Antibacterial Effect of Stainless Steel Surfaces Treated with a Nanotechnological Coating Approved for Food Contact

Stainless steel, widely present in the food industry, is frequently exposed to bacterial colonization with possible consequences on consumers’ health. 288 stainless steel disks with different roughness (0.25, 0.5 and 1 μm) were challenged with four Gram-negative (Escherichia coli ATCC 25922, Salmonella typhimurium ATCC 1402, Yersinia enterocolitica ATCC 9610 and Pseudomonas aeruginosa ATCC 27588) and four Gram-positive bacteria (Staphylococcus aureus ATCC 6538, Enterococcus faecalis ATCC 29212, Bacillus cereus ATCC 14579 and Listeria monocytogenes NCTT 10888) and underwent three different sanitizing treatments (UVC, alcohol 70% v/v and Gold lotion). Moreover, the same procedure was carried out onto the same surfaces after a nanotechnological surface coating (nanoXHAM® D). A significant bactericidal effect was exerted by all of the sanitizing treatments against all bacterial strains regardless of roughness and surface coating. The nanoXHAM® D coating itself induced an overall bactericidal effect as well as in synergy with all sanitizing treatments regardless of roughness. Stainless steel surface roughness is poorly correlated with bacterial adhesion and only sanitizing treatments can exert significant bactericidal effects. Most of sanitizing treatments are toxic and corrosive causing the onset of crevices that are able to facilitate bacterial nesting and growth. This nanotechnological coating can reduce surface adhesion with consequent reduction of bacterial adhesion, nesting, and growth.


Introduction
Stainless steel is widely present in the food industry as the main component of work surfaces, cookware, pasteurizers, homogenizers, separators, decanters, mixing, process and storage tanks, fittings, valves, pumps, and pipework [1][2][3]. Based on the high resistance to corrosion by acidic or sulfur dioxide-containing foods and cleanability, AISI 316 or 316 L stainless steel should be preferred to AISI 302 or AISI 304 [2,4]. Although naturally passivated by air or other oxidizers, stainless steel needs to undergo additional surface treatments (pickling, electropolishing, and mechanical cleaning) to improve its strength [5].
values named R 0.25, R 0.5, and R 1, respectively. After microbiological and microscopic analyses all disks were covered with a surface treatment named nanoXHAM ® D.
The disks were then equally divided into three groups of 96 each by roughness average values named R 0.25, R 0.5, and R 1, respectively. After microbiological and microscopic analyses all disks were covered with a surface treatment named nanoXHAM ® D.

NanoXHAM Coating
NanoXHAM ® D [Moma Nanotech s.r.l, Brugherio (MB), Italy] [30] is a coating compliant with regulation 1935/2004/CE and National Sanitation Foundation (NSF) standard 51, and therefore suitable for contact with food products (Figure 1). Moreover, it is also certified as not cytotoxic according to the ISO 10993-5:2009 [30]. It is a transparent thin film of amorphous SiOxCyHz, deposited via PECVD at room temperature [31]. The thickness is about 1 μm, perfectly flexible [26], adherent to the substrate, corrosion, and wearresistant. The surface of this nanotechnological coating is hydrophobic and the surface tension is about 28 mN/m according to ISO 8296-2003.

FT-IR Spectra
To compare the surface modification introduced by plasma deposition, FT-IR spectra of treated and untreated samples were acquired by means of a Thermo Avatar 370 spectrometer (Thermo Nicolet Corp., USA), equipped with an attenuated total reflection (ATR) accessory. A diamond crystal was used as an internal reflectance element on the ATR accessory. Spectra were recorded at room temperature, in the range from 4000 to 650 cm −1 , acquiring 32 scans per set data of 4 cm −1 resolution. Two spectra were recorded for each sample.

FT-IR Spectra
To compare the surface modification introduced by plasma deposition, FT-IR spectra of treated and untreated samples were acquired by means of a Thermo Avatar 370 spectrometer (Thermo Nicolet Corp., Madison, WI, USA), equipped with an attenuated total reflection (ATR) accessory. A diamond crystal was used as an internal reflectance element on the ATR accessory. Spectra were recorded at room temperature, in the range from 4000 to 650 cm −1 , acquiring 32 scans per set data of 4 cm −1 resolution. Two spectra were recorded for each sample. All strains were grown in tryptic soy broth (TSB, bioMérieux, Florence, Italy), incubated at 37 • C for 24 h, and activated by two successive transfers.

