Photocatalytic Degradation of Levofloxacin and Inactivation of Enterococci Levofloxacin-Resistant Bacteria Using Pure Rare-Earth Oxides

Abstract

In this study, La₂O₃ and CeO₂ nanopowders were prepared using a simple and cost-effective precipitation method. Wide-angle X-ray diffraction (WAXD), UV-Visible reflectance diffuses (UV-Vis DRS), Raman spectroscopy, and specific surface area were used to characterize the photocatalysts, evidencing that the used preparation method was effective in the generation of crystalline CeO₂ and La₂O₃. In particular, WAXD results showed that the average crystallite size of the achieved La₂O₃ and CeO₂ samples were about 22 nm and 28 nm, respectively. The photocatalytic performances of the prepared catalysts were investigated in the degradation of levofloxacin (LEV) and the inactivation of a waterborne pathogen levofloxacin resistant (Enterococcus faecalis ATCC 29212) by using a photoreactor equipped with a solar simulator (SS). After 120 min, the CeO₂ and La₂O₃ photocatalytic treatments allowed us to achieve between 75% and 83% of levofloxacin removal, respectively. A complete removal of 10⁶ CFU/mL Enterococcus faecalis ATCC 29212 was achieved after 5 and 60 min of La₂O₃ and CeO₂ photocatalytic processes, respectively.

1. Introduction

To avert the global water crisis and the shortage of natural resources, the development of sustainable and innovative strategies to treat and reuse wastewater is urgently required. Wastewater contains a wide category of contaminants, both organic and inorganic, including pathogens which could be released into the environment, threatening human health directly or indirectly when they are not properly removed. Among several contaminants of emerging concern (CECs), the occurrence of antibiotics in wastewater is gaining growing attention due to the limit of conventional wastewater treatment to remove such kinds of compounds [1].The chronic release of antibiotics into receiving water bodies leads to direct and indirect effects in human beings by bioaccumulation and antibiotic resistance. Recently, multiple genetic and genomic studies proved that wastewater treatment plants are sinks of resistant genes and organisms [2,3], increasing the spread of resistant strains into the environment. Enterococcus faecalis is a familiar urinary tract infection (UTI) pathogen of bacterial origin, recently recognized as an emerging concern due to its increasing resistance to antimicrobials, mainly vancomycin and levofloxacin [4]. To date, several conventional disinfection methods have been used to remove pathogens in wastewater, the most common being chlorination [5]. However, several potential ecological risks have been associated with chlorine, including the production of toxic disinfection byproducts (DBPs) [6]. Thus, research is pushing towards alternative treatments, including advanced oxidation processes (AOPs), which present excellent capability to degrade CECs and to inactivate waterborne pathogens through highly reactive oxidizing species [7,8,9]. Heterogeneous photocatalysis, with the application of semiconductor materials as photocatalysts, is one of the most prominent approaches to remove several contaminants of emerging concern, including waterborne pathogens from wastewater [10,11,12]. Titanium oxide (TiO₂) and zinc oxide (ZnO) are considered to be excellent photocatalysts due to their high stability higher efficiency, low cost, easy availability, lower toxicity, eco-friendliness, and highly oxidizing photo-generated holes [13]. However, there are several drawbacks to utilizing metal oxide for photocatalytic degradation, such as incomplete mineralization and a wide band gap [14], which limit their photocatalytic activity. Over the last few years, substantial efforts have been made towards study of extremely effective photocatalysts for the oxidative decomposition of organic pollutants to end products, such as carbon dioxide (CO₂) and water (H₂O) [15]. Specifically, rare earth (REEs) metal oxides have generated great interest as a practical way to hike the efficiency of photocatalytic processes. Studies focused on the use of lanthanides have demonstrated an improvement in wastewater treatments, as their use can produce, on average, 30% less sludge compared to other alternatives such as iron and aluminum [16]. The use of REEs, such as lanthanum (La), cerium (Ce), gadolinium, etc., could represent an alternative for reducing wastelands generated by sludge [16] and could improve the removal of organic pollutants by enhancing photocatalytic activity [17]. Specifically, rare earth semiconductors have gained great attention for their tunable band alignments, modulated atomic configurations, singular electronic states, and efficient optoelectric properties. For example, the particular electronic band structure of La₂O₃ allows for the formation of complexes with various Lewis bases, assuring an improvement of optical collection, which consequently promotes photocatalytic efficiency and stability [18,19]. On the other hand, the presence of crystal defects in the CeO₂ lattice due to the high amount of oxygen vacancies prevents the recombination of electron-hole ions, thus improving the photocatalytic degradation performance. Various physical and chemical processes can be used to produce REE nanoparticles; thermal decomposition, mechanical milling, and fame spray are examples of physical processes, [20,21] whereas hydrothermal, sol–gel, combustion methods, microemulsion, chemical co-precipitation, and vapor deposition are chemical processes [22]. Among the synthetic methods, the chemical precipitation process has several benefits, including low cost and the absence of hazardous substances [23]. In this work, lanthanum oxide (La₂O₃) and cerium oxide (CeO₂) were synthesized using a simple precipitation method for the photocatalytic degradation of levofloxacin (LEV) and the inactivation of a levofloxacin Enterococcus-resistant strain (Enterococcus faecalis ATCC 29212). To the best of our knowledge, this is the first study investigating the LEV degradation and the inactivation of a levofloxacin Enterococcus-resistant strain using pristine rare earths oxides.

