Feasibility of Using H3PO4/H2O2 in the Synthesis of Antimicrobial TiO2 Nanoporous Surfaces

Ti6Al4V alloys are the primary materials used for clinical bone regeneration and restoration; however, they are substantially susceptible to biomaterial-related infections. Therefore, in the present work, we applied a controllable and stable oxidative nanopatterning strategy by applying H3PO4, a weaker dissociating acid, as a substitute for H2SO4 in the classical piranha reaction. The results suggest that our method acted as a concomitant platform to develop reproducible diameter-controlled TiO2 nanopores (NPs). Interestingly, our procedure illustrated stable temperature reactions without exothermic responses since the addition of mixture preparation to the nanopatterning reactions. The reactions were carried out for 30 min (NP14), 1 h (NP7), and 2 h (NP36), suggesting the formation of a thin nanopore layer as observed by Raman spectroscopy. Moreover, the antimicrobial activity revealed that NP7 could disrupt active microbial colonization for 2 h and 6 h. The phenotype configuration strikingly showed that NP7 does not alter the cell morphology, thus proposing a disruptive adhesion pathway instead of cellular lysis. Furthermore, preliminary assays suggested an early promoted osteoblasts viability in comparison to the control material. Our work opens a new path for the rationale design of nanobiomaterials with “intelligent surfaces” capable of decreasing microbial adhesion, increasing osteoblast viability, and being scalable for industrial transfer.


Introduction
Titanium (Ti) and its alloy (Ti6Al4V) are the main biocompatible metallic options currently used to promote bone formation and restoration [1]. However, contamination by microbial adhesion can negatively compromise Ti effectiveness and clinical success. Promising strategies have been reported to generate nanopatterned surfaces to control microbial adhesion. For example, roughness in texture, deposition of antimicrobial nanocoatings, and nanoscale tuned surfaces [2,3]. Of particular interest, the fabrication of controlled-sized TiO 2 NPs has emerged as a current trend for controlling microbial adhesion and colonization [4]. Interestingly, chemical oxidative nanopatterning has proven to be a versatile strategy for the development of controlled NP tuned surfaces [5]. us far, the H 2 SO 4 /H 2 O 2 system (piranha solution) is the optimal etching/oxidative protocol for the generation of reproducible NPs and the activation of metallic surfaces. Nonetheless, the mixture with strong acids exacerbates an extreme exothermic reaction that acts violently after organic matter contact (at low concentrations) and on metallic surfaces [6]. Consequently, it is substantial to develop a stable and nonexothermic etching solution for chemical nanopatterning capable of producing NPs on Ti6Al4V surfaces. us, by applying H 3 PO 4 , a weaker dissociation acid of pKa lower than H 2 SO 4 [7,8], we could reduce the exothermic reactivity on Ti6Al4V without disturbing the formation of reproducible NPs. erefore, our work aims to synthesize diameter-controlled and reproducible NPs on Ti6Al4V to reduce microbial adhesion and promote early osteoblast growth using an oxidative nanopatterning procedure with H 3 PO 4 /H 2 O 2 . is strategy is an excellent alternative to the piranha solution since it is more stable, ecofriendly, and nonexothermic suspension without requiring heating conditions.

Surface Physicochemical
Characterization. e surface morphology was analyzed using field-emission scanning electron microscopy (FE-SEM; Tescan LYRA 3) at a 20 kV accelerating voltage with a secondary electron detector. e NP distribution was generated from 50 NPs randomly measured from a FE-SEM micrograph. Energy dispersive X-ray spectroscopy (EDX, Brucker XFlash) coupled to the FE-SEM was used for the chemical analysis. Raman spectroscopy (Raman Station 400F Perkin-Elmer) was applied at RT using a 785 nm diode laser beam at a power of 15 mW. e water contact angle (WCA) was quantified using an automated tensiometer ( eta Attension; Biolin Scientific), placing a 5 µL droplet of deionized water at RT and 45% relative humidity.

Microbial Characterization.
We prepared fresh overnight grown cultures of Staphylococcus aureus (S. aureus, ATCC 25923), Escherichia coli (E. coli, ATCC 25922), and an isolated C. albicans strain as previously described [9]. e active fungal suspension was adjusted to 2 × 10 4 CFU/mL with Sabouraud dextrose (SD) broth. en, 50 µL of the working C. albicans were cultured over the surfaces, which were individually placed in a 12-well plate (Corning, USA). Similarly, the S. aureus and E. coli inoculums were tailored to 1 × 10 7 CFU/mL using tryptic soy (TS) broth and cultured. e materials were incubated for 2 h and 6 h (defined as initial and late adhesion, respectively) at 37°C in static conditions, washed thrice with 1 × phosphate-buffered saline (PBS) for 5 min and ultrasonicated in 2 mL of SD or TS broth [10]. e remaining suspensions were serially diluted and cultured in SD (C. albicans) or TS (bacterial cells) agar for 24 h at 37°C.

