The current state-of-the-art techniques for restoring decayed teeth with aesthetic resin composites require demineralisation of enamel and dentine with phosphoric acid in etch-and-rinse systems (E&R), whilst phosphoric acid esters find use in self-etch systems (SE). The completely or partially demineralised collagen matrix allows the intrusion of adhesive resins that, upon polymerisation, render the surface compatible with the placement of restorative resin composite materials [1]. However, bonded restorations have a limited lifespan and remain susceptible to hydrolytic, enzymatic, and/or fatigue degradation over time. Current SE systems were primarily developed to overcome post-operative and moisture sensitivity of E&R systems, reduce the number of clinical steps, and decrease the discrepancy between etching and resin infiltration, creating a more homogenous hybrid layer within dentine [2]. However, SE systems are unable to create desirable etch patterns on enamel, especially with milder formulations [3] and exposed enamel prisms are only observed with stronger systems [4]. In addition, SE systems are hydrophilic since the aqueous component enables ionisation of acidic functional monomers, which can lead to phase separation, incomplete polymerisation, and hydration of the hybrid layer [5,6]. This has led clinicians to use a selective etch approach, where the enamel is preferentially etched with phosphoric acid prior to the use of a SE system [7], though this remains technique sensitive as the phosphoric acid rinse step can cause over or under-drying of dentine.

Resin composite failures are primarily dependent on the quality of adhesion with the tooth structure. The most compelling problems associated with SE systems are attributed to sub-optimal enamel etching which has been correlated with significant marginal discoloration clinically [8], hydrolysis of the hydrophilic resin components [9], degradation of collagen fibrils via endogenous matrix metalloproteinases (MMPs) [10], as well as degradation and/or poor hybridisation of resin tags, all being detrimental to the durability of adhesive bonds, especially under cyclic stresses [11,12].

In current SE systems, 10-methacryloyloxydecyl dihydrogen phosphate (10-MDP) is the most commonly used functional monomer, with a carbon spacer chain that confers hydrophobicity, methacrylate groups enabling polymerisation and etching by virtue of the phosphate group [13]. However, 10-MDP has a high pKa, hence behaving as a weak acid that limits the demineralisation of dental tissues. SE adhesives with 10-MDP have shown promising results in adhesion and superior performance compared to other acidic functional monomers, both in vitro and in vivo [3,14,15], however, the hydrophilic end bound to the calcium may still hydrolyse [16]. Hence the interface remains prone to both hydrolytic and enzymatic degradation compromising the longevity of the restoration [17]. Moreover, 10-MDP-based SE systems demonstrate lower etching efficacy if there is a predominance of enamel and only lead to superficial dentine demineralisation, with shallow resin tags approximately 1 µm deep, indicating the somewhat superficial interactions of 10-MDP-Ca salts [18].

Bis[2-(methacryloyloxy) ethyl] phosphate (BMEP) is an acid-stable polymerisable monomer with a lower pKa than 10-MDP, which was found to self-polymerise upon interaction with hydroxyapatite improving the resin polymer network [19] as the phosphate group of BMEP is flanked by two methacrylate groups that enable crosslinking with other methacrylate monomers. The low pKa, hydrophilicity, and short carbon chain in BMEP were found to promote effective etching of both enamel and dentine, resulting in longer resin tag formation whilst offering potent MMP inhibition [19]. The aim of this study was to develop and characterise a two-step SE bonding system with a primer formulated with BMEP, tested in conjunction with an adhesive with or without BMEP. These systems were compared to a commercial 10-MDP-containing bonding system. The null hypotheses were that there would be no difference in (1) the enamel surface roughness on treatment with the experimental systems and commercial reference, (2) the enamel bond strengths of the experimental systems and commercial reference, (3) the dentine bond strengths, interfacial nanoleakage and flexural fatigue of the experimental systems and commercial reference, and (4) the MMP enzyme inhibition of the experimental bonding systems compared to an inhibitor control.