe images obtained by the stereomicroscope camera, from which the different evaluators issued their scores (Figure 8).

4. Discussion

The alternative hypothesis of this study was accepted, as statistically significant differences in marginal microleakage were observed among the restorative materials evaluated. The study investigated the microleakage behavior of overlay type restorations fabricated from five indirect materials subjected to simulated functional stress. A standardized methodology, including a single sagittal section per specimen, allowed reproducible evaluation at four reference points. This approach is consistent with previous evidence indicating that excessive sectioning may induce structural weakening and affect leakage interpretation [1,9,26,28].

The findings of this study must be interpreted in light of its inherent in vitro limitations. Laboratory conditions cannot fully replicate the complexity of the oral environment, where factors such as salivary flow, pH fluctuations, biofilm activity, and thermal–mechanical fatigue occur simultaneously. The loading protocol, although intentionally designed to simulate a parafunctional high-stress scenario, does not reproduce long-term cyclic fatigue. Furthermore, each restoration was cemented following the material-specific adhesive system recommended by the manufacturer, which means that the results represent the performance of each material–cement combination rather than the isolated behavior of the material itself. These limitations should be considered when extrapolating the results to clinical situations.

It should also be acknowledged that each material was cemented using its manufacturer-recommended adhesive and resin cement. Therefore, the results represent the performance of the material cementation system as a whole, and the effect of the restorative material cannot be completely isolated from the influence of the corresponding adhesive protocol. Cementation thus acts as an inherent confounding variable that should be considered when interpreting the differences observed among groups.

Differences in microleakage may also be influenced by the manufacturing technique of each material. Milled restorations generally present greater dimensional stability due to the subtractive nature of the process, which reduces the risk of polymerization-related distortions. In contrast, the additive manufacturing workflow of VarseoSmile Crown^Plus, based on layer-by-layer photopolymerization, may be more sensitive to factors such as degree of conversion or interlayer accuracy, potentially affecting marginal adaptation. Although marginal fit was not specifically measured in this study, these well-known characteristics of the manufacturing processes offer a plausible explanation for the variations observed among materials.

The microleakage patterns observed across materials can be partly explained by their intrinsic microstructural characteristics. ZirCAD Prime, with its high tetragonal content in the internal region and transformation-toughening behavior, exhibited superior marginal integrity under sequential loading and thermal cycling. In contrast, the semi-crystalline structure of BioHPP^®, despite offering excellent resilience, may permit greater interfacial deformation under compressive stress, explaining its higher leakage values.

Lithium-disilicate (IPS e.max^® CAD), characterized by its dense network of elongated crystals, provides high strength but limited elasticity, which may increase stress concentration along margins under cyclic thermal expansion. Similarly, the hybrid composition of VarseoSmile CrownPlus, with polymer-based matrices and ceramic microfillers, may be influenced by polymerization shrinkage and degree of conversion inherent to 3D-printed materials, resulting in a less favorable marginal seal after aging.

G-CAM’s performance could be related to its graphene-reinforced polymer matrix, which enhances the material’s toughness and elastic recovery; however, its exact proprietary composition prevents definitive mechanistic conclusions. Overall, the relationship between composition, elastic modulus, and microstructure offers a coherent explanation for the differences observed in microleakage.

One of the main limitations in the microleakage literature is the lack of standardization among experimental protocols. Wide variations exist in dye type, margin location (enamel versus dentin), finish line configuration, cavity design, and thermocycling conditions [1,9,27,38]. Likewise, leakage evaluation can be qualitative, semi-quantitative, or quantitative, which complicates direct comparison across studies [1,9]. The present work adopted a 0–4 ordinal scale, frequently used and clinically meaningful, which provided reproducible data [1,10].

Tracer selection is another critical methodological factor. Although methylene blue is commonly employed due to its simplicity, cost-effectiveness, and availability, alternatives such as basic fuchsin or silver nitrate offer enhanced visualization under electron microscopy [1,10,13,26,27,32,40,41]. Costa et al. [42] emphasized that silver nitrate requires photographic development and precise filtration, which increase sensitivity but also procedural complexity. The use of 2% methylene blue for 24 h at 37 °C, as in this study, aligns with validated protocols by Heintze et al. [43] and AlHabdan et al. [9].

All samples were coated with nail varnish leaving 1 mm of exposed margin, limiting dye penetration to the tooth–restoration interface [1,10,30,31,32,44,45]. Although some authors suggest removing the varnish before analysis, previous evidence with methylene blue indicates no significant influence on leakage scores [42]. The choice of a single cut per specimen balanced analytical reliability with sample integrity, as multiple cuts can compromise structural cohesion [13,42].

4.1. Material-Dependent Differences

The results revealed significant differences among materials. IPS e.max^® ZirCAD Prime and G-CAM exhibited the lowest microleakage values, even after compressive loading and thermocycling. These outcomes may be attributed to their high dimensional stability and robust internal structure: zirconia oxide for the former and graphene-reinforced polymer for the latter. In contrast, BioHPP, a PEEK-based composite with ceramic fillers, showed the highest leakage across all testing phases.

