*3.2. Surface Characterization of Modified and Control Surfaces*

Regarding surface hydrophobicity of the transferred surfaces, CA measurements from plates modified with the different topographies showed higher values than the control surface. The highest CA was found for the experimental surface modified using the *C. esculenta* leaf model, followed by *S. molesta y C. aurea,* respectively. The difference between control and modified surfaces was statistically significant (*p* < 0.001). In addition, the difference in CA between the experimental surface modified with the *C. esculenta* leaf (highest CA value) and *C. aurea* (lowest CA value) was statistically significant (*p* = 0.004, Figure 4).

**Figure 4.** Contact angle measurements from control (SS 316L) and transferred (experimental) surfaces.

When comparing surface roughness from modified surfaces versus SS 316L control, the former showed higher *R*a values, and the plates modified using the *C. aurea* leaf showed the highest values, followed by *C. esculenta* and SS 316L control. The difference between modified surfaces and SS 316L control was statistically significant (*v* < 0.001, Figure 5). In addition, the Ra values from the plates modified using the *C. aurea* leaf were lower than the respective natural leaf. In contrast, the Ra values from the plates modified using the *C. esculenta* leaf were higher than the respective natural leaf (data not shown). Such comparison could not be made for *S. molesta* since Ra values could not be obtained for the natural leaf as explained above.

**Figure 5.** Average roughness of control (SS316L) and transferred (experimental) surfaces.

#### *3.3. Evaluation of S. mutans Adhesion*

The values from the plates modified using the *C. aurea* leaf showed the lowest adhesion (1.3 × 106 ± 2.5 × 105 CFU/surface) when compared to control surface (1.9 × <sup>10</sup><sup>6</sup> ± 2.5 × 105 CFU/surface) and the other two experimental surfaces (1.9 × 106 ± 2.4 × 105 CFU/surface and 2.3 × 106 ± 5.0 × 105 CFU/surface for *C. esculenta* and *S. molesta*, respectively). The difference was statistically significant (*p* < 0.001). The modified surface using the *C. esculenta* leaf showed lower adhesion than the plate modified with the *S. molesta* topography. However, the values from *C. esculenta* modification and SS 316L were similar and the difference was not statistically significant (Figure 6). In addition, when comparing

the *S. mutans* adhesion to the plates modified using the *S. molesta* topography versus the other treatments, including the control surface, higher values were found

**Figure 6.** *S. mutans* adhesion to control (SS 316L) and transferred (experimental) surfaces.

#### **4. Discussion**

In recent years, different papers have addressed the subject of how topographic modifications on the surface of biomaterials may assist in reducing bacterial adhesion and biofilm formation [27–29]. The current investigation assessed *S. mutans* adhesion to the surface of surgical-grade stainless steel plates that were modified using different topographies following biomimetic inspiration. Natural patterns were selected based on their ability to self-clean and the apparent high hydrophobicity exhibited in their natural environment (water-repellency).

The topography of the natural leaves used as models was assessed and relevant properties were measured. They showed high hydrophobicity (130.4◦ for *C. esculenta*, 140.5◦ for *C. aurea* and 150.3◦ for *S. molesta,* on average). According to hydrophobicity values presented by Kim and Choi [36] and Falde et al. [37], only the *S. molesta* leaf could be classified as superhydrophobic (CA > 150◦), while *C. esculenta* and *C. aurea* were classified as hydrophobic (CA between 90◦ and 150◦). Jaggessar et al. [38] reported a contact angle between 90◦ and 150◦ for *C. esculenta*, which is in agreement with the values obtained in the present work. For comparison purposes, no values could be found for *C. aurea* and *S. molesta* in the scientific literature.

Then, a comparison between the hydrophobicity values from the natural surfaces and the values from the transferred surfaces was performed and a reduction in the contact angle measurement after the transference was found in all cases. Even though there was a reduction in hydrophobicity values, they were still higher than SS 316L. This finding may have different explanations. Biological surfaces, including natural leaves, may have protective coatings made of natural waxes that increase the hydrophobicity and such coatings could not be transferred to stainless steel plates using soft lithography [36–41]. In addition, silica sol-gel was used to transfer the topography from each leaf to polished SS 316L. Several authors [42–44] have found that when silica sol synthesized with similar TEOS:MTES ratios as the ones in the current investigation is used to coat SS surfaces, an increase in hydrophobicity is found due to the presence of methyl groups from the silica that reduce the ability of the surface to absorb water [44]. Therefore, the presence of silica may explain why the hydrophobicity values from the transferred surfaces are higher than the value from polished SS 316L. However, this effect of silica on stainless steel is not as strong as the effect that a protective wax coating has in the natural plants and leaves, hence, the values from natural sources are significantly higher than those from the transferred surfaces, which are, in turn, higher than silica-free SS 316L.

