Nanoporous aerogel as a bacteria repelling hygienic material for healthcare environment

Nanoporous silica (SiO₂) aerogels were synthesized via sol–gel polymerization of tetraethyl orthosilicate (TEOS; Sigma-Aldrich Co., St. Louis, MO, USA) at room temperature (23 °C). Gelation of 5 ml TEOS dissolved in 11 ml of ethanol (200 proof; Koptec, King of Prussia, PA, USA) was initiated by mixing solution with 0.37 ml of catalyst solution which was prepared by adding 1.85 g of ammonium fluoride (NH₄F; Sigma-Aldrich Co., St. Louis, MO, USA) and 22.8 ml of ammonium hydroxide (NH₄OH; Sigma-Aldrich Co., St. Louis, MO, USA) to 100 ml of water. The reaction was allowed to take place for 24 h. Once the gel had set, ethanol inside the gel was extracted by supercritical carbon dioxide (CO₂; Brazos Valley Welding Supply, Inc., Bryan, TX, USA) at the critical point (31.1 °C, 72.9 bar). After the supercritical CO₂ extraction, the obtained hydrophilic silica aerogels were submerged in 6 wt% trimethylsilyl chloride (TMCS; Sigma-Aldrich Co., St. Louis, MO, USA)/hexane (ACS grade; Avantor Performance Materials, Inc., Center Valley, PA, USA) solution for 24 h to functionalize silica aerogel as schematically illustrated in figure 1(a). Resultant methylated silica aerogel was rinsed with hexane and dried at 60 °C until hexane evaporated completely.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Since silicon dioxide (SiO₂) was the main constituent of the developed samples, quartz (SiO₂) was selected as a control surface to ensure a chemical similarity between these. Quartz slides (Ted Pella, Inc., Redding, CA, USA) cut into 1 cm × 1 cm × 1 mm were first rinsed with Milli-Q water (resistivity ≥18.2 MΩ cm; EMD Millipore Corp., Billerica, MA, USA), then dried under a stream of nitrogen (N₂; Brazos Valley Welding Supply, Inc., Bryan, TX, USA). Subsequently, oxygen plasma treatment was applied at a power of 20 W, pressure of 80 mTorr, oxygen flow rate of 20 sccm, and a time of 1 min by using CS-1701 reactive-ion etcher (Nordson March, Concord, CA, USA) for cleaning purposes. We also note that oxygen plasma treatment can effectively sanitize surfaces from pre-existing bacteria [50]. Methylated quartz slides were prepared by placing quartz slides in solution of 6 wt% TMCS in hexane for 24 h, as shown in figure 1(b). The samples were then purged under a stream of nitrogen before use.

Surface morphology of all samples was characterized by scanning electron microscope (SEM, JSM-7500F; JEOL, Tokyo, Japan) and atomic force microscopy (AFM, Dimension Icon; Bruker, Santa Barbara, CA, USA). Before SEM examination, the samples were coated with 8 nm of platinum/palladium to minimize possible charging effects. The SEM micrographs were obtained at an accelerating voltage of 1 kV and emission current of 20 μA.

AFM topographic micrographs were obtained by using tapping mode in air at room temperature. Specifications of the silicon tip cantilever are nominal tip radius of 2 nm, nominal spring constant of 0.4 N m⁻¹, and nominal resonant frequency of 70 kHz. Extremely lightweight HNSA samples were mounted on a glass slide by using instant glue and then placed on the stage to decrease data noise.

Specific surface area, pore diameter, and pore volume distribution of HNSA were characterized by BET and Barrett–Joyner–Halenda (BJH) methods [51]. To achieve these values, nitrogen adsorption isotherms were measured with an ASAP2010 (Micromeritics Instrument Co., Norcross, GA, USA) at liquid nitrogen temperature of 77 K.

ATR-FTIR spectroscopy was used to characterize TMCS-functionalized quartz and TMCS-functionalized silica aerogel (i.e., HNSA) surfaces at ambient conditions. ATR-FTIR spectra were recorded with an IRPrestige-21 (Shimadzu Corp., Kyoto, Japan) system and data were analyzed using IRsolution (Shimadzu Corp., Kyoto, Japan) software version 1.40.

To confirm TMCS coverage, XPS spectra of TMCS-functionalized quartz and TMCS-functionalized silica aerogel (i.e., HNSA) surfaces were recorded with a PHI VersaProbe II Scanning XPS Microprobe (Physical Electronics, Chanhassen, MN, USA). XPS measurements were carried out using an Al Kα radiation source (1486.6 eV) operating at 25 W, and at a working pressure of 10⁻⁷ Pa.

