Microstructure and color stability of calcium silicate-based dental materials exposed to blood or platelet-rich fibrin

To investigate the effects of blood and platelet-rich fibrin (PRF), commonly used scaffolds in regenerative endodontic treatment (RET), on the hydration, microstructure, and color stability of three hydraulic calcium silicate cements (HCSCs), OrthoMTA, RetroMTA, and TotalFill-BC-RRM. The HCSCs were prepared and placed into polyethylene molds and transferred to Eppendorf tubes containing PRF, blood, or PBS and then incubated for 1 week or 1 month. The microstructure and hydration of the cements were studied by scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), and X-ray diffraction (XRD). The chromatic alteration of materials was also measured using a spectrophotometer. The data for color stability were analyzed using 2-way analysis of variance and Tukey post hoc tests. There was no significant difference between the color stability of cements exposed to PBS (p > 0.05). The chromatic alteration of cements exposed to blood was significantly greater than those exposed to PRF and PBS (p < 0.001). In the presence of blood and PRF, the color change of OrthoMTA was significantly greater than that of RetroMTA and TotalFill (p < 0.05), with no significant difference between RetroMTA and TotalFill (p > 0.05). XRD analysis of all cements revealed a calcium hydroxide peak after 1-week and 1-month exposure to the media; however, OrthoMTA and TotalFill exposed to blood and PRF for 1 month showed weaker calcium hydroxide peaks. SEM images revealed cements exposed to PBS had a different surface microstructure compared to those exposed to blood and PRF. Furthermore, the surface microstructure of HCSCs was influenced by the type of cement radiopacifier (bismuth oxide or zirconium oxide). EDS analysis of the elemental composition in all groups displayed peaks of Ca, O, C, Si, P, and Al. Color stability, hydration behavior, and microstructure of HCSCs were affected by exposure to PRF and blood and the type of cement radiopacifier. As some important physicochemical properties of HCSCs could be influenced by the environmental conditions and the type of radiopacifier, alternatives to blood clot and HCSCs containing substitutes for bismuth oxide might be more suitable in RETs.