2. Materials and Methods

The precursors for the design of the REE catalyst were cerium nitrate hexahydrate (Ce(NO₃)₃∙6H₂O, 99%; Sigma Aldrich) and lanthanum nitrate hexahydrate (La(NO₃)₃∙6H₂O, 99.9%; Sigma Aldrich). For the preparation of the Enterococci-selective agars, Tryptic Soy Broth (TSB, Difco, Becton-Dickenson Labs) and Slanetz and Bartley agar without TTC (Biolife, Monza, Italy) were used. Levofloxacin (98.0%) was purchased from Sigma Aldrich (Milan, Italy). All chemicals were analytical grade and were used as received without further purification. Deionized water was used throughout the experiments.

2.1. Catalysts Preparation Procedure

The precipitation method was used to synthesized single cerium and lanthanum oxide nanopowders (CeO₂, La₂O₃) (Figure 1) [24]. Ce(NO₃)₃∙6H₂O or La(NO₃)₃∙6H₂O were dissolved in Milli Q water up to the point where they reached a concentration of 0.2 M for each salt. Then, to the obtained solutions were added dropwise ammonium solution up to pHs 8–8.5. The stirring was continued for 15 min after the completion of precipitation. The obtained colloidal solutions were washed with Milli Q water followed by four cycles of centrifugation at a speed of 3000 rpm (for 15 min). The resultant products were then dried in an oven at 110 °C for 8 h and crushed to powders. The oven-dried precursors were calcined at 850 °C for 2 h in air atmosphere to obtain CeO₂ and La₂O₃ nanopowders.

2.2. Characterization Techniques

An automatic Bruker D8 Advance diffractometer (Billerica, MA, USA) with reflection geometry and nickel-filtered Cu-Kα radiation was used to obtain the wide-angle X-ray diffraction (WAXD) patterns. The lattice parameter values were determined using the following equations for the CeO₂ cubic structure (Equation (1)) and for the La₂O₃ hexagonal structure (Equation (2)), respectively:

$${\displaystyle \frac{1}{d_{hkl}^2}}={\displaystyle \frac{\left(h^2+k^2+l^2\right)}{a^2}}$$

(1)

$${\displaystyle \frac{1}{d_{hkl}^2}}={\displaystyle \frac{4}{3}}{\displaystyle \frac{\left(h^2+k^2+hk\right)}{a^2}}+{\displaystyle \frac{l^2}{c^2}}$$

