*2.4. Evaluation of the Antibacterial Activity of the Geopolymeric Materials*

The geopolymer samples were exposed to the bacteria under study by activating the materials with UV-A light for 48 h. The UV-A radiation was provided by two mercury lamps (Electrolux T8 20 W BLB) located inside a black acrylic dome. The lamps emitted light at an intensity of 10.3 W·m<sup>−</sup>2, which was measured with a Delta Ohm HD 2102.2 Photoradiometer using the filter for UV-A light range (*k* = 360 nm) at 5 mm [14]. The growth inhibition capacity of the bacteria in the materials was studied using *E. coli*, *S. aureus*, and *P. aeruginosa* bacteria. The tests were performed in two stages. The first stage followed the standard to evaluate the bactericidal capacity with the bacteria to be tested in the GP (with 10 wt.% TiO2) and mGP (with 5 wt.% CuO) samples. This initial stage was performed by following the bacterial growth inhibition zone method (disc diffusion agar test). This method consisted of placing the sample, i.e., the GP and mGP disc, in a Petri dish and spreading nutrient agar on the sample. The Petri dishes were then covered for 24 h at 25 ◦C, and the bacterial growth was verified by the formation of an "inhibition halo" around the disc and under the disc. If the material is favorable to bacterial proliferation, more bacteria would be found at the site where the material rests (Figure 3).

**Figure 3.** Assembly of the nutrient agar method.

By considering the results of the initial stage, the ability of the geopolymer paste (GP) and the mortar with glass (GP-G) to inhibit bacterial growth was evaluated in the second stage. During this stage, the methodology used was based on the Hasmaliza study [9], with a few modifications. The bacteria were cultured in tryptic soya broth (Merck, Darmstadt, Germany) for 24 h. At the beginning of the test, a count was performed in each of the broths by performing six serial dilutions in plate count agar (PCA, Scharlau 01-161). Subsequently, 100 μL of each of the selected bacteria were placed on discs, which were exposed for 2 to 24 h at room temperature (25 ◦C) under conditions isolated from contamination. After each of the exposure times, the discs were washed with 5 mL of a 1% peptone-buffered saline solution (Difco, Waltham, MA, USA), which was performed with vigorous stirring for 5 min; finally, the six dilutions were performed in a type-2 laminar flow cabinet (C4, Colombia) to count the number of colony-forming bacteria using the pour plate procedure with PCA. Finally, each agar with its respective inoculum was left in the incubator (WTC Binder, Germany) for 24 h at 37 ◦C to subsequently count the bacteria grown for each material in the different dilutions.

### **3. Results and Analysis**

### *3.1. Physicochemical and Microstructural Characterization of the Starting Materials and the Produced Geopolymers*

The chemical composition of MK and G is shown in Table 2. MK was basically composed of silica and alumina oxides (approximately 97 wt.%) and contains TiO2 impurities (1.73 wt.%). This oxide was in the form of anatase, as seen in the XRD results (Figure 4). The results shown in Table 2 indicate that the glass waste used in G was from the SiO2–Na2O–CaO system (calcium sodium glass) and was highly amorphous, as corroborated by the XRD results based on the elevation of the baseline for 2θ angles between 20◦ and 30◦ (Figure 4).

**Figure 4.** X-ray diffraction (XRD) patterns of the raw materials and GP.


**Table 2.** Chemical composition of metakaolin (MK) and glass waste (wt.%).

Titanium oxide existed in the anatase phase (TiO2, Inorganic Crystal Structure Database ICSD # 154604), as identified by the peaks observed in the diffractogram (Figure 4) at approximately 2θ = 25.3◦, 37.8◦, 48.1◦, 53.9◦, and 55.1◦, which also indicated the high photocatalytic potential of TiO2. The XRD pattern of the copper oxide nanoparticles synthesized from copper acetate by the Pechini method showed peaks at 2θ = 34◦, 36◦, 38◦, 49◦, 53◦, and 58◦, which corresponded to pure monoclinic CuO (ICSD code 67850), a phase considered to have bactericidal properties [42]. Similar results were found by Román et al. [15]. The XRD pattern obtained for the GP indicated that the material was amorphous (see the lifted baseline for 2θ between 25◦ and 35◦), with the only crystalline phase being the anatase phase, which was contributed to a high degree by the TiO2 particles and to a lesser extent by MK.

