Corundum

The starting material was a diatomite, i.e. a tripolaceous siliceous rock (Tripoli rock) cropping out in the Crotone Basin in southern Italy. Tripoli was analysed by X-ray diffraction (XRPD) with a Siemens D5000 operating with a Bragg-Brentano geometry; CuKα = 1.518 Å, 40 kV, 40 mA, 2–35° scanning interval, step size 0.020° 2θ with a scan rate of 13 sec/step. Characterization of this material revealed a mineralogical assemblage mainly consisting of an amorphous siliceous fraction (diatoms and sponges, visible as the bulge in the range 17–25°2theta) with minor presences of quartz, montmorillonite, chlorite, kaolinite, K-micas and small amounts of calcite (Fig. 1). Chemical analysis of Tripoli rock is reported in Table 1 and was performed by X-ray fluorescence analysis (Axios-Max Advanced Panalytical; 60KV; 160 mA; 4000 W; 0.0001°2 θ). We considered the value of LOI of 1 g of sample obtained in ceramic meltpots running in an oxidizing furnace.Figure 1

XRPD pattern of “Tripoli rock”.

Table 1

Chemical composition of “Tripoli rock” analysed by X-ray fluorescence.

	Tripoli rock
SiO2	81.07 (0.45)
TiO2	0.26 (0.02)
Al2O3	5.03 (0.02)
Fe2O3	2.14 (0.03)
MnO	0.07 (0.01)
MgO	1.11 (0.01)
CaO	1.72 (0.03)
Na2O	0.25 (0.02)
K2O	0.68 (0.02)
P2O5	0.07 (0.01)
LOI	7.73 (0.03)
Tot.	100.13 (0.18)

The standard deviation values calculated for three analyses are reported in brackets. LOI: loss on ignition.

The synthesis of leucite was conducted through the mixing of silicate and aluminate solutions. These solutions were prepared according to the procedure already described in Novembre et al.^{24 (link)} In the present study, 5.19 g of the ground and powdered Tripoli material were treated with HNO₃ (65%), in order to dissolve the calcite fraction in order to remove the soluble calcite fraction from the starting material. The diatomitic sample (ca. 5 g after the HNO₃ treatment) was added to 50 mL of KOH (6.8%). This solution was thoroughly mixed with a magnetic stirrer for 2 h and then put in a teflon reactor/bomb and heated in an oven at 80 °C for 24 h. After filtration, the remnant solid and insoluble fraction, which consisted of clay minerals and quartz, was separated from the silicate solution. The resulting molar composition of the solution was 0.060 K₂O–0.026SiO₂–0.625H₂O with traces as follows: 2.01 ppm Mg, 2.11 ppm Ca and Al, Ti and Mn lower than 0.1 ppm. Based on a mass balance calculation following this step-wise chemical separation process, the Tripoli rock was determined to be composed of: 63.27 wt% amorphous silica (diatoms and sponges), 32 wt% of clay minerals and quartz, and 3.83 wt% calcite.
The aluminate solution was prepared as follows: 0.45 g of Al(OH)₃ (65%) was mixed with 50 mL of KOH (6.8%). The obtained aluminous solution with a composition of 0.060 K₂O–0.0076Al₂O₃–0.625 H₂O (Mn, Ti and Mg < 0.01 ppm; Fe < 0.4 ppm; K, Ca and Si < 0.2 ppm) was then heated at 100 °C for one hour.
A series of three syntheses were carried out by varying the volume ratio of the two solutions according to Table 2.Table 2

Starting mixture and relative obtained mineralogical assemblages of experimental runs.

