1

Introduction

The interaction between restorative dental materials and tooth tissue encompasses multiple aspects of dental anatomy and materials science. Until relatively recently, many adhesive dental restorative materials were thought to have a passive hard tissue interaction based on simple infiltration with the enamel or dentin upon which they were placed. However, there is increasing interest in mapping the interactions between materials and tooth tissue, where the former has a more aggressive interaction with the latter, while promoting ‘bioactivity’; it can be argued that materials such as glass ionomer cements have had such interactions for over thirty years, but only recently have mechanisms of adhesion and adaptation been elucidated indicating the potential for chemical interactions with both resin and water-based cement-like materials . Bioactivity can be defined as ‘materials that elicit a specific biological response at the interface between tissues and the material, which results in the formation of a bond’. In this paper we will concentrate on two classes of water-based cement-type restorative materials, glass ionomer and calcium-silicate based cements, and examine their interactions with the hard tissues in health and disease, in particular examining their potential for bio-mineralization in dentin. These materials are promoted as dentin replacements, mimicking many of the physical properties of this composite biological material, but they do not, as yet, have the wear resistance and mechanical properties to make them suitable as long-term enamel replacements.

2

Glass ionomer cements

Glass ionomer cements (GICs) were first introduced to dentistry in 1975 and since then they have been used in a wide range of clinical applications. Conventional GICs are dispensed in a powder form supplied with its own liquid. The powder is formed of fluoro-aluminosilicate glass, while the liquid is an aqueous solution of a polyalkenoic acid, such as polyacrylic acid, although in later formulations, the acid may be added to the powder in a dried polymer form . Strontium has been added in some commercial GICs, such as GC Fuji IXGP (GC Corporation, Tokyo, Japan), to substitute calcium due to its radiopaque properties . This substitution does not have any effect on the setting products or cement’s remineralizing capability .

Upon mixing, an acid–base reaction takes place between the polyalkenoic acid and the ion-leachable glass particles , which occurs in two phases. The first phase is a dissolution phase, in which the acid attacks the surface of the glass particles to release ions such as aluminum, fluoride, and calcium or strontium. Following ion release, the polyacid molecules become ionized and adopt a more linear form. This renders the polyacid’s carboxylic groups more accessible for the ions and facilitates their cross linking in the later stage of gelation . Eventually, the set cement will be formed of a composite of un-reacted glass particle cores, encapsulated by a siliceous gel and embedded in a polyacid–salt matrix which binds the components together . Within the polyacid–salt matrix of set GICs, water is distributed in two forms; loose water, which can be removed through desiccation, and bound water which is chemically locked into the matrix . Water plays an essential role during the maturation of the cement as well as the diffusion of ions .

Further material developments have included a newly named material: ‘Glass Carbomer ^® ’ (GCP Dental, Mijlweg, Netherlands) which is claimed to contain nano glass particles, hydroxyapatite/fluorapatite (HAp/FAp) nano particles and liquid silica. The nanocrystals of calcium fluorapatite (FAp) may act as nuclei for the remineralization process and initiate the formation of FAp mineral as well as nanocrystals of hydroxyapatite (HAp). The glass has a much finer particle size compared to conventional GICs, giving properties that are thought to aid its dissolution and ultimate conversion to FAp and HAp. However, using “magic angle spinning” nuclear magnetic resonance spectroscopy, Zainuddin et al. have shown that the HAp in the powder is consumed during the cement formation process in this material and so may in fact have reduced availability for bio-mineralization.

When a freshly mixed GIC is placed on wet dentin, an interaction between the two materials takes place, in the form of an ion exchange . Aluminum, fluoride, and calcium or strontium leach out of the cement as the glass is being dissolved by the polyacid, while calcium and phosphate ions also move out of the underlying dentin as a result of the self-etching effect of the setting cement on mineralized dentin . This ion exchange process creates an intermediate layer composed of ions derived from both substrates . The release of fluoride and calcium/strontium ions has provided GICs with the potential for remineralization of carious tissues , where ion exchange could replenish the demineralized tissues’ ions, thus tipping the balance in favor of apatite (re-)formation.

