Cold atmospheric plasma treatment

Samples were processed using a miniaturized mono channel jet plasma device developed by the Laplace-PRHE research group (Laplace lab (“Plasmas Réactifs Hors Equilibre” team), University of Toulouse, Toulouse, France) and already described in previous studies [12, 13]. The di-electric barrier discharge (DBD) configuration employed for this plasma jet comprises a quartz tube with an internal diameter of 4 mm and an external diameter of 6 mm, to which two aluminum electrodes (20 mm wide and spaced 10 mm apart) have been welded (Fig. 1).

Fig. 1The alternative text for this image may have been generated using AI.

Experimental set-up of the cold atmospheric plasma device. a 2D internal schematic representation of the di-electric barrier discharge plasma jet configuration in Helium; b plasma jet in operation in contact with a tooth sample

High-voltage pulses were applied with a 10 kV input voltage amplitude. The high-voltage pulses were provided by a direct-current (DC) voltage generator (Technix HV SR10-R-300, Technix, France) coupled to a high-voltage chopper (DEI PVX-4150, Berkeley Nucleonics – DEI Division, USA). A signal generator (Armexel PDG-2520, USA) controlled the chopper in order to set the parameters of the pulse high-voltage signal: repetition rateat 10 kHz and pulse width at 1 µs.

The choice of plasma gas depends on the desired CAP properties and the intended application. In the present study, Helium gas was used due to its lower ionization energy and inert nature. The excited Helium can efficiently generate a variety of reactive oxygen and nitrogen species (RONS) when it interacts with the surrounding air (Supplementary data 1). In our study, the Helium flowed through the quartz at a flow rate of 3L.min−1.

Depending on the analyses to be carried out, the samples were exposed to the plasma jet for times varying from 15 to 180 s at a gap (d) distance of 5 mm or 20 mm. All the parameters are summed up in the PLASDENT grid (Table 1) [14].

Table 1 PLASDENT grid of the study, according to [14]Contact angle measurement

In order to evaluate the influence of the plasma jet on the modification of dentin surface’s wettability, 91 healthy human molars have been collected in Toulouse Hospital (Service d’Odontologie, Hôpitaux de Toulouse, Toulouse, France) in compliance with the Hospital’s rules for tissue collection. Indeed, in accordance with international regulations since the Declaration of Helsinki, this tissue bank has been registered since 2022 by the department of Research and Innovation of the French Ministry of Higher Education, Research and Innovation (Bioethics Unit) under the acronym DENTABOUCHE (n°DC-2022-5010) in order to respect the traceability of surgical waste and the non-opposition of patients to entrust it to a tissue bank. Before the beginning of the study, the teeth were conserved in a 1% chloramine solution in order to decontaminate them as well as maintain their hydration and their integrity [15]. To be suitable for use in the study, the teeth had to present healthy crown and root—i.e., those who presented a decay, a restoration of any type, a crack or a fracture were excluded. From the beginning of the tooth collection for this study, it was decided that they would not be stored in the chloramine solution for more than one month before being used. This parameter guided the sampling for the experiments presented below. None of the samples were used for two different types of experiments. Teeth were randomly assigned to the various experimental groups by adding one tooth per group in turn as they were retrieved from the tissue bank.

To expose the dentin on which performing the bonding procedure, the crowns of the teeth were sectioned perpendicularly to their longitudinal axis, 2 mm above the cement-enamel junction, with a low-speed diamond disc under irrigation (IsoMet 2000, Buehler, Leinfelden-Echterdingen, Germany). Once their dentin was exposed, it was etched, rinsed and slightly dried before CAP application at a distance of 20 mm (for 30 s, 60 s and 120 s) or 5 mm (for 15 s, 30 s and 60 s). The control group (CA0) was not exposed to CAP. The contact angle of distilled water was measured by sessile drop technique with the use of a goniometer (Surftens Contact angle measuring instrument, OEG GmbH, Frankfurt, Germany). A 2 μL drop of distilled water was syringed onto the center of the dentin. Images were captured and processed directly after the drop deposition for analysis using Surftens4 software (OEG GmbH, Frankfurt, Germany).

SEM observations

Human molars were observed using scanning electron microscopy (SEM) (FEI Quanta FEG 250, ThermoFisher Scientific, Waltham, MA, USA) after sputter-coat with platinum (Leica EM MED020, Leica Microsystems GmbH, Wetzlar, Germany). They were both etched with 35% orthophosphoric acid for 20 s and one of them was exposed to CAP (at 5 mm for 30 s) whereas the other was not.