Introduction
Management of immature teeth with necrotic pulps is one of the most challenging endodontic treatment modalities [1]. Chemo-mechanical root canal disinfection and subsequent filling of the root canal using conventional techniques is difficult and has led to the introduction of regenerative endodontic treatments (RET) [2]. The main goal of RET is to resolve periradicular disease, while ensuring continuing thickening of the dentine making up the root canal walls, increasing the length of the root, and allowing the root apex to continue developing. RET is based on recruiting stem cells into the bacteria-free root canal to populate in a resorbable scaffold inside the root canal system [3]. This is then followed by protecting the scaffold through the placement of a coronal barrier, normally a hydraulic calcium silicate cement (HCSC), followed by a conventional coronal restoration [4].
A blood clot, created by provoking bleeding from the periapical tissues into the root canal system, has been used as a biological scaffold [5]. Over time, the advantages of platelet derivatives such as platelet-rich plasma (PRP) and platelet-rich fibrin (PRF), as an important source of growth factors that have been used to enhance the regeneration of various tissue defects, including the dentine-pulp complex, have been reported in RET [6]. PRF, as an easy-to-use autologous scaffold [7,8] with a strong three-dimensional fibrin matrix, contains a high concentration of growth factors that are gradually released [9] that encourages the migration, proliferation, and differentiation of stem cells [7,10]. PRF has several advantages compared to blood and PRP [11], and clinical studies [11,12] have used PRF as a scaffold in RET and achieved successful results.
Sealing and protecting the scaffold with an appropriate dental material is crucial to create a suitable environment for regeneration. Mineral trioxide aggregate (MTA), an HCSC, has been used as a coronal barrier in most of the studies in the field of RET [13]. ProRoot MTA (Dentsply, Tulsa Dental Specialties, Tulsa, OK, USA) is the first commercial representative of the endodontic HCSCs. Although the introduction of MTA has revolutionized the field of endodontics, it suffers from a number of disadvantages such as handling difficulties, long setting time, and the potential for tooth discoloration [14]. Therefore, other HCSCs were then introduced to overcome these shortcomings. The radiopacifier bismuth oxide found in the formulation of some HCSCs has been associated with tooth discoloration [15]. OrthoMTA (BioMTA, Seoul, Korea) is a HCSC delivered as powder composed of tricalcium silicate, dicalcium silicate, tricalcium aluminate, tetracalcium aluminoferrite, free calcium oxide, and bismuth oxide. According to the manufacturer, the powder of RetroMTA (BioMTA, Seoul, Korea) consists of calcium carbonate, silicon oxide, aluminum oxide, and hydraulic calcium zirconia complex [16]. TotalFill BC RRM putty (FKG Dentaire, La Chaux-de-Fonds, Switzerland) is a premixed HCSC, and according to the manufacturer, it consists of calcium silicates, zirconium oxide, tantalum peroxide, calcium phosphate monobasic, and filler agents that set in the presence of moisture.
As HCSCs are in close contact with scaffolds, the interaction between them might affect the physical properties of the cement [17,18]. Indeed, the adverse effects of blood contamination on the physical and biological properties of HCSCs, such as bioactivity, hydration, and setting time as well as tooth discoloration, have been investigated [17][18][19]. Several studies have reported the effect of PRF, as a beneficial natural scaffold, on color stability [20] and bioactivity potential [18] of some HCSCs. However, there is insufficient information on the effect of platelet-rich fibrin on the physical properties and color stability of HCSCs containing various radiopacifiers. This study was designed to evaluate the effects of blood and PRF compared with those of PBS on the microstructure and color stability of HCSCs containing bismuth oxide (OrthoMTA) to those that do not contain bismuth oxide (RetroMTA and TotalFill BC RRM). The null hypothesis was that would be no difference regarding the microstructure and color stability of OrthoMTA, RetroMTA, and TotalFill exposed to PBS, PRF, or blood.
OrthoMTA and RetroMTA cements were prepared according to the manufacturer's instructions. TotalFill was available as a ready-to-use putty and did not require prior preparation. The materials were placed into polyethylene cylindrical molds (3 mm diameter and 3 mm height) with minimal pressure. The polyethylene cylinders containing the material were then transferred to 0.2-mL Eppendorf tubes that were filled with either blood, PRF, or PBS so that the lower surfaces of the test materials were just in contact with blood, PRF, or PBS. The Eppendorf tubes were then incubated at 37 °C and fully saturated humidity for 1 week or 1 month.
The whole fresh human blood used in this study was collected from a healthy consented volunteer. This study was approved by the Ethics Committee of Tehran University of Medical Sciences (Ethics code: IR.TUMS.DENTISTRY. REC.1398.027) and University of Jordan. To prepare PRF, fresh human blood was poured into a test tube without the addition of anticoagulants and immediately centrifuged (PRF Centrifuge DUO Quattro, Nice, France) at 1300 rpm for 8 min. To remove the exudates, the PRF clot was separated from the platelet-rich plasma and red blood cell layers and compressed in a compression box.
The surface microstructure and elemental composition of specimens were qualitatively assessed using scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopic (EDS), respectively. Also, the hydration process and phase composition of specimens were evaluated using X-ray diffraction analysis (XRD). Assessment of color stability was conducted by spectrophotometry.

SEM and EDS analysis
Specimens exposed to blood, PRF, or PBS for 1 week (n = 2 for each group) and 1 month (n = 2 for each group) were mounted on aluminum stubs, carbon-coated, and analyzed using a scanning electron microscope (Mira3 XMH SEM; TESCAN, Brno, Czech Republic) fitted with an energy-dispersive X-ray detector (SEM-EDS; TESCAN). The microstructure of surfaces exposed to blood, PRF, or PBS were examined with several magnifications (× 5000 to × 25,000) and element analysis was performed by EDS.

