Article Highlights

  1. 1.

    The degradation of pyrazinamide in aqueous solution by electron beam irradiation was highly efficient.

  2. 2.

    The addition of H2O2 or K2S2O8 improved the degradation and mineralization of pyrazinamide.

  3. 3.

    The degradation mechanism of pyrazinamide was studied by the experimental results and quantum chemical calculations.

Introduction

In recent years, the presence of antibiotics in water has received much attention [1,2,3]. Conventional wastewater treatments have limited effectiveness in treating most antibiotics in water and lead to the persistence of antibiotics and their metabolites in an aqueous environment [4]. In the past, the concentrations of antibiotics in the various environment samples ranging from ng L−1 to µg L−1 have been detected worldwide [5]. As a communicable antibiotic, PZA is used to treat tuberculosis. In 2019, there are predictable 10 million people developed tuberculosis [6]. Usually, most people who develop tuberculosis need about six months to be cured with first-line antibiotics [6], so tones of PZA is consumed to treat tuberculosis. It is reported that approximately 30% of PZA is absorbed in the body during the 24 h [7]. PZA is detected in a wastewater treatment plant near a pharmaceutical factory in India [8]. It is reported that the predicted environmental concentration of PZA is 7.5 ug/L in China [9]. Moreover, PZA and its metabolites can induce bacterial resistance at low concentrations [10, 11]. Therefore, it is necessary to achieve the effective removal of PZA in the water treatment.

The advanced oxidation processes (AOPs) based on rapidly generating powerful hydroxyl radicals (·OH) which react non-selectively with most organic compounds, are considered a potential technique to reduce pharmaceuticals, including antibiotics from water compared to the conventional treatment process [12, 13]. It is reported that TiO2 shows photocatalytic activity on the degradation of PZA under UV irradiation [14, 15]. However, the efficiency of UV degradation may be reduced in treating surface water due to the absorption of UV energy by suspended solids [16]. Photo-Fenton [17] and electron-Fenton [18] techniques are studied to degrade PZA in the aqueous medium, although the issues about iron sludge and higher current need to be solved. In addition, all the above degradation techniques require a long time.

EBI is particularly suitable for the degradation of persistent organic compounds due to its simplicity and high degradation efficiency [19]. High radiation energy can effectively degrade organic compounds in aqueous, such as pharmaceuticals personal care products [20] (PPCPs), antibiotics [21] and dyes [22]. Besides, EBI is a special AOP due to the simultaneous generation of the reductive species (eaq−, H·) and oxidative species (·OH) in water radiolysis (Eq. 1) [23]. Therefore, some organic pollutants in water challenging to be degraded by ·OH can be removed through the reductive species generated by water radiolysis [24].

$${\text{H}}_{2} {\text{O}} \rightsquigarrow {\text{e}}_{{{\text{aq}}}}^{ - } \left( {2.6} \right) + {\text{H}}^{ \cdot } \left( {0.6} \right) +{}^{ \cdot }{\text{OH}}\left( {2.7} \right) + {\text{H}}_{2} {\text{O}}_{2} \left( {0.7} \right) + {\text{H}}_{2} \left( {0.45} \right) + {\text{H}}^{ + } \left( {2.6} \right)$$
(1)

The degradations of most organic pollutants include a set of complex reactions [25, 26]. It is hard to comprehend the detailed degradation mechanism of organic compounds only through experiments. Quantum chemistry calculation is a standard method to research the degradation mechanism of organic pollutants [26,27,28,29]. It is reported that quantum chemistry calculation is used to successfully explain the degradation mechanism of cyanuric acid [27]. Therefore, to forecast the predominant reaction pathway and understand the degradation mechanism of PZA by EBI, quantum chemistry calculations and experiments analysis are used concurrently.