Regarding roughness, the values from the *S. molesta* leaf could not be obtained using AFM due to the topography of this natural surface consisting of multiple macrometric hair-like structures that prevented the tip from making close contact with the surface, which is necessary to obtain a correct reading. The topography of a given sample is one of the limitations exhibited by AFM, especially when the sample has steeply inclined surfaces [45], such as the topographic features shown by *S. molesta,* even at higher scales. The roughness of the remaining natural leaves (*C. aurea* and *C. esculenta*) was similar, but when comparing the roughness from the natural leaves and the transferred surfaces, an increase in roughness was observed for *C. esculenta* and a reduction was found for *C. aurea*. Such reduction in roughness may be explained by the fact that biological surfaces could have a hierarchical structure and intricate architecture [40,46] and the transfer process employed in this protocol used smooth silica sol, which may have filled some of the irregularities on the surface, hence the reduction that was observed. An explanation for the increase in roughness in the transfer of *C. esculenta* remains to be elucidated.

As for bacterial adhesion, surface roughness and hydrophobicity play a major role in how bacterial species adhere to a surface and form a biofilm. De la Pinta et al. [47] found that more abundant biofilm was formed on rougher surfaces, but such results could not be correlated when hydrophobicity was considered. Raspor et al. [48] assessed bacterial adhesion to five SS 304 surfaces to determine the influence of roughness on adhesion. They found that adhesion increases as roughness increases. Bohinc et al. [49] also obtained similar results on glass. Díaz et al. [50,51] demonstrated that roughness at the nano scale reduces bacterial adhesion, while roughness at the micro scale increases it. Xu et al. [52] confirmed such findings and explained that such reduction at the nano scale is due to the narrow space available for the bacterium and the obstacle that this represents for bacterial aggregation, which had been previously established by Hochbaum and Aizenberg [27]. These findings are conflicting with the results of the current work, since the rougher topography (*C. aurea*) showed the lower bacterial adhesion. This may be explained by the fact that, even though both the modified surface using the *C. aurea* model and the actual leaf showed higher roughness values, its apparently more organized topography and/or the size of its surface features related to the size of the bacterial species used [27] were responsible for the reduction in bacterial adhesion.

When analyzing the relation between hydrophobicity and bacterial adhesion, the higher the CA value, the lower the *S. mutans* adhesion. The most hydrophobic surface (*S. molesta*) showed an increase in bacterial adhesion, while the less hydrophobic surfaces (*C. aurea* and *C. esculenta*) exhibited lower adhesion of this bacterial species. Since it was not possible to obtain the roughness value from *S. molesta*, it is not possible to determine that the increase in bacterial adhesion is solely ascribable to its hydrophobicity, but a combination of high hydrophobicity and an apparent high roughness. It is important to notice that the difference in the relation between hydrophobicity and bacterial adhesion between the current investigation and the work by De la Pinta et al. [47] may be due to the bacterial species under evaluation, since more hydrophobic species prefer more hydrophobic surfaces. In the current work, *S. mutans* showed lower adhesion to more hydrophobic surfaces, which is in agreement with the results by Satou et al. [53], who demonstrated that *S. mutans* is a hydrophilic bacterial species that show more affinity to hydrophilic surfaces.

Adhesion to topographically modified surfaces has been addressed in the literature for a few years. Vladillo-Rodriguez et al. [54] created engineered nano surfaces and assessed bacterial adhesion. They concluded that different surface patterns caused reduction of bacterial adhesion ranging from 40% to over 95%. Bhardwaj and Webster [55] modified titanium substrates and found a 95% reduction in *Staphylococcus aureus* adhesion, a 90% reduction in *Pseudomonas aeruginosa* adhesion and a 81% reduction in *Escherichia coli* adhesion.

When modifications were based on natural models (biomimetics), Carman et al. [56] used a surface based on the sharkskin, known as Sharklet, and found a reduction in the aggregation of spores from green algae. May et al. [28], Chung et al. [29] and Reddy et al. [57]

found a reduction in bacterial adhesion to modified surfaces, using patterns from the Sharklet model, on different materials. Bixler et al. [24] evaluated anti-fouling properties of microstructures based on butterfly wings and rice leaves and obtained promising results. However, most works on surface modification using bio-inspired or biomimetic approaches are based on surfaces obtained from animal models, while plants offer many possibilities that need to be evaluated. Previous works using surface modification based on a natural leaf (*C. esculenta*) have also demonstrated a reduction in bacterial adhesion to SS and titanium surfaces [30,34]. Since *C. aurea* showed better results than *C. esculenta* for reduction in bacterial adhesion in the current work, future investigations using botanical materials as models to modify biomaterials surfaces must be continued.
