Oxidative degradation of some antibiotics by permanganate ion in alkaline medium: A kinetic and mechanistic approach

Purpose: To investigate the kinetics of oxidative removal of two β-lactam antibiotics (A), namely, ampicillin and flucloxacillin. Methods: In this study, permanganate ion (MnO4-) was used as an oxidant in an alkaline medium at fixed ionic strength of 0.1 mol dm-3 and a temperature of 298 K utilizing a spectrophotometric technique. The obtained oxidation products were characterized using spot tests and FT-IR spectra. Results: The stoichiometry of the reactions was 1:4 (A : MnO4-). The reactions were a first order credence in [MnO4-] and fractional-first order kinetics in antibiotic and hydroxyl ion. Influence of ionic strength was successfully explored. Dependence of reaction rates on temperature was studied and the activation parameters were computed and discussed. A plausible mechanism for the oxidation reactions has been elucidated. A consistent rate-law expression was also derived. Conclusion: This study introduces a significant treatment method for antibiotic removal, thus helping to protect the environment and human health.


INTRODUCTION
Antibiotics are amongst the most noteworthy group of pharmaceutical drugs utilized in curing human and animals from fungal and bacterial inflammations [1]. These antibiotics are used also in food manufacturing [2]. Although, antibiotics are presumed to be drugs for human and animals, they are foreign matter to the body and must be eliminated from the body after their medical action via drug metabolism process.
Through drug metabolism, antibiotics undergo oxidation, reduction and hydrolysis, resulting in biotransformation of antibiotics in the body so that they can be eliminated more easily.
This process may lead to an active, inactive or pharmacological metabolite entering the environment and water cycle, and this can have a negative and harmful impact, disturbing the ecological balance [3,4]. Therefore, it is of utmost importance to find green, safe and powerful treatment methods to eliminate antibiotic residues present in the contaminated environment and water to protect the environment and human health [5,6]. Biotechniques are still limited and are not effective because of the complicated chemical structures of antibiotics [7]. It is very important to use improved treatments, or combine several biological, chemical and physical methods [8].
Antibiotics are susceptible to oxidation, leading to their degradation, termination of their action and/or their elimination from the body [9][10][11][12]. Although, oxidation is a relatively common method for drug decomposition, it has not been studied extensively because oxidative degradation can often be reduced to reasonable levels through various ways. One of the most presumed techniques for the removal of antibiotics is chemical oxidation, which is a widespread path for their decomposition, and they have a principal role in water treatment operations and in understanding the metabolism of such medications. The role of oxidizing agents during the oxidation process is the conversion of the poisonous substances into less harmful compounds in order to make them safe when they are discharged into the environment [11,12].
A review of literature revealed little reported investigations into the kinetics of oxidative degradation of antibiotic drugs in aqueous acidic or alkaline media [10][11][12][13]. In view of the previously mentioned aspects, the existing investigation is concerned with the kinetic mechanism of oxidative removal of two β-lactam antibiotics, viz, ampicillin and flucloxacillin (their structures shown in Figure  1) using permanganate ion (MnO4 -), which is a significant, powerful and inexpensive oxidant [14][15][16] in aqueous alkaline medium. This study aims to find out the reactive species of antibiotic bioreductants, as well as permanganate ion oxidant in alkaline medium, in order to to discover the selectivity of antibiotics towards MnO4in such medium. This study is also extended to delineating a mechanistic picture of the existing redox reactions and to derive the rate-law expression consistent with the acquired investigational kinetic outcomes.

EXPERIMENTAL Materials
Ampicillin and flucloxacillin antibiotics were purchased from Glentham Life Sciences. All other chemicals were purchased from Merck or Sigma in spectroscopic grade and were utilized as supplied. Antibiotic solutions were freshly prepared by dissolving their weighted samples (ampicillin, flucloxacillin sodium) in doubledistilled water. Potassium permanganate solutions have been prepared and consolidated as mentioned earlier [14]. NaOH solutions were employed to supply the necessary alkalinity. Sodium perchlorate solution was used to fix and explore the influence of the ionic strength (I). The temperature of all reactions were under control to within ± 0.1 o C.