XRD analysis
For phase composition analysis of unhydrated cements (n = 2 for each group), 1-week (n = 2 for each group), and 1-month specimens (n = 2 for each group) exposed to blood, PRF, or PBS, the cements were removed from the polyethylene cylindrical molds, dried under a vacuum, and then crushed into a very fine powder. The powders were tested using a Philips X'Pert Pro diffractometer (PANalytical, Netherlands) which monochromatized Cu Kα radiation conditions (λ = 1.54 Å, operated on 35 mA and 40 kV current). The diffraction angles (2θ) were scanned from 15 to 65°. Phase identification was accomplished using Xpert HighScore software for XRD analysis. In addition to hydrated specimens, the phase composition of unhydrated OrthoMTA and Ret-roMTA powder and TotalFill putty were determined.

Assessment of color stability
The color of the lower surface of HCSCs exposed to blood, PRF, or PBS (n = 12 for each cement exposed to each medium) was measured by spectrophotometry (VITA Easyshade V; Zahnfabrik, Bad Säckingen, Germany). The color assessment was performed by the same investigator under steady laboratory conditions and the device was calibrated before use for each specimen. Color measurements were performed prior to exposure of HCSCs to blood, PRF, or PBS as the baseline color and 1 month after exposure.
The chromatic alteration (ΔE) between the initial and the second measurements was calculated using the measurement of the L*, a*, and b* values and the following formula: In this color measurement, L* indicates the value of lightness-darkness, a* indicates greenness-redness, and b* indicates blueness-yellowness.

Statistical analysis
Data were evaluated using SPSS software (PASW Statistics 18; SPSS Inc., Chicago, IL). Two-way ANOVA and Tukey post hoc tests were used to evaluate the effects of cements and media variables on chromatic alteration (ΔE). The level of significance was set at p < 0.05.

XRD analysis
The results of XRD analysis of three HCSCs are shown in Fig. 1.

OrthoMTA
XRD analysis of the unhydrated powder demonstrated the phases present, namely tetra-calcium aluminoferrite, tri-calcium aluminate, bismuth oxide, dicalcium silicate, and tricalcium silicate. One-week specimens exposed to PBS, PRF, and blood and 1-month specimens exposed to PBS displayed dicalcium aluminate and calcium hydroxide, in addition to components presented in the unhydrated powder. However, an obvious drop in peak intensity of calcium hydroxide in the specimens exposed to PRF and blood for 1 month was shown.

RetroMTA
XRD analysis of the unhydrated powder revealed peaks of calcium carbonate, calcium zirconium complex, zirconium oxide, dicalcium silicate, and tricalcium silicate. In addition to the ingredients of the unhydrated powder, calcium hydroxide was observed in the specimens exposed to PBS, PRF, and blood for 1 week and 1 month. Also, the specimens exposed to PBS for 1 month had carbonate apatite depositions.

TotalFill BC RRM
Analysis of the unhydrated putty revealed peaks of monobasic calcium phosphate, tantalum pentoxide, zirconium oxide, dicalcium silicate, and tricalcium silicate. In addition to the phases present in the unhydrated putty, calcium hydroxide was identified in all hydrated specimens after 1 week and 1 month, except the specimens exposed to PRF and blood for 1 month in which the intensity for calcium hydroxide peaks decreased.

SEM analysis
OrthoMTA SEM images revealed accumulations of coral-like particles after 1 week of exposure to PBS. One-month specimens exposed to PBS revealed hexagonal and globular particles within clusters of crystalline structures at higher magnifications. OrthoMTA specimens exposed to PRF or blood for 1 week and 1 month had no prominent crystalline structures (Fig. 2).