In this paper, the degradation of PZA in aqueous solution by EBI and the influence of different environmental factors on the degradation were investigated. Moreover, the degradation mechanism was illustrated through degradation products analysis and quantum chemical calculations. The experiments were performed: (1) the effects of PZA initial concentrations, irradiation doses, and water matrices; (2) the effects of anions and organic matter that were possibly present in water, such as Cl−, CO32−, HCO3−, NO3−, NO2−, SO42−, and fulvic acid; (3) the effects of combining hydrogen peroxide or potassium persulfate with EBI; (4) the effects of initial solution pH and some scavengers. Finally, the calculations of chemical reaction thermodynamic and chemical reaction kinetic are performed.

Materials and methods

Chemicals

Pyrazinamide (MW = 123.11) and fulvic acid (MW = 308.24) with 98% purity were purchased from J&K Chemical Ltd. (Shanghai, China). Sodium chloride, sodium carbonate, sodium bicarbonate, sodium sulfate, sodium nitrite, and sodium nitrate were analytic grade reagents purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. Tert-butanol, sulfuric acid, sodium hydroxide, potassium persulfate, and hydrogen peroxide were analytic grade reagents purchased from J&K Chemical Ltd. (Shanghai, China). Methanol (HPLC), methanol (HPLC–MS), and formic acid (HPLC–MS) were purchased from ANPEL Laboratory Technologies (Shanghai) Inc. Deionized water (resistance > 18.2 MΩcm) was prepared using the Milli-Q-Plus ultra-pure water system from Millipore (Sartorius 611, Germany).

Irradiation source and methods

The electron accelerator (GJ-2-II) with a beam energy of 1.8 meV and variable current (0–10 mA) was manufactured by Xianfeng electrical plant and belonged to the Institute of Applied Radiation, Shanghai University, China. The PZA solution was packed in 0.1 mm thick polyethylene bags (12 cm × 10 cm) and irradiated at different irradiation doses at room temperature (25 °C). The absorbed dose was determined using a Fricke dosimeter, which showed 176 Gy absorbed dose per 500 Gy irradiation dose. In addition, all sample solutions except for the N2O-saturated experiments were purged with N2 before irradiation.

Analytical methods

The analytical instruments in this paper include the ultra-high-performance liquid chromatography (UHPLC), the high-performance liquid chromatography–mass spectrometry (LC–MS), the ion chromatography (IC), and the total organic carbon (TOC) analysis. In this study, all analyzed solutions were filtered through a 0.45 µm polytetrafluoroethylene (PTFE) filter membrane. The detailed analysis is available in the supplementary material.

Computational details

All calculations of solution-phase free energy used the following formula [30]:

$$\begin{array}{*{20}c} {G_{soln} = G_{gas} + \Delta G_{solv} + RT\ln \left( \frac{RT}{P} \right)} \\ \end{array}$$
(2)

where the \(G_{gas}\) is the value of the free energy in the gas phase, the \(\Delta G_{solv}\) is the value of the free energy of solvation, and the \(RT\ln \left( \frac{RT}{P} \right)\), the change of energy from the gas phase standard state to the solution phase standard state, is 1.8 kcal mol−1. In this study, \(G_{gas}\) is equal to the sum of the high precision single-point energy and the low precision free energy correction, and \(\Delta G_{solv}\) is equal to the difference between the single-point energy under the SMD [31] solvation model and the gas phase single-point energy.

First, Gaussian 09 program [32] was used to carry out the calculations of geometrical optimizations and vibrational frequency analyses. The calculations were performed at the M06-2x/6-311 g(d,p) level of theory, which was more cost-effective for studying organic compounds [33], with SMD in water as the continuum solvation model and the Grimme’s DFT-D3 correction [34, 35] as dispersion corrected method. All stationary points (no imaginary frequency) and transition states (only single imaginary frequency) were identified by vibrational frequency analyses. Secondly, to obtain free energy correction, Shermo 2.0.5 [36] was used to calculate molecular thermodynamic properties under the conditions of 298.15 K and 1 atm. In order to improve the accuracies of the zero-point energy (ZPE) and internal energy (U), the frequency scale factors for calculating ZPE and U were 0.970 and 0.983[37], respectively. Thirdly, the high precision single-point energy calculations were using Orca 4.2 program [38, 39] with the PWPB95-D3(BJ)/def2-QZVPP level of theory on the optimized structures. Fourthly, the calculations of solvation free energy were performed through the difference of the single-point energy between the M05-2x/6-31G(d) level with the SMD model and M05-2x/6-31G(d) on the same optimized structures. It is reported that the M05-2 × functional with the 6-31g(d) basic set is more suitable for the SMD model [31]. Finally, the final solution-phase free energy was obtained by Eq. (2).

