2.1

Synthesis of niobium silicate

Niobium silicate particles were synthesized by the sol-gel route. Briefly, Tetraethylorthosilicate (TEOS – Si(OC ₂ H ₅ ) _{4 –} Aldrich Chemical; St Louis, MO, USA) was used as silica precursor, and niobium pentachloride (NbCl ₅ – CBMM, Araxá, MG, Brazil) was used as an inorganic modifier. Water: Teos molar ratio was set at 10:1 while Nb:Si molar ratio was set at 1:9. After heat treatment, particles were grounded in an analytical mill (M 20 Universal mill- IKA, Campinas, Sp, Brazil) at 28,000 rpm. Milled particles were dispersed in absolute ethanol and sieved to controlling particle size distribution. The structure of the particles was evaluated by x-ray diffraction (X’PertPRO, PANalytical MPD, Netherlands) and Fourier-transform infrared spectroscopy (VERTEX 70, Bruker Optics, Ettingen, Germany). The morphology of the particles was analyzed by scanning electron microscopy (JSM-6060, Jeol, Akishima, Japan), laser diffraction (Mastersizer Malvern PANalytical, Almelo, Netherlands), and nitrogen adsorption (Quantachrome Instruments Corporate Headquarters, USA). Glassy structured particles were produced with a mean particle size of 2.054 μm with a specific surface area of 616.96m ² /g. The characterization of synthesized particles is shown on the supplementary material ( Fig. 1 ).

Polymerization kinetics of experimental composite resins. Degree of conversion (%) (A) and polymerization rate (Rp.[s ⁻¹ ]) (B) are shown over time. The relationship between polymerization rate and degree of conversion is plotted in 2C.

2.2

Formulation of experimental composite resins

Composite resins were formulated with bisphenol A-glycidyl methacrylate (BisGMA – 70 wt%) and triethylene glycol dimethacrylate (TEGDMA – 30 wt%). Monomers were weighed and mixed in a sonication bath for 8 min. As the photoinitiator system, the trimethyl benzoyl-diphenylphosphine oxide (TPO – 1 mol%) and 0.01 wt% butylated hydroxytoluene (BHT). All reagents were purchased by Aldrich Chemical, St Louis, MO, USA.

Niobium silicate was used as filler in 50 wt% concentration to produce the SiNb group. Before the application of niobium silicate in the composite resin, particles were silanized with an acetone solution of 3-(trimethoxysilyl)propyl methacrylate (Merck KGaA, Darmstadt, Germany) at 95 wt% acetone concentration. A control group (SiBa) was produced using barium borosilicate glass in the same concentration (Esstech-Essington, Pennsylvania, USA) with a mean particle size of 4.58 μm. Colloidal silica (Aerosil 200- Evonik, Hanau, Germany) was used in both groups in 1.5 wt% to adjust the handling of the experimental composites.

2.3

Experimental composite resin properties

2.3.1

Refractive index

The refractive index for the experimental composite resins was analyzed by a spectral ellipsometer (Sopra GES-5E, Texas, USA) in a spectral range between 320 and 480 nm. Unpolymerized SiNb and SiBa composite resins were evaluated (n = 1). A filler-free resin was analyzed, as well. The amplitude of the refractive index between the spectral range was used and compared to the filler-free resin.

2.3.2

Degree of conversion and polymerization kinetics

The degree of conversion was analyzed in a spectrometer (FTIR – VERTEX 70, Bruker Optics, Ettingen, Germany) equipped with a reflectance device (ATR) by Fourier Transformed Infrared Spectroscopy (FTIR). Unpolymerized resins (n = 3) were placed in the ATR crystal in polyvinylsiloxane molds measuring 4 mm diameter x 1 mm height. Samples were polymerized with a multiwavelength light-emitting diode (Valo Cordless, Ultradent, UT, USA) at 1000 mW/cm ² for 40 s in a standardized distance of 1 mm between the sample and the light-curing unit. The DC of the composite resins was determined by the ratio of the absorbance peak of aliphatic carbon-carbon double bonds (1640 cm ⁻¹ ) to the peak of aromatic double bonds (1610 cm ⁻¹ ), an internal standard.

For polymerization kinetics measurement, the same apparatus was used. Samples (n = 3) were placed above the ATR crystal and polymerized for 40 s. The measurements were performed from the moment the photoactivation started until its end in standardized parameters of scanning range (4000–800 cm ⁻¹ ), resolution (4 cm ⁻¹ ), and scanning time (40 s). The data were calculated based on the peak height in the 1640cm ⁻¹ and 1610 cm ⁻¹ as a function of time. The data was submitted to a sigmoidal curve fitting in a linear regression with Sigma Plot 12.0 for Windows (Systat Software Inc, San Jose, CA, USA). The degree of conversion and the polymerization rate was plotted over time, and descriptive analysis was performed.