The surface hydrophobicity was determined by a sessile drop technique, a 5 μl water droplet was placed on different types of surfaces. The static contact angle values were measured by ImageJ (National Institutes of Health, Bethesda, MD, USA) software via Low-Bond Axisymmetric Drop Shape Analysis (LBADSA) plug-in [52]. The values reported are an average of six measurements.

Working cultures of Escherichia coli O157:H7 (ATCC 700728) and Staphylococcus aureus were inoculated into 9 ml of tryptic soy broth (TSB; Becton, Dickinson and Co., Sparks, MD, USA) by transferring a loopful of culture from tryptic soy agar (TSA) slant. Both strains were incubated aerobically without agitation for 24 h at 37 °C. A loopful of culture was then transferred every 24 h for 2 days to fresh TSB and incubated aerobically at 37 °C. The final concentration ranging from 8.7 to 9.1 log CFU ml⁻¹ were reached by E. coli O157:H7 and S. aureus in the growth media, as determined by plate count method.

Prior to inoculation experiments, the samples (i.e., hydrophilic bare quartz, hydrophobic quartz, and HNSA) were sterilized by washing with 70% (v/v) ethanol followed by sterile Milli-Q water rinsing. After the sterilization procedure, the samples were inoculated with bacteria by submerging them in 9 ml bacterial suspensions (8.7–9.1 log CFU ml⁻¹) for 4 h at room temperature. The treated samples were then gently removed from the bacterial suspensions to count bacteria attached to surfaces. All inoculation experiments were replicated four times.

SEM was used to quantify the attachment of E. coli O157:H7 and S. aureus to various surfaces by direct counting. After acrolein (Sigma-Aldrich Co., St. Louis, MO, USA) inactivation, 10 nm thickness of gold coating was applied to sample surfaces to ensure electrical conductivity required by SEM technique. For quantitative analysis, at least ten different 100 μm × 100 μm scan areas (total scan area larger than 100 000 μm²) were analyzed to count the number of attached bacteria.

Angle dependent ellipsometer (Nanofilm EP3-SE; Nanofilm Technology GmbH, Göttingen, Germany) was utilized to obtain the refractive index of HNSA. The measurements were carried out before and after 4 h exposure to bacterial suspensions.

Bacterial suspensions of E. coli O157:H7 and S. aureus strains were grown in the presence of HNSA and in the presence of 1% (v/v) bleach solution for 4 h at room temperature. Bacterial suspensions without any treatment were used as negative control. The number of bacteria remaining in suspension was determined by pour plate method. Each condition was replicated three times.

Microbiological data obtained by SEM were converted to logarithm of a cell density (log cells mm⁻²). All analyses were performed by using statistical package for Microsoft Office Excel (Microsoft Corp., Redmond, WA, USA) software. The data were analyzed by one-way and two-way analysis of variance with Tukey's test to determine significant differences at a p-value of <0.05.

It is well-established that surface texture and roughness play an important role in bacterial adhesion on surfaces [53, 54]. Hence, before proceeding with bacterial attachment experiments, we first characterized the surface topography of materials prepared for this study to better understand and compare their adhesion behavior. Figures 2(a)–(c) display SEM micrographs of hydrophilic bare quartz, hydrophobic quartz, and HNSA surfaces, showing the surface texture of each material. SEM micrographs visually highlight the differences and reveal the highly nanoporous structure of HNSA.

When the length scale of surface roughness is larger than bacteria size, bacterial colonization is enhanced because bacteria prefer surfaces/textures that increase the total contact area of bacteria-material interfaces such as valleys, depressions, pits, and edges [55]. To quantitatively compare the characteristic sizes of bacteria and surface roughness, AFM measurements on three different material surfaces were performed over a 5 μm × 5 μm area, as shown in figures 2(d)–(f). The root-mean-square (RMS) roughness value of bare quartz surfaces were found to be 0.96 ± 0.05 nm. After surface functionalization with TMCS, RMS roughness reached 1.52 ± 0.11 nm. The small difference in RMS roughness can be attributed to the surface reaction, which introduces new Si–O bond (∼148 pm) connected to Si–C bond (∼318 pm) on top of bare quartz surfaces. RMS roughness of 118.58 ± 19.02 nm was obtained from HNSA surfaces. While surface roughness of HNSA surfaces was more than fifty times rougher than bare quartz and hydrophobic quartz surfaces, the scale of surface roughness was still much smaller than length and diameter of bacteria (i.e., rod-shaped E. coli O157:H7 is 0.5–1.0 μm wide by 1.0–4.0 μm long and spherical-shaped S. aureus is 0.7–1.0 μm in diameter). Thus, the physical attachment such as penetration and trapping of bacteria to HNSA surfaces can be prevented. In addition, these roughness values of HNSA meet the criterion of surface roughness (≤1.0 μm) outlined for hygienic materials [56].

