**3. Results**

#### *3.1. Optical Microscopic Images*

All of the specimens were observed by a technique that used optical microscopy combined with the Raman spectroscopic analyzer. In Figure 1, optical microscopic images before and after the swelling process were shown for DEMM polymer brush specimens. All of them were observed in the perpendicular direction to specimens' surfaces. Before swelling (Figure 1a), the surface was pretty smooth. However, the roughness increased after swelling and the top of fiber-like brushes could be observed in Figure 1b. In the same way, PMMA specimens showed the top of brushes after swelling (Figure 2b), while they did not show any brushes before swelling (Figure 2a). These photos clearly show that polymer brush specimens were made on glass surfaces.

**Figure 1.** Optical microscopic images for N,N-diethyl-N-(2-methancryloylethy)-N-methylammonium bis(trifluoromethylsulfonyl) imide (DEMM-TFSI) specimens (**a**) before and (**b**) after selling process.

**Figure 2.** Optical microscopic images for Polymethyl methacryrate (PMMA) specimens (**a**) before and (**b**) after selling process.

These specimens were immersed in liquid cultures with bacteria. Figure 3 shows the surfaces of glass substrates as references after biofilm formation by *E. coli* and *S. epidermidis*, respectively. It was not easy for biofilms to form on glass specimens. However, biofilms were observed in some parts of the substrates. The scattered individual bacteria could be observed. However, some aggregates composed of bacteria were also observed. The latter should be biofilms. The light points (in the photos) are reflections of laser beams on the specimens. In both cases of bacteria (Figure 3a for the case of *E. coli* and Figure 3b for the case of *S. epidermidis*), the photos show biofilm formations on glass specimens.

**Figure 3.** Optical microscopic images for glass substrates after biofilm formation: (**a**) *E. coli* and (**b**) *S. epidermidis*.

Figure 4 shows the optical microscopic images for surfaces of DEMM polymer brush specimens after biofilm formations by *E. coli* and *S. epidermidis*, respectively. Being compared with the results shown in Figures 1 and 2, surfaces of DEMM polymer brush specimens appear to be filled with contaminants. The results clearly show that biofilms were formed on the entire surfaces of the specimens.

**Figure 4.** Optical microscopic images for DEMM polymer brush specimens after biofilm formation: (**a**) *E. coli* and (**b**) *S. epidermidis*.

Figure 5 shows the optical microscopic images for the surfaces of PMMA polymer brush specimens after biofilm formations (by these same bacteria). Figure 5a, corresponds to the result of *E. coli*, while Figure 5b refers to that of *S. epidermidis*. In both cases, biofilms can be observed in the same way. However, they seem to exist locally. There were still some places where no biofilms were observed. Even brushes could be observed in the backgrounds. The brushes in Figure 5b appear swollen through the biofilm formation process. These results show that it is more difficult for the biofilms to form on PMMA polymer brush specimens than on those of DEMM.

**Figure 5.** Optical microscopic images for PMMA polymer brush specimens after biofilm formation: (**a**) *E. coli* and (**b**) *S. epidermidis*.

#### *3.2. Raman Scpetroscopy*

All of the specimens were subjected to Raman spectroscopy and the results were analyzed to confirm the presence of biofilm by "finger print method" (Identification by comparing the data with those in the previous papers [71–81]). For each case, the same measurements were repeated three times for three specimens and the three results for each case were overlapped in the Raman shift charts basically, except for some problematic exceptional cases. Even though the measurement points on each specimen were chosen randomly, bacteria assembled and EPS was confirmed by the optical microscopy for all of those measurement points.

Before biofilm formation, the DEMM coating and PMMA coating did not show any characteristic Raman peaks. In both cases, Raman displayed small peaks of glass substrate that were observed at 1080 and 560 cm−1. On the contrary, many characteristic peaks appeared for the specimens after the biofilm formation processes. Figure 6 shows the results for glass substrates. Figure 6a corresponds to the results of *E. coli.* In this case, unfortunately, the third specimen gave no robust data due to a signal failure. The first specimen showed a slightly different result from the second. From the two results, we could judge that biofilms were formed on glass substrates, since most of them could be attributed to polymers derived from biofilms or bacteria. The peak around 570 and 800 cm−<sup>1</sup> could be attributed to polysaccharides or lipids [71,72]. The peak around 1100 cm−<sup>1</sup> could be assigned also to that of polysaccharides or lipids [71,73]. Also, the peak around 2800 cm−<sup>1</sup> could be assigned to polysaccharides [74] and lipids [75]. Figure 6b shows the results of *S. epidermidis* for glass specimens. The peak relating to nucleic acids around 2400 cm−<sup>1</sup> [73] was observed in the case of *S. epidermidis.* The peak of polysaccharides and/or lipids at 2800 cm−<sup>1</sup> was observed clearly. Lipid's peak was observed also at 1400 cm−<sup>1</sup> for the first specimen [75]. The peak for polysaccharides/lipids at 1100 cm−<sup>1</sup> [71,73] was also observed. Two peaks between 500 and 800 cm−<sup>1</sup> would be related to polysaccharides/lipids [71,72].

