A Comprehensive Study of the Degradation of Veterinary Antibiotics by Non-Thermal Plasma: Computational, Experimental, and Biotoxicity Assessments

Abstract

Water quality is threatened by numerous pollutants, among which antibiotics are of great concern due to their widespread use and unaltered excretion, leading to water contamination and fostering antibiotic resistance. To comprehensively address sustainable water remediation, herein, the susceptibility to non-thermal plasma degradation of two veterinary antibiotics (Oxytetracycline (OTC) and Lincomycin (LNC)) are compared in an integral approach, including computational analyses, plasma irradiation assays, and a byproduct toxicity assessment. The computational assessment was performed by evaluating the ionization potential (IP) obtained from Density Functional Theory calculations and determining the antibiotics’ susceptible sites for radical attack. Plasma irradiation achieved nearly complete degradation (~100%) of both compounds with the initial concentration of 1 mg L⁻¹, while 60% degradation was observed when the starting concentration was 10 mg L⁻¹. The mineralization rates were 21% and 31% for OTC and LNC, respectively. The degradation profiles followed similar trends, as expected from their comparable IP values. After treatment, the solution exhibited lower biotoxicity compared to the original antibiotics. Therefore, this work represents a step forward in addressing one of the key environmental challenges of our time and encourages further extending the analysis towards the remediation of water polluted with many other organic compounds.

1. Introduction

Nowadays, water quality is threatened by numerous pollutants, such as pharmaceutical products, agrochemicals, and plastics [1,2]. In recent years, the concern about the presence of antibiotics in surface waters has intensified. Inadequately treated effluents from pharmaceutical industries and hospitals, and residues from livestock farms, account for the main sources of these emerging contaminants in water. Antibiotics used in veterinary practice are in the majority because of their growing use in preventive treatment for infectious diseases, including their incorporation into animal feed. In fact, veterinarians prescribe 50% of all antibiotics globally [3]. The implications of their excessive use in veterinary care become even more alarming considering that approximately 20% of antibiotics are unaltered when excreted in urine and feces, ultimately leading to the contamination of soils, rivers, and lakes and fostering antibiotic resistance [4,5]. Antibiotics in natural waters can alter the microbial flora, even if present at trace levels, leading to a significant negative environmental impact on ecosystems. Therefore, the development of technologies for remediating water impacted by antibiotics has emerged as one of the key challenges of our time. It is crucial that these technologies do not involve the use of unsafe or non-environmentally benign substances and that they are sustainable from both energetic and economic perspectives [6,7].

Technologies for the effective removal of these pollutants from water include advanced oxidation processes (AOPs) such as ozonation, Fenton and photo-Fenton processes, photochemical oxidation, photocatalysis, and plasma [8,9]. Among the AOPs, plasma offers numerous advantages: (i) it consumes significantly less energy than ozonation and UV radiation; (ii) its efficiency surpasses that achieved with TiO₂ photocatalysis; (iii) it does not require the addition of chemical compounds such as TiO₂ for TiO₂ photocatalysis or ions like Fe for Fenton and photo-Fenton techniques; (iv) it has a higher degradation power than UV radiation; and (v) it does not face limitations such as the separation of catalysts used in photocatalysis [10,11].

Plasma is generated by the ionization of a gas when sufficient energy is supplied, allowing an electric current to pass through it. When an aqueous system is irradiated with plasma, $OH•$ radicals are formed at the plasma–water interface, which are the main species responsible for organic compounds’ degradation [12]. The reactions of $OH•$ with organic compounds produce organic radicals ($R•$). Additionally, the radicals generated can react with oxygen to form superoxide radicals ($O_2^{-}•$), peroxides ($ROO•$), alkoxides ($RO•$), and others. This complex sequence of chain reactions leads to processes that can involve both the formation and breaking of covalent bonds. The progressive oxidation eventually yields CO₂ and H₂O as final products (mineralization) [13,14].