Moreover, intrinsic reaction coordinate (IRC) calculations verified the correctness of the transient state (TS). The calculation of the Reactions of radicals’ combination used the downhill method in IRC. The traditional transition state theory [40] (TST) was used to calculate chemical reaction kinetics kTST:

$$k^{TST} = n\kappa \frac{{k_{B} T}}{h}\frac{{Q_{TS} }}{{Q_{R} }}e^{{ - \frac{{\Delta G^{0} }}{RT}}}$$
(3)

where n is the reaction path degeneracy, к is the transmission coefficient calculated by the Skodje-Truhlar method [41], kB is Boltzmann constant, T is the thermodynamic temperature (K), h is Planck constant, QTS and QR are the partition function of transition state and reactants respectively, ΔG0 is the free energy difference between the reactants and the transition state, and R is gas constant. In addition, to avoid confusion, k represents the reaction rate constant obtained from the reference, and kTST represents the reaction rate constant by calculated.

Results and discussion

Degradation kinetics

Figure 1 shows the distinction of PZA removals following different irradiation doses and initial concentrations. The efficiency of PZA removals are positively correlated with irradiation doses at all initial concentrations and negatively correlated with initial concentrations at the same irradiation dose. It can be attributed to the relative concentrations of PZA and reactive species. At higher doses and lower initial concentrations, the higher relative concentration of PZA increases the removal efficiency.

Fig. 1
figure 1

Degradation of PZA at different initial concentrations and irradiation doses

Full size image

Radiation-chemical yield (G-value) is the number of molecules produced or consumed with the absorption of 100 eV of radiation energy. The calculation formula is as follows:

$${\text{G}} = \frac{{\left( {{\text{C}}_{{0 }} - {\text{ C}}} \right){\text{ NA}}}}{{{6}.24 \times 10^{{{19}}} {\text{ D}}}}$$

where C0 and C are the PZA concentration (M) the initial concentration and at the absorbed dose of D (kGy), respectively; NA is Avogadro constant; 6.24 × 1019 is the conversion factor of kGy to 100 eV L−1.

The G-value are showed in Table S1. G-value is positively correlated with the initial PZA concentration at the same irradiation dose and negatively correlated with the irradiation dose at the same initial PZA concentration. The reason for the decrease in G-value may be the competition of by-products with PZA molecules for reactive radicals.

The PZA degradation kinetics accords with the pseudo-first order kinetics (Eq. 4).

$$\begin{array}{*{20}c} {\ln \frac{C}{{C_{0} }} = - k_{d} D} \\ \end{array}$$
(4)

where kd is the dose constant (kGy−1), C and C0 are the PZA concentration (mM) at the irradiation dose of D (kGy) and the initial concentration, respectively.

The reaction dose constants are 4.39 (± 0.06), 2.77 (± 0.03), 0.89 (± 0.03) and 0.57 (± 0.02) kGy−1 with initial concentrations of 0.05, 0.1, 0.2 and 0.3 mM respectively, suggesting that the reaction dose constant decrease with the increased of the initial PZA concentration. The reason for this is that the increased intermediates of PZA degradation compete with PZA molecules for the reactive species generated by water radiolysis, and the amount of reactive species is proportional to the irradiation dose.