Quantitative details of porosity characteristics of HNSA such as specific surface area (646.70 ± 5.04 m² g⁻¹), pore diameter (19.85 ± 0.10 nm), and pore volume (3.21 ± 0.02 cm³ g⁻¹) were determined by BET and BJH methods. These values are comparable to functionalized silica-based aerogels reported previously [57]. Here, we note that pore diameter (∼19 nm) of HNSA is much smaller than the bacteria dimensions ranging from 700 nm to 4000 nm as described above, thereby preventing bacterial penetration. This explanation is consistent with the reported data that pore diameter of 0.1 μm was not sufficient to allow bacteria to penetrate [58]. Furthermore, HNSA showed extremely low thermal conductivity due to its high surface area and low density (see supplementary data, section 1 and figure S1). We calculated the thermal conductivity of HNSA to be 0.047 ± 0.015 W m⁻¹  K⁻¹ using time-resolved temperature measurements, which is about an order of magnitude smaller than common thermal insulation materials [59].

Chemical modification of quartz and silica aerogel surfaces by methylation were characterized by ATR-FTIR spectroscopy, as shown in figures 3(a) and (b). ATR-FTIR spectra of pure (unreacted) TMCS, TMCS-functionalized quartz, and TMCS-functionalized silica aerogel (i.e., HNSA) surfaces showed multiple peaks that are significant, while bare quartz and bare silica aerogel surfaces had no peak between 2800 cm⁻¹ and 3000 cm⁻¹. Chemically unbound TMCS molecules had symmetric and asymmetric C–H stretching peaks at 2900 cm⁻¹ and 2962 cm⁻¹. On the other hand, methylated quartz had peaks at 2850 cm⁻¹ and 2916 cm⁻¹, and HNSA had peaks at 2900 cm⁻¹ and 2970 cm⁻¹, respectively. The presence of these spectrum peaks are due to overtones and combinations of symmetric and asymmetric C–H stretching vibrations upon functionalization reactions. Furthermore, peak shifts in C–H stretching region can be explained by replacement of chlorine (Cl) atoms by oxygen (O) atoms during methylation and phase transformation of samples from the liquid state to the crystalline state [60]. In addition, Si–Cl stretching vibrations near 620 cm⁻¹ only existed for TMCS, providing clear evidence of methylation [61].

Zoom In Zoom Out Reset image size

After successful functionalization of silica aerogel with TMCS, chemical stability test was conducted. HNSA samples were submerged in DI water and 10% (w/v) hydrogen peroxide (H₂O₂) for two weeks to prove chemical resistance against common environmental conditions that material may be exposed. Any chemical release, leaching, and degradation was not detected from ATR-FTIR measurements with detection limit of <1 ppm (see supplementary data, section 2 and figure S2).

It is known that bacterial adhesion is sensitive to the surface chemistry [62]. Therefore, it is essential to determine the degree of methylation on quartz and silica aerogel surfaces to better understand their bacterial adhesion behavior. To this end, XPS technique was used to obtain chemical information about the surface of solid materials. As shown in figure 4(a), O 1s, C 1s, Si 2s, and Si 2p were the main peaks observed in the XPS spectra. The integration of area under these peaks was conducted to determine the relative atomic concentration of O, C, and Si atoms. The atomic percentages for C, O, and Si were ∼16%, 26%, and 58%, respectively, for hydrophobic quartz. While the atomic percentages for C, O, and Si were ∼15%, 25%, and 60%, respectively, for HNSA. Given that the penetration depth of XPS is about 8 nm to 10 nm and RMS roughness for HNSA was about 118 nm, the effect of porosity (i.e., methyl groups in the pores of HNSA) on the XPS spectra is expected to be small [63]. Therefore, these findings suggest that the degree of methylation is similar for both quartz and silica aerogel.

Zoom In Zoom Out Reset image size

It is well-established that surface hydrophobicity of the material has an effect on bacterial attachment. As most of the bacteria surfaces are fairly hydrophilic, it is important to study hydrophobicity of material surfaces because hydrophilic materials tend to aggregate more on hydrophilic surfaces [64]. Therefore, we have investigated surface hydrophobicity of samples to better understand bacterial attachment to material surfaces. As shown in figure 4(b), while contact angle measured on bare quartz was hydrophilic (θ < 10.0°), methylated quartz (θ = 95.1° ± 1.0°) and methylated silica aerogel (θ = 134.4° ± 1.1°) were hydrophobic. The contact angle difference in these three types of surfaces can be explained on the basis of previous studies showing that effect of structural properties. Although materials have the same surface chemistry, physical factors such as roughness, texture, or crystal structure can result in variations in water contact angle values [65].