2

Glass ionomer cements

Glass ionomer cements (GICs) were first introduced to dentistry in 1975 and since then they have been used in a wide range of clinical applications. Conventional GICs are dispensed in a powder form supplied with its own liquid. The powder is formed of fluoro-aluminosilicate glass, while the liquid is an aqueous solution of a polyalkenoic acid, such as polyacrylic acid, although in later formulations, the acid may be added to the powder in a dried polymer form . Strontium has been added in some commercial GICs, such as GC Fuji IXGP (GC Corporation, Tokyo, Japan), to substitute calcium due to its radiopaque properties . This substitution does not have any effect on the setting products or cement’s remineralizing capability .

Upon mixing, an acid–base reaction takes place between the polyalkenoic acid and the ion-leachable glass particles , which occurs in two phases. The first phase is a dissolution phase, in which the acid attacks the surface of the glass particles to release ions such as aluminum, fluoride, and calcium or strontium. Following ion release, the polyacid molecules become ionized and adopt a more linear form. This renders the polyacid’s carboxylic groups more accessible for the ions and facilitates their cross linking in the later stage of gelation . Eventually, the set cement will be formed of a composite of un-reacted glass particle cores, encapsulated by a siliceous gel and embedded in a polyacid–salt matrix which binds the components together . Within the polyacid–salt matrix of set GICs, water is distributed in two forms; loose water, which can be removed through desiccation, and bound water which is chemically locked into the matrix . Water plays an essential role during the maturation of the cement as well as the diffusion of ions .

Further material developments have included a newly named material: ‘Glass Carbomer ^® ’ (GCP Dental, Mijlweg, Netherlands) which is claimed to contain nano glass particles, hydroxyapatite/fluorapatite (HAp/FAp) nano particles and liquid silica. The nanocrystals of calcium fluorapatite (FAp) may act as nuclei for the remineralization process and initiate the formation of FAp mineral as well as nanocrystals of hydroxyapatite (HAp). The glass has a much finer particle size compared to conventional GICs, giving properties that are thought to aid its dissolution and ultimate conversion to FAp and HAp. However, using “magic angle spinning” nuclear magnetic resonance spectroscopy, Zainuddin et al. have shown that the HAp in the powder is consumed during the cement formation process in this material and so may in fact have reduced availability for bio-mineralization.

When a freshly mixed GIC is placed on wet dentin, an interaction between the two materials takes place, in the form of an ion exchange . Aluminum, fluoride, and calcium or strontium leach out of the cement as the glass is being dissolved by the polyacid, while calcium and phosphate ions also move out of the underlying dentin as a result of the self-etching effect of the setting cement on mineralized dentin . This ion exchange process creates an intermediate layer composed of ions derived from both substrates . The release of fluoride and calcium/strontium ions has provided GICs with the potential for remineralization of carious tissues , where ion exchange could replenish the demineralized tissues’ ions, thus tipping the balance in favor of apatite (re-)formation.

3

Calcium-silicate based cements

Calcium-silicate based cements were first introduced to dentistry in 1993 when Torabinejad developed a formula based on ordinary Portland cement (OPC) to produce the mineral trioxide aggregate, or the gray MTA . This material was principally composed of tri-calcium silicate, di-calcium silicate, tri-calcium aluminate, and tetra-calcium aluminoferrite, in addition to calcium sulphate and bismuth oxide, added as a radiopaquer for clinical applications . In 2002, a white MTA version was developed, which was identical to the gray form but lacked the tetra-calcium aluminoferrite and had reduced aluminate levels . Since their introduction, MTAs have been principally used for endodontic applications such as repairing perforated roots, apexification, or pulp capping due to their relatively long working and setting times.

In 2011, Biodentine™, a quick-setting calcium-silicate based dental cement, was introduced by Septodont (Saint Maur des Fosses – France). Henceforth this commercial name will be used for representation and brevity. Biodentine™ was developed as a dentin replacement material, a novel clinical application of this family of materials, intending it to function as a coronal restoration. The relatively short setting time (around 12 min), can enable the use of this cement for restorative procedures; impossible with MTAs that achieve an initial setting 3–4 h . Biodentine™ is principally composed of a highly purified tri-calcium silicate powder that is prepared synthetically in the lab de novo, rather than derived from a clinker product of cement manufacture. Additionally, Biodentine™ contains di-calcium silicate, calcium carbonate and zirconium dioxide as a radiopacifer. The di-calcium and tri-calcium silicate phases form around 70% of the weight of Biodentine’s de-hydrated powder, which is close to that of white MTA and white Portland cement . Unlike MTA, Biodentine does not contain calcium sulphate, aluminate, or alumino-ferrate. The powder is dispensed in a two part capsule to which is added an aliquot of hydration liquid, composed of water, calcium chloride, and a water reducing agent.