**Figure 6.** Raman peaks on glass substrates after biofilm formation: (**a**) *E. coli* and (**b**) *S. epidermidis*.

Figure 7 shows the results for DEMM polymer brush specimens. Figure 7a shows the results in the case of *E. coli*. All of the specimens show similar tendencies for Raman peaks, except for the broad one around 2000 cm−<sup>1</sup> (peaks for stretching or vibration of triple bonds between carbon-carbon or carbon-nitrogen) [76]. However, the peak was not observed for the third specimen. For other peaks, most of them were overlapped, even though the intensities were different from each other. The peak around 2800 cm−<sup>1</sup> could be attributed to those for polysaccharides and/or lipids [74,75]. The peak around 1400 cm−<sup>1</sup> was considered that for lipids [74]. The small peak at 1300 cm−<sup>1</sup> is highly likely that for amid III (protein) [72], while that at 1100 cm−<sup>1</sup> could be assigned to that for lipids or proteins [71,73]. The broad peak around 2000 cm−<sup>1</sup> corresponding of triple bonds between carbon atoms or carbon-nitrogen has been still undecidable. It suggests nitrile compounds or alkyne molecules would exist. Both might be formed through some components derived from biofilms, since the broad peaks were sometimes found in the past by us [79]. However, in this case, we used acetonitrile to swell the brush coating. It was highly likely that the chemical was remained at the upper side of coating. Figure 7b shows the results in the case of *S. epidermidis*. In this case, peaks for three specimens were almost overlapped, even though the strengths were different for each of them. The broad peak was also observed around 2000 cm−<sup>1</sup> [76]. The peak around 2800 cm−<sup>1</sup> could be attributed to that for lipids/polysaccharides [74,75], while lipids or proteins peaks were observed at 1100cm−<sup>1</sup> [75].

**Figure 7.** Raman peaks on DEMM polymer brush specimens after biofilm formation: (**a**) *E. coli* and (**b**) *S. epidermidis*.

Figure 8 shows the results for PMMA specimens. Also in these cases, most of the peaks were overlapped in both cases. In the case of *E. coli* (Figure 8a), the second specimens could not be shown due to the unintentional signal failures. In the case of *S. epidermidis*, more complicated peaks (than those in other cases) appeared, as shown in Figure 8a,b. Peaks around 2800 cm−<sup>1</sup> could be attributed to either of polysaccharides [74] (at higher wave number) or lipids [75] (at lower wave number). Both peaks were almost overlapped around 2800 cm−1. The peak around 1600 cm−<sup>1</sup> was considered the one for polysaccharides [74]. The peak at 1400 cm−<sup>1</sup> could be assigned to that for lipids [75]. Three peaks seen from 1100 to 1400 cm−<sup>1</sup> belong to lipids or proteins [72,79]. We presume that the new peak just over 3000 cm−<sup>1</sup> would belong to lipid [79] (However, we could not deny the possibility of nucleic acids [73]). Other new ones at 1000 cm−<sup>1</sup> would be assigned to amino acid (phenylalanine) [71]. Small peaks from 550 to 800 cm−<sup>1</sup> could usually be assigned to polysaccharides or lipids [71,72]. Therefore, we presume that biofilms formed on the specimen (particularly PMMA) and the sticky surface as a result of having the components of culture media incorporated into them. All of these Raman spectroscopic analyses suggest that biofilms formed on specimens' surfaces. However, they indicate that biofilms were more difficult to form on PMMA than on DEMM polymer brush coated specimens. This tendency was also supported by the results of microscopic observations. Particularly, the tendency was very remarkable for nucleic acids derived from biofilms.