Effects of solution pH and water components

Figure 2a indicates the effects of solution pH on PZA degradations. The PZA degradation is decreased in alkaline condition comparing with neutral condition, while negligible effect in acidic condition. The kd values are 0.93 (± 0.02), 0.89 (± 0.03), 0.31 (± 0.03) kGy−1 at pH = 4, 7.5, and 11, respectively. Solution pH is related to the G values of reactive species generated by water radiolysis, which decide the degradation efficiency of PZA. It is reported that the sum (Geaq− + GH·) and G·OH are essentially unchanged at pH = 3–11 but decreased at pH > 11 with the electron accelerator as irradiation source [42]. Thus, the explanation of the decrease in PZA degradation at alkaline condition is that ·OH with the high reduction potential (Eo(·OHaq/OH−aq) = 1.77 V) is converted to ·O− with the lower reduction potential (Eo(·O− aq/OH−aq) = 1.64 V) in the reaction (5) [23]. In fact, ·OH is an electrophile while ·O− is a nucleophile in the reactions with organic matter [23]. Due to the conjugated π-bond in the pyrazine ring, pyrazinamide is nucleophilic and reacts more readily with ·OH compared to ·O−. The above facts indicate that ·OH may be the main reactant in PZA degradation. Besides, the PZA degradation at pH 4 is similar to pH 7.5 because of the constant G·OH.

$$\begin{array}{*{20}l} {^{ \cdot } {\text{OH}} + {\text{OH}}^{ - } \to{}^{ \cdot } {\text{O}}^{ - } + {\text{H}}_{2} {\text{O}}} \hfill &\quad {{\text{k}} = 1.3 \times 10^{10} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(5)
$$\begin{array}{*{20}l} { {\text{e}}_{{{\text{aq}}}}^{ - } + {\text{H}}^{ + } \to {\text{H}}^{ \cdot } } \hfill &\quad {{\text{k}} = 2.3 \times 10^{10} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(6)
Fig. 2
figure 2

a Degradation of PZA at different initial pH (C0 = 0.2 mM); b degradation of PZA in deionized water and surface water (C0 = 0.2 mM); c effects of inorganic anions (5 mM) on PZA degradation (C0 = 0.2 mM); d effects of the different concentrations of fulvic acid on PZA degradation (C0 = 0.2 mM)

Full size image

The effects of different water matrices, such as deionized water and surface water on PZA degradations are shown in Fig. 2b. It indicates the vital distinction on PZA degradations in surface water in comparison to deionized water. The kd values are 0.89 (± 0.03) and 0.23 (± 0.03) kGy−1 at deionized water and surface water, respectively. It can be explained that the ·OH generated in water radiolysis reacts with organic matter and inorganic anions in water, which leads to the decrease of ·OH concentration reacting with PZA. Thus, a higher irradiation dose is required to degrade PZA in treating surface water compared to deionized water.

Inorganic anions such as Cl−, CO32−, HCO3−, NO3−, and SO42− are the common ingredients in realistic waters. As shown by Eqs. (7)–(20), the anions exhibit high chemical reactivity with the reactive species formed in water radiolysis. As shown in Fig. 2c, CO32−, NO3−, and HCO3− inhibit the removal compared with the N2 saturated solution. The kd values are 0.18 (± 0.05), 0.22 (± 0.01), 0.63 (± 0.02), 0.86 (± 0.03) and 0.91 (± 0.02) kGy−1, in the presence of CO32−, NO3−, HCO3−, Cl− and SO42−, respectively. CO32−, NO3− and HCO3− obviously decrease the concentration of ·OH reacting with the PZA. It can be attributed that CO32− and HCO3− exhibit faster reaction rates with ·OH, which lead to the reduction in the chance of the ·OH reacting with PZA. It is reported that NO3− can inhibit the degradation of erythromycin [21]. NO3− easily combines with H+ to form HNO3, which further reacts with ·OH [44]. However, the effect of Cl− on the concentration of ·OH is complex due to the reactions (7) and (8). The reaction rate of forming ·ClOH− is 2.15 × 107 s−1, which is lower than 6.1 × 109 s−1, the reaction rate of ·ClOH− decomposing into Cl− and ·OH when the concentration of Cl− is 5 mM. Thus, the presence of Cl− has almost no effect on the removal. SO42− is the only anion that enhances the removal slightly, although the impact is negligible. The little impact can be explained by Eq. (13), the addition of SO42− is converted to ·SO4− which may be a stronger oxidizer for PZA compared to ·OH (Table 1).