SEM micrographs of E. coli O157:H7 strain attached to three different types of silica-based materials upon dip-inoculation are shown in figures 5(a)–(c). Bacterial adhesion was greatest to hydrophilic bare quartz with a mean density of 5.79 ± 0.01 log cells mm⁻². Inoculation of hydrophobic quartz with bacteria resulted in a mean density of 5.02 ± 0.03 log cells mm⁻², which corresponds to 82.95% ± 0.90% reduction in bacterial attachment in comparison to bare quartz. Bacterial adhesion on HNSA was much less, with a mean density of 2.67 ± 0.24 log cells mm⁻², indicating a reduction of 99.91% ± 0.05% in bacterial attachment in comparison to bare quartz, as shown in figure 5(d). Similar reduction trends were also confirmed by fluorescent E. coli O157:H7 EDL933 strain (see supplementary data, section 3 and figure S3).

Zoom In Zoom Out Reset image size

We also investigated the attachment of Gram-positive bacteria, S. aureus, to each silica-based surface using the same inoculation conditions described above (figure 6). Mean densities of bacteria present on hydrophilic bare quartz, hydrophobic quartz, and HNSA were 6.05 ± 0.02 log cells mm⁻², 5.23 ± 0.04 log cells mm⁻², and 2.89 ± 0.10 log cells mm⁻², respectively. These data indicate that bacterial adhesion was reduced by 84.90% ± 1.52% on hydrophobic quartz compare to bare quartz. The reduction reached up to 99.93% ± 0.02% on HNSA. Statistical analysis revealed that bacterial (i.e., E. coli O157:H7 and S. aureus) reduction on each surface with respect to bacterial types were not statistically significant (p ≥ 0.05), which means they show similar reduction behavior.

Zoom In Zoom Out Reset image size

Differences in bacterial adhesion behavior were observed on hydrophilic and hydrophobic surfaces. This phenomenon can be explained in terms of the hydrophobic effect. It is known that bacteria with high surface hydrophobicity adhere more extensively on hydrophobic surfaces while bacteria with low surface hydrophobicity prefer hydrophilic surfaces [66]. Hence, considering that E. coli O157:H7 and S. aureus are hydrophilic with water contact angles of 15°–32° [67], our observed trends are in accordance with prior studies.

However, the reduction in bacterial adhesion to HNSA compared to hydrophobic quartz cannot be entirely explained by the hydrophobic effect. We attributed the superior bacterial antiadhesive characteristics of HNSA to two phenomena: (i) the formation of air pockets (bubbles) upon contacting with aqueous suspensions and (ii) the reduction of van der Waals interactions between bacteria and HNSA due to the porosity. First, considering the highly hydrophobic nature of HNSA, the trapping of air pockets (bubbles) inside surface roughness is likely to occur. Such gas pockets have previously been reported to prevent bacteria to reach the crevices and valleys of surfaces [68]. This phenomenon reduces the effective contact area between bacteria and solid surfaces, thereby leading to the reduced probability of adhesion. Second, since both hydrophobic quartz and HNSA were functionalized with methyl groups, van der Waals interactions are expected to make a significant contribution to the bacterial adhesion on these surfaces. As a first approximation, van der Waals interactions are additive (body forces), and can, hence, be calculated by integrating the pairwise potentials from individual dipoles that are distributed over the volume of bacteria and HNSA [69, 70]. Hence, one can expect that the attractive van der Waals interactions between bacteria and HNSA will significantly weaken due to the presence of pores in HNSA. However, this is only true when pores are filled with air (or empty), not filled with water. To check this point, using spectroscopic ellipsometry, the refractive index of HNSA was measured under dry conditions and after submerging in bacterial suspensions for 4 h and found to be 1.008 ± 0.001 and 1.013 ± 0.001, respectively. These values are much lower than the refractive index of nonporous silica materials (i.e., 1.460–1.540), indicating that almost all pores are filled with air.

To confirm that our observation is due to the antiadhesive properties of material not antimicrobial activity, additional screening tests were performed. Studies of E. coli O157:H7 and S. aureus growth were carried out in the presence of HNSA and 1% bleach solution by pour plate method (table 1). Comparison of bacterial colony-forming unit (CFU) revealed that 8–9 log reduction was observed for samples containing 1% bleach solution. On the other hand, bacterial growth behavior was similar in the absence (negative control) and in the presence of HNSA. Overall, antimicrobial activity was not detected in samples tested with HNSA, showing bacterial antiadhesive properties of HNSA is indeed responsible for the observed inhibitory effects.