(2)

where the value of $d_{hkl}$ for a WAXD peak was determined using Bragg’s law (Equation (3)):

$$2d_{hkl}\mathit{sin}\theta =\lambda$$

(3)

$h$, $k$, and $l$ are the crystal plane indices, $d_{hkl}$ is the spacing corresponding to the crystal plane ($h$ $k$ $l$), while $a$ and $c$ are the lattice parameters (for pure hexagonal phase of La₂O₃: $a$ = $b$ ≠ $c$, $\alpha$ = $\beta$ = 90° and $\gamma$ = 120°; for pure cubic fluorite structure of CeO₂: $a$ = $b$ = $c$, $\alpha$ = $\beta$ = $\gamma$ = 90°). To determine the lattice parameters’ values, all of the planes for the hexagonal phase of La₂O₃ and for the cubic fluorite structure of CeO₂ were considered. The average crystallite size of CeO₂ and La₂O₃ were calculated using the Scherrer equation [25] (Equation (4)):

$$D={\displaystyle \frac{K\lambda }{\beta \mathit{cos}\theta }}$$

(4)

where $D$ is the average crystallite size (nm), $K$ is the particle shape factor and taken as 0.89, $\lambda$ is the X-ray wavelength corresponding to the Cu-Kα irradiation (1.5418 Å), $\beta$ is the full width at half maximum (FWHM) of the diffraction peaks (radiant), and $\theta$ is the angle of Bragg diffraction [26]. The diffuse ultraviolet–visible reflectance (UV-Vis DRS) spectra of the samples, recorded with an RSA-PE-20 reflectance spectroscopy accessory (Labsphere Inc., North Sutton, NH, USA), were obtained using a Perkin Elmer Lambda 35 spectrophotometer (Waltham, MA, USA). The reflectance data were reported as the F(R∞) values, from Kubelka–Munk theory, versus wavelength. To estimate the indirect (Eg_i) and direct (Eg_d) band gap energies of the samples, [F(R∞) × hν] ^1/2 versus photon energy (hν) and [F(R∞) × hν]² versus photon energy (hν) were calculated. From the intersection of the straight line with the x-axis, the Eg_i and Eg_d values of the photocatalysts were determined. Raman spectra were obtained using a Dispersive MicroRaman spectrometer (Invia, Renishaw, Wotton-under-Edge, UK) using a laser with a radiation emission with a wavelength equal to 514 nm in the range of 100–1600 cm⁻¹. Fourier-transform infrared (FTIR) spectra were obtained with an FTIR (BRUKER Vertex70, Bruker, Karlsruhe, Germany) spectrometer equipped with a deuterated triglycine sulfate (DTGS) detector and a KBr beam splitter using KBr pellets. The spectra were obtained at a resolution of 2.0 cm⁻¹ in the range of 400–4000 cm⁻¹. The BET specific surface area (S_BET) of photocatalysts were measured using dynamic N₂ adsorption measurements at −196 °C using the Sorptometer KELVIN 1042 instrument (Milano, Italy). The pore volume and mean pore radius were obtained by using a Nova Quantachrome 4200 e Instrument analyzer (Rome, Italy), applying the Barrett–Joyner–Halenda (BJH) method. Before the analysis, the samples were pretreated in He flow at 150 °C for 0.5 h.

