Influence of Hydroxyapatite Nanospheres in Dentin Adhesive on the Dentin Bond Integrity and Degree of Conversion: A Scanning Electron Microscopy (SEM), Raman, Fourier Transform-Infrared (FTIR), and Microtensile Study

An experimental adhesive incorporated with different nano-hydroxyapatite (n-HA) particle concentrations was synthesized and analyzed for dentin interaction, micro-tensile bond strength (μTBS), and degree of conversion (DC). n-HA powder (5 wt % and 10 wt %) were added in adhesive to yield three groups; gp-1: control experimental adhesive (CEA, 0 wt % HA), gp-2: 5 wt % n-HA (HAA-5%), and gp-3: 10 wt % n-HA (HAA-10%). The morphology of n-HA spheres was evaluated using Scanning Electron Microscopy (SEM). Their interaction in the adhesives was identified with SEM, Energy-Dispersive X-ray (EDX), and Micro-Raman spectroscopy. Teeth were sectioned, divided in study groups, and assessed for μTBS and failure mode. Employing Fourier Transform-Infrared (FTIR) spectroscopy, the DC of the adhesives was assessed. EDX mapping revealed the occurrence of oxygen, calcium, and phosphorus in the HAA-5% and HAA-10% groups. HAA-5% had the greatest μTBS values followed by HAA-10%. The presence of apatite was shown by FTIR spectra and Micro-Raman demonstrated phosphate and carbonate groups for n-HA spheres. The highest DC was observed for the CEA group followed by HAA-5%. n-HA spheres exhibited dentin interaction and formed a hybrid layer with resin tags. HAA-5% demonstrated superior μTBS compared with HAA-10% and control adhesive. The DC for HAA-5% was comparable to control adhesive.


Introduction
Dentin is a calcified tissue that forms the majority of the tooth structure; it is comprised of microscopic channels called dentinal tubules and an intra-tubular matrix [1,2]. Dentin hypersensitivity (DH) results from the exposure of these tubules due to erosion, abrasion, attrition, abfraction, and dental caries [3][4][5]. To restore dental caries, multiple restorative materials are available; however, dental

Synthesis of Experimental and HA Resin Adhesives
Employing the methods from the study by Ye et al., the experimental adhesive was synthesized [28]. A combination of monomers including bisphenol A glycol dimethacrylate (BisGMA), triethylene glycol dimethacrylate (TEGDMA), 2-hydroxyethyl methacrylate (HEMA), and ethyl 4-dimethylamino benzoate and camphorquinone (Esstech Inc., Essington, PA, USA) were used. The formulation included a mixture of 50% Bis-GMA, 25% TEGDMA, and 25% HEMA (60%) by weight with ethanol as a solvent (30%-m/m). First, 0.5% (n/n) ethyl 4-dimethylamino benzoate and 0.5% camphorquinone photo-initiators were incorporated according to the monomer moles. In addition, as an electron initiator, 1.0% (n/n) diphenyliodonium hexafluorophosphate (DPIHP) was also added to the adhesive mixture. The mixture was prepared in a three-necked flask with a magnetic stirrer and condenser (SA300; Sansyo, Tokyo, Japan). To prevent photo-polymerization, the prepared adhesive was kept in an isolated dark chamber covered with foil. This adhesive was regarded as control experimental adhesive (CEA).
Nano-spheres of HA (hydroxyapatitie, Sigma Aldrich, St. Louis, MO, USA) were obtained and silanized. In order to enhance the adhesion of n-HA to the adhesive, 5% silane was added to the solvent of 95% acetone for silanization. Silanized n-HA spheres were added in 10% and 20% concentrations (m/m) to the control adhesive to produce HA adhesive 5 wt % (HAA-5%) and HAA 10 wt % (HAA-10%). To achieve the homogenization of n-HA spheres in the adhesive components, the dispersion of HA using sonication in a centrifuge was performed. The formulated adhesives were stored at 37 • C for 24 h for allowing solvent evaporation. The adhesives were stored at 4 • C and were to be used in 20 days due to having a limited shelf life.