SEM was used to determine the particle size of the TiO2 and CuO particles (Figure 5); the average TiO2 and CuO particle sizes were 0.615 μm and 0.198 μm, respectively. Figure 5a shows the elongated and irregular shape of the TiO2 particles, while Figure 5b shows that the CuO particles has a rounded and more homogeneous shape.

**Figure 5.** SEM images of the (**a**) TiO2 and (**b**) CuO particles.

The FTIR results obtained for MK, GP, and GP-G are shown in Figure 6. In the MK spectrum, a large band corresponding to asymmetric vibrations, specifically, vibrations of the Si–O–Si and O–Si–O groups, was centered at 1089 cm−<sup>1</sup> [43]. The band located at 814 cm−<sup>1</sup> was associated with the vibrational mode of the Al–O bond of the AlIV present in MK [43,44].

When MK was activated under alkaline conditions to form a geopolymer, the main bands at wavenumbers of 1007 and 1015 cm−<sup>1</sup> shifted in both the GP and GP-G mortar spectra; this band corresponded to the asymmetric vibration of the Si–OT bonds in the GP gel potassium aluminum silicate hydrates (KASH) (where T corresponds to Si or Al tetrahedrons) [45]. According to Tchakouté et al. [46], the infrared absorption bands at wavenumbers close to 1100 cm−<sup>1</sup> and 950 cm−<sup>1</sup> could indicate that the geopolymers contained more SiQ3 and SiQ<sup>2</sup> species, respectively. Additionally, the band observed in the GP-G mortar spectrum was more intense than that observed in the GP spectrum, indicating the presence of more undissolved silica in the alkaline system [43]. This result was expected, since the mortar was designed with particles of G. The band located at approximately 703 cm−<sup>1</sup> came from symmetric vibrations of the presence of Ti–O–Ti photoactive species [10].

**Figure 6.** Fourier-transform infrared (FTIR) spectra obtained for MK, GP, and GP-G.

The peaks at approximately 3434 cm−<sup>1</sup> and 1645 cm−<sup>1</sup> in the MK, GP, and GP-G spectra corresponded to stress vibrations of the H–OH group, indicating that water molecules were associated with free water [38,47–49]. Finally, the peaks corresponding to the bending vibration of the Si–O–Si group [47,49] were identifiable at 473, 463, and 470 cm−<sup>1</sup> for the MK, GP, and GP-G samples, respectively.

The SEM images of GP and GP-G are shown in Figure 7. Figure 7a shows that the GP had a smooth and homogeneous surface, indicating the presence of large amounts of the KASH gel. Unreacted MK that was not dissolved by the alkaline activation conditions was also identified [11,44,47]. GP showed a homogeneous distribution of the TiO2 nanoparticles inside the matrix. In a previous study [11], the effect of the incorporation of TiO2 in the geopolymeric matrix was investigated, and it was noteworthy that there was no significant difference between the morphologies of the systems with and without TiO2, and that the formation of the KASH gel was not impeded by the addition of TiO2, as these coexisted in the structure of the material. Comparing the physical properties (Table 3) of the geopolymer matrix [11], with and without TiO2 addition, we found that the water absorption decreased by just 3.9%, and the density and volume of the permeable pores increased by 5.73% and 1.52%, respectively.

**Figure 7.** SEM images of the samples with 10 wt.% TiO2: (**a**) GP and (**b**) GP-G mortar.

The glass particles of the GP-G mortar (Figure 7b) were identifiable by their large sizes and angular smooth surfaces [41,47], which were due to the milling processes that these particles underwent. The presence of micro-cracks in the sample was also observed with the incorporation of glass (Figure 7b); similar results were reported by Lin et al. [47] and Mejía de Gutiérrez et al. [41]. The GP-G had a more homogeneous and denser surface similar to that reported by Lin et al. [47] and a higher density (see Table 3); the material was more compact, which was reflected in its lower water absorption capacity and lower volume of permeable pores.