Shear bond strength

One hundred and twenty healthy human molars, extracted for orthodontic, periodontal or wisdom tooth reasons, have been collected and prepared as described above. The roots of each tooth were included in hard plaster cylinders to prepare the future shear bond strength tests (Fig. 2). Depending on the conditions of CAP application, 8 groups were constituted (n = 15). Apart from the use of CAP, all samples were treated using a standard clinical treatment protocol of 15 s etching with 35%-orthophosphoric acid (Ultra-Etch 35% orthophosphoric acid, Ultradent Products, Utah, USA), rinsing and slight drying. The surface was then exposed to the plasma jet at a distance of 20 mm (during 30 s, 60 s, 120 s and 180 s) or 5 mm (during 15 s, 30 s and 60 s). The control group wasn’t exposed to CAP. After CAP application, the surface was immediately coated with a universal adhesive used on 2-steps etch-and-rinse mode (FuturaBond U, VOCO, Cuxhaven, Germany) and photopolymerized during 10 s at a power of 1200mW/cm2 (D-Light Pro, GC Europe, Leuven, Belgium). Each sample was positioned into a specific support on which a plastic mold was centrally perforated and placed in contact with dentin to receive a calibrated resin pin. The dental composite resin (Estelite Asteria, Tokuyama Dental Corporation, Tokyo, Japan) was thus injected in the mold and light-cured for 20 s at a power of 1200mW/cm2. After mold removal, the samples could be placed horizontally on the platform of a universal testing machine (UltraTester Bond Strength Testing Machine, Ultradent, South Jordan, UT, USA) whose indented tip was positioned in contact with the resin pad. The shear bond strength (SBS) test was carried out at a speed of 2 mm/min until the restoration was separated from the tooth and the resistance value was registered (in MPa).

Fig. 2The alternative text for this image may have been generated using AI.

Experimental protocol applied for shear bond strength testing. a The teeth were cut to expose dentin surfaces and b then included in hard plaster cylinders. c Dentin etching was performed, followed by (d) water cleaning and slight drying. e After CAP treatment or not, the adhesive was applied and light-cured, f as well as the 2 mm-side piece of dental composite resin. Finally, g the shear bond test was performed at a crosshead speed of 2 mm/min

Raman spectroscopy

The use of Raman spectroscopy was decided to determine and understand the effects of CAP on dentin and collagen. One human molar was collected and stored in a 1% chloramine solution. The crown of the tooth was sectioned perpendicularly to its longitudinal axis, 2 mm above the cement-enamel junction, with a low-speed diamond disc under irrigation (IsoMet 2000, Buehler, Leinfelden-Echterdingen, Germany). Then, using the same instrument, 2 cubic samples measuring 2 mm x 2 mm x 2 mm were cut from the dentin and then analyzed using a confocal Raman microscope (Labran HR800, Horiba Scientific, Palaiseau, France) equipped with a 633 nm laser excitation beam, a 600-line grating, a *100 objective, a 100% transmission filter (220 µW), a 10 s exposure time and an accumulation of 3. Calibration was carried out using order 1 of Silicon at 520.7 ± 1 cm−1. Spectra were obtained over a range from 400 to 4000 cm−1 and analyzed using LabSpec6 software (Horiba Scientific, Palaiseau, France). The baseline corrections were performed to remove the fluorescence background. The sets of peaks constituting dentin corresponded to the PO43−, CO32− and C–C groups; those of the collagen corresponded to Amide I, Amide III, C=O, C–H, N–H and also C–C groups. On the subsequent day, following the preliminary analysis, the samples were subjected to a treatment of cold atmospheric plasma (5 mm, 30 s) and the Raman spectroscopy was launched again. In order to obtain a representative Raman answer, triplicate of micro-spot (1 µm diameter) and macro-spot (50 µm diameter) were performed on each sample.

X-ray photoelectron spectrometry (XPS)

The same protocol of tooth preparation was applied here than for Raman spectroscopy. Two others cubic samples, dimensions 2 mm x 2 mm x 2 mm, were prepared in the dentin and photoelectron emission spectra were recorded using a monochromatized Al Kalpha (hν = 1486.6 eV) source on a XPS Kalpha system (ThermoFisher Scientific, Waltham, MA, USA). The X-ray Spot size was about 400 µm. The Pass energy was fixed at 30 eV with a step of 0.1 eV for core levels and 160 eV for surveys with a step of 1 eV. The spectrometer energy calibration was done using the Au 4f7/2 (83.9 ± 0.1 eV) and Cu 2p3/2 (932.8 ± 0.1 eV) photoelectron lines. XPS spectra were recorded in direct mode N (Ec) and the background signal was removed using the Shirley method. The flood Gun was used to neutralize charge effects on the top surface.

The day after the first analysis, samples were submitted to CAP treatment (5 mm, 30 s) and XPS analysis performed again.

Statistical analyses

Quantitative variables are presented with mean ± standard deviation. Comparison between the groups was performed with an Anova test (for contact angle measurements) under the double hypothesis of values normal distribution (verified with a Shapiro–Wilk test) and variances equality (verified with a Levene’s test). Otherwise, especially for SBS tests, a non-parametric Kruskal–Wallis test was preferred. Post-hoc Tukey tests were used after the Anova to compare groups by pairs, or Mann–Whitney non-parametric tests after the Kruskal–Wallis.

Due to the constraint of sampling with biological tissues, a power calculation for this study was only performed after the experiments, based on the results of Shear Bond Strength tests. With a result of 98.3%, it confirmed that the sampling was sufficient to detect clinically meaningful differences. Database was built on Microsoft Excel® (Microsoft 365, Microsoft Corporation, WA, USA) then the analyses and figures carried out using Stata v.13® (Stata Corp, TX, USA) and GraphPad Prism 5® (GraphPad Software Inc., CA, USA) softwares.