Microbiological Analysis
One hundred µL of the overnight cultures of each bacterium were transferred to 10 mL TSB and incubated for 24 h at 37 • C with shaking at 150-250 RPM. Then a further 10-fold dilution in 1 mL of saline 0.9% was done. Cultures were spectrophotometrically measured at 600 nm and the viable cell count was determined by plating onto Tryptic soy agar (TSA).
Suspensions of approximately 10 6 CFU/mL of each strain were used for inoculating onto stainless steel disks.
One hundred µL of the inoculum were firstly placed at the center of the stainless steel disk and then spread on the whole surface by means of a sterilized spatula, covered by the petri dish lid and placed in the incubator at 37 • C 24 h to render attachment before being processed with sanitizers.
Thirty-six Petri dishes (12 for R 0.25, 12 for R 0.5, and 12 for R 1) containing one stainless steel disk each were used for each bacterial strain. For each roughness, nine out of twelve stainless steel disks underwent one of three different sanitizing treatments i.e., UV (UVC, 253 nm) direct exposure under the hood (70 cm distance), alcohol 70% v/v, and gold lotion (GL, Miyauchi Citrus Research Center, Shigoka-Machi Takasaki Gunma, Japan) for 12 h while 3 out of 12 stainless steel disks were not sanitized and worked as negative controls.
One mL of ethanol and GL was applied directly on the disk surface with friction in circular movements for 30" by means of a sterile loop after the inoculum spreading.
A sterile swabbing (Sterile swabs without culture medium, Incofar s.r.l., Modena, Italy) was carried out after 12 h of a challenge with sanitizers by rubbing of the surface at room temperature. Then, the tip of the swab was placed in the sterile swab tube with 1 mL of saline 0.9% and vortexed for one minute. Then we took 0.1 mL from the sterile swab tube and plated them onto the agar plate with a sterile spatula (final concentration 10 −1 ).
Serial tenfold dilutions (0.1 mL + 0.9 mL) of re-suspensions were spread onto appropriate agar plates for the viable cell count. The colonies were physically counted on the plate following the incubation at 37 • C for 24 h.
GL is a commercially available natural product made of peels derived from navel oranges, Citrus hassaku, Citrus limon, Citrus natsudaidai, Citrus miyauchi, and Satsuma, with a total content of flavonoids equal to 0.45 mg/mL [32].

Atomic Force Microscopy Analysis
A BioScope I microscope equipped with a Nanoscope IIIA controller (Veeco Metrology, Plainview, NY, USA) was used to acquire AFM topography images. In order to reduce vibrations or movements which negatively affect the tip scan, all the samples were sticked on the underlying substrate.
The BioScope head was then mounted on the top of the samples. AFM imaging was performed in non-contact mode at room temperature, in air; triangular doped silicon cantilevers (Veeco, NTESP) with nominal spring constants between 20 and 80 N/m and a resonance frequency around 270 KHz were used. AFM images post-processing and root mean square (RMS) roughness quantification were obtained by using the free software Gwyddion (v. 2.41).

Environmental Scanning Microscopy Analysis
Morphological analysis of the nanoXHAM ® D surface-treated stainless steel disks was performed by scanning electron microscopy (Nova Nano SEM 450, ThermoFisher Scientific, Rodano (MI), Italy) equipped with an energy-dispersive X-ray microanalysis system (X-EDS, QUANTAX-200, Bruker Nano Analytics, Berlin, Germany) using secondary electrons. Each sample was mounted onto a sample stub via double-sided adhesive tape and images were taken at an accelerating voltage of 15 kV.

Statistical Analysis
All the experiments were carried out in triplicate. Data were analyzed using GraphPad Prism 7 software (GraphPad Software, Inc., La Jolla, CA, USA). All data are presented as the means ± standard error of the mean (SEM) and were first checked for normality using the D'Agostino-Pearson normality test. Differences in bacterial growth for each strain at different roughness, both on untreated and nanoXHAM ® D-treated disks and after different sanitizing treatments, were analyzed using a two-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test. The difference among controls of each strain at different roughness, both on untreated and nanoXHAM ® D-treated disks, was analyzed using a Kruskal-Wallis test followed by Dunn's multiple comparison test.