Kinetic measurements
Kinetic measurements were ascertained in fractional-first order circumstances, where the antibiotics existed in significant excess, higher than that of [MnO4 -] at a fixed I and T. The measured UV-Vis absorption readings were done on a Shimadzu UV-1800 PC double-beam spectrophotometer. The reactions were tracked by recording the reduction of the absorbance of MnO4 -(at λ = 526 nm) with time. All experiments were performed three times and the rate constants were found to be reproducible in the range of ± 3 %.

Reactions stoichiometry and products identification
It is important to determine the stoichiometries of the overall reactions before considering the acquired kinetic data in more details, because of the complexity of the kinetics of these reactions due to construction of transient species of manganate (VI) and/or manganate (V) intermediate complexes [15]. Stoichiometric measurements were performed at [OH ─ ] = 5.0x10 -3 mol dm -3 with slightly excess permanganate concentration. The reaction mixtures were left in a dark place for 1 day. where, A refers to ampicillin and flucloxacillin ( Figure 1), and B and C are their identified oxidation products; the corresponding carboxylic acids and 5,5-dimethyl-thiazolidine-2,4dicarboxlic acid, respectively. Such oxidation products were identified employed spot tests and FT-IR spectra. The FT-IR spectra showed appearance of sharp band at around 1660 cm -1 which is characteristic for carboxylic C=O stretch in the product B.
Also, appearance of new broad band at around 3370 cm -1 which is due to carboxylic OH with N-H stretching confirming the product C. Ammonia and CO2 byproducts were identified by spot tests [36]. The same oxidation products were identified earlier for ampicillin [11,12] and flucloxacillin [13].

Spectroscopic changes
Initial experiments showed that the existing oxidation reactions were of such rates that can be monitored using the traditional spectrophotometer. Spectroscopic changes (200 -700 nm) during the oxidation of ampicillin (Amp) and flucloxacillin (Flx) by permanganate ion are shown in Figure 2 (a,b).
As manifested in these graphs, there is a progressive decay of MnO4 ─ ion band at  = 526 nm with a concurrent appearance of new bands at about 606 and 435 nm especially in case of ampicillin.

Dependence of oxidation rate on permanganate concentration
The pseudo-first order plots (ln Abs. vs. time) in the oxidation of ampicillin and flucloxacillin by permanganate ion in alkaline medium was examined at various [MnO4 ─ ]0, ranging between 1.0 x 10 -4 and 8.0 x 10 -4 mol dm -3 while other constituents were kept constant. The results indicated that the obtained pseudo-first order rate constant plots are straight lines for about 80% of the reactions accomplishment as illustrated in Figure 3. Also, changing the initial concentration of the oxidant didn't affect much on the observed rate constant values (kobs) as included in Table 1

Dependence of oxidation rates on [OH -]
The rates of oxidation of the investigated antibiotics by alkaline permanganate was determined at different pH values in order to clarify the reactions mechanism. The experimental observation indicated that the rates of oxidation reactions were increased with rising [OH -] as indicated from the kobs values included in Table 1. The plots of kobs versus [OH -] gave good straight lines with positive intercepts on the kobs axes, Figure 5(a). Also, the plots of log kobs vs. log [OH -] yielded straight lines with slopes of 0.894 and 0.883 as illustrated in Figure 5(b), indicating that such reactions were fractional-first orders in [OH -].

Dependence of oxidation rates on ionic strength (I) of reactions medium
In order to show the nature of the interactive species in the rate-determining step of the redox reactions and the suggested mechanism of the oxidation reactions, kinetic experiments were accomplished at fixed alkali and antibiotic concentrations, while the concentration of sodium perchlorate was increased. The obtained results indicated that the rates of the oxidation reactions were increased as I increased, as noticed from the values of kobs listed in Table 1. The Debye-Hückel plots were set to be linear with positive slopes as illustrated in Figure 6.