Table 1 Reactions of Cl−, CO32−, HCO3−, NO3− and SO42− with ·OH, ·H and eaq−
Full size table

Fulvic acid (FA) shows a potent inhibition on the degradation of PZA owing to the competition for ·OH. From Fig. 2d, at the dose of 5 kGy, the degradation percentages of PZA are 77.1%, 74.9% and 75.8% at the initial FA concentration of 1, 3 and 5 mM, respectively. It implies that PZA degradation inhibitions in different FA concentrations tend to be similar at more than 5 kGy.

Effects of oxidizer

As we can see from Fig. 3a, it is promoted the removal of PZA at lower H2O2 concentrations, such as 1 and 3 mM. However, 5 mM H2O2 inhibits the degradation at lower doses but promotes when the dose is more than 2.5 kGy. One possible explanation of the result is that the addition of H2O2 increases the ·OH concentration (Eqs. 21 and 22) [23] at lower H2O2 concentration. In addition, excess H2O2 acts as the ·OH scavenger (Eq. 23) [23] at higher H2O2 concentration, which leads to a decrease in the removal of PZA. It can be observed in Fig. 3b, in contrast to H2O2, high K2S2O8 concentrations always promote the removal. As shown in reactions (24) and (25) [23], eaq− and H· can induce S2O82− to product ·SO4−, a powerful oxidant, which increases the degradation of PZA. Moreover, ·SO4− can be converted to ·OH by reactions (26) and (27) [46,47,48], which is beneficial for the removal.

$$\begin{array}{*{20}l} {{\text{e}}_{{{\text{aq}}}}^{ - } + {\text{H}}_{2} {\text{O}}_{2} \to {}^{ \cdot } {\text{OH}} + {\text{OH}}^{ - } } \hfill &\quad {{\text{k}} = 1.1 \times 10^{10} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(21)
$$\begin{array}{*{20}l} {{\text{H}}^{ \cdot } + {\text{H}}_{2} {\text{O}}_{2} \to {}^{ \cdot } {\text{OH}} + {\text{H}}_{2} {\text{O}}} \hfill &\quad {{\text{k}} = 9.0 \times 10^{7} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(22)
$$\begin{array}{*{20}l} {^{ \cdot } {\text{OH}} + {\text{H}}_{2} {\text{O}}_{2} \to {\text{HO}}_{2}^{ \cdot } + {\text{H}}_{2} {\text{O}} } \hfill &\quad {{\text{k}} = 2.7 \times 10^{7} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(23)
$$\begin{array}{*{20}l} {{\text{e}}_{{{\text{aq}}}}^{ - } + {\text{S}}_{2} {\text{O}}_{8}^{2 - } \to {}^{ \cdot } {\text{SO}}_{4}^{ - } + {\text{SO}}_{4}^{2 - } } \hfill &\quad {{\text{k}} = 1.1 \times 10^{10} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(24)
$$\begin{array}{*{20}l} {{\text{H}}^{ \cdot } + {\text{S}}_{2} {\text{O}}_{8}^{2 - } \to {}^{ \cdot } {\text{SO}}_{4}^{ - } + {\text{SO}}_{4}^{2 - } + {\text{H}}^{ + } } \hfill &\quad {{\text{k}} = 2.5 \times 10^{7} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(25)
$$\begin{array}{*{20}l} {^{ \cdot } {\text{SO}}_{4}^{ - } + {\text{ H}}_{2} {\text{O}} \to {}^{ \cdot } {\text{OH}} + {\text{SO}}_{4}^{2 - } } \hfill &\quad {{\text{k}} = 3.6 \times 10^{2} \,{\text{s}}^{ - 1} } \hfill \\ \end{array}$$
(26)
$$\begin{array}{*{20}l} {^{ \cdot } {\text{SO}}_{4}^{ - } + {\text{OH}}^{ - } \to {}^{ \cdot } {\text{OH}} + {\text{SO}}_{4}^{2 - } } \hfill &\quad {{\text{k}} = 1.4 \times 10^{7} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(27)
Fig. 3
figure 3

a Effects of the different concentrations of H2O2 on PZA degradation (C0 = 0.2 mM); b effects of the different concentrations of K2S2O8 on PZA degradation (C0 = 0.2 mM); c effects of H2O2 and K2S2O8 on the PZA mineralization (C0 = 0.2 mM)