**Table 3.** Physical properties of the GP and GP-G samples.

#### *3.2. Assays with Bacteria*

Figure 8 shows the results of the first experimentation stage, corresponding to the standardization method using Petri dishes and exposure in nutrient agar with the GP and mGP. As shown in Figure 8a, for the mGP, no inhibition halo was observed for *E. coli*. Although authors such as Román et al. [15] showed the capacity of CuO nanoparticles to eliminate *Ochrobactrum anthropi*, in the present study and under the composition used (5 wt.% CuO), CuO geopolymeric compounds with antibacterial capacity for *E. coli*, *S. aureus*, and *P. aeruginosa* were not obtained.

**Figure 8.** Standardization test for *Escherichia coli*: (**a**) GP (10 wt.% TiO2) and mGP (5 wt.% CuO); (**b**) inhibition of growth of bacteria under GP sample.

Previous studies demonstrated that the quantity of nanoparticles is an important parameter. The entrapment and inhibition of bacteria on surfaces was evaluated using CuO and Ag nanoparticles [2,22,33,34,42], and a good compatibility was reported in orthodontic adhesives using CuO in high quantities [17]. The amount of CuO nanoparticles used in this preliminary study was probably low and, for this reason, it did not allow for bacterial elimination. This is according to Guo et al. [14], who evaluated different percentages of TiO2 in coatings fabricated by immersion and reported that an incorporation of 5 wt.% TiO2 particles failed to eliminate *E. coli*, which was attributed to the processes of initial photocatalytic inactivation step being slow with this proportion of titanium oxide.

These nanoparticles were incorporated into materials because of their high surface/volume ratio, which allowed interactions with bacterial cell membranes, preventing these microorganisms from attaching to the material surface (bacteriostatic effect); when this occurred, the material was able to eliminate these microorganisms (bactericidal effect) [13,22–24,34]. Table 4 shows the results of the initial standardization assay for GP and mGP with the three bacteria tested.


**Table 4.** Results of the first bactericidal test phase: standardization method.

In the case of the geopolymer with TiO2 nanoparticles (GP), the standardization results were satisfactory for *E. coli*, in which a bactericidal effect was in fact observed for the sample GP. The inhibition was satisfactory, and a halo did form around the sample (Figure 8a); additionally, when sample GP was removed from the agar nutrient, no bacteria were found under the sample, as shown in Figure 8b. On the contrary, for the mGP, the halo formation was not evidenced; when the mGP sample was removed from the agar nutrient, an appreciable number of colonies grew below it for the three bacteria tested.

Based on the preliminary results of the first phase of standardization, the study was continued exclusively with GP samples; additionally, a GP-G mortar was fabricated and tested. The decision to use a geopolymer mortar with glass to measure the bactericide capacity of the geopolymers was based on the results reported by Mejía de Gutiérrez et al. [41], who observed a higher photocatalytic efficiency for B-rhodamine degradation in compounds functionalized with TiO2 when glass waste was used to prepare the geopolymer mortars. Likewise, studies by Sikora et al. [35] showed that the use of glass enhanced the photocatalytic activity for *E. coli* colony degradation, and elimination occurred in 30 min.

In the second phase, the number of CFUs was analyzed in different solutions for short times (4–5 h) and after one day in the agar with the *E. coli*, *P. aeruginosa*, and *S. aureus* bacteria.