Despite similar constituents, there is significant variation in calcium-silicate dental cement manufacturing processes. This affects the purity of their constituents and hydration products, as well as their behavior . Studies have shown that cements such as ProRoot MTA (Dentsply, Tulsa Dental Products, Tulsa, OK) and MTA Angelus (Angelus Soluções Odontológicas, Londrina, Brazil) share almost the same composition as white OPC, except for the addition of bismuth oxide to the MTAs, for radiopacity . However, the MTAs also include impurities and contaminating heavy metals such as chromium, arsenic, and lead . This suggests their manufacture is similar to OPCs but less segregated and refined as the particle sizes also vary more widely . On the other hand, other calcium silicate based dental cements, such as Biodentine™ and MTA-bio (Angelus Indústria de Produtos Odontológicos Ltda, Londrina, Brazil) have been produced under more stringent production conditions from raw materials, in an attempt to avoid any potential contamination of the basic constituents, and to avoid the incorporation of aluminum oxide .

Similar to OPC, calcium-silicate based dental cements set through a hydration reaction . Although the chemical reactions taking place during the hydration are more complex, the conversion of the anhydrous phases into corresponding hydrates can be simplified as follows:

This setting reaction is a dissolution–precipitation process that involves a gradual dissolution of the un-hydrated calcium silicate phases (C ₃ S and C ₂ S) and formation of hydration products, mainly calcium silicate hydrate (CSH) and calcium hydroxide (CH). The CSH is a generic name for any amorphous calcium silicate hydrates which includes the particular type of CSH that results from the hydration of tri- and di-calcium silicate phases . The CSH precipitates as a colloidal material and grows on the surface of un-hydrated calcium silicate granules, forming a matrix that binds the other components together, gradually replacing the original granules. Meanwhile, calcium hydroxide is distributed throughout water filled spaces present between the hydrating cement’s components . Compared with OPC, the hydration of MTA is affected by the bismuth oxide, which forms 20% of its weight and acts as a radiopacifier . Hydrated MTA and tri-calcium silicate cements were found to release more calcium ions than hydrated OPC .

For Biodentine™, the setting reaction is expected to be similar to the hydration of pure tri-calcium silicate cement and is therefore expected to produce the same hydration products as WMTA and OPC, except for the absence of alumina and gypsum hydration products. The zirconium oxide present in the Biodentine™ powder as a radiopacifier also acts as an inert filler and is not involved in the setting reaction , unlike the bismuth oxide present in MTA . The hydration reaction starts with the fast dissolution of the tri-calcium silicate particles, which therefore can explain the fast setting of Biodentine™ , in addition to the presence of calcium chloride in the liquid, which is known to speed up the hydration reaction , and the absence of calcium sulphate that acts as a retarder.

Despite similar contents, the manufacturing variations and absolute constituent ratios influence clinical behavior within the materials group. Therefore, it is essential to study and characterize these cements individually, in order to understand their nature and clinical behavior. Such studies have been limited for Biodentine™ . Therefore, further characterization of Biodentine™ and its setting reaction is required. However, a number of studies have been conducted on the interaction of Biodentine™ with dental tissues and pulpal tissues .

The interfacial properties of Biodentine™ and a glass-ionomer cement (GIC Fuji IXGP) with dentin have been studied using confocal laser scanning microscopy (CLSM), scanning electron microscopy (SEM), micro-Raman spectroscopy, and two-photon auto-fluorescence and second harmonic-generation (SHG) imaging by Atmeh et al. . Their results indicated the formation of tag-like structures alongside an interfacial layer called the “mineral infiltration zone”, where the alkaline caustic effect of the calcium silicate cement’s hydration products degrades the collagenous component of the interfacial dentin. This degradation leads to the formation of a porous structure that facilitates the permeation of high concentrations of Ca ²⁺ , OH ⁻ , and CO ₃ ²⁻ ions, leading to increased mineralization in this region. Comparison of the dentin–restorative interfaces shows that there is a dentin-mineral infiltration with the Biodentine™, whereas polyacrylic and tartaric acids and their salts lead to the diffuse penetration of the GIC; consequently a new type of interfacial interaction, “the mineral infiltration zone”, is suggested for these calcium-silicate-based cements ( Fig. 1 a–d ).