2.3. Experimental Set-Up for Photocatalytic Tests

All of the photocatalytic tests were carried out in a jacketed custom-made photoreactor. The reactor consisted of a 100 mL cylindrical pyrex glass reactor vessel equipped with solar simulator (SS) (Xenon lamp) that provides illumination approximating natural sunlight (light power of 250 W/m², with a spectral wavelength range of 320–430 nm). The UV-A part of the solar spectrum is responsible for the activation of REE nanopowders in photocatalytic processes. The temperature was kept constant at 25 °C by circulating the cooling water continuously into the modified double-hinge photoreactor. Furthermore, the solutions were kept in constant magnetic mixing during the execution of the photocatalytic oxidation experiments (60 min dark + 180 min light for E. faecalis inactivation, 60 min dark + 120 min light for LEV removal). The photocatalytic tests were carried out by adding 0.5 g/L of photocatalysts (CeO₂ or La₂O₃) to a solution of the selected target, starting from an initial concentration of 10⁶ CFU/mL for E. faecalis ATCC 29212 and 1 mg/L for LEV. The stirred mixture was left for 60 min in the dark to establish an equilibrium of adsorption between drug/bacteria and the catalyst. At defined time intervals (dark: 0, 5, 10, 15, 30, 60, 120, and 180 min), aliquots were sampled and centrifuged at 3000 rpm for 20 min.

In the case of the photocatalytic tests with LEV, a phosphate buffer with sodium dihydrogen phosphate/disodium hydrogen phosphate (NaH₂PO₄/Na₂HPO₄) was used to keep the pH constant at 7.00. For LEV detection, MeOH:10 mM CH₃COONH₄ (70:30 v/v) mixture was used for the mobile phase in the HPLC column, flowing at 1 mL/min through a Luna Phenomenex^® C18 (250 4.6 mm; 5 m) column. Per run, a volume of 20 μL of the sample was injected and disclosed with a UV detector set at 295 nm.

2.4. Bacterial Count and Inactivation Test

LEV resistance phenotypes were tested using the Kirby–Bauer method according to standard recommendations [27]. The test allowed for the selection of the Enterococcus faecalis ATCC 29212 strain as resistant to levofloxacin. Briefly, the colonies, before the treatment, were transferred to appropriate culture broth for growing. Bacterial strains were grown to the exponential phase in Tryptic Soy Broth at 37 °C overnight. After 24 h, via spectrophotometric method at 590 nm, the density was adjusted to obtain a suspension of 0.5 McFarland (standard turbidity), corresponding approximately to 1–2 × 10⁸ CFU/mL suspension. The inactivation experiment was conducted by spiking an aliquot of bacterial suspension with an initial density of 10⁸ CFU/mL to the reaction solution, already added with the catalyst, in order to dilute it up to a concentration of 10⁶ CFU/mL. The stirred mixture was then sampled after 60 min in the dark and then at time intervals of 0, 5, 10, 15, 30, 60, and 180 min. Subsequently, through the technique for inclusion of the inoculum in a solidifiable substrate (“Pour Plate”), the sample to be analyzed was inoculated into an empty sterile 90 mm Petri dish, then, adding the agarized substrate (12–15 mL), was kept in fusion at 45–46 °C. For this procedure, a selective medium for the isolation and enumeration of fecal streptococci (Slanetz and Bartley agar) was used. The inactivation of total (initial concentration 1.0 × 10⁶ CFU/mL) Enterococci was evaluated after 24 h of incubation on this substrate. Bacterial count was performed in triplicate.

3. Results

3.1. Photocatalytic Materials Characterization

WAXD patterns of La₂O₃ and CeO₂ are shown in Figure 2.

The main diffraction peaks for La₂O₃ at $2\theta$ equal to 25.42°, 28.42°, 29.26°, 38.87°, 45.40°, 51.49°, 54.77°, 59.81°, 66.00°, 74.74°, and 78.60°, which refer to the crystalline planes (100), (002), (101), (102), (110), (103), (112), (004), (104), (211), and (114), respectively, are detectable. All of these signals are indicative of a hexagonal crystalline and correspond to the ICDD Card No. 01–083-1349 [26,28,29]. The diffraction peaks at 27.58°, 32.09°, 46.54°, 55.40°, 58.17°, 68.53°, 75.83°, and 78.22° are typical of CeO₂ cubic fluorite, and all diffraction signals are referred to (111), (200), (220), (311), (222), (400), (331), and (420) crystalline planes according to ICDD Card No (00-004-0593) [30,31].