Characterization of HA and Synthesized Adhesive
The morphology of n-HA spheres was evaluated using Scanning Electron Microscopy (SEM). The presence of n-HA spheres within the experimental adhesives was identified with SEM, Energy-Dispersive X-ray spectroscopy (EDX), and Micro-Raman spectroscopy. EDX identified the elements and chemical groups within the adhesives. Prior to assessment, the HA adhesive was placed on a glass slide and photo-polymerized using a light-curing unit (Eliphar S10; 600-800 mW·cm −2 output; 3M ESPE, St. Paul, MN, USA) for 20 s. Polymerized specimens were mounted on aluminum stubs and sputter-coated with gold for 100 s (Baltec sputter, Scotia, NY, USA). SEM evaluations were performed at different magnifications (FEI Quanta 250, Scanning Electron Microscope, Hillsboro, OR, USA), at 10 kV accelerating voltage.

Tooth Preparation and Bonding Procedure
Ninety extracted teeth were cleansed using an ultrasonic scaler (Superior Instruments Co, New York, NY, USA) and were treated with chloramine trihydrate solution (Merck, Darmstadt, Germany) for 48 h. Teeth were mounted in orthodontic resin (Opti-Cryl, South Carolina, Columbia) at the cement-enamel junction (CEJ) level within 15 mm (height) sections of polyvinyl pipes (4 mm) and stored in distilled water. Tooth dentin was exposed by removing occlusal enamel with a slow speed diamond saw (Buehler Isomet 2000 Precision saw, Lake Bluff, IL, USA). Dentin was conditioned with 35% phosphoric acid (Ultra etch Econo Kit-Optident-Yorkshire, UK) for 10 s followed by washing and drying with cotton pellets. Thirty teeth each were treated with CEA (Gp:1, control), HAA-5% (Gp:2), and HAA-10% (Gp:3) adhesives. Adhesives were dispensed on a paper pad and using a micro-brush smeared on a dentin surface with agitation for 10 s followed by air thinning for 3 s. Following another similar adhesive application, polymerization was done with a light-curing device (Eliphar S10; 600 mW·cm −2 output; 3M ESPE, St. Paul, MN, USA) for 20 sec at 10 mm distance. For each specimen, a resin composite build-up was performed on the bonded adhesive using a resin mold and metal condenser in 2 mm increments (Filtek Universal; 3M ESPE, St. Paul, MN, USA).
The composite build-up was light cured for 20 s each side after the removal of excess. All bonded specimens in each adhesive group were stored in distilled water for 24 h at 37 • C. Twenty specimens in each adhesive group (CEA, HAA-5%, and HAA-10%) were used for the µTBS testing. Five (out of ten remaining) specimens in each group were assessed for bond integrity of the resin-dentin interface using SEM, and the last five bonded specimens in each adhesive group were utilized for Micro-Raman spectroscopy analysis.

Micro-Tensile Bond Strength (µTBS)
Prior to specimen sectioning, 10 bonded specimens in each adhesive group (CEA, HAA-5%, and HAA-10%) were thermo cycled (TC) in water baths at 5 • C and 55 • C for 30 s each and 5 s dwelling time (THE-1100, SD Mechatronik GmbH, Germany). However, the other 10 specimens remained non-thermocycled (NTC) in each adhesive group and were stored in distilled water for 1 week. The sectioning of all bonded specimens in the study was performed with a slow speed diamond saw (Buehler Isomet 2000 Precision saw, Lake Bluff, IL, USA) to form 1 mm × 1 mm, composite-dentin bonded beams. Six beams from each tooth were used for µTBS assessment (sixty) in each adhesive subgroup (CEA-TC, CEA-NTC, HAA-5%-TC; HAA-5%-NTC, HAA-10%-TC, HAA-10%-NTC). Bonded beams were secured to the micro-tensile tester (Bisco Inc., Richmond, VA, USA) jaws with cyanoacrylate (Superglue, Minneapolis, MN, USA) and were assessed under tension at a crosshead speed of 0.5 mm/minute until failure. Failure modes were evaluated among the groups, classified as adhesive, cohesive, and mixed types using a digital microscope (Hirox KH 7700, Tokyo, Japan). Means and standard deviations of µTBS among the groups were analyzed using ANOVA and a post hoc multiple comparisons test.