RetroMTA
After 1 week of exposure to PBS, accumulations of platelike crystals among coral-like shaped aggregates on the Fig. 1 XRD analysis of unhydrated OrthoMTA, RetroMTA, and TotalFill (black), exposed to PBS (green), PRF (blue), and blood (red) for 1 week and 1 month surface of RetroMTA were formed. Spherical aggregates composed of minute particles along the periphery within a multi-globular matrix were observed on the surface of 1-month specimens of RetroMTA. Fused globular aggregates were observed on the surface of the RetroMTA specimens exposed to PRF for 1 week. Smaller more compacted fused globular particles were demonstrated on 1-month RetroMTA samples exposed to PRF. After 1 week of exposure to blood, islands of unstructured surfaces mixed with small crystalline particles were observed.

TotalFill BC RRM
Bundles of string-like aggregates and aggregates of corallike particles were present after 1-week and 1-month exposure to PBS, respectively. The material had an unstructured surface after 1-week exposure to PRF. However, aggregates of single and multiple globular structures on the surface of specimens exposed to PRF for 1 month were found. Fig. 2 SEM images of OrthoMTA specimens exposed to PBS, PRF, and blood after 1 week and 1 month Specimens exposed to blood revealed small globular aggregates after 1 week; however, larger and more fused globular particles were seen on the surface of 1-month samples (Fig. 4).

EDS analysis
Analysis of the elemental composition of precipitates formed on the cement surfaces in all groups displayed high peaks of Ca, O, C (Fig. 5 a-c). In addition, a high peak of Si was observed on the surface of OrthoMTA specimens. Precipitates on the surface of OrthoMTA exhibited moderate peaks of Bi and Al and low peaks of P.
EDS analysis of RetroMTA specimens revealed moderate peaks of Si and low peaks of Zr, P, and Al, whereas precipitates on TotalFill BC RRM displayed moderate peaks of Si and Zr, and low peaks of Al, P, and Ta.

Color stability assessment
The mean values for the chromatic alteration in each subgroup are illustrated in Fig. 6. Analysis of the chromatic alteration in the various groups after 1-month exposure to PBS, PRF, and blood revealed a significant difference between the types of cement in the presence of different media (p = 0.016).

Chromatic alteration (ΔE) of samples after 1 month by type of media
In the present study, there was no significant difference between the chromatic alterations (ΔE) of the three types of cement exposed to PBS (p > 0.05). There was a significant difference between the chromatic alterations produced by the different cements exposed to blood and PRF. In the presence Fig. 4 SEM images of TotalFill BC RRM specimens exposed to PBS, PRF, and blood, after 1 week and 1 month of blood, the chromatic alteration of OrthoMTA specimens was significantly greater than RetroMTA and TotalFill specimens (p < 0.001) as well as in the presence of PRF (p < 0.005 and p < 0.003, respectively, for RetroMTA and TotalFill). No significant difference was found between RetroMTA and TotalFill (p > 0.05).

Chromatic alteration (ΔE) of samples after 1 month by cement
The chromatic alteration of all cements exposed to blood was significantly greater than those exposed to PRF and PBS (p < 0.001). No significant difference was found between cements exposed to PBS or PRF (p > 0.05).
Photos of a specimen from each group are shown in Fig. 7.