(a) Reflection mode confocal image of the dentin–Biodentine™ interface showing the highly reflective band (bracket). The micro-permeability of the Rhodamine-B dye solution was reduced which was demonstrated by the fluorescence mode image of the same area (b). (c) The mineral infiltration zone (MIZ) is characteristic for the interfacial dentin underneath the Biodentine™ cement: tag-like structures forming within the dentinal tubules (tip of arrow) have interrupted the dye solution permeating from the pulp chamber (base of arrows). (d) The ion exchange layer in the dentin–GIC interface with the scalloped appearance in the inter-tubular dentin (arrows) as a result of the acidic nature of the setting cement demineralizing the intra-tubular dentin.

4

Dentin remineralization

For remineralization of dentin, different approaches have been applied, which can be classified as classical and non-classical , or top-down and bottom-up approaches . In the classical (top-down) approach , dentin remineralization is based on the epitaxial growth of residual crystallites, which act as nucleation sites for the calcium phosphate minerals to precipitate when dentin is stored in a solution rich with calcium and phosphate ions . However, recent studies have indicated that such an approach would result in an incomplete and non-functional remineralization of dentin . The classical remineralization approach results in extra-fibrillar remineralization of the collagen matrix of dentin, without the mineralization of the collagen’s intra-fibrillar compartments . This is due to the lack of control on orientation and size of the apatite crystals formed during this process. The non-classical approach was suggested as an alternative in vitro remineralization system, which attempted to achieve a hierarchical biomimetic remineralization of the organic matrix of dentin . This approach involves the use of synthetic substitutes for certain dentin matrix proteins that play an essential role during the bio-mineralization process. Two types of analogs were suggested: the first is a sequestration analog, which requires polyanionic molecules such as polyacrylic acid to allow the formation and stabilization of amorphous calcium phosphate (ACP) . These nano-aggregates of ACP are thought to form flowable nano-precursors which can infiltrate the water filled gap zones in dentinal collagen fibrils, where they precipitate as polyelectrolyte-stabilized apatite nano-crystals . This precipitation is guided by the second analog, which is a dentin matrix phosphoprotein substitute . This analog is usually a polyphosphate molecule, such as sodium metaphosphate, which acts as an apatite template, encouraging crystalline alignment in the gap zones , leading to a hierarchical dentin remineralization .

In the biomimetic bottom-up remineralization approach, calcium silicate based cements, such as MTA, were used as the calcium source . Upon hydration, these cements release calcium in the form of calcium hydroxide over a long duration . In addition to calcium provision, these cements release silicon ions into underlying dentin . Silica was found to be a stronger inducer of dentin matrix remineralization compared to fluoride . Furthermore, calcium silicate cements have the advantage of high alkalinity, which favors apatite formation and matrix phosphorylation , thereby providing a potential caustic proteolytic environment which could enhance dentin remineralization .

The non-classical approach is thought to be advantageous over the classical, as the former provides continuous replacement of water, which occupies the intra-fibrillar compartments, by apatite crystals. The non-classical approach also provides a self-assembly system that does not require the presence of nucleation sites . However, the idea of using matrix protein analogs could be challenging clinically and difficult to apply. Recent developments have attempted to take this concept a further step into clinical application . Therefore, the classical approach remains a closer and more applicable approach to the clinical situation.

5

Imaging dentin mineralization

Many different techniques have been used to evaluate dentin mineralization. Scanning and transmission electron microscopy (SEM and TEM) , Fourier transform infrared spectroscopy (FTIR) , Raman spectroscopy , X-ray diffraction (XRD) , energy dispersive X-ray spectroscopy (EDX) , microradiography , micro-CT scanning , and nano-indentation . However, they do not enable high-resolution observation of this process within the organic matrix of dentin. The capability of these techniques in detecting mineralization at high resolution is limited to surface characterization, whether morphological, chemical, or mechanical. Hence, they do not enable the observation of the mineralization process and its morphological features deeply within the structure of the remineralizing organic matrix and may not allow a return to the same sample position over a period of time. For this purpose, combining a microscopic technique with the capability of deep tissue imaging, such as two-photon fluorescence microscopy, with a selective mineral labeling fluorophore, such as Tetracycline, could provide a useful method .