Full size image

TOC is regarded as a reliable indicator to estimate the level of organic pollution in water [49]. Figure 3c indicates that the addition of H2O2 and K2S2O8 enhance the mineralization of PZA at high doses. The effect of H2O2 on the PZA mineralization is complex. In the presence of 5 mM H2O2, there is the slight inhibition at the lower doses but a little promotion at the higher doses, and the TOC value can be decreased from 13.65 to 9.88 mg L−1 at 10 kGy. However, the addition of K2S2O8 always promotes the PZA mineralization, and decreases the TOC from 13.65 to 8.05 mg L−1 at 10 kGy, proving that K2S2O8 is more potential to decrease TOC.

Effects of scavengers

In Fig. 4, we compare the effects of radicals’ scavengers such as N2O and tert-butanol (t-BuOH) on PZA degradations. The addition of N2O gas significantly contributes to the degradation of PZA compared to the N2. Thus, it is reasonable to suppose that ·OH is the primary reactant in PZA degradation for N2O is beneficial for increasing the concentration of ·OH by Eq. (28) [23]. Moreover, t-BuOH is a typical ·OH scavenger by Eq. (29) [23], and the addition of it reduces the efficiency of PZA removal. It also confirms that ·OH is a more dominant reactive species in all possible reactants generated by water radiolysis from the opposite perspective.

$$\begin{array}{*{20}l} {{\text{e}}_{{{\text{aq}}}}^{ - } + {\text{N}}_{2} {\text{O}} + {\text{H}}_{2} {\text{O}} \to {}^{ \cdot } {\text{OH}} + {\text{N}}_{2} + {\text{OH}}^{ - } } \hfill &\quad {{\text{k}} = 9.1 \times 10^{9} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(28)
$$\begin{array}{*{20}l} {^{ \cdot } {\text{OH}} + {\text{t}} - {\text{BuOH}} \to {}^{ \cdot } {\text{CH}}_{2} {\text{C}}\,\left( {{\text{CH}}_{3} } \right)_{2} {\text{OH}} + {\text{H}}_{2} {\text{O}} } \hfill &\quad {{\text{k}} = 6.0 \times 10^{8} \,\left( {\text{M s}} \right)^{ - 1} } \hfill \\ \end{array}$$
(29)
Fig. 4
figure 4

Degradation of PZA at different scavengers (C0 = 0.2 mM)

Full size image

Analysis of degradation mechanism and quantum chemistry calculation

Table S2 (Supplementary Material) shows the detected compounds according to the data of the LC-QTOF-MS analysis and the IC analysis. The results of the IC analysis show decreased formic acid concentrations and increased ammonium cation concentration following the increasing irradiation dose. It can be attributed to the mineralization of formic acid in the presence of ·OH and that ammonium cation is hard to oxidize to nitrate. Compared to the experiment of Fenton, 5-hydroxypyrazinamide and NH3 are detected in this experiment, but pyrazin-2-ylmethanol, pyrazine, acetamide, and acetic acid are not found. It shows that detailed degradation mechanism of PZA using EBI is different from Fenton [18].

From Scheme 1 and Table S4, the first reaction is free radicals attacking the aromatic ring or amide group. Four intermediates, IM1, IM2, IM3 and IM4, are formed via the reaction between ·OH and the C of the aromatic ring, while IM5 is hardly formed due to both a higher free energy barrier (21.73 kcal mol−1) and an unfavorable reaction free energy change (+ 15.24 kcal mol−1). A lower reaction rate constant (1.01 × 10–3 (M s)−1) also indicates that the generation of IM5 is impractical. In all reactions between PZA and ·OH