The study protocol was approved (number SRC/REG/2015-2016/63) by the Institutional Ethical Committee, College of Dentistry, King Khalid University; it was conducted during the first academic semester of 2016. Forty non carious, sound human maxillary premolar teeth extracted for orthodontic reasons were used in this in vitro study. The exclusion criteria for teeth samples included caries, large restorations, abrasion, microcracks, discolouration and white spot lesions. After removal of dental plaque, calculus, and periodontal fibres, the teeth were stored in distilled water at room temperature until the preparation for the study. The teeth samples were randomly divided into four groups (n=10) of ten each according to the adhesive system used for luting the PLV. All teeth samples were mounted in acrylic resin blocks to provide better control during tooth preparation. The facial surface reduction was initiated by preparing 0.6 mm horizontal grooves with the help of depth preparation diamond bur (Diatech, Coltane. AG, Switzerland). The facial surface was painted with waterproof paint to enable the uniform reduction. The painting of the surface was helpful in differentiating the contrast between prepared and unprepared teeth. The anatomical reduction of the facial surface was accomplished by using the round ended diamond bur in three different planes until the colour at the bottom of grooves disappears. The chamfer finish line at the gingival margin was prepared. The proximal area preparation was completed with a round ended tapered bur along the long axis of the tooth. Cusp overlap was accomplished with 1.5 mm cusp reduction and extending the reduction onto the lingual incline of the buccal cusp by 2 mm. Single operator prepared all the teeth used in the study.

A total of forty PLV were fabricated from lithium disilicate glass ceramic ingots (IPS Empress Esthetic, Ivoclar, Schaan/Liechtenstein) with ETC1 shade. The PLV samples were made by burnout and heat pressing of 0.7 mm thickness wax pattern at 920⁰C. The neutral glaze was applied over the outer surface of the PLV by firing ceramic samples at 765⁰C. Since smooth and polished porcelain is considered necessary for porcelain’s colour stability, surface polishing was performed using 3 mm and 6 mm diamond polishing burs and paste.

All PLV fitting surfaces were etched with 5% hydrofluoric acid (IPS ceramic etching gel) for 20 seconds, subsequently washed thoroughly with water and dried with oil free air. The samples were ultrasonically cleaned for 10 minutes, following this, a monocomponent silane (Monobond Plus) was applied to the conditioned surface. It was allowed to react for one minute and thoroughly scattered with air. The corresponding tooth surfaces were conditioned with four different adhesive system following the manufacturer’s instructions [Table/Fig-1].

The colour determination of each sample was accomplished by the colour spectrometer (easy shade-VITA Zahnfabrik, Bad Säckingen, Germany). The ceramic veneer was divided into three equal parts with the help of a digital caliper as cervical, middle and incisal portions. The central part of each segment was selected for measuring the colour and used as reposition demarcation for future reference. The CIELAB detail of each location was tabulated; the average of all three areas was regarded as the reference colour of the PLV specimen.

Following the colour determination of each sample, PLV samples were subjected to accelerated aging by thermocycling process in water between 5°C and 55°C for 5000 cycles with a dwell time of 30 seconds. The samples were further exposed to the xenon lamp of 7000⁰K and 150000 Lux. The light exposure duration was for 100 hours and during the light exposure the temperature was maintained at 37⁰C with 100% humidity.

Following the accelerated ageing process, follow up colour measurements were performed following a similar procedure at the identical areas for each sample.

The colour differences (ΔE) were determined using the coordinates L*, a* and b* by using formula [12]:

Where ΔL* is the variation of L*, Δa* is the variation of a*, and Δb* is the variation of b*. The statistical analysis was performed with SPSS 19.0 software (IBM Corporation, Armonk, New York, USA). The values obtained for L*, a*, b* and ΔE* were subjected to ANOVA and LSD post-hoc comparison at the significance level p=0.05.

The luting cements used in the study had various ranges of colour change after the accelerated ageing process. The mean ∆L, ∆a, ∆b values before and after the accelerated ageing for all the groups along with the mean colour change (∆E) are summarised in the [Table/Fig-2]. The mean colour changes observed in the Group I (etch wash-dual cure) was highest with ∆E at 2.491. The Group II (etch wash-light cure) and Group III (self-etch) had mean ∆E value at 1.110 and 2.357 respectively. The Group IV (self–adhesive) presented the lowest colour change (∆E) values at 0.614. The results of ANOVA [Table/Fig-2] showed the statistically significant difference in ∆E among the luting cements tested in the study with a p-value at 0.041.

The statistical analysis with LSD post-hoc multiple comparison test [Table/Fig-3] showed the presence of statistically significant difference between etch wash-dual cure cement (Group I) and self-adhesive cement (Group IV) with a p-value of 0.017. The statistical analysis also revealed the presence of the significant difference between self-adhesive (Group IV) and self-etching cement (Group III) (p=0.026).

