Two commercial hydraulic calcium silicate-based dental materials were tested (Biodentine; Septodont, Saint Maur des Fosses, France; MTA Angelus; Angelus, Londrina, Brazil) together with two experimental tricalcium silicate-based materials including: tricalcium silicate (TCS; Mineral Research Processing, Meyzieau, France), tricalcium silicate containing 20% zirconium dioxide (TCS-ZrO; Sigma-Aldrich, Gillingham, U.K.) as a radiopacifier. The TCS and radiopacified TCS were tested to assess the effect of the additives in the commercial materials.
The commercial materials were mixed according to the respective manufacturer’s instructions whilst the TCS-based prototypes were mixed with water at a water/powder ratio (by mass) of 0.35 for both the TCS and TCS-ZrO. Test samples 10 mm in diameter and 2 mm thick were prepared of each material type and were allowed to set in a humid environment in an incubator at 37 °C for 24 h. The end of setting was verified when an a final set Gilmore needle 1.06 mm in diameter and weighing 453.6 g (Impact Test Equipment, Stevenston, UK) failed to leave a mark on the material surface. Solubility was determined after immersion in either water, Hank’s balanced salt solution (HBSS; Sigma Aldrich, Gillingham, UK) or HBSS containing 10% fetal calf serum (Sigma-Aldrich, Gillingham, UK).
The material microstructure and chemical composition were assessed using scanning electron microscopy, energy dispersive spectroscopy and X-ray diffraction (XRD) analysis immediately after setting without any immersion or storage in media.
For scanning electron microscopy, samples were embedded in a cold-cure epoxy resin (Epoxy-fix; Struers, Ballerup, Denmark) and the surfaces were polished using an automatic polishing machine (Buhler, Lake Buff, IL, USA) using diamond discs (MD Piano; Struers, Ballerup, Denmark) under water coolant using 250, 500 and 1200 grit followed by polishing cloths MD Largo, MD Dac and MD Nap; (Struers, Ballerup, Denmark) using 9, 3 and 1 µm diamond impregnating polishing liquids. Polished samples were mounted on aluminium stubs (Agar Scientific, Stansted, UK) with double-sided carbon tape. An ultra-thin conductive gold coating (Emitech K550X; Ashford, UK) was sputtered on the polished surfaces, which were then viewed using the SEM (EVO MA10; Zeiss, Oberkochen, Germany) with an accelerating voltage of 20 kV and a working distance of 8.5 mm. The materials were then examined using back-scattered electrons to obtain elemental contrast at different magnifications and energy dispersive spectroscopy over an area was performed to assess the elemental distribution within samples.
For the X-ray diffraction (XRD) the materials were ground using an agate mortar and pestle to a fine powder. XRD was performed with a diffractometer (Bruker D8 Advance; Bruker, Billerica, MA, USA) with a CuKα radiation at 40 mA and 45 kV was set to rotate between 10° and 60° with a 0.02° 2θ step and a step time of 0.6 s. Phase identification was undertaken using a search-match software (DIFFRAC.EVA; Bruker, Billerca, MA, USA) using the ICDD database (International Centre for Diffraction Data, Newtown Square, PA, USA).
The solubility assessment was performed using a method recommended by ISO 6876; 20129 with water and also with variations in the liquid used. A new µCT-based method was also used by calculating the solubility by weight and volume changes after exposure to the solution. For both methods, the assessment was performed by 3 operators (CW, JH, SK) independently to have an intra-laboratory comparison. Each operator prepared and tested their own specimens using the same batches of materials and same equipment.
ISO 6876 method
The materials were mixed as indicated in the materials section. Six specimens measuring 20 mm in diameter and 1.5 mm high were prepared for each material type and immersion medium. The materials were allowed to set for 24 h at 37 °C in 100% humidity before weighing to the nearest 0.001 g (TS400D, Ohaus, Florham Park, NJ, USA). The samples (n = 2) were placed in a shallow dish and 50 ± 1 mL of either deionized water as suggested by the ISO standard, or alternatively in HBSS or HBSS containing 10% fetal calf serum. The use of alternative solutions is a deviation from the ISO standard. The container was covered and allowed to stand for 24 h before transferring all contents to a second dish after filtering. The liquid was evaporated at 110 ± 2 °C until a constant mass was obtained and the containers were placed in a desiccator at room temperature to cool before weighing. The difference in mass of the dish before and after drying as the amount of material removed calculated as a percentage of the original combined mass of the two specimens expressed the material solubility in the different solutions. The experiment was repeated three times for each material and each solution and the experiments are undertaken by three operators working independently.
Perspex blocks measuring 20 × 20 mm were prepared containing a standard cavity 4 mm in diameter and 3 mm deep drilled in the centre of each block. All Perspex blocks were weighed and the mass recorded as M0. The four hydraulic calcium silicate materials were mixed and compacted into the cavities using a stainless-steel condenser and a microscope slide was then placed over the mould to ensure a flat sample surface. The materials were allowed to set at 37 °C for 24 h in 100% humidity. The materials and blocks were weighed again and mass recorded as M1. All the weights were taken to the accuracy of 0.0001 g. Six blocks were prepared for each material and each solution tested.
Microcomputed tomography was performed and images of samples were obtained using a µCT scanner (SkyScan 1172; Bruker, Billerca, MA, USA) at 70 kV and 142 µA in the presence of a 0.5 mm aluminium filter at ambient temperature (22 °C). A flat field correction was taken on the day, prior to scanning to correct for variations in the pixel sensitivity of the camera. Images were reconstructed using software (NRecon Version 1.4.0; Bruker, Billerca, MA, USA) with ring artifacts reduction of 13 and beam-hardening correction of 20%. The volume of the materials was determined (CTAn, Version 18.104.22.168+; Bruker, Billerca, MA, USA).
After the initial measurements, the blocks were then immersed in either 15 mL of water, or HBSS or HBSS + FCS at 37 °C. After 1 week the blocks and materials were retrieved and surface dried using a filter paper before weighing (recorded as M2). The volume of the materials was determined using µCT. All the measurements were undertaken by three operators to have intra-laboratory comparisons and changes in mass and volume were determined.
Surface characterization and leachate analysis
Surface characterization was performed using SEM after the completion of the solubility assessment using the Perspex blocks. For SEM, the material surfaces were coated with gold and imaged using secondary electrons. Images were captured at 500, 100 and 2000 magnification to assess surface microstructural changes. The leachates collected from samples immersed in the different solutions were analysed for calcium, silicon, zirconium and tungsten with inductively coupled plasma mass spectroscopy (ICP-MS; Optima 8000, Perkin Elmer, Waltham, (MA) USA).
Statistical analysis was performed by one operator using Predictive Analytics Software (PASW version 18; IBM, Armonk, NY, USA). One-way ANOVA was used to determine whether there were significant differences among data sets. The data was tested to ensure it was normally distributed and then with analysis of variance with p = 0.05, the Tukey post-hoc test was used.