Results
During the PECVD treatment, due to inelastic collision with electrons, monomers are dissociated to smaller chemical species or radicals, describing principally via breaking of the Si-O-Si and Si-C bonds. Since plasma deposition modifies the surface of a material only at the micron level, infrared analysis, rather than transmission spectra, is found to be an appropriate technique to characterize the induced chemical modification. The ATR FT-IR spectra of treated and untreated samples, registered in the range between 4000 cm −1 and 600 cm −1 , are shown in Figure 2. The frequencies listed in Table 1 are the main IR absorption peaks detected and assigned to characteristic vibrational modes.

Statistical Analysis
All the experiments were carried out in triplicate. Data were analyze GraphPad Prism 7 software (GraphPad Software, Inc., La Jolla, CA, USA). All presented as the means ± standard error of the mean (SEM) and were first chec normality using the D'Agostino-Pearson normality test. Differences in bacterial for each strain at different roughness, both on untreated and nanoXHAM ® D-treat and after different sanitizing treatments, were analyzed using a two-way analysi iance (ANOVA) followed by Tukey's multiple comparison test. The difference controls of each strain at different roughness, both on untreated and nanoXH treated disks, was analyzed using a Kruskal-Wallis test followed by Dunn's multip parison test.

Results
During the PECVD treatment, due to inelastic collision with electrons, monom dissociated to smaller chemical species or radicals, describing principally via bre the Si-O-Si and Si-C bonds. Since plasma deposition modifies the surface of a only at the micron level, infrared analysis, rather than transmission spectra, is fou an appropriate technique to characterize the induced chemical modification. The A IR spectra of treated and untreated samples, registered in the range between 40 and 600 cm −1 , are shown in Figure 2. The frequencies listed in Table 1 are the absorption peaks detected and assigned to characteristic vibrational modes. .

Untreated Stainless steel Samples
In Figure 3 differences among the three sanitizing methods (UV, alcohol 70% v/v, and GL) and control in different surface roughness (R 0.25, R 0.5 and R 1) against four Gram-positive bacteria (Staphylococcus aureus ATCC 6538, Enterococcus faecalis ATCC 29212, Bacillus cereus ATCC 14579, and Listeria monocytogenes NCTT 10888) are represented.
To better address the antibacterial effect possibly exerted by the surface, we further compared the bacterial count of each strain for each surface roughness without sanitizing methods ( Figure 4).
We then screened also the Gram-negative bacteria, (E. coli ATCC 25922, S. typhimurium ATCC 1402, Y. enterocolitica ATCC 9610, and P. aeruginosa ATCC 27588) and evaluated differences among the three sanitizing methods (UV, alcohol 70% v/v and GL) and control in different surface roughness (R 0.25, R 0.5 and R 1 μm) ( Figure 5).
As noted for Gram-positive bacteria, no bacterial count was detectable after UV and alcohol 70% v/v treatment for all strains regardless of roughness ( Figure 5A-D).
We then screened also the Gram-negative bacteria, (E. coli ATCC 25922, S. typhimurium ATCC 1402, Y. enterocolitica ATCC 9610, and P. aeruginosa ATCC 27588) and evaluated differences among the three sanitizing methods (UV, alcohol 70% v/v and GL) and control in different surface roughness (R 0.25, R 0.5 and R 1 µm) ( Figure 5).
As noted for Gram-positive bacteria, no bacterial count was detectable after UV and alcohol 70% v/v treatment for all strains regardless of roughness ( Figure 5A-D).
No significant difference was observed among bacterial strains regardless of roughness with respect to the initial inoculum (10 6 CFU/mL) ( Figure 6). Microorganisms 2021, 9, x FOR PEER REVIEW 9 of 18 No significant difference was observed among bacterial strains regardless of roughness with respect to the initial inoculum (10 6 CFU/mL) ( Figure 6).