The increased expectation of aesthetic outcome of the society has led to the wider application of ceramic laminates in modern dentistry. The colour of the restoration in harmony with adjacent natural teeth and its permanency are important criteria for the satisfactory aesthetic outcome. The resin cements with different adhesive strategy and polymerization techniques are routinely used for the luting purpose. The dentin substrate underneath the composite resin cement is composed of hybrid layer [15]. The hybrid layer consists of polymerized monomer and demineralized collagen resulting from adhesive treatment [16]. The structural integrity of hybrid layer components plays a significant role in the luting cement colour. The tooth samples in this study were subjected to the accelerated ageing to anlyse the effect of different adhesive bonding techniques on the colour stability of luting cements used in the PLV. In the present study, the spectrophotometer was used to identify the colour, since it is considered as more objective, quantifiable and superior to visual method [17]. The Vita Easy shade used in the study consists of embedded fiber-optic light source at the instrument tip. Hence, the influence of external light and its induced colour changes due to metamerism were excluded. The colour was recorded in CIELAB system. It has multiple advantages; it includes all perceivable colours, device independence, perceptual uniformity similar to human perception and ability to denote the colours in numerical units. The in vitro ageing process of composites resin cements was achieved by immersion in the water and thermocycling to simulate the oral environment. Researchers, furthermore, advocated the xenon light exposure in controlled temperature to evaluate the colour stability for resin cement [18].

All the tested adhesive systems during the study showed the colour changes to varying degree. Among the tested adhesive system, etch-wash dual cure luting cement recorded the highest mean colour (∆E) change. This study confirms the observation from earlier studies that the dual cure resins have more ageing induced colour changes in comparison to light cure resins [19,20]. The increased colour changes can be attributed to multiple factors like degradation of residual amines, and oxidation of remaining unreacted carbon double bonds. These structural changes in the cement lead to the formation of yellow compound. Etch wash-light cure resin cement had lesser mean colour change (∆E) at 1.110. Previous researchers have also found better colour stability of the light cure cements [21]. The aliphatic amines are used as polymerization initiators in light cure resins. The aromatic tertiary amines have more tendencies for oxidization than the aliphatic amines. Hence, the light cure resins are reported to have better colour stability [8].

The colour changes are also attributed to the properties of the resin matrix and filler salinization [22]. The water absorption is mainly influenced by the polarity of monomer like methacrylate, TEGDMA and UDMA with polar functional groups like carboxyl, phosphate, hydroxyl in Bis-GMA and HEMA [23]. The water absorption results in disruption of interchain hydrogen bonding of polymer matrix [24]. The altered molecular change results in the altered refractory index. The dentin bonding surface adjoining the margins of the restorations are exposed to saliva and other oral fluids. The dentin fluid within the dentinal tubules is also likely to interact with the adhesive agents. Both water absorption and the degradation of the adhesive lead to the discolouration of the bonded area. The total etch and wash procedure removes all inorganic content in dentin matrix and produces the lower surface tension at the dentin surface. Hence, it may lead to more water absorption during the ageing process.

Self-etch adhesives have become popular among the dentist due to quick and easy manipulation steps. The self-etch adhesives rely upon both on micro-mechanical interlocking and chemical bonding to the functional monomers [25]. The ∆E value for the self-etching adhesive was comparatively less than etch-wash dual cure resin cement. The self-etch adhesive has pH values of 1-2. Though, smear layer is not completely removed, the low pH results in deep demineralisation. The higher constituent hydrophilic monomers, lower carbon-carbon double bond conversion in self-etch adhesive contributes to the higher water absorption, greater degradation process [26] and consequent colour change.

The benefit of self-adhesive cements is quick and simple application procedure, as it requires no preconditioning of the tooth structure. This dentin bonding technique showed the least mean colour change (∆E). The acid functionalised monomers are utilised to accomplish the demineralisation and bonding to the tooth structure [27]. Predominantly used acid functionalised monomer in the cement is methacrylate monomer with carboxylic or phosphoric acid groups. The pH of acidic monomer is balanced to avoid excessive hydrophilicity in final polymer and to achieve sufficient self –etching character. The polymerisation reaction lead to increased hydrophobicity as the acid functionality is neutralised with calcium in tooth structure and other metal oxides in filler [28]. The swelling and discolouration of the polymer are less due to the hydrophobic nature of the final set polymer.

Many researchers are of the opinion that the colour changes (∆E) less than 3.5 is imperceptible and clinically acceptable [29,30]. The mean colour changes observed among all the groups in the present study were lower than 3.5. The limitations of the study include the small sample size, and the exact simulation of ageing in oral cavity is difficult to replicate. The effect of other external factors like dietary colours, plaque accumulation, and exposure to the various pH environments was not considered. Further studies are required to analyse the colour changes using different shades of luting cements, and effect of different levels pH fluids on mean colour changes.

Within the limitation of the present in vitro study, following conclusions were drawn: