Preparation, characterisation and microbiological examination of Pickering nano-emulsions containing essential oils, and their effect on Streptococcus mutans biofilm treatment

Essential oils (EOs) are commonly applied in mouth care products like mouthwashes, mostly as an ethanolic solution or by usage of surfactants as solubilising agents. In this study, we present a formulation for preparation of Pickering nano-emulsions (PnE) of EOs as a novel form for application of EOs in mouth care. For the preparation of PnE, we have synthesised surface-modified silica nanoparticles with a mean diameter of 20 nm, as well as we have examined the effect of EOs concentration on PnE droplet size and stability. In vitro study of their effect on the Streptococcus mutans biofilm as the main pathogen of dental health problems has been performed. We have found that EOs in the PnE form has the highest effectiveness against biofilm formation. Diffusion through the biofilm model membrane was studied to explain this observation. We have found that PnEs have a better performance in the transportation of EOs trough model membrane than the ethanolic solutions and conventional emulsions (CEs).

The EOs based commercial products that are used for treatment or prevention of dental diseases are containing surfactants, solvents or co-solvents to enhance the water solubility of EOs 7 . Among the solvents, different alcohols (e.g. ethanol, propylene glycol, glycerine) are applied in mouthwash or gargles. Commonly used surfactants are polysorbates (20, 60 or 80) 8 and SLS (Na-lauryl-sulphate) 9 , which are used as emulsion stabilisers or EO solubilising agents. Unfortunately, the surfactants can cause mucous membrane irritation by damaging their cell membranes; additionally, at long-term usage, they can get into the blood circulation and cause other side effects 10 . The alcohols can cause dehydration of the mouth, which makes the mucous membrane more sensitive to infections or other diseases 11 .
Additionally, the application of solvents and surfactants in microbiological tests can lead to misinterpretation of the experimental data, because alcohols and surfactants also have antimicrobial activity 12 that is why the effect of essential oils cannot be unambiguously determined. To avoid the use of solvents and surfactants, yet to provide the availability of EOs on the test or treatment site, Pickering emulsions can be prepared, which are emulsions stabilised with solid particles 13 . For this purpose, we can use biologically inert, non-toxic particles, e.g. cellulose 14 , silica 15 or PLGA 16 particles, which are widely used in pharmaceutical technology.
Pickering emulsions can have the same or better stability than the conventional, surfactant stabilised emulsions interface 17 . The adsorption of the solid particles and surfactants on the liquid-liquid interface is a spontaneous, reversible process. The adsorption energy of solid particles on the liquid-liquid interface is higher than the adsorption energy of surfactants. In some cases, where the adsorption energy of solid particles is extremely high, adsorption process can be considered as irreversible 17 . Because of their stability, Pickering emulsions can be used in pharmaceutical and medical applications.
Beside emulsion stabilisation, the solid particles may interact with biofilm, and targeted EO transportation can also be achieved 18 . The effectiveness of targeted EO delivery depends on the emulsion type, the emulsion stabilisation agent and emulsion droplet size 19,20 . Usually, the biofilm consists of polysaccharide matrix, and it is impermeable for many pharmaceutically active ingredients, that is why the treatment of stiff bacterial colonies is challenging 21 . Because of the hydrophilic properties of Streptococcus mutans biofilm 22 , an O/W type emulsion should be applied. When the emulsion is stabilised by the particles with appropriate hydrophilic/lipophilic surface character, which can adsorbed preferably on the biofilm, targeted delivery can be reached. The emulsion droplet size also plays an important role in the delivery of EOs: the mean pore size of biofilms in the top layer ranges from approximately 1.7-2.7 µm and 0.3-0.4 µm in the bottom layers 23 . Because of the pore size of biofilms we can assume, that if the emulsion droplet size is less than 300 nm, the droplets can penetrate deep into the biofilm matrix.
We aimed to prepare O/W type Pickering emulsions with four EOs, such as cinnamon bark (Cinnamomum verum J. Presl.), clove (Syzygium aromaticum (L.) Merr. and Perry), peppermint (Mentha x piperita L.) and thyme EO (Thymus vulgaris L.) and to examine the influence of the EOs concentration on droplet size and stability of Pickering emulsions. Their antimicrobial activities against Streptococcus mutans have been studied before 24 , moreover several researchers have found that these EOs can act as biofilm inhibitors with other bacterial strains [25][26][27][28] . Their ethanolic solutions and conventional emulsions (CEs) have also been tested to compare the effectiveness of different EO delivery forms.
We have also examined the diffusion properties of different EO forms through model membrane, to explain the difference in their antibacterial or inhibitory effects. In the case of cinnamon EO we have prepared and examined PnE with SNPs that have different hydrophilic/lipophilic surface character to determine its influence on the antibacterial activity and diffusion properties as model membrane agar gel was used, which is a suitable Streptococcus mutans biofilm model because they have similar hydrophilic properties and tortuous pore structure 29 .

Results and Discussion
Synthesis and characterisation of silica nanoparticles. Hydrophilic silica nanoparticles (HS) were synthesised by a method established by Stöber, Fink and Bohn, the mean diameter was 20 nm, and the PDI was 0.008, determined by DLS. The stability of the PnEs can be influenced by the hydrophilic/lipophilic surface properties of silica nanoparticles 30 . In our previous work 31 , we have ascertained that the PnEs with appropriate stability and tailored size can be prepared with silica nanoparticles, which were partially surface modified with ethyl groups. For this reason, we have modified the surface of HS by ethyl functional groups with a theoretical surface coverage of 20% (20ET) and 40% (40ET).
The TEM examinations showed that the mean diameter of silica samples was mean 20 nm; they are highly monodispersed, nearly spherical and have a smooth surface (see Fig. 1). In the case of HS high negative zeta potential is expected, and the measurements confirm it. High number of free silanol groups at the surface causes the negative zeta potential, as their pK value is approximatively 4.5, which means that some of them are dissociated in the water suspension. The surface modification will decrease the number of free silanol groups on the surface, which should cause the decrease of the zeta potential under the same conditions. The zeta potential of HS suspended in water was ζ = −116 mV. After surface modification the zeta potential values decreased with increasing surface coverage. The values were ζ = −79 mV for 20ET and ζ = −63 mV for 40ET.
The surface modification of SNPs was examined with FT IR spectroscopy. No significant difference in the position of the vibrational frequency for the SiO 2 functional groups was observed (see Fig. 2). The surface modification of HS with ethyl functional groups caused the decrease of the number of Si-OH groups on the surface of SNPs, hence intensity of peaks belonging to ν as Si-O-Si (1105 cm −1 ), ν defrom Si-OH (1395 and 1645 cm −1 ) and ν strech Si-OH (3095-3685 cm −1 ) decreased, while the intensity of peaks of ν s Si-O-C increased (815 cm −1 ). New peak attributed to the ν-CH 2 -have appeared for the surface modified samples. Their vibrational frequencies are the same in both cases, 2855 and 2930 cm −1 , but intensities are higher for the sample 40ET. The results clearly GC-MS and GC-FID analysis of essential oils. The exact composition of EOs was determined with gas chromatography. The components were identified by comparing their retention times and relative retention factors with standards and oils of known composition. Two parallel measurements have been performed. The main components are the follows eugenol 78.64% in clove EO, cinnamaldehyde 63.77% in cinnamon bark EO, menthol 50.4% and menthon 19.8% in peppermint EO, thymol 39.88% and p-cymene 19.2% in thyme EO. In Table 1. other major components can be seen. We did not indicate or identify the compounds which were present significantly under 1%; this is the reason why the totals are not 100%.