synthesis run	starting mixture	SiO2/Al2O3	mineralogical assemblage
1	10 ml siliceous sol + 10 ml aluminous sol	3.40	KAlSi2O6 + KAlSIO4-O1 (1.5–20 h); KAlSi2O6 (24 h)
2	12.5 ml siliceous sol + 7.5 ml aluminous sol	5.70	KAlSi2O6 + KAlSIO4-O1 (1.5–15 h); KAlSi2O6 (20 h)
3	10 ml siliceous sol + 5 ml aluminous sol	6.80	KAlSi2O6 + KAlSIO4-O1 (3 h); KAlSi2O6 (15–20 h)

The reactants were vigorously mixed for two hours with a magnetic stirrer. Each mixture was heated inside an autoclave at 150 °C and ambient pressure for a duration of one hour. The hydrothermally derived gel precursors were recovered from the reactors, filtered from the solution, thoroughly washed with distilled water and dried in an oven at 40 °C for 24 hours. These gel products were examined by XRPD analysis (Fig. 2) in order to assess their amorphous character. The three gel precursors were then calcined at 1000 °C with periodic sampling carried out at scheduled intervals.Figure 2

XRPD patterns of the hydrothermal gel precursors. (a): synthesis run 1; (b): synthesis run 2; (c) synthesis run 3.

All intermediate and final products of the three syntheses were analysed by XRPD under the same operating conditions as those for the “Tripoli rock” analysis. Identification of phases and relative peak assignment were made with reference to the following JCPDS codes: 00-038-1423 for leucite and 00-011-0579 for KAlSIO₄-O1. The amounts of both the crystalline and amorphous phases in the synthesis powders were estimated using Quantitative Phase Analysis (QPA) applying the combined Rietveld and Reference Intensity Ratio (RIR) methods; corundum NIST 676a was added to each sample, amounting to 10%, and the powder mixtures were homogenized by hand-grinding in an agate mortar. Data for the QPA refinement were collected in the angular range 5–70° 2θ with steps of 0.02° 2θ and 10 s step⁻¹, a divergence slit of 0.5° and a receiving slit of 0.1 mm.
Data were processed with the GSAS software²⁸ and the graphical interface EXPGUI^{29 (link)}. The unit cell parameters were determined, starting with the structural models proposed by Dove et al.³⁰ for leucite and Kremenovic et al.^{31 (link)} for KAlSiO₄-O1. Parameters were refined following Novembre et al.^{25 (link)}.
Analysis of synthesized powders was performed by inductively coupled plasma optical emission spectroscopy (ICP-OES, Perkin Elmer Optima 3200 RL) after alkaline fusion of the sample in a Pt crucible (lithium meta-tetra borate pearls, at 3/2 ratio) and subsequent acid solubilization^{27 (link)}.
Scanning electron microscope (SEM) analyses were carried out with a JEOL JSM-840 with operating conditions of 15 kV and window conditions ranging from 18 to 22 mm, following the procedure as explained in Ruggieri et al.^{32 (link)}.
Vibrational spectra of the synthesized products were obtained with an Infrared spectrometer FTLA2000, equipped with SiC (Globar) filament source, KBr beamsplitter and DTGS detector. Samples were prepared according to the method of Robert et al.^{33 (link)} using powder pressed pellets (sample/KBr ratio of 1:100); spectra were processed with the program GRAMS-Al.
Thermal behaviour of gel precursors were studied by differential thermal analysis and thermogravimetry (DTA-TG) by means of a Mettler TGA/SDTA851^e instrument (10°/minute from 30° to 1100 °C, using an approximate sample weight of 10 mg in Al₂O₃ crucible).
Density of leucite was measured by He-picnometry using an AccuPyc 1330 pycnometer.