It is a well-known fact that the efficiency of plasma irradiation processes for the degradation of organic pollutants from water depends on the chemical structure of the contaminants involved. To assess the susceptibility of organic compounds to radical attack and predict the efficiency of plasma irradiation for organic compounds’ degradation in water, molecular modeling can be performed. In this context, the use of DFT (Density Functional Theory) for the calculation of the ionization potential (IP) and Fukui indexes have been reported to accurately predict the relative sensitivity and identify the most susceptible sites of the target compounds to radical attack in AOPs [15,16,17]. According to Koopmans’ Theorem, the HOMO energy can be related to the ionization potential (IP), which is associated with the degradation rates of systems subdued to oxidation processes, as reported in the literature [18]. In addition, Fukui indexes are frequently used to determine the most susceptible sites to radical attack, since they allow the identification of the atoms in a molecule with a greater tendency to accept or donate an electron [19,20,21]. The higher the tendency to lose or accept an electron, the greater the tendency of a molecule to become polarized in the presence of an external field or due to an electronic density change. The information obtained from both calculations (i.e., IP and Fukui indexes) provides valuable information about the susceptibility of organic compounds to AOPs.

However, full mineralization of the target pollutants is not commonly achieved in AOPs, and byproducts may exhibit toxicity levels similar to or higher than the parent compounds. Hence, it is crucial to carry out biotoxicity studies to assess whether the degradation process harms or benefits the water ecosystem.

In this work, the feasibility of using non-thermal plasma irradiation for the treatment of water polluted with two antibiotics widely used in the veterinary field (Oxytetracycline (OTC) and Lincomycin (LNC) (Figure 1)) was computationally and experimentally studied. Their relative sensitivity to plasma irradiation was calculated using the Density Functional Theory (DFT). In addition, aqueous solutions of both antibiotics were subjected to plasma irradiation, and the degree of degradation and mineralization was evaluated. Furthermore, since byproducts generated by plasma irradiation of organic compounds may exhibit increased biotoxicity compared to the original substances, biotoxicity assays were also conducted before and after the treatment [22,23,24]. To our knowledge, it is the first time that these two antibiotics are compared in an integral approach that includes molecular modeling of the feasibility of being degraded by plasma irradiation, an experimental corroboration of the computational result, and a biotoxicity evaluation of the byproducts.

All in all, this work represents one step forward in the application of non-thermal plasma using a simple, easily scalable reactor to remediate water polluted with organic compounds.

2. Computational and Experimental Details

2.1. Computational Modeling Details

Computational modeling was used to perform a conformational search of both antibiotics [25], optimize the geometries of the obtained conformers by applying the Density Functional Theory (DFT) method at B3LYP/6-311G (d,p) level [26,27,28], calculate their condensed Fukui indexes [29,30], and determine their ionization potentials [18,31]. Complete computational details, including methodology and parameters, are provided in Supplementary Materials, Section S1.

2.2. Experimental Details

2.2.1. Plasma Irradiation Experiments

The antibiotics were irradiated with non-thermal plasma generated in a trielectrode reactor operated in ambient air (Figure 2), which was designed by some of the authors of this work. The reactor is based on a surface dielectric barrier discharge (DBD) extended to a third electrode. The solution to be treated flows through a low-impedance channel, which acts as a third electrode connected to the ground, away from high-voltage DBD electrodes. A detailed description of the reactor can be found in the reference [32].

Briefly, the experimental setup consisted of a surface DBD developed between electrodes E1 and E2 by applying a high AC voltage signal V_ac. The third electrode E3, connected to the ground, consisted of an aluminum adhesive tape attached to a channel through which the water to be treated flowed. The V_ac voltage applied to the E1 and E2 is elevated with a high DC voltage set at 10 kV to provide the necessary electric field for the streamers to propagate towards E3. The reactor was operated at V_ac peak-to-peak voltage values around 11 kV and a frequency of about 10 kHz, which corresponds to the resonance of the system.

In the degradation experiments, 1 and 10 mg L⁻¹ aqueous solutions of OTC (Oxytetracycline hydrochloride—Sigma-Aldrich O5875, 3050 Spruce Street, Saint Louis, MO, USA) and LNC (Lincomycin hydrochloride—Sigma-Aldrich 62143, 3050 Spruce Street, Saint Louis, MO 63103, USA) were prepared using Milli-Q water (resistivity > 18 MΩ, Elga Classic UV MK2, Lane End Business Park, Lane End, High Wycombe, United Kingdom). Assays were performed in duplicate, employing 100 mL of the corresponding solution for each one. Samples were continuously recirculated through the reactor using a peristaltic pump at a 70 mL min⁻¹ rate. Aliquots of the samples were taken at specific times for their analysis. Throughout the experiments, pH and ORP were measured using an Adwa AD8000 pH/mV/EC/TDS/Temperature Bench Meter (Adwa, Szeged, Hungary).