2.3.3

Flexural strength

Flexural strength was measured according to ISO 4049 [ ]. Five rectangular specimens (n = 10) with 25 mm x 2 mm x 2 mm were prepared for each group in a metallic mold. Polymerization was performed in 2 windows for 20 s on each side of the specimen. Specimens were stored in distilled water at 37 °C for 24 h before the tests. Flexural strength was determined with a three-point test at a crosshead speed of 0.75 mm/min in a mechanical testing machine (EZ-SX – Shimadzu, Kyoto, Japan) until the specimens fractured. Flexural strength was calculated from the following equation:

where F is the maximum load exerted on the specimen, Ɩ is the distance (mm) between the supports ± 0.01 mm, b is the width (mm) of the specimen immediately before testing, and h is the height (mm) of the specimen measured with a digital caliper immediately before testing.

2.3.4

Cytotoxicity

Primary gingival fibroblasts were used to test the effect of experimental composites on cell behavior. Cells were cultivated in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 100 U/mL penicillin and 10% of FBS. Cells were cultured at 37 °C, and under 5% CO ₂ / 95% air atmosphere.The medium was changed every 2–3 days. Before cell culture, samples (n = 3–4 mm diameter × 1 mm height) were immersed in 1 mL of supplemented DMEM and kept in a 37 °C environment for 24 h. For cell treatment, gingival fibroblasts (5 × 10 ³ ) were seeded in 96-well plates, and after 24 h, each well was treated with 100 μl of DMEM medium conditioned with composite resin specimens. Cell treatments were performed in triplicate. After 72 h in culture, cells were fixed and stained with 50 μl SRB 0.4% (Sigma Aldrich, St. Louis, EUA). Quantification of viable cells was performed for absorbance at 560 nm in a Microplate Spectrophotometer (Multiskan™ GO, Thermo Fisher Scientific, USA). Medium without treatment was used as a control to normalize the percentage of viable cells.

2.3.5

Radiopacity

Radiopacity was tested according to ISO 4049 [ ]. Samples (n = 5–6 mm diameter x 1 mm height) were exposed to x-ray during 0.8 s at a standardized distance of 400 mm. The X-ray source (Dabi Atlante model Spectro 70X) was operated with a tungsten anode at 70 kV, and 8 mA. Samples were placed in phosphorous plates next to an aluminum step wedge. A digital system (VistaScan- Durr Dental, Bissingen- Germany) was used to obtain the images. The pixel density in the images was measured in a standardized area in an image software (ImageJ, NIH, Maryland, USA) in the samples and the aluminum step wedge. The pixel density values were converted to mmAl according to the pixel density in each image.

2.3.6

Softening solvent

Specimens (n = 3) measuring 4 mm diameter x 1 mm height were prepared in polyvinylsiloxane molds and photoactivated for 20 s on each side. Discs were polished before an initial Knoop microhardness measurement (KHN1) in a Knoop Hardness Tester (HMV 2, Shimadzu, Japan). Three indentations were performed in each specimen with a 10 g load for 5 s. Specimens were then immersed in 70% ethanol solution for 2 h, as described previously [ , ], and a new KHN measurement was performed (KHN2). The differences between KHN1 and KHN2 were used to calculate de ΔKHN% for softening solvent analysis.

2.3.7

pH

Samples (n = 3) measuring 6 mm diameter x 1 mm height were immersed in distilled water for pH analysis in a digital pHmeter (DM 22- Digimed, São Paulo, Brazil). The initial pH of 5 mL of distilled water was initially measured, and after sample immersion, the pH was measured at 30 min, 1, 2, 3, 4, 24 h, 48 h, 72 h, 7, 14, 21 and 28 days. The water was not changed during the experiment.

2.3.8

Mineral deposition

For mineral deposition analysis, samples were produced measuring 4 mm diameter and 1 mm height. The experimental composite resins were immersed in simulated body fluid (SBF) [ ] for 7 and 14 days at 37 °C. After the determined time points, Raman spectroscopy (Senterra, Bruker Optics, Ettingen, Germany) was used to determine the amount of phosphate on the sample’s surface based on the PO ₄ ³⁻ intensity at the 960 cm ^-1 region. The surfaces of the analyzed samples were imaged using scanning electron microscopy (JSM-6060, Jeol, Akishima, Japan).

2.4

Statistical analysis

A descriptive analysis of the characterization of niobium silicate particles was performed. The amplitude was used to describe refractive index data. The normality of data was assessed by Shapiro-Wilk. Student’s T-Test for independent samples was used to compare differences between groups in the degree of conversion, flexural strength, cytotoxicity, pH, KHN1, and %ΔKHN. The differences between KHN1 and KHN in each experimental group were assessed by the Student’s T-Test for dependent samples. Descriptive analysis was performed for polymerization kinetics and for mineral deposition results obtained by Raman spectroscopy and SEM. All tests were performed at a 5% significance level.