Micro-Raman Spectroscopy of Dentin-Adhesive Bonded Specimen
Five bonded specimens in each adhesive group were assessed using Micro-Raman. Using a Raman spectrometer (ProRaman-L Analyzer; TSI, Shoreview, MN, USA) and Raman microscopic software (Ramanreader), Raman spectra were identified. The specimen positioning was adjusted to target dentin, n-HA spheres, and the interface, and correction for dark counts was done to obtain adequate signals. By using a laser beam with 0.9 objective lens and 600 mW power, the interface was observed. Specimens were divided into zones of interest (1 mm sections), and 3 scans of 60 s were performed for each zone. The details of the spectra were obtained between the range of 800 cm −1 and 2000 cm −1 with noise filtration.

SEM Observation of Bonded Adhesive/Dentin Interface
Bonded beam specimens of CEA, HAA-5%, and HAA-10% (n = 5) were polished (Beuhler Polisher, Lake Bluff, IL, USA) and cleaned in an ultrasonic bath (Bandelin Digital-Sigma-Aldrich Darmstadt, Berlin, Germany) for 4 min. Bonded beams were further conditioned at the interface with 35% phosphoric acid gel (Ultra etch Econo Kit-Optident-Yorkshire, UK) for 10 s and washed with distilled water (15 s). Specimens were immersed in 5.25% of sodium hypochlorite (NaOCl) solution (5 min) and washed. An ethanol treatment at 80%, 90%, and 100% concentration was performed to desiccate the specimens. The specimens were mounted on aluminum stubs and sputter-coated with gold under vacuum (Baltec SCD sputter, Scotia, NY, USA) and visualized at multiple magnifications under SEM (FEI Quanta 250, Scanning Electron Microscope, Hillsboro, OR, USA) at 10 kV voltage.

Degree of Conversion (DC)
Employing transmission Fourier Transform-Infrared Spectroscopy (FTIR) the degree of conversion (DC) of the adhesives was assessed. Adhesives including, CEA, HAA-5%, and HAA-10% were evaluated before and after polymerization. Homogenized adhesive was smeared on the potassium bromide disc of the spectroscope (Shimadzu, Kyoto, Japan). The absorbance peaks for the C-C double bonds were recorded for the uncured resin, as the adhesives remained in contact with the FTIR sensors (Thermo Scientific Nicolet iS20 FTIR spectrometer, Waltham, MA, USA). The adhesive resins were polymerized using a dental light curing unit for 40 s, and the FTIR peaks were recorded again. Using a standardized baseline method [29], the C-C (aromatic) reference peaks (1607 cm −1 ) and C=C (aliphatic) absorbance peak (1638 cm −1 ) were obtained. The FTIR spectra were observed between 400 and 4000 cm −1 for determining the DC. The conversion rate for adhesives was identified by calculating the ratios of (C=C and C-C) absorbance intensities (percentage of unreacted double bonds) prior to and post photo-polymerization using the below equation [30].
where C aliphatic is 1638 cm −1 absorption peak of polymerized resin, C aromatic is 1607 cm −1 absorption peak of polymerized resin, U aliphatic is 1638 cm −1 absorption peak of unpolymerized resin and U aromatic is 1607 cm −1 absorption peak of unpolymerized resin.

SEM/EDX Outcomes
On SEM analysis, the n-HA powder demonstrated an average powder size of <100 nm ( Figure 1). With regard to the morphology, the n-HA powder particles were spherical in shape without any coarse edges ( Figure 1). The EDX mapping analysis revealed the presence of oxygen, calcium, and phosphorus in the 5 wt % (Gp: 2) and 10 wt % (Gp: 3) HA samples ( Figure 2). It was also observed that upon the mixing of n-HA powder in the experimental adhesive, a uniform distribution showing a homogenous dispersion of HA particles in the adhesive for HAA-5% ( Figure 3A) and HAA-10% with agglomerates ( Figure 3B) was observed. The Field Emission (FE)-SEM micrograph demonstrated the formation of a hybrid layer, and an admirable integration was observed between the dentin and HAA-5% ( Figure 4A). Upon further magnification, widely distributed n-HA particles were seen in the newly formed hybrid layer as well ( Figure 4B).
Polymers 2020, 12, x FOR PEER REVIEW 5 of 16 (aliphatic) absorbance peak (1638 cm −1 ) were obtained. The FTIR spectra were observed between 400 and 4000 cm −1 for determining the DC. The conversion rate for adhesives was identified by calculating the ratios of (C=C and C-C) absorbance intensities (percentage of unreacted double bonds) prior to and post photo-polymerization using the below equation [30].