2.2.2. Analytical Methods

The concentration of the antibiotics in the aliquots taken at different times during each experiment was determined as follows. For OTC, concentrations were determined by high-performance liquid chromatography (HPLC), using an HP1100 Series apparatus (Agilent Technologies, 76337 Waldbronn, Germany) with a Zorbax Eclipse XDBC18 column (4.6 mm × 250 mm, 5 µm) at 25 °C, coupled with an Agilent Technologies 1200 UV-Vis detector, working at λ = 354 nm, with a mobile phase of H₂O with TFA at pH 2 and Acetonitrile–Methanol (65:17.5:17.5).

For LNC, the concentrations were determined by a Waters Alliance 2695 High-Performance Liquid Chromatography (HPLC) system coupled with Waters Acquity TQ triple quadrupole mass spectrometer (MS/MS) (Waters Corporation, 34 Maple Street, Milford, MA 01757, USA). The column used was a reverse-phase Phenomenex (Torrance, CA, USA) Kinetex C18 (100 mm × 4.6 mm; 2.6 µm particle size). The mobile phase was prepared using 0.1% formic acid in water (A) and 100% acetonitrile (B), with the following gradients: 0–0.5 min, 2% B; 0.5–7 min, 2–80% B; 7.0–9.0 min, 80–98% B; 9.0–10.0 min, 2% B; 10.0–15.0 min, 2% B at a flow rate of 0.5 mL/min. The retention time of LNC was 5.7 min. Electrospray ionization (EI) was operated in the positive ion mode with capillary voltage of 1.5 kV, programmed at 150 °C. Desolvation temperature was programmed at 450 °C. The molecular ion m/z 406.98 [M + H]⁺ and product ion m/z 126.2 were selected with an optimized collision energy of 30 eV. The ionization energy, Multi-Reaction Monitoring (MRM) transition ions (precursor and product ions), capillary and cone voltage, desolvation gas flow, and collision energy were optimized by the Waters IntelliStart™ optimization software package (Empower 3 v1.71).

The extent of mineralization of the irradiated antibiotics was determined by using a Total Organic Carbon Analyzer Shimadzu TOC-VCPH (Jenck S.A., Av. Alvarez Thomas 228, Buenos Aires C1427 CCP, Argentina) (Software TOC-Control V 2.00, Shimadzu).

2.2.3. Data Analysis

The degradation D% was defined as

$$D\%={\displaystyle \frac{C_0-C}{C_0}}\times 100$$

(1)

where $C_0$ is the initial concentration of the antibiotics and $C$ is the final concentration after the irradiation time. Since including all the reactants and the byproducts can be difficult to do in AOPs, a simplified rate law can be used, which is defined as follows:

$$v={\displaystyle \frac{\partial C}{\partial t}}=-k_{\mathrm{a}}·C^{n_{\mathrm{a}}}$$

(2)

where $v$ is the degradation rate, $k_{\mathrm{a}}$ is the apparent rate constant, $C$ is the concentration at a given time $t$, and $n_{\mathrm{a}}$ is the apparent reaction order. The integration yields Equation (3) for $n_{\mathrm{a}}=$1 and Equation (4) for $n_{\mathrm{a}}\ne$1 [33].

$$C=C_0·\left(1-e^{-k_{\mathrm{a}}·t}\right)$$

(3)

$$C={\left({\displaystyle \frac{1}{\frac{1}{C_0^{n_{\mathrm{a}}-1}}+(n_{\mathrm{a}}-1)·k_{\mathrm{a}}·\mathrm{t}}}\right)}^{{\displaystyle \frac{1}{n_{\mathrm{a}}-1}}}$$

(4)

In addition, the energy yield ($Y(\mathrm{m}\mathrm{g}·{\mathrm{k}\mathrm{W}}^{-1}{·\mathrm{h}}^{-1}$), defined as the ratio between the degraded mass during plasma treatment and the energy spent, was calculated by the following equation:

$$Y={\displaystyle \frac{V·C_0·D}{P·t·100}}$$

(5)

where $V=0.1 \mathrm{L}$ is the treated volume, $C_0$ (mg ${\mathrm{L}}^{-1}$), is the initial concentration of the antibiotic, P $\left(\mathrm{k}\mathrm{W}\right)$ is the average input power which was around 1 $\mathrm{W}$, and t is the treatment time in hours.