These findings agree with other studies reporting superior marginal behavior in ceramics and highly reinforced materials [46,47,48,49,50]. Juloski et al. [13] and Mousavinasab et al. [51] also observed greater leakage at dentin margins compared to enamel, an influence controlled in the present study by standardizing all margins in enamel. IPS e.max CAD and VarseoSmile Crown^Plus showed intermediate outcomes, with significant increases in leakage following functional and thermal aging. For VarseoSmile Crown^Plus, Dilan Seda Metin et al. [29] linked marginal misfit to printing angle, supporting the variable performance observed here. In addition to the manufacturing workflow, other intrinsic factors of 3D-printed resin materials may contribute to their marginal behavior. Polymerization shrinkage can occur during both the initial layer-by-layer printing and the post-curing stage, potentially affecting interfacial adaptation. Moreover, the degree of conversion may vary depending on light penetration, restoration geometry, and post-curing efficiency, leading to heterogeneous polymer networks. These characteristics may partly explain the less favorable marginal performance observed in the 3D-printed group compared with the milled materials.

IPS e.max CAD, although clinically established and esthetically favorable, exhibited moderate marginal integrity, likely due to its lower resistance to compressive stress compared with zirconia. This finding underscores the importance of functional stability in material selection for posterior restorations.

4.2. Functional and Mechanical Interpretation

The simulated occlusal forces (three 500 N compressive cycles with 15 s loading and resting intervals) aimed to reproduce realistic clinical conditions such as hard food mastication or parafunctional habits. While in vitro conditions cannot fully replicate intraoral variability, they provide valuable insight into functional resistance under controlled parameters. The superior performance of IPS e.max ZirCAD Prime and G-CAM may also relate to their elastic modulus and viscoelastic behavior. Proper compatibility between the modulus of the restorative material and dentin helps optimize stress distribution and minimize marginal disruption [22,23,24,25,52,53,54,55]. The graphene reinforcement in G-CAM provides a unique balance between rigidity and flexibility, which may attenuate interfacial stresses and prevent microfractures [56,57,58]. Although the favorable performance of G-CAM may be related to the presence of graphene-reinforced polymer matrices, the specific contribution of graphene cannot be confirmed in this study. The material’s detailed formulation is proprietary, and no direct assessment of its microstructural reinforcement mechanisms was performed. Therefore, any interpretation regarding graphene-related effects must be considered speculative and based solely on general principles of polymer reinforcement rather than on evidence generated by the present experiment.

In contrast, BioHPP demonstrated inferior stability under mechanical and thermal loading. Although its polymeric matrix has advantages such as light weight and favorable elasticity, its limited deformation resistance may explain the greater marginal leakage observed. Similar results have been reported for polymer-based restorative systems [20,59,60,61].

4.3. Influence of Cementation

Despite standardized adhesive procedures, interfacial compatibility between cement and substrate likely influenced the results. Ceramic materials like IPS e.max CAD, which can be silanized, often exhibit superior bonding with resin cements. Conversely, materials such as BioHPP^® require specific surface conditioning (e.g., air abrasion with alumina or application of dedicated primers) [19] Even subtle differences in bonding affinity may explain variations in marginal sealing among groups.

4.4. Clinical Relevance

From a clinical perspective, IPS e.max^® ZirCAD Prime and G-CAM may be the most reliable choices for patients with high occlusal load or parafunctional habits, providing superior marginal sealing and dimensional stability. Conversely, BioHPP^® could be indicated in low-stress situations or when minimal weight and biocompatibility are prioritized. The study reinforces the importance of balancing esthetics, functional demand, and adhesive performance when selecting indirect restorative materials.

4.5. Limitations and Future Directions

This in vitro design cannot fully reproduce intraoral conditions such as saliva, bacterial biofilm, or dynamic occlusal movement. The leakage evaluation, although clinically relevant, relied on an ordinal qualitative scale. Future studies could benefit from quantitative image-based analysis, larger material diversity, and clinical trials to validate long-term performance.

Another limitation is that each material was cemented using its own manufacturer-recommended adhesive system. As a result, the outcomes reflect the behavior of the material–cement combination rather than the isolated influence of the restorative material.

5. Conclusions

• IPS e.max^® ZirCAD Prime and G-CAM exhibited the lowest marginal microleakage values under all experimental conditions, indicating superior marginal sealing compared with the other tested materials.
• VarseoSmile Crown^Plus and IPS e.max^® CAD showed intermediate performance, while BioHPP^® presented the highest microleakage, particularly after mechanical loading and thermal aging.
• These findings suggest that zirconia-based and graphene-reinforced CAD/CAM materials provide greater marginal stability under functional stress, making them potentially more suitable for patients with high occlusal loads or parafunctional habits.
• Although this in vitro protocol simulated thermomechanical fatigue, clinical studies are necessary to confirm the long-term behavior of these restorative materials under real intraoral conditions.