NanoXHAM ® D-Coated Stainless steel Samples
After microbiological and microscopic analyses all disks were treated with the nan-oXHAM ® D and both analyses were repeated on treated disks.
In Figure 7, differences among the three sanitizing methods (UV, alcohol 70% v/v and GL) and control at different surface roughness (R 0.25, R 0.5 and R 1 μm) against the same four previous Gram-negative bacteria (E. coli ATCC 25922, S. typhimurium ATCC 1402, Y. enterocolitica ATCC 9610 and P. aeruginosa ATCC 27588) are summarized.

NanoXHAM ® D-Coated Stainless steel Samples
After microbiological and microscopic analyses all disks were treated with the nanoXHAM ® D and both analyses were repeated on treated disks.
In Figure 7, differences among the three sanitizing methods (UV, alcohol 70% v/v and GL) and control at different surface roughness (R 0.25, R 0.5 and R 1 µm) against the same four previous Gram-negative bacteria (E. coli ATCC 25922, S. typhimurium ATCC 1402, Y. enterocolitica ATCC 9610 and P. aeruginosa ATCC 27588) are summarized.
Contrary to the uncoated surfaces, a significant mean decrease in the bacterial count of the control was observed for almost all strains (26 ± 0.78 CFU/mL) with respect to the initial inoculum (10 6 CFU/mL), regardless of roughness ( Figure 7B,C,D). Interestingly, no E. coli ATCC 25922 count was detected in R 0.25 and R 0.5, while only in R 1 it reached 16.67 ± 8.81 CFU/mL ( Figure 7A). Further, as previously reported for the uncoated surfaces, no bacterial count was detectable after UV and alcohol 70% v/v treatment for all strains regardless of roughness ( Figure 7A-D). As for GL treatment, it showed the same trend for all strains regardless of roughness, in fact, an overall mean significant bacterial count inhibition was observed (27.31 ± 2.25 CFU/mL) when compared with the initial inoculum (10 6 CFU/mL) ( Figure 7A-D).
A significant decrease in bacterial count was observed in R 1 for S. typhimurium ATCC 1402 and Y. enterocolitica ATCC 9610 (20 ± 0 and 10 ± 0 CFU/mL, respectively) when compared with their control (25 ± 2.23 and 26.67 ± 1.66 CFU/mL, respectively), **** p < 0.0001 and * p < 0.05, respectively, after GL treatment ( Figure 7C,D). However, a significant bacterial count decrease was also observed in R 0.25 for Y. enterocolitica ATCC 9610 (18.33 ± 1.05 CFU/mL) with respect to its control (30 ± 0 CFU/mL), * p < 0.05 ( Figure 7D). Contrary to the uncoated surfaces, a significant mean decrease in the bacterial count of the control was observed for almost all strains (26 ± 0.78 CFU/mL) with respect to the initial inoculum (10 6 CFU/mL), regardless of roughness ( Figure 7B,C,D). Interestingly, no E. coli ATCC 25922 count was detected in R 0.25 and R 0.5, while only in R 1 it reached 16.67 ± 8.81 CFU/mL ( Figure 7A). Further, as previously reported for the uncoated surfaces, no bacterial count was detectable after UV and alcohol 70% v/v treatment for all strains regardless of roughness ( Figure 7A-D). As for GL treatment, it showed the same trend for all strains regardless of roughness, in fact, an overall mean significant bacterial count inhibition was observed (27.31 ± 2.25 CFU/mL) when compared with the initial inoculum (10 6 CFU/mL) ( Figure 7A-D).