SEM/EDX Outcomes
On SEM analysis, the n-HA powder demonstrated an average powder size of <100 nm ( Figure 1). With regard to the morphology, the n-HA powder particles were spherical in shape without any coarse edges (Figure 1). The EDX mapping analysis revealed the presence of oxygen, calcium, and phosphorus in the 5 wt % (Gp: 2) and 10 wt % (Gp: 3) HA samples ( Figure 2). It was also observed that upon the mixing of n-HA powder in the experimental adhesive, a uniform distribution showing a homogenous dispersion of HA particles in the adhesive for HAA-5% ( Figure 3a) and HAA-10% with agglomerates ( Figure 3b) was observed. The Field Emission (FE)-SEM micrograph demonstrated the formation of a hybrid layer, and an admirable integration was observed between the dentin and HAA-5% ( Figure 4A). Upon further magnification, widely distributed n-HA particles were seen in the newly formed hybrid layer as well ( Figure 4B).

Microtensile Bond Strength (µTBS)
The µTBS MPa values (Mean ± SD) observed for the samples (both TC and NTC) of three groups in this study along with their respective failure modes are shown in Table 1. It can be observed that for the NTC samples, the HAA-5% group had the greatest mean µTBS values (32.50 ± 3.47) followed by CEA (27.36 ± 2.80) and HAA-10% (24.41 ± 3.72) groups. For the TC samples, the same pattern of µTBS values as observed with NTC samples was detected, and HAA-5% again had the greatest values (25.31 ± 3.84). On an intergroup comparison for NTC samples, the differences observed for CEA (27.36 ± 2.80) compared with HAA-5% (32.50 ± 3.47) and HAA-5% (32.50 ± 3.47) compared with HAA-10% (24.41 ± 3.72) were statistically significant (p < 0.01). For the intergroup comparison of TC samples, only the differences observed between CEA (21.71 ± 3.85) and HAA-10% (17.60 ± 4.36), and HAA-5% (25.31 ± 3.84) matched with HAA-10% (17.60 ± 4.36) were statistically significant (p < 0.01). On intragroup comparison for TC and NTC samples, differences between all groups were statistically significant (p < 0.01). Overall, if the µTBS values of TC and NTC specimens are compared, the NTC samples revealed higher mean values than the TC samples. There was no clear distribution pattern observed for the failure modes observed for the three groups; however, the majority of the failures observed were of adhesive type. Adhesive failure depicts a failure in adhesion with fractures, which are not observed in dentin or the resin [31]. In our study, although adhesive type failures were the most common, among them, the smallest percentage of adhesive-type failures was observed in NTC-HAA-5% followed by NTC-HAA-10%.

FTIR Analysis
The representative FTIR spectra of n-HA particles, CEA, HAA-5% (cured and uncured), and HAA-10% (cured and uncured) are shown in Figure 5. The characteristic peaks at 563 and 600 cm −1 show the evidence of the presence of apatite in the HAA. The apatitic phosphate groups usually demonstrate a characteristic split band at 560 and 600 cm −1 , as observed for HAA in our study [32]. In addition, the presence of apatite could also be confirmed by the presence of a sharp single phosphate band [33], observed at 1040 cm −1 for HAA in our study.

Micro-Raman Spectroscopy Outcomes
The representative Micro-Raman spectra of n-HA particles, CEA, and HAA-5% is shown in Figure 6A-C, respectively. The phosphate and carbonate groups could be identified in Figure 6A at 960 cm −1 and 1070 cm −1 correspondingly. For CEA, the presence of phosphate and carbonate groups is not clear ( Figure 6B). The Micro-Raman spectra of HAA-5% group again demonstrates the presence of the phosphate group at 960 cm −1 and carbonate group at 1070 cm −1 ( Figure 6C).