Table 1. Crystallite size, lattice parameter, specific surface area (S_BET), direct band gap (Eg_d), and indirect band gap (Eg_i) of the samples.

Sample	D, nm	Lattice Parameters, A°		SBET, m2 g−1	Egd, eV	Egi, eV
		a = b	c
La2O3	22 ± 5	4.04 ± 0.05	6.17 ± 0.04	12	5.46	5.32
CeO2	28 ± 5	5.51 ± 0.05	−	2.4	2.81	2.59

reports the average crystallite size, the cell parameters, the specific surface area evaluated using the BET method (S_BET), the direct energy of band gap (Eg_d), and the indirect energy of band gap (Eg_i) for the two oxides. The lattice parameter for CeO₂ is about 5.51 Å (Table 1) [32]. The average crystallite size of the CeO₂ is about 28 nm, in agreement with the literature, in which the CeO₂ sample prepared with the co-precipitation method reported a similar value of crystallite size of 20 nm [30]. On the other hand, the average crystallite size of La₂O₃ is equal to 22 nm, a value lower than that reported in the literature (41 nm). $a$ and $c$ lattice parameters for La₂O₃ are about 4.04 Å and 6.17 Å, respectively (Table 1). The S_BET value of La₂O₃ is 12 m² g⁻¹ [33], while CeO₂ exhibits a very low S_BET value of about 2.4 m² g⁻¹ [34].

The evaluation of band gap energy (Figure 3) showed that the Eg_i and Eg_d values of the La₂O₃ sample are 5.32 eV and 5.46 eV (Table 1), respectively. On the other hand, CeO₂ sample exhibits lower Eg_i and Eg_d values, equal to 2.59 and 2.81 eV, respectively, indicating that both the catalysts are able to absorb only UV light [35].

Raman spectra in the range of 100–1600 cm⁻¹ for CeO₂ and La₂O₃ photocatalysts are displayed in Figure 4.

The CeO₂ spectrum exhibited a strong band at about 465 cm⁻¹, corresponding to the symmetrical stretching mode of the Ce-O8 units, typical of CeO₂ with fluorite type structures [36]. It is worthwhile to note the appearance of a red shift of the main peak from 461 cm⁻¹ (reported in the Raman spectrum of bulk CeO₂ nanoparticles [37]) to 465 cm⁻¹ detected in the CeO₂ Raman spectrum of Figure 4. This effect can be attributed to oxygen vacancies in the CeO₂ lattice leading to defective structures, inducing an enhancement of the photocatalytic activity. The La₂O₃ spectrum evidenced bands located at 101, 192, 407, 1087, 1205, and 1481 cm⁻¹. The most intense Raman signal located at 407 cm⁻¹ corresponds to the La–O vibration [38].

The FTIR spectra in the range of 400–4000 cm⁻¹ for CeO₂ and La₂O₃ photocatalysts are shown in Figure 5.

The spectrum of the La₂O₃ photocatalyst shows a band at 645 cm⁻¹ associated with the La-O stretching vibration [39]. The bands at about 1462 and 1387 cm⁻¹ are linked to COO⁻ functional groups [28]. Another sharp band sharp band at about 856 cm⁻¹ could be associated with associated with C–O bending vibrations [28]. The band observed at about 3600 cm⁻¹ evidences the presence of O–H stretching vibrations associated with water adsorbed on the La₂O₃ surface [40]. The spectrum of CeO₂ shows a broad band extending in the range of 3000–3610 cm⁻¹ due to the O-H stretching vibration of OH–groups. The signal at around 1579 cm⁻¹ is ascribed to the bending vibration of C-H stretching. Moreover, the stretching vibration of the Ce—O bond was observed at about 555 cm⁻¹ [30,41,42].