Discussion
The current study exposed cements for two periods of time, 7 days and 1 month, in order to assess the progress of the hydration process and microstructure of the materials exposed to various media. To simulate the intracanal coronal barrier used in RET, HCSCs were placed in polyethylene cylindrical molds and transferred to Eppendorf tubes so that their lower surface was in contact with fresh human whole blood or PRF, which is used as a natural scaffold in regenerative treatments.
Hydraulic calcium silicate-based cements need moisture to set, which they acquire from physiological tissue fluids and blood. Calcium hydroxide, as a hydration by-product of HCSCs, combines with phosphate in the environment and forms hydroxyapatite as the key element in inducing hard tissue formation [21]. PBS is a simulated tissue fluid containing phosphate that mimics clinical conditions in laboratory studies and was considered an ageing medium in the control group of the present study [18,20,[22][23][24].
The present results revealed that in contrast to PBS and PRF, blood contamination significantly increased chromatic alteration associated with all materials. This finding is in accordance with the results of studies that reported increased discoloration following blood contamination of HCSCs [18,22,25].
Fe 3+ , a dark brown by-product of natural oxidation and reduction of erythrocyte ferrous (Fe 2+ ), may cause chromatic alteration of materials [25]. Also, it has been demonstrated that diffusion of blood components into the porosities of partially hydrated cements may exacerbate chromatic alteration [26]. The partial absence of erythrocytes in platelet derivatives such as PRF might be the cause for less color change compared to whole human blood [20].
The chromatic alteration of OrthoMTA specimens exposed to blood and PRF was significantly more pronounced than TotalFill and RetroMTA. This finding may be linked to the content of tetracalcium aluminoferrite and bismuth oxide in OrthoMTA, which is not present in other test cements. Presence of metal components such as iron and bismuth oxide is one of the important factors affecting the color of endodontic cements and subsequently tooth crown discoloration. Moreover, relating to the mechanism of discoloration caused by bismuth oxide, the theory of bismuth oxidation has been proposed. This reaction results in the creation of unstable oxygen, the reaction of oxygen with carbon dioxide, and then the production of bismuth carbonate as a discoloration agent [27,28]. In this regard, some studies suggest that MTA chromatic alteration may be related to bismuth oxide, which has been added to both white and gray MTA as a radiopacifier [29]. While several studies have shown less tooth discoloration in the cement containing zirconium oxide and/or tantalum oxide than bismuth oxide containing cements [22,27], it should be noted that the final discoloration of the tooth can be due to the discoloration of the cement itself, the interaction of bismuth oxide in some types of cement with dentin, or the penetration of blood and its components into dentinal tubule. Therefore, discoloration caused by blood may mask the effects of cements with or without bismuth oxide in tooth discoloration and prevent an accurate study of the color stability of HCSCs.
In the current study, XRD analysis of the three cements revealed high peaks of calcium hydroxide, after 1-week exposure to PBS, PRF, and blood. However, the analysis of OrthoMTA and TotalFill exposed to blood and PRF after 1 month showed a drop in calcium hydroxide peaks. Like the present study, it has been reported that no peaks of calcium hydroxide formed on ERRM cements, which is similar to TotalFill exposed to blood, after 28 days, but contrary to cements in contact with water and HBSS [19]. Regarding the reduction of calcium hydroxide in the specimens of OrthoMTA and TotalFill exposed to PRF and blood for 1 month, it seems that the hydration process of calcium silicate cements arrested, and dissolution of calcium hydroxide commenced. Nekoofar et al. [17] found no peaks of calcium hydroxide in specimens mixed entirely with blood, which is considered to be due to the inhibition of calcium hydroxide formation and/or its dissolution. In addition, the formation of amorphous calcium silicate hydrate (CSH) following to the hydration of HCSC should be considered, because only crystalline compounds are traceable in the XRD patterns [17]. Also, the lack of this amorphous content as well as absence of ettringite, as hydration reaction indicators, could be due to the short time of incubation and incomplete hydration [17,30]. On the other hand, the effects of blood and PRF on the amounts of calcium hydroxide at different time intervals may be related to differences in the chemical composition of HCSCs. In assessing the hydration behavior of bismuth contained HCSCs, bismuth remains an