Micro-Raman Spectroscopy Outcomes
The representative Micro-Raman spectra of n-HA particles, CEA, and HAA-5% is shown in Figure 6A-C, respectively. The phosphate and carbonate groups could be identified in Figure 6A at 960 cm −1 and 1070 cm −1 correspondingly. For CEA, the presence of phosphate and carbonate groups is not clear ( Figure 6B). The Micro-Raman spectra of HAA-5% group again demonstrates the presence of the phosphate group at 960 cm −1 and carbonate group at 1070 cm −1 ( Figure 6C).

Discussion
The present study investigated the effect of inclusion of n-HA spheres and its different content levels (5 wt % and 10 wt %) in an experimental resin adhesive on its bond strength and DC. Based on the outcomes, 5 wt % n-HA showed higher µTBS and comparable DC to controls. However, µTBS and DC of 10 wt % n-HA adhesive (HAA-10%) was significantly lower than control and 5 wt % n-HA (HAA-5%) adhesive. Therefore, the hypothesis was partly rejected. A multitude of explanations including HA particle size, opacity, chemical composition, and adhesive viscosity can be attributed to the current observations.
Dental adhesives are critical in the clinical success of restorations, as a compromised res-adhesive bond is associated with microleakage, DH, and secondary caries development [9]. The use of nanoparticles as reinforcing filler improves the mechanical, physical, and working properties of dental composites, as opposed to macro or micro-sized fillers [34]. The researchers have incorporated various therapeutic ions including silver nanoparticles (AgNPs) [35], zirconium oxide-titanium dioxide (ZrO 2 -TiO 2 ) nanoparticles [36], copper-doped bioactive glass nanoparticles (CuBGn) [37], and n-HA to improve the properties of adhesives [38]. The n-HA offers superior biocompatibility to act as a source of calcium and phosphate, in addition to possessing a strong ability to bond with tooth proteins [38,39]. It is also evidenced that n-HA particles are a better source of free calcium ions, which are required for remineralization [39]. These benefits of n-HA encouraged its incorporation to the experimental dental adhesive and its further influence on resin-dentin bond integrity, DC, and µTBS.
The µTBS test was employed in the present study, as it shows better discriminative competence [40], precise estimation of bond strength, and standardized relatable measurements [41]. It is suggested that the lower n-HA particle size (<100 nm) with spherical morphology, in comparison to micro HA, increases the exposed surface area for adhesion and mineral release [42]. In addition, a lower n-HA particle size also ensures adequate calcium and phosphate ion release (essential for remineralization) while retaining excellent mechanical properties [43], although this particular aspect was not investigated in our study. In the present study, <100 nm n-HA particles enhanced the µTBS of HAA-5% group as compared with the CEA group (without n-HA particles). Another probable reason for the improved µTBS for HAA-5% is its ability to biomineralize with the collagen network of tooth tissue [44], resulting in enhanced remineralization and bond integrity. However, it should be noted that the µTBS of HAA-10% was lowest among all groups, suggesting a compromise in dentin interaction and penetration, hybrid layer formation, mechanical properties, and dentin wetting. A study by Wagner et al. reported a decrease in adhesive bond strength with an increase in filler wt % of >10%, which was due to an increase in the viscosity of the material [45]. In the present study, agglomerate formation in HAA-10% specimens was detected in SEM micrographs; such clusters can lead to defects in the bonding complex concentrating applied stresses, decreasing bond strength [34]. The µTBS outcomes for HAA-5% are in agreement with the findings of Leitune et al., as they demonstrated an improvement in bond strength with the inclusion of n-HA in the experimental adhesive resin in comparison to control group [21]. In order to mimic the oral cavity's dynamic conditions, bonded dentin-adhesive beams were thermocycled to depict bond durability with temperature changes encountered in an in vivo environment. Earlier, Helvatjoglu-Antoniades et al. demonstrated that thermocycling could reduce the bond strength of dentin adhesives [46]. Their findings are in conformity with the present study, as all adhesive specimens (CEA, HAA-5%, and HAA-10%) showed a decrease in µTBS due to TC [46].