Tetracyclines are polycyclic naphthacene carboxamides, composed of four carboxylic rings, which facilitate binding of the molecule to surface available calcium ions of the mineralized tissue. This binding process, helpfully enhances the fluorescence of Tetracycline after binding to calcium in these tissues . Therefore, this labeling agent has been widely used for the study of mineralization process in bone , and teeth : it is normally excited in the deep blue (400 nm) excitation range.

Two-photon excitation fluorescence microscopy is an advanced imaging technique that enables high-resolution observation of deep tissues due to the reduced scattering of high-intensity pulsed infrared laser light compared with conventional confocal microscopy . Our home-built two-photon lifetime fluorescence microscope uses a femtosecond Ti-Sapphire laser to excite the sample and a time-correlated single-photon counting card to record the lifetime of the emitted fluorescence, providing extra intensity-independent imaging contrast. Measurement of the subtle fluorescence lifetime variations across a dentin–restoration interface allows the sensitive detection of newly-formed Tetracycline-bound minerals.

In a study reported by Atmeh et al. 1.5 mm thick demineralized dentin disks were apposed to set calcium silicate dental cement: Biodentine™. The samples were stored in PBS solution with Tetracycline and kept in an incubator at 37 °C. Disks of demineralized dentin were kept separately in the same solution but without the cement to be used as control samples. After 6 weeks, one half of each sample was examined using a two-photon fluorescence microscope, illuminating at 800 nm and with a 40× objective ( Fig. 2 ). A highly fluorescent Tetracycline band was detectable beneath both the apposed and under-surface of the disks ( Fig. 2 c and d). Small globular structures were also noticed in the tubular walls adjacent to highly fluorescent substances within the dentinal tubules ( Fig. 2 a and b).

Mineralization detection with Tetracycline labeling and two-photon fluorescence imaging. (a) Intra-tubular mineralization in the form of highly fluorescent structures within the dentinal tubules (arrows). (b) Intra-tubular mineralization with globular appearance. A highly fluorescent band is present within the dentin just underneath the cement (c), as well as on the opposing dentin surface of the demineralized disk (d).

With the enhanced understanding of the dental caries process and improvements in both dental materials and diagnostic devices, more interest has been directed toward minimally invasive approaches for the treatment of dental caries. Such approaches eventually aim to minimize the excavation of dental tissues and instead, to encourage their recovery and repair. Dentin caries results from a bacteriogenic acid attack followed by enzymatic destruction of the organic matrix. These lesions can be classified into caries-infected and -affected dentin based on the extent and reversibility of the damage induced. In the caries-infected dentin, the organic matrix is irreversibly damaged, while the deeper caries affected dentin is hypomineralized with sound organic matrix, which could be repaired and remineralized. The slow progression of caries allows a reparative intervention, which can restore the mineralized architecture after excavating the infected layer.

6

Carious dentin remineralization models

Carious dentin remineralization has been studied extensively using different in vitro models. In these models, the dentin caries process has been simulated by partial demineralization of sound dentin using pH cycling or short duration application of phosphoric acids or ethylenediaminotetraacetic acid (EDTA) and in some studies total demineralization and . These methods were actually aiming to simulate the caries affected “hypomineralized” dentin, which has the potential to be remineralized. However, these models all show shortcomings in that they do not model the effect of the pulpal response to the carious lesion and the tissue fluid dynamics/re-mineralization phenomena within the dentin tubules. The challenge of using real carious dentin as a substrate is its inherent variability, so methods of characterizing the extent of caries removal and the position of the tooth restoration interface within the treated lesion would be beneficial for introducing some degree of consistency to a variable substrate. The use of confocal endoscopic fluorescence detection methods can aid in this depth determination and discrimination , while other optical techniques such as Raman spectroscopy and fluorescence lifetime imaging may also help to characterize the dentin caries substrate .