2.2.4. Biotoxicity Assay Details

The biotoxicity of the samples before and after the plasma treatment was assessed according to the Dutka method [34,35] using Lactuca sativa seeds. Briefly, for each assay, 20 seeds were placed on a filter paper in a Petri dish and put into contact with 5 mL of one of the following solutions: Na₂SO₄ 0.1 M (positive control), ultrapure water (negative control), or the antibiotic solution (with or without being subdued to plasma irradiation). The Petri dishes were kept in the dark for 5 days, after which seed germination and root elongation were measured, and the following indexes were calculated: the relative seed germination ($RSG$), growth relative to the radicle ($GRR$), and germination index ($GI$):

$$RSG={\displaystyle \frac{number of germinated seeds in the sample}{number of germinated seeds in the negative control sample}}·100$$

(6)

$$GRR={\displaystyle \frac{Length average radicle in the sample}{Length average radicle in the negative control sample}}·100$$

(7)

$$GI={\displaystyle \frac{RSG·GRR}{100}}$$

(8)

$GI$ values can be used as indicators for low or “L” (near 100%), moderated or “M” (below 50%), and high or “H” (near 0%) toxicity.

3. Results and Discussion

3.1. Computational Modeling

After the optimizations, the number of conformers that were within the feasibly accessible energy window (i.e., 6 kcal/mol, Supplementary Material S1 [25]) was two for OTC and five for LNC. All conformers were verified to correspond to energy minimum points and were weighted according to Boltzmann factors to determine which had the most significant contribution (Supplementary Materials, Figures S1 and S2 and Table S1). For each antibiotic, the conformer with the lowest energy is shown in Figure 3, together with their corresponding electrostatic potential mapped on the isodensity surface at a 0.004 value. In these maps, a color spectrum is used to highlight the intensity of the electrostatic potential of each region of the molecules, ranging from red (lowest electrostatic potential, associated with the maximum negative charge density) to blue (highest electrostatic potential, associated with maximum positive charge density) [36]. As expected, the maximum positive electrostatic potential corresponds to the location of the positively charged ammonium ions.

On the other hand, as an electron-deficient species, free radicals are a short-lived species with a high tendency to react and stabilize [37,38,39], leading to processes that involve the breaking and formation of covalent bonds. In the case of plasma irradiation, since radicals generated by this AOP are mainly ROS (e.g., $OH•$), the target molecules could undergo either degradation processes or the incorporation of oxygenated functional groups.

In order to identify the sites with the highest potential reactivity towards radical attack for each antibiotic, their Fukui functions ${(f}^0\left(r\right))$ were calculated and plotted (Figure 4). In addition, the corresponding condensed Fukui indexes ($f_x^0$) were also calculated and are presented in

Table 1. Condensed Fukui index values using Mulliken charges for OTC and LNC.

Antibiotic	Atom		${\mathit{f}}_{\mathit{x}}^{+}$	${\mathit{f}}_{\mathit{x}}^{-}$	${\mathit{f}}_{\mathit{x}}^{\mathbf{0}}$
OTC	12a	C	0.023	0.010	0.016
	6	C	0.005	0.017	0.011
	4a	C	0.008	0.008	0.008
	16	N	0.010	0.001	0.005
	11a	C	−0.004	0.007	0.002
	29	C	0.003	0.000	0.001
LNC	20	N	0.138	0.039	0.089
	29	C	0.097	0.003	0.050
	26	N	0.117	−0.049	0.034
	16	O	0.007	0.026	0.017
	3	O	0.008	0.019	0.014
	24	C	0.061	−0.034	0.013

. The most susceptible sites to radical attack obtained from the Fukui functions are shown in purple in Figure 4, and, in all cases, these sites have highly condensed Fukui indexes [18,40].