In order to evaluate contraction stress, which generates during photopolymerization of resin composites, transparent and photosensitive plates made of epoxy resin (Epidian 53, Organika-Sarzyna SA, Nowa Sarzyna, Poland) were used. The calibrated orifices (3 mm in diameter and 4 mm in thickness) in resin plates were prepared. The circular shape and size of orifices was meant to mimic an average tooth cavity. To obtain higher micromechanical retention, surface of the plates was sandblasted with a 50-μm grain corundum Cobra (Renfert, Hilzingen, Germany). Thus, prepared plates were immersed in distilled water for 3 months to eliminate errors associated with water sorption of resin. Next, dedicated bonding system was applied and cured with Elipar S10 lamp (3M ESPE, Landsberg am Lech, Germany) (Table 3). The orifices were filled with tested material in one layer. Three samples were prepared for each material. The polymerization was performed according to the manufacturer’s instructions (Table 2 and Table 3). Both light curing units (Mini L.E.D and Elipar S10) had an output irradiance of 1250 mW/cm² and 1450 mW/cm², respectively, as stated by the manufacturer. To ensure consistent irradiance values, the light curing units were calibrated with radiometer system (Digital Light Meter 200 from Rolence Enterprice Inc., Taoyuan, Taiwan).
Next, samples were stored in distilled water at room temperature. After selected period of time (30 min, 24 h, 72 h, 120 h, 168 h, 240 h, 336 h, 504 h, 672 h, and 1344 h), the generated strains in the plates were visualized in circular transmission polariscope FL200 (Gunt, Hamburg, Germany). Photoelastic images were registered by digital camera (Canon EOS 5D Mark II/Canon Inc., Tokyo, Japan), both in parallel and perpendicular orientation of filter polarization planes. Met-Ilo computer program (J. Szala, 2012, Poland) was applied to determine the arrangement and the dimension of interference fringes. The analysis of stress and strain was carried out in a two-dimensional state of the stress and three-dimensional state of deformations. The analysis of stress and strain was carried out in a two-dimensional state of the stresses and three-dimensional state of deformations. Additionally, the calculation was conducted following this assumption: the relative change in composite volume caused the extension of composite and the extension of base material being “tooth model” (epoxy resin plate). Accordingly, it was possible to determine the radial and circumferential stresses based on the Equations (4) and (5) given by Timoshenko [43 ]: ${\sigma }_r=\frac{a^2·p_s}{b^2-a^2}·\left(\frac{b^2}{r^2}-1\right)$
${\sigma }_{\theta }=\frac{a^2·p_s}{b^2-a^2}·\left(\frac{b^2}{r^2}+1\right)$
where σ_r—the radial stress, σ_θ—the circumferential stress, p_s—the shrinkage stress around composite filling, a—the radius of the internal orifices in the plate, b—the radius of the largest of isochromatic fringe, and r—the radius contained in the region from a to b.
Upon calculating the shrinkage stress on the circumference of the orifices, the radial and circumferential stresses were determined on the basis of Equations (2) and (3).

The thermogravimetric instrument (STA449F3, Selbu, Bavaria, Germany) was employed to measure the composition and relative content of hydration products of different binders. The hydrated sample powder weighed about 20 mg and was placed in a corundum crucible. The sample together with the Al₂O₃ crucible was elevated from room temperature of 25 °C to 800 °C at N₂ with a controlled heating rate of 10 °C/min. The content of chemically bound water (BW) and portlandite (CH) was calculated according to Equation (4) and Equation (5) [56 (link)], respectively.
$$W_{BW}=\frac{W_{50}-W_{550}}{W_{550}}\times 100$$
$$W_{CH}=\frac{\left(W_{400}-W_{500}\right)}{W_{550}}\times \frac{74}{18}\times 100$$
where the W₅₀, W₄₀₀, W₅₀₀, and W₅₅₀ was the weight of specimens at 50 °C, 400 °C, 500 °C, and 550 °C.

Using XRD analysis and the Rietveld refinement technique, the phase composition of cement paste was quantitatively investigated. XRD patterns were employed a D8 Advance diffractometer at 40 kV and 100 mA with CuKa radiation. The scanning range of 5~70° with a step length of 0.01°. Corundum (α-Al₂O₃, 10%) was used as an internal standard, and mixed with the stop cement powder. The Rietveld calculation was carried out by using TOPAS Academic V5.