FE-SEM Microbial Analysis.
For FE-SEM analysis, each material was rinsed thrice with warm PBS, fixed in 3% glutaraldehyde (Sigma-Aldrich, USA) at 4°C overnight, rinsed thrice with PBS, and postfixed with 3% glutaraldehyde for 2 h at RT. e samples were dehydrated in a graded series of ethanol for 2 h and placed into a desiccator until the analysis.
2.5. Cytotoxicity Assessment Using MTT. In order to analyze the cytotoxicity of the experimental surfaces, we applied the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) viability assay [11]. We used MG-63 human osteoblast-like cells (ATCC CRL-1427). Prior to cell culture, each experimental material was placed in an individual well of a 12well polystyrene plate (Corning, USA). e initial cell density was 1 × 10 4 cells/surface in passage three. ey were harvested and cultured in complete medium constituted of Dulbecco's modified Eagle's medium (DMEM, ermo Fisher Scientific, USA) supplemented with 10% heat-inactivated fetal bovine serum ( ermo Fisher Scientific, USA) and 100 units/mL of penicillin-streptomycin ( ermo Fisher Scientific, USA) at 37°C in a humidified 5% CO 2 incubator for 24 h. Afterward, the cells were washed thrice with warm PBS. 2 mL of MTT (Sigma-Aldrich, USA) in DMEM (5 mg/mL) was added into each well and further incubated at 37°C in a humidified 5% CO 2 incubator for 3 h. e resulting formazan crystals were dissolved after discarding the medium containing MTT and transferring the 12-well plate into an orbital shaker at 200 rpm, 37°C with dimethyl sulfoxide (Sigma-Aldrich, USA) for 20 min. en, the dissolved crystals were deposited into a 96-well polystyrene plate (Sigma-Aldrich, USA), and the optical density (O.D.) was recorded at 590 nm using a microplate reader ( ermoskan, ermo Fisher Scientific, USA).

Statistical Analysis.
Numerical data of three independent studies performed each in triplicate were assessed by one-way analysis of variance followed by Tukey's multiple comparison test using GraphPad Prism 7. A P < 0.05 was considered statistically significant. Figure 1(a) illustrates diameter-controlled NPs on Ti6Al4V surfaces after the nanopatterning protocol. Moreover, the highzoom revealed the formation of homogeneous and ordered nanostructures for each reaction. Figure 1(b) represents NPs of ≈7 nm (1 h), ≈14 nm (30 min), and ≈36 nm (2 h), indicating that time could be the predominant thermodynamic parameter for NP diameter control at room temperature using the H 3 PO 4 system. e EDX showed the materials' elemental values (Figure 1(c)), highlighting that phosphorous (P) was not incorporated as a doping complexing element in any treatment. However, EDX is a technology that enables chemical profiling of the X-ray photons generated from the beamed electrons of the deeper layers of the surface [12]. e X-ray photoelectron spectroscopy (XPS) is recommended for clarifying this interesting trend. It is important to note that phosphate coatings could generate electrostatic interactions that might be favorable for promoting bacterial adhesion [13]. Furthermore, the low carbon levels suggest the absence of any organic pollutants ( Figure 1(c)). us, the lower carbon level and the nanostructured distribution detected on the NPs, especially on NP7, could be attributed to the reduced WCA (Figure 1(d)). e reactions achieved Ti 4+ by an initial oxidation, followed by the dissolution of the oxide layer, and the nucleation of a thin oxide layer.