The most susceptible regions to both antibiotics and plasma irradiation (i.e., bond cleavage and/or radical attack) predicted by the computational analyses carried out in this work are in agreement with the degradation pathways reported in the literature for OTC [41,42] and LNC [43], evidencing the usefulness of computational modeling as a predictive tool for the degradation of organic molecules subjected to AOPs.

Regarding the degradation rates, taking into account that lower degradation times are associated with lower ionization potential values [18,29], in this work, the vertical first ionization potential ($IP$) of both antibiotics was calculated from the HOMO energy values ($E_{HOMO}$) according to the Koopmans’ Theorem (Equation (9)) [18,31], and the results are shown in

Table 2. $HOMO$ energies and ionization potential ($IP$) for the antibiotics.

Antibiotic	${\mathit{E}}_{\mathit{H}\mathit{O}\mathit{M}\mathit{O}}$		$\mathit{I}\mathit{P}$
	[kcal/mol]	[eV]	[eV]
OTC	−149.2	−6.5	9.0
LNC	−145.6	−6.3	8.8

.

$$IP=1.3124 \times (-E_{HOMO})+0.514 \mathrm{e}\mathrm{V}$$

(9)

Since the IP values for both antibiotics are comparable, similar degradation times for both OTC and LNC can be expected in AOPs.

3.2. Experimental Results

3.2.1. Plasma Irradiation Assays

Antibiotic solutions with initial concentrations of 10 and 1 mg L⁻¹ were irradiated with non-thermal plasma generated in a trielectrode reactor open to the air. The pH and the Oxidative Reduction Potential (ORP) were monitored during the experiment. The antibiotics solutions had an initial pH of 6.5 and decreased during the experiment until a final value of 3.5 (Figure 5) due to the generation of acidic species as a result of the antibiotics’ degradation [44,45]. In addition, the Oxidative Reduction Potential (ORP) remained positive during treatment (20 to 200 mV), which is consistent with the presence of reactive radicals of oxidative behavior [46].

The concentration of the antibiotics in aliquots taken at different times during each experiment was determined, and the degradation percentage (D%) was calculated (Section 2.2.3, Equation (1)). The results for OTC and LNC are shown in Figure 6. Both OTC and LNC follow comparable degradation trends, with similar D% at every time. For C₀ = 10 mg L⁻¹, the maximum D% reached was approximately 60% at 90 min of treatment, whereas for C₀ = 1 mg L⁻¹, nearly complete degradation was achieved at 60 min.

It is worth noting that comparable degradation trends for both antibiotics assessed in the irradiation experiments were predicted by the molecular modeling studies, whose calculations resulted in similar IP values. These results highlight the accuracy of the molecular modeling predictions and the advantage of using these tools to anticipate the outcome of irradiation experiments.

The degradation curves were fitted to pseudo-order kinetic models (Equations (3) and (4)), and the results are shown in Figure 7. As illustrated in the figure, for both antibiotics, the degradation rate decreased as the initial concentration increased.

The dependence of the degradation rate and treatment efficiency of plasma irradiation on the initial antibiotic concentration can be attributed to an increase in the byproduct’s concentration as the initial concentration rises, resulting in competition with the target compounds for reactive radical species [41,45,47].

It is noteworthy that despite the lower rate constant, the amount of antibiotics degraded by plasma is larger in the case of the higher initial concentration. Herein, an increase in energy yield was observed as the concentration of the antibiotic in the irradiated solution increased. Specifically, the energy yield for the OTC and LNC at C₀ = 1 mg L⁻¹ was Y = 100 mg/kWh, and for the higher concentration, the yields were 390 and 403 mg/kWh for OTC and LNC, respectively.

On the other hand, the mineralization extent of OTC and LNC solutions with an initial concentration of 10 mg L⁻¹ was analyzed after 90 min of irradiation. The data showed 21% and 31% of mineralization for OTC and LNC, respectively. These results are consistent with literature reports indicating that complete mineralization of these antibiotics using AOPs has not been achieved yet [48,49,50]. In fact, some studies report that β-lactam antibiotics, such as OTC, only reach 25–30% mineralization, reflecting a slow progress toward complete degradation [8,45].