To better address the antibacterial effect possibly exerted by the nanoXHAM ® Dtreatment surface, we further compared the bacterial count of each strain for each surface roughness without the 12-h sanitization with UV, alcohol 70% v/v, or GL (Figure 8). It is noteworthy the significant difference, in terms of bacterial count, between the control (0 CFU/mL) and the GL treatment (29.17 ± 1.16 CFU/mL) observed in R 0.5 for E. coli ATCC 25922 (** p < 0.01) and the control in R 0.25 and R 0.5 (20 ± 0 and 25 ± 0 CFU/mL, respectively) and the GL treatment (40 ± 0 and 48.33 ± 1.66 CFU/mL, respectively) for S. typhimurium ATCC 1402 (**** p < 0.0001 and * p < 0.05, respectively) ( Figure 7A,C).
To better address the antibacterial effect possibly exerted by the nanoXHAM ® Dtreatment surface, we further compared the bacterial count of each strain for each surface roughness without the 12-h sanitization with UV, alcohol 70% v/v, or GL ( Figure 8).
As for the untreated disks we also screened Gram-positive bacteria, L. monocytogenes NCTT 10888, E. faecalis ATCC 29212, B. cereus ATCC 14579, and S. aureus ATCC 6538, on the nanoXHAM ® D-surface treatment (Figure 9). As noticed for Gram-negative bacteria, no bacterial count was detected after UV and alcohol 70% v/v treatment regardless of roughness ( Figure 9A-D).
Stainless steel surfaces exhibiting different roughness were also examined in a small range by AFM before and after the nanoXHAM ® D surface treatment ( Figure 11).
The small range RMS roughness analysis reveals for the uncoated stainless steel surface ( Figure 11A-C) a tendency similar to that detected on a larger scale, namely with a progressive roughness increasing from R 0.25 to R 1. Interestingly, after the nanoXHAM ® D surface treatment (panels C, D, and F) RMS roughness appears to be almost constant. This represents a clear indication that the coating deposition procedure does not affect the initial surface roughness and results in a homogeneous covering layer regardless of the initial values of roughness.
As observed for the AFM, ESEM images acquired on nanoXHAM ® D-treated stainless steel disks confirmed the presence of slanting lines over the surface as well as the presence of a homogeneous layer made of film amorphous SiOxCyHz ( Figure 12A). The X-EDS microanalysis confirmed the presence of the aforementioned elements ( Figure 12B). Further, a significant decrease in bacterial count was also observed for E. faecalis ATCC 29212 in R 0.5 (20 ± 0 CFU/mL) with respect to R 1 (25.33 ± 0.33 CFU/mL), * p < 0.05 ( Figure 10B). Conversely, a significant decrease in bacterial count was also observed for B. cereus ATCC 14579 in R 1 (20 ± 0 CFU/mL) with respect to R 0.25 (25 ± 0 CFU/mL), * p < 0.05 ( Figure 10C).
Stainless steel surfaces exhibiting different roughness were also examined in a small range by AFM before and after the nanoXHAM ® D surface treatment ( Figure 11). The small range RMS roughness analysis reveals for the uncoated stainless steel surface ( Figure 11A-C) a tendency similar to that detected on a larger scale, namely with a progressive roughness increasing from R 0.25 to R 1. Interestingly, after the nanoXHAM ® D surface treatment (panels C, D, and F) RMS roughness appears to be almost constant. This represents a clear indication that the coating deposition procedure does not affect the initial values of roughness.
As observed for the AFM, ESEM images acquired on nanoXHAM ® D-treated stain less steel disks confirmed the presence of slanting lines over the surface as well as the presence of a homogeneous layer made of film amorphous SiOxCyHz ( Figure 12A). The X-EDS microanalysis confirmed the presence of the aforementioned elements ( Figure  12B).