preparation and characterisation of pickering nano-emulsions. The maximum concentration of
EOs was set to MIC value for all examined emulsions. The MIC values of pure EOs in ethanolic solutions were previously determined against Streptococcus mutans with broth macrodilution test (see Part 3.4), for different EOs these were: clove EO 1.02 g/L, cinnamon EO 0.80 g/L, peppermint EO 1.96 g/L, thyme EO 0.40 g/L. We have prepared PnEs with HS, 20ET or 40ET stabilising agents, for CEs Tween80 surfactant was used, the concentrations of stabilising agents were 1 g/L for all experiments. The emulsions were stored at room temperature; t = 25 °C. Each experiment was made in triplicates. Stability of emulsions was determined from periodical droplet size determination with DLS (see Table 2.). The emulsions were considered to stable when the droplet size did not change within 24 hours, and creaming, sedimentation or disproportionation did not occur. Because of the large number of experimental data, we did not indicate data for the standard deviations of droplet size in Table 2. All values were in 1.2-8.3% range.
Because the volume fraction of EO (Θ o ) was under 0.01 in every case, we could assume, that all PnEs and CEs were O/W type emulsions. We have performed filter paper tests with CoCl 2 and dye test with Sudan red G to confirm this assumption 29 .
The results confirmed that the PnEs could have same or better stability than CEs (see Table 2), because of the high adsorption energy of solid nanoparticles on the liquid-liquid interface. The most stable emulsions are the thyme EO containing ones, in this case the stability of PnEs was 4 months, while its CEs were stable for only 1 month.    www.nature.com/scientificreports www.nature.com/scientificreports/ Biofilm formation and treatment. The EOs concentration was MIC/2 value in all examined emulsions or solutions, which is a standard concentration for biofilm inhibition tests 32 . Our results of the crystal violet assay showed that the absolute ethanol and Tween80 solution have antibacterial effect, and they reduced the biofilm mass, while the HS, 20ET and 40ET nanoparticle suspensions had no antibacterial effect, and they did not reduce the biomass significantly.
Even so, the PnEs prepared with 20ET were the most effective forms in biofilm inhibition. For each EOs, the ethanolic solutions and CEs showed less biofilm inhibition (see Fig. 3.); e.g. in the case of thyme EO, the inhibitory rates (IR) were 26.9, 47.4 and 72.1% for ethanolic solution, CE and PnE respectively.
We have found cinnamon EO have the highest inhibition effect among tested EOs. The inhibitory rates for its different forms are 59.7 for ethanolic solution, 69.0% for CE and 85.9% for PnE respectively. In this case we have performed tests on the influence of the hydrophilic/lipophilic surface properties of stabilising SNPs on the inhibitory effect on PnE. HS and 40ET nanoparticles stabilised PnEs were also tested in biofilm inhibition. The results showed (Fig. 4), that in the case of 20ET nanoparticles stabilised PnE had better biofilm inhibition effect (IR 85.9%) than 40ET (IR 81.5%), and HS stabilised ones (IR 69.4%).
In Fig. 5 we can see the SEM images of untreated and treated biofilms. Figure 5A,B are the images of untreated control biofilms. The in vitro adherence of Streptococcus mutans colonies can be clearly seen, and thick coherent biofilm has formed on the surface. The SEM images bear out the results of biofilm inhibition experiments. The PnE of peppermint EO reduced the biofilm mass (Fig. 5C) minimally, while in the images of biofilms treated with PnE of clove (Fig. 5D) and thyme EO (Fig. 5E) we can see bacterial colonies on the surface, but the adherence of coherent biofilm was reduced.
The SEM images clearly show, high inhibitory effect of the PnEs of cinnamon EO. On the images of biofilms treated with HS (Fig. F) and 40ET (Fig. H) stabilized PnEs of cinnamon EO we can see some coherent biofilm spots, while on the image of biofilm treated with 20ET stabilized PnE of cinnamon EO (Fig. G), no coherent biofilm formation can be observed.
In all the cases, a very good correlation of the inhibitory rate obtained by the crystal violet assay and the data obtained from SEM images can be found.