3.2.2. Biotoxicity Assays

To evaluate the potential adverse effects of these antibiotics on biota under extreme exposure conditions, biotoxicity assays using the Dutka method were carried out [34]. The assay evaluates the inhibition of germination and the growth of the radicle and hypocotyl of Lactuca sativa seeds as indicators of sublethal toxicity, making it particularly responsive to low levels of toxic compounds. While it is important to acknowledge that not all contaminants impact organisms in the same way, this method has proven to be a reliable early indicator of environmental toxicity, supported by extensive literature validating its application across various environmental monitoring matrices [35,51,52,53,54,55].

An initial study was carried out employing high concentrations of both OTC and LNC (1000 mg L⁻¹). Despite these values being higher than typical concentrations found in wastewater, they are able to show minimal differences in biotoxicity [34,35,45]. This approach was used to determine if such high levels significantly affect the germination and growth of Lactuca sativa seeds. The results of the seeds exposed to the solutions are shown in Figure 8, together with the negative control (H₂O) and the commercial seed.

The figure clearly illustrates how both antibiotics inhibited the normal development of Lactuca sativa.

Additionally, as mentioned previously, numerous reports have pointed out that byproducts generated during plasma irradiation can be equally or even more toxic than the original compounds [24]. Therefore, the biotoxicity assessment of the OTC and LNC samples before and after the plasma treatment (for both C₀ = 10 mg L⁻¹ and C₀ = 1 mg L⁻¹) was carried out. Figure 9 shows the appearance of Lactuca sativa seeds’ growth after the 5-day experiment.

In

Table 3. Dutka method results for OTC and LNC (1 and 10 mg L⁻¹) before and after plasma irradiation. *** indicates statistically significant differences compared to water (p ≤ 0.001) based on Games–Howell post hoc test. No statistically significant differences were observed for the other samples compared to water or between each other.

Sample	Number of Non-Germinated Seed	Average Radicle Length [cm]
H2O	1 ± 1	3.1 ± 0.5
OTC 10 mg L−1 non-irradiated	4 ± 1	1.9 ± 0.3 ***
OTC 10 mg L−1 irradiated 90 min.	2 ± 1	2.7 ± 0.3
OTC 1 mg L−1 non-irradiated	3 ± 1	3.0 ± 0.4
OTC 1 mg L−1 irradiated 90 min.	2 ± 1	3.5 ± 0.4
LNC 10 mg L−1 non-irradiated	2 ± 1	3.1 ± 0.5
LNC 10 mg L−1 irradiated 90 min.	2 ± 1	3.6 ± 0.5
LNC 1 mg L−1 non-irradiated	1 ± 1	3.3 ± 0.5
LNC 1 mg L-1 irradiated 90 min.	1 ± 1	4.0 ± 0.5

, the number of non-germinated seeds and average radicle length are shown. As can be seen in the table, the results obtained for the irradiated samples of OTC and LNC for both 10 and 1 mg L⁻¹ have no significant differences compared with those obtained for the negative control (H₂O).

Based on the results obtained from the experiment, the germination indexes (GI%) were calculated according to Equation (8) (Experimental Details Section 2.2.3) and are shown in Figure 10.

It is noteworthy that the biotoxicity of the solutions after the plasma irradiation was closer to the lowest toxicity reference value (100% GI).

4. Conclusions

Molecular modeling accurately predicted the degradation rates of the veterinary antibiotics OTC and LNC under non-thermal plasma irradiation, identifying their most susceptible degradation sites. Both antibiotics had similar ionization potentials and degradation curves and reached comparable final degradation percentages. Such findings highlight the potential of molecular modeling as a valuable tool for predicting the plasma degradation susceptibility of organic compounds.

Regarding plasma irradiation, this simply designed reactor offered several advantages: (i) no need for inert gases, (ii) room temperature operation, (iii) easy scalability akin to a flowing water system, and (iv) low energy requirements. It provided promising results using an environmentally friendly technology that could be accurately predicted through molecular modeling.

The TOC results confirmed the greater stability of the tetracycline four-ring structure in oxytetracycline (OTC) compared to lincomycin (LNC). In addition, none of the solutions after plasma irradiation showed biotoxicity against Lactuca sativa.

Overall, the outcome of this work offers valuable insight into the remediation of organic compounds-polluted water using a comprehensive approach, including molecular modeling, plasma irradiation, and byproducts biotoxicity studies.