Discussion
In this study, we reported the significant bactericidal effect exerted by all the three sanitizing treatments tested against all bacterial strains, with respect to the initial inocu lum (10 6 CFU/mL), regardless of roughness and surface coating. At the same time, it is

Discussion
In this study, we reported the significant bactericidal effect exerted by all the three sanitizing treatments tested against all bacterial strains, with respect to the initial inoculum (10 6 CFU/mL), regardless of roughness and surface coating. At the same time, it is worth pointing out the more significant effect exerted by UV and alcohol 70% v/v treatments as no viable bacterial cells were detected under the same conditions.
Results concerning the lack of differences among bacterial strain growth on uncoated stainless steel disks, regardless of surface roughness, are in agreement with other literature reports [6,15,37,38].
Conversely, the nanoXHAM ® D induced some results that were as interesting as conflicting. In fact, if on one hand a bactericidal effect was also observed both in the control and after all sanitizing treatments regardless of roughness, a quite puzzling, but not significant, increase in bacterial count was observed after GL treatment, with respect to the control, in most Gram-negative bacteria.
Although we did not perform any punch-through experiments to investigate the possible presence of forces that could take place at the surface/bacterial membrane interface, we believe that such differences might be ascribed to a different pH gradient created by GL and to the incubation temperature (37 • C).
In fact, we previously demonstrated that the physical status of supported (on silicon) lipid bilayers, which share the same physicochemical features with Gram-negative bacteria, can be strongly influenced by environmental conditions such as pH and temperature [39,40]. By means of the AFM, we observed that a pH value of 3.5 was able to promote a bilayer stiffening through a surface charge and electrostatic free energy modification with a consequent formation of small holes, which is the first step towards bacterial cell disruption, although this process is slowed at temperatures higher than 22 • C.
Being GL pH value close to 4.3, it is reasonable to hypothesize a process similar to that described above and therefore the slight increase in count observed in Gram-negative bacteria is due to a delayed death process induced by pH and temperature.
Thanks to the versatility of AFM, some authors also underlined the ease of removal of two different kinds of bacteria (S. aureus and P. aeruginosa) depending on the surface features [41]. They reported the ease of removal of S. aureus from smooth surfaces and of P. aeruginosa from 0.5µm surfaces, addressing such difference to the different cell/surface contact area. Indeed, cocci had a smaller cell/surface contact area on smooth surfaces with respect to rods. However, this aspect was observed only for E. coli ATCC 25922 in R 0.25 and R 0.5 of nanoXHAM ® D-treated control disks.
As far as concerns the comparison among nanoXHAM ® D-treated control disks, the growth of two Gram-negative (S. typhimurium ATCC 1402 and P. aeruginosa ATCC 27588) and of two Gram-positive bacteria (L. monocytogenes NCTT 10888 and E. faecalis ATCC 29212) resulted to be significantly reduced in R 0.25 with respect to R 0.5 and in particular to R 1.
Moreover, E. faecalis ATCC 29212 growth was also significantly reduced in R 0.5 with respect to R 1 while B. cereus ATCC 14579 growth was reduced in R 1 with respect to R 0.5 and in particular R 0.25.
It is worth noting that cytotoxicity assays conducted on nanoXHAM ® D-coated stainless steel surfaces revealed the lack of any cytotoxic effect of the coating and, therefore, the significant bacterial count decrease observed on all nanoXHAM ® D-coated disks might be ascribed to a synergistic effect of low bacterial attachment forces and an increased hydrophobicity of the treatment.
Moreover, FT-IR-ATR, SEM, and EDX elemental analysis measurements confirmed that the organic silicon thin film (nanoXHAM ® D) was successfully deposited on stainless steel disks. In particular, the ATR technique of treated samples, compared to the steel sample reference, showed absorption characteristics assigned to C-H stretching, Si-(CH3)x bending, Si-O stretching and Si-O bending modes, respectively.
We also confirmed previously achieved results concerning the potential use of a new, natural, and non-corrosive sanitizing product (GL) to achieve significant bactericidal effect against all tested bacterial strains [29,42]. In particular, GL is rich in flavonoids, which are natural compounds that have been extensively studied for their antibacterial properties [29,[43][44][45], and can be considered a valuable alternative candidate to commercially available sanitizing products for stainless steel such as iodine, biguanide, quaternary ammonium compounds, peracetic acid and sodium hypochlorite that showed evident corrosive action [46].

Conclusions
We can conclude that untreated stainless steel surface roughness is poorly correlated with bacterial adhesion and only sanitizing treatments can exert significant bactericidal effects. Unfortunately, most sanitizing treatments are toxic and corrosive in the long run causing the onset of crevices that are able to facilitate bacterial nesting and growth.
With the advent of this new nanotechnological coating i.e., nanoXHAM ® D, it has been possible to change the surface physicochemical characteristics obtaining an overall bactericidal effect possibly due to a synergistic effect of low bacterial attachment forces, increased hydrophobicity, and less toxic and corrosive sanitizing treatments such as UV ethanol and GL.
It is therefore necessary now to accurately undertake time-course experiments with such sanitizing treatments to better exploit the potential of nanoXHAM ® D and achieve the best result within the shortest time to fulfill food industry regimes.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.