Dependence of oxidation rates on temperature
To determine activation parameters, the reactions were conveyed at a variety of temperatures, as well as at a fixed concentration of the other reaction ingredients. The experimental results showed that the rates of the oxidation reactions were found to increase by raising the temperature as listed in Table 2. On the other hand, Eyring and Arrhenius plots of the second order rate constant values (k2) were set to be linear as illustrated in Figure 7 (a) and (b), correspondingly. The activation parameters were assessed from the slopes and intercepts of these plots and are introduced in Table 3.

Polymerization
The eventual formation of free radicals in the oxidation reactions between the examined antibiotics and permanganate ion in an alkaline medium, was examined by conducting a polymerization test. This test was carried out by the addition of a definite acrylonitrile quantity to the reaction mixture for about 4 h. No polymerization appeared in all reaction mixtures (there were no white precipitates formed), implying that the existing oxidation reactions did not proceed via formation of the free radicals mechanism.

Suggested reaction mechanism
The obtained increase in the oxidation rates as a result of increasing alkali concentration, as well as the structures of the examined antibiotics [18], suggested a deprotonation of antibiotic molecules (AH) according to the following equation: The deprotonated form (A ─ ) appears to be the reactive species in the rate-controlling stage of the proposed reactions mechanism.
It is well known [19] that permanganate ion in strong alkaline solutions is reduced by either one-electron change to give manganate (VI) ion, or by simultaneous two-electron changes to form hypomanganate (V). But it is difficult to decide whether permanganate is reduced by one-or by two-electron transfer mechanism owing to the following fast reactions (Eqs 4 and 5).
MnO4 ─ + Mn V O4 3-= 2 Mn VI O4 2-…………. (4) 2 Mn V O4 3-= Mn VI O4 2-+ Mn IV …………… (5) A further reaction of Mn V O4 3will proceed at a much lower rate owing to its lower reactivity [20]. This fact may be considered as good evidence against the assumption of accumulation of Mn V transient species, since its formation is followed by one of the fast reactions 4 and 5.
Many investigators [21,22] proposed that supreme oxidation reactions by MnO4 ─ in alkaline and neutral media occurred via construction of complexes between reductant and oxidant. In the present study, complex construction was indicated spectrophotometrically by the achieved UV-Vis spectra as shown in Figure 1, as well as kinetically, as the graphs of 1/kobs vs. 1/[A] were linear with positive slopes [23] as shown in Figure 8 (a). Increasing rates of oxidation with increasing ionic strength suggests that oxidation reactions occurred between ions of similar charges [24,25].
In the light of the experimental observations, the most acceptable suggested mechanism for the present oxidation reactions, includes a rapid deprotonation of the antibiotic (Eq. 3), followed by the attack of MnO4 ─ on the deprotonated antibiotic to construct a complex, [A -Mn VI O4 2-] (C), Eq. (6): Then, the formed transient complex decays in the rate-controlling stage to yield the preliminary oxidation products as follows: Preliminary oxidation products ….. (7) The latter interacts with three moles of MnO4 ─ ions in subsequent fast steps to yield the final oxidation products of the antibiotic.
The rate of decay of MnO4 ─ could be expressed by the rate-law (8) Figure 8 (a) and (b). Values of the rate constant of slow step of the proposed reactions mechanism (k1) and the equilibrium constants (K1 and K2) could be computed from the slopes and intercepts of these plots and are inserted in Table 4. Table 3 includes the computed activation parameters which were in accordance with the suggested oxidation mechanism. The obtained higher negative values of S ≠ reaffirms construction of complexes amongst the reacting species [24]. The positive values of H ≠ and G ≠ signifies that the complexes formation was endothermic and non-spontaneous [25]. The higher values of Ea ≠ indicated that the ratecontrolling stage was the decay of the constructed complexes.