Results and Discussion
us, decomposing H 2 O 2 into O 2 and H + generates nanodefects [14].  [16]. Similarly, the authors detected the presence of Ti 2p 3/2 and O1 s signals that have been ascribed for the Ti-O bond gap, thus proposing the formation of a consistent thick layer mainly of TiO 2 . Interestingly, the Raman analysis (Figure 2) suggests that a thinner TiO 2 layer could be generated after the nanopatterning process, as there was no bandgap between 800 and 200 nm corresponding to amorphous TiO 2 [17,18]. Although the presence of NPs was detected for all the experimental materials, the EDX results also supported this interesting finding, as no oxygen levels were detected. More surface chemistry analyses are recommended in order to support those interesting findings. On the other hand, the study by Pisarek reported that the 24 h treatment phase resulted in the deposition of phosphate ions [16], and far more critical is the fact of possible corrosion detriments by the extensive reaction period. Although this oxidative strategy was applied in previous studies, the role of    generating new single nanocavities and pits growing together increasing the diameter. Moreover, it has been described that oxidative etching procedures favor the attack of ß-grains over the α-grains, therefore altering the surface microtexture over time [1,19] [22]. e authors also advocated that H 3 PO 4 associates with H 2 O 2 via H-bonding to form a stabilized complex, which may inhibit the H 2 O 2 reduction and decrease the enthalpy required to conduct an exothermic reaction, as observed here, thus far proposing that H 3 PO 4 is safer and more manageable than Medical implant contamination is a paramount concern that negatively compromises the biomaterials' "gold success": achieving complete clinical healing and restoration [23], thus highlighting that nanotextured surfaces are an important and acceptable strategy to reduce microbial adhesion. Our results suggested that smaller NPs (NP7) could avoid the S. aureus adhesion for each growing phase (Figure 3). Interestingly, similar outcomes of bacterial colonization were detected for the NP14, NP36, and control. Moreover, we can focalize that SEM micrographs illustrated an analogous spherical phenotype commonly observed for coccus bacteria (Figure 3, insets). Furthermore, the micrographs suggest that higher biofilm colonization agrees  with the increased viability detected. A similar behavior was detected for the E. coli model evaluated on the experimental materials ( Figure 4). e E. coli growing ability was reduced on the NP7 material in comparison with the study surfaces. Of particular interest is the fact of similar bacterial morphology conducted by the Ti6Al4V alloy and the NP7. e high-zoom micrographs (Figure 4, insets) clearly show that bacilli configuration is present in plenty on the evaluated surfaces. However, from the low-zoom micrographs, we can highlight that the NP7 reduced the bacterial adhesion, as mainly detected at the late adhesion phase. Importantly, the NP7 supported antifungal behavior ( Figure 5), following comparable results to those of S. aureus and E. coli. In the early adhesion, we detected that the control caused enlarged cell alterations, which were further transformed into hyphae and pseudohyphae morphologies. Meanwhile, NP7 conserved a downregulated proliferating phenotype. Principally, we discovered a substantially growing fungal viability for the larger NPs, further proposing that smaller NPs can present detrimental fungal outcomes, as previously reported [24]. Notably, these results suggest that the smaller NPs could disrupt the formation of nanoscale bonds required to conduct a proper microbial adhesion, in accordance with previous works [9,25]. In particular, the significant hydrophilicity on NP7 may positively influence the reduced electrostatic interactions for microbial bonding [26] and the amorphous nature of the TiO 2 thin coating [27]. However, we recommend more physicochemical studies to explain the current antimicrobial results.
It is well known that the modification of a conventional surface material to its nanostructured counterpart can alter the cellular activity [2,28,29]; a critical result presented from this work is the finding that nanoporous surfaces developed by etching with an H 3 PO 4 /H 2 O 2 mixture can improve osteoblast activity. In Figure 6, it is presented the osteoblast viability on the experimental materials after 24 h of culture. e results suggested that the NP7 and NP36 nanostructured surfaces promoted higher cellular proliferation in contrast to the control alloy and the NP14. It is essential to consider that MTT assays take advantage of mitochondrial activity, pointing toward the fact that a higher quantification of resulting formazan crystals is directly proportional to a healthy and active osteoblast growing population [30,31]. erefore, we can hypothesize that nanoporous distribution may play a more critical role in promoting early osteoblast proliferation instead of the NP size. is information can be in part supported by the fact that NP7 and NP36 share a more distributed porous size than NP14. On the other hand, NP7 showed higher osteoblast activity and more hydrophilicity, indicating improved surface energy and enhanced early cellular growth. is current trend opens the concept that a high surface-area-to-volume ratio and improved surface energy can establish a stimulating microenvironment that can accelerate the bone-growing functionality, as depicted in previous studies of different size-controlled nanostructured coatings [32][33][34][35].

Conclusions
Our results established an oxidative nanopatterning protocol using H 3 PO 4 /H 2 O 2 as a feasible, safer, and controllable system for developing homogeneous nanotextured surfaces on Ti6Al4V. e H 3 PO 4 /H 2 O 2 resulted in a stable mixture that did not show violent exothermic reactions during the preparation of the solutions and the alloy surface modification. Inherently, our protocol resulted in NPs of 7, 14, and 36 nm outlining the reaction time as the main variable for diameter control under the studied synthetic conditions. Importantly, we have demonstrated that smaller NPs can reduce the early adhesion of S. aureus, E. coli, and more strikingly, C. albicans. us, more attractively, NP7 tailored a long-lasting antimicrobial action for 6 h of incubation, particularly with the absence of cellular phenotype alterations. On the other hand, the cytotoxicity analysis suggested that the surfaces might not disrupt the initial osteoblasts' proliferation, tailoring the nanostructures as a stable surface for osteoactive conditions. Our work opens a new path for the rationale design of nanobiomaterials with "intelligent surfaces" capable of decreasing microbial adhesion and being scalable for industrial transfer.

Data Availability
All data generated or analyzed in this study are included in this work.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
EBP and BVS contributed to conceptualization, funding acquisition, writing of the original draft, review, and editing.