**Figure 8.** Raman peaks on PMMA polymer brush specimens after biofilm formation: (**a**) *E. coli* and (**b**) *S. epidermidis*.

#### **4. Discussion**

In these experiments, we vividly observed the biofilms formed on DEMM brush coated specimens. However, the contaminants derived from biofilms could be washed away easily by rinsing with water. From this viewpoint, the ionic liquid type polymer brush like DEMM would be convenient and favorable. This is because it could easily attract biofilms and the related contaminants on the materials' surfaces.

Figure 9 schematically illustrates the mechanism of antifouling for a polymer brush coating. In water and some other liquids, the polymer brush would contain water droplets between brushes (Figure 9a,b). Due to the static electrical force or mechanical one caused by geographical configurations, bacteria could get trapped and result in the formation of biofilm. However, the hydrophobic parts of the biofilms act repulsively against water drops trapped between brushes. Then they could be washed away easily as a whole (Figure 9b,c). In addition, brushes would be swollen in the aqueous solution

further. The swollen effect would be added to the buoyancy and repulsion forces. Thus, biofilms could be removed. However, being compared with PMMA polymer brush coated specimens, why would DEMM polymer brush coating specimens attract bacteria-biofilms-contaminants so much?

**Figure 9.** The mechanism of antifouling for polymer brush coating. (**a**) Polymer brush coating before immersion into a liquid solution. (**b**) Water drops are trapped in polymer brush coating, while biofilms formed by bacterial activities are attached to the top of polymer brush. (**c**) Repulsion forces are produced between water drops and hydrophobic parts of biofilms. (**d**) Water drops push biofilms up with its buoyancy and swelling action of polymer brush.

The most likely answer is due to the polar characteristics and the differences between the two types of polymer brush coated specimens. The surface of a PMMA brush coated specimen is generally neutral, while that of a DEMM brush coated specimen is polarized. Therefore, a DEMM polymer brush would tend to attract biofilms more.

From the viewpoint of merits as the ionic liquid, we could conclude that ionic liquid type polymer brush coating could attract biofilms and contaminants as a result much more in the vicinity of surfaces. Then they could be washed away by water washing [57]. The characteristic would lead to various applications such as automobiles and medical instruments. In addition, according to the practical purposes, some persons may want to remove any contaminants from the beginning. The attachment and attraction of bacteria and the following biofilm formation seem to depend on the polar characters [82]. Fortunately, the ionic liquid could arrange the polar characteristics easily. Therefore, the appropriate combination of ionic liquids and polymer brush production processes might lead to some changes of anti-biofouling effect and properties in the future, if they would be required and needed. When we come to think about those characteristics (about the application of ionic liquids to a polymer brush coating), this type of polymer brush would have a promising future as an advanced coating and material.

#### **5. Conclusions**

DEMM polymer brush coating as a liquid ionic polymer brush was investigated from the viewpoint of its biofilm formation and antifouling effect. It was also compared with the results for the PMMA polymer brush coating. Our results are demonstrated.


**Author Contributions:** Planning, experiments for biofilms, writing drafts, data curation, H.K. and A.O. (Atsuya Oizumi); Producing polymer brush coating specimens, discussion, advice, T.S., T.K. (Toshio Kamijo) and S.H.; Writing-review & editing, proof-reading, discussion, D.M.B.; Discussion and advice for biofilm formation and evaluation, N.H., A.O. (Akiko Ogawa), T.K. (Takeshi Kogo), D.K. and K.S., Discussion and advice for ionic liquids, K.T.; Discussion and advice from the viewpoint of application, S.-H.L. and M.-H.L.

**Funding:** This research was funded by JSPS KAKENHI (Grants-in-Aid for Scientific Research from Japan Society for the Promotion of Science) Grant Number 17K06826. This study was funded also by Mitsubishi Electric Corporation and the Society of Industrial Technology for Antimicrobial Articles (SIAA).

**Acknowledgments:** The current investigation was carried out under the project entitled Network Formation Project, 2017 by the headquarters of the National Institute of Technology (Kosen) financially (Network Formation Project, 2017). We would like to thank them for their constant encouragement. We very highly appreciate Mitsubishi Electric Corporation and its research center of advanced technology for their useful advice and financial support. Also we would like to thank the Society of Industrial Technology for Antimicrobial Articles (SIAA) as well as the Japan Food Research Laboratory (JFRL).

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