Analyzing the GP-G mortar exposed to the different solutions for short times (Table 5) showed that, after 5 h, CFUs were not present the in the 10−<sup>7</sup> mL solutions for the three bacteria evaluated. Evaluating *S. aureus* exposed to GP (Figure 9) for 4 h in all solutions showed CFU growth, while, for the GP-G in the 10−<sup>7</sup> solutions, zero CFUs were evident. However, the GP exhibited significant antimicrobial activity for only the 10−<sup>7</sup> solutions containing *P. aeruginosa* (Figure 9). The results reported in this study corresponding to GP and GP-G are satisfactory. The toxicity of nanoparticles for inhibiting bacterial growth was previously demonstrated for *S. aureus* using ZnO nanoparticles by Sikora [35], and for *E. coli* using CuO by Kumar [20] and TiO2 by Sunada et al. [50]. Sunada et al. [50] explained that the photokilling reaction of *E. coli* cells using TiO2 is initiated by a partial decomposition of the outer membrane, followed by an attack of the cytoplasmic membrane, resulting in cell death. In general, the adhesion of nanoparticles onto the cell wall and the membrane of the microorganisms produces morphological changes characterized by shrinkage of the cytoplasm and membrane detachment, finally leading to rupture of cell wall [51]. Additionally, the photocatalysis generates reactive oxygen species (ROS) and free radical species, contributing to an increase in the antibacterial potential of the nanoparticles [34,51,52].

For the agar solutions exposed to GP for longer times (Table 6), the greatest inhibition occurred in the 10−<sup>6</sup> solutions for *Pseudomonas aeruginosa* (Figure 10) and *E. coli* (Figure 11). The highest efficacy for the inhibition of CFU bacterial growth was evidenced by the GP-G mortar in the 10−<sup>6</sup> solutions for the three bacteria evaluated. This result confirms that the incorporation of glass wastes in geopolymeric compounds increased the photocatalytic capacity and the bactericidal effect of the MK-based geopolymer supplemented with TiO2 and, therefore, improved the bacterial growth inhibition capacity on surfaces.

*Coatings* **2020**, *10*, 157

Finally, according to Tables 5 and 6, samples GP and GP-G showed a bacteriostatic effect depending on the bacteria and time.

Although TiO2 and other nanoparticles previously attracted a lot of interest due to their antibacterial properties in different applications, according to the results, in general, samples containing glass wastes showed a higher bacteriostatic effect. The use of ground glass particles maximized the capacity for inhibiting CFU growth after 5 h for the GP-G mortars in solutions of 10−<sup>7</sup> mL containing the three bacteria evaluated (*S. aureus*, *P. aeruginosa*, and *E. coli*). Furthermore, GP-G seems to be particularly effective against *P. aeruginosa*.


**Table 5.** Bacterial growth inhibition capacity of GP geopolymers (4 h) and GP-G (5 h) in bacteria-rich agar solutions.

\* (+), bacterial growth determined by CFUs; (++), bacterial growth determined by the formation of large colonies in all areas; (−), inhibition of bacterial growth.


**Table 6.** Bacterial growth inhibition capacity of GP (24 h) and GP-G (25 h) in bacteria-rich agar solutions.

**Figure 9.** Reduction in the number of viable bacteria over time in the *S. aureus* agar for (**a**) GP and (**b**) GP-G.

**Figure 10.** Reduction in the number of viable bacteria over time in the *P. aeruginosa* agar for (**a**) GP and (**b**) GP-G.

**Figure 11.** Reduction in the number of viable bacteria over time in the *E. coli* agar for (**a**) GP and (**b**) GP-G.

#### **4. Conclusions**

Geopolymer composites based on the alkaline activation of MK and nanoparticles of TiO2 (10 wt.%) and CuO (5 wt.%) were produced. This study evaluated the physical and microstructural properties of the geopolymers and the functionality of these geopolymers for inhibiting the growth of the bacteria *S. aureus, P. aeruginosa*, and *E. coli.* Based on the results obtained, the following conclusions can be drawn:


**Author Contributions:** Conceptualization, R.M.-d.G.; methodology, R.M.-d.G., M.V.-C. and M.A.; formal analysis, S.R.-B., M.A. and D.M.; investigation, R.M.-d.G., M.V.-C. and M.A.; resources, project administration and funding acquisition, R.M.-d.G.; writing—original draft preparation, M.V.-C. and S.R.-B.; writing—review and editing, R.M.-d.G. and M.V.-C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Universidad del Valle (Cali, Colombia), grant number 139-2017.

**Conflicts of Interest:** The authors declare no conflict of interest.