In the present study, the uniform mixing and dispersion of n-HA particles within the experimental adhesive, as shown by the SEM images, is critical to positively alter the properties of dentin adhesive. If the particles are not properly diffused in the adhesive, it could lead toward observation of gaps or cracks, hindering the optimal performance of the adhesive [47]. It should also be noted that numerous resin tags with varying depths were observed on SEM micrographs with a similar distribution for CEA and HAA specimens, indicating an optimal removal of smear layer and adequate adhesive penetration at the dentin interface. This is usually the case when the performed etching is sufficient to remove the smear layer [47]. However, it should be noted that the resin tag depth in the tubules does not necessarily impact the bond strength and intactness of the adhesive, as reported by Anchieta et al. [48].
Micro-Raman spectroscopy is considered to be a vibrational spectroscopic diagnostic technique used to study the polarization of molecules [49]. We employed Micro-Raman spectroscopy to study n-HA as a bonding component in the hybrid layer, as this technique involves scattering instead of absorption; therefore, specimen of any thickness can be visualized without any destruction [50]. In our study, n-HA demonstrated intense phosphate and carbonate peaks, whereas for the HAA-5% group, although these peaks were visible, their intensity was decreased. The plausible reason of this particular finding could be that there is a transitional change in the intensity of the Micro-Raman bands when the minerals are replaced with resin, as reported previously [46]. Our study results are in line with a previous study as the peak intensity for phosphate and carbonate groups decreased for HAA-5% group [51].
The representative FTIR spectra in our study demonstrated a characteristic split band and sharp phosphate band for n-HA particles. Then, the same particles were added to the experimental adhesive to achieve HAA-5% and HAA-10% adhesive. There was a linear inverse relationship between the addition of n-HA particles and DC observed in the study. The DC of HAA-5% and HAA-10% group decreased compared with CEA group, and this finding is contradictory to Zhang and Wang, who reported that the DC increases when HA particles are added to the experimental adhesive [52]. The possible reason for this finding could be attributed to the fact that they used low HA concentration (1-4 wt %) in their study [52], in contrast to the 5-10 wt % used in the present study. The n-HA particles are opaque in nature to the visible light and can affect polymerization adversely [53]. Therefore, a high n-HA concentration in experimental adhesive could provide a less favorable medium for the light penetration and polymerization of the monomers to polymers, hence, leading toward less DC. Therefore, it is suggested that n-HA should be added to the adhesive with a diluent that could regulate its increased viscosity, leading toward more DC. Another reason could be that the addition of a higher amount of HA in the adhesive causes higher viscosity, leading toward a resultant low DC [21], although in the present study, the viscosity of the adhesives was not investigated.
In the present study 10 wt % HAA incorporation in experimental adhesive showed an agglomeration of particles with lower µTBS and DC, and these findings could critically compromise the dentin adhesion overtime. Treatments are available to improve the interface between the nanoparticles and the surrounding polymer to improve filer dispersion, including the addition of reactive groups, silane modifications, and polymer grafting Moreover, the addition of diverse ratios of n-HA in the adhesive could assist in establishing precise µTBS values in the dentin [27]. Therefore, further studies investigating the influence of incorporating surface-modified HA nano-spheres at higher concentrations and diverse ratios on mechanical properties, adhesive bonding, and the degree of conversion are recommended. In addition, the leaching of HA particles with possible cytotoxic effects such as free radical generation is a possible concern. As the biological interaction of HA nanoparticles with cell types and tissues were not assessed in the present study, future studies on the organic response of surface-modified HA particles would be revealing.

Conclusions
Within the limitations of the study, incorporated n-HA spheres uniformly dispersed in the experimental adhesive to exhibit dentin interaction and the formation of a hybrid layer with resin tags at varying dentin depths. HAA-5% demonstrated superior µTBS as compared with HAA-10% and control adhesive. The degree of conversion for HAA-5% was comparable to the control adhesive; however, HAA-10% was the lowest. Therefore, the inclusion of 5 wt % n-HA could improve the adhesive's mechanical properties as it proved to be promising filler in our study and could have practical applications in clinical dentistry.