BET plots of La₂O₃ and CeO₂ are shown in Figure S1 in the Supplementary Materials and the obtained S_BET are reported in Table 1. It is possible to observe that the S_BET of La₂O₃ (12 m² g⁻¹) was significantly higher than of CeO₂ (2.4 m² g⁻¹). The pore volume and mean pore radius are reported in Table S1 in the Supplementary Materials. These results were in agreement with S_BET values. Indeed, the La₂O₃ pore volume and mean pore radius were higher than CeO_2.

3.2. Levofloxacin Photodegradation Results

Preliminary experiments were carried out under dark conditions to evaluate the possible extent of the adsorption process. Under these conditions, 19% of LEV removal was observed in the presence of CeO₂ (Figure 6A), whereas no removal was observed in the presence of La₂O₃ (Figure 6B).

After the dark step, the SS light was turn on and the photocatalytic process was brought ahead for 120 min with regular sampling at a determined interval time. The collected sample was then analyzed using HPLC-UV showing a gradual decrease in LEV concentration using both CeO₂ and La₂O_3.

In detail, after 120 min of irradiation time, the CeO₂ and La₂O₃ processes allowed us to achieve between 75% and 83% of LEV removal, respectively. The better photocatalytic performance observed in the presence of La₂O₃ can be ascribed to its higher BET surface area and pore volume.

It is worth noting that LEV photodegradation utilizing pure CeO₂ and La₂O₃ was never studied in the literature. Indeed, only papers dealing with the coupling of CeO₂ with other compounds for photocatalytic removal of this drug have been reported [43,44,45,46].

3.3. Photocatalytic Inactivation Results

The removal of the levofloxacin-resistant E. faecalis ATCC 29212 strain starting from an initial concentration of 1.0 × 10⁶ CFU/mL is shown in Figure 7. Preliminary tests were carried out in dark to assess the natural decay of the target bacteria. No significant effects could be observed. The CeO₂ system yielded a complete inactivation (2.5 log bacterial decrease) after 60 min. The La₂O₃ system showed a higher degradation rate given that the Enterococci inactivation occurs only after 5 min through a drastic decrease of approximately 2.1 log unit enterococci starting from the dark phase. Similarly, to the tests conducted on the photodegradation of LEV, this analysis on bacterial inactivation showed that the most effective system was the La₂O₃.

Based on these preliminary results, the combination of rare earth elements (REEs) with catalysts in AOPs is a promising treatment for the removal of both chemical and microbiological CECs, due to the exceptional properties of rare earth elements. Moreover, the possibility to use REE-combined catalysts deserve to be investigated in order to further improve the process behaviors.

4. Conclusions

Pure La₂O₃ and CeO₂ photocatalysts were synthesized by using a co-precipitation process, and the physical, chemical, and optical properties were studied with several characterization techniques. WAXD analysis confirmed the crystalline structure for both samples, showing the presence of all diffraction peaks related to the cubic fluorite and hexagonal crystalline structure for CeO₂ and La₂O₃, respectively, while UV–vis DRS spectra showed that both of the samples had a higher absorption contribution in the UV region. This optical feature of the samples was confirmed using Eg_i (2.59 eV) and Eg_d (2.81 eV) values for CeO₂ and Eg_i (5.32 eV) and Eg_d (5.46 eV) values for La₂O₃.

The preliminary study on the degradation of the drug levofloxacin highlights how REE-based catalysts can play an important role in the removal of CECs. After 120 min, the SS/ La₂O₃ process allowed for a removal (83%) higher than the SS/CeO₂ process (75%). Overall, the La₂O₃ system showed the best degradation performance in kinetic terms both with regard to the chemical and biological target. Moreover, the CeO₂ and La₂O₃ processes allowed us to achieve a complete inactivation of Enterococcus faecalis ATCC 29212 after 60 min and 5 min, respectively. The use of powder nano-catalysts represents one of the main limitations to full-scale applications of these processes, due to the difficulty to recover the catalysts at the end of the treatment. To overcome this drawback and optimize process behaviors, the immobilization of nanoparticles on macroscopic supports should be attempted.