In vitro diffusion studies. After the discussion on the results obtained from biofilm inhibition tests, an assumption has been made that there should be a correlation between inhibitory rate and diffusion properties   www.nature.com/scientificreports www.nature.com/scientificreports/ emulsions was similar (see Table 3), and we could assume that the diffusion properties depend only on the type of the emulsion or surface properties of emulsion stabilising agent. The diffusion profiles of different EOs are very similar; this is why we have graphically presented only the in vitro diffusion study curve of cinnamon EO as it has shown the highest inhibitory rate (see Fig. 6). The results of cumulative amount after 6 hours for all samples can be seen in Table 3. The cumulative amount of EO means the diffused EO amount after 6 hours. The diffusion curves of all the other samples can be seen in Supplementary Information. Generally, we can conclude that the cumulative EO amounts were highest for PnEs. In the case of cinnamon EO these values are: 51.4% for 20ET stabilised PnE, 30.7% for CE and 10.5% for ethanolic solution respectively. At thyme EO, cumulative EO amount for ethanolic solution was under the limit of detection value.

conclusions.
We have successfully prepared O/W type PnEs of EOs with droplet size tailored for diffusion trough biofilm porous structure (210-370 nm). For the stabilisation of PnEs, we have synthesised and surface-modified spherical silica nanoparticles. These emulsions remained stable for at least 1 week, in some cases more, then 4 months.
We have determined MIC values against Streptococcus mutans for all examined EOs, and we have performed a biofilm inhibition tests applying EOs in MIC/2 amount. We have tested tree forms for all examined EOs, ethanolic solution, CE and PnE. The test results shown that the PnE form has highest antimicrobial effectiveness compared to CE or ethanolic solutions for all examined EOs. We assume that the difference in effect should correlate with the penetration of EOs in different form trough polysaccharide biofilm structure. To confirm this idea we have studied the diffusion trough model membrane (agar gel). The results have shown that for all samples, there is a direct relation of antimicrobial or inhibitory effect with diffusion properties of EOs in different form. The PnE form had the highest cumulative amount for all EOs, and ethanolic solution had the lowest cumulative amount. This result is an explanation why the PnE have antibacterial effect while ethanolic solution with the same EO concentration have inhibitory effect, which is clearly seen on the SEM images. The highest difference in the inhibitory   www.nature.com/scientificreports www.nature.com/scientificreports/ rate of different forms has been observed in the case of thyme oil. The thyme oil has the highest antimicrobial effectiveness in comparison with other examined EOs as its MIC value is lowest. The ethanolic solution did not provide a measurable amount of thyme EO trough model membrane, and it had a very low inhibitory rate (27%). The CE form had a cumulative amount of 9%, and the inhibition rate has increased to 47%, while the PnE form had 19% of cumulative amount, and the inhibition rate increased to 72%. These results indicate that the delivered amount of an EO with high antimicrobial effectiveness to the biofilm has the greatest effect on the inhibition activity. Beside the droplet size, the appropriate hydrophilic/lipophilic surface character is very important for effective delivery of EO to biofilm. We have found that PnEs stabilised with silica that has a moderate hydrophobic character (20ET), shows the highest delivery of EOs to biofilm that is accompanied with highest inhibitory effect.

Essential oil c oil g/l Formula
We can conclude that O/W type PnEs stabilised by nanoparticles with appropriate hydrophilic/hydrophobic surface properties, provide a new possibility for the application of EOs in pharmaceutical treatment against Streptococcus mutans biofilm formation. Preparation and characterisation of O/W type PnEs. The concentration of emulsifiers was set to 1 mg/ml and was kept constant for all experiments. The influence of EOs concentration on the emulsion droplet size was examined; it was varied until the minimum inhibitory concentration against Streptococcus mutans (see Table 2.) The exact emulsification process, the droplet size, and stability measurements can be seen in the Supplementary Information. Preparation of the biofilm samples for Scanning Electron Microscopy. The biofilm was imaged by SEM (JEOL JSM-6300) as previously described 37 with some modification (briefly see Supplementary  Information). To determine the effect of EOs and different formulations, we have used control samples that were treated with ethanol and Tween80 solutions as well with the suspension of HS, 20ET or 40ET nanoparticles. Their concentration was the same as it was in the ethanolic solution, CEs, and PnEs 38 .