unreacted powder in the hydrated composition of these cements, affecting the MTA hydration mechanism. This element enters the structure of hydrated calcium silicate and forms calcium silicate hydrate-bismuth (CSH-Bi), which can affect the formation and dissolution rate of calcium hydroxide and, consequently, the bioactivity of the hydrated material [31], thus preventing complete hydration [17,32]. It is worth to mention that during the initial stages of HCSCs hydration, the amounts of calcium hydroxide produced as a product of hydration increased followed by a drop in its production at later stages reflecting the slowing in hydration reaction [32]. Furthermore, the reduction of calcium hydroxide could be attributed to the carbonation in HCSCs that converts calcium hydroxide into calcium carbonate in the presence of atmospheric carbon dioxide [33]. However, in the present study, decrease in calcium hydroxide intensity was more evident in 1-m specimens of OrthoMTA and TotalFill exposed to blood and PRF which might be due to the effect of blood and PRF on hydration process of these cements. All cements tested in the study had a different surface microstructure after 1 week and 1 month of exposure to PBS, which may be due to differences in chemical composition. This finding might be related to the bioactivity of HCSCs and the deposition of apatite crystals in the presence of fluid containing phosphate [18,23], as seen in the present study where RetroMTA cement was exposed to PBS for 1 month. Indeed, over time, the size of the precipitated crystals and surface microstructure as well as the chemical composition of the PBS-exposed specimens changed.
Also, all cements exposed to PBS had different surface microstructures compared to cements exposed to blood and PRF, indicating that the cements were affected by blood and PRF. The crystalline microstructure was seen in all cements exposed to PBS while it was not observed in the cements exposed to blood and PRF. This result is in accordance with Nekoofar et al. [17] who revealed the unfavorable effects of blood contamination on the microstructure and hydration behavior of MTA. Furthermore, the surface microstructure of specimens exposed to PRF and blood for 1 month varied compared to the 1-week samples. This indicates changes in the hydration process, over time, which resulted in different surface microstructures [23,30].
EDS analysis of the material revealed that the surface precipitations in all groups mainly contained high peaks of carbon (C), oxygen (O), and calcium (Ca), which reflect the nature of HCSCs, as reported in similar studies [19,23,34]. It has also been reported that high peaks of calcium, oxygen, and silicon occur during EDS analysis of pure Portland cement exposed to PBS and PRF [18], which might be due to the release of hydrated calcium silicate and calcium hydroxide following hydration reaction of the HCSCs [35]. In the present study, the peaks of bismuth (Bi) were observed specifically in OrthoMTA specimens, while the peaks of zirconium (Zr) in RetroMTA and TotalFill and the peaks of tantalum (Ta) were detected only in TotalFill specimens, which can be justified by the radiopacifier content of these cement [19,23,34].
Aluminum (Al) was seen in all specimens, which can play a role in the formation of the ettringite crystalline microstructure [17]. Aluminum contributes to the formation of tetra-calcium aluminoferrite and tricalcium aluminate in the OrthoMTA cement [27], this may be considered a reason for greater amounts of Al in the EDX analysis. However, Camilleri stated that the presence of aluminum in HCSCs is generally rare [30].
In the present study, phosphorus was observed in the EDS analysis of all specimens, which could be due to the presence of phosphates in PBS, blood, and PRF. Previous studies have also shown the presence of phosphorus in blood and PRF [18,19,25]. Furthermore, TotalFill BC RRM contains phosphorus as monobasic calcium phosphate [19,23,34]. High peaks of oxygen can be attributed to the presence of oxygen (O) in the chemical composition of all calcium silicate materials [34]. The presence of carbon (C) may also be related to carbon dioxide and carbon in blood and its derivatives, and subsequently the presence of calcium carbonate deposits, which are formed by the reaction of calcium and carbonate ions in the environment [19]. In addition, carbon is related to use of the carbon grid before SEM-EDS analysis.

Conclusions
The surface microstructure of three studied cements exposed to PBS was different from the cements exposed to blood and PRF. The hydration behavior and microstructure of HCSCs were affected by exposure to PRF and blood.
The chromatic alteration of all cements exposed to blood was significantly greater than those exposed to PRF and PBS. RetroMTA and TotalFill specimens had less color change when exposed to blood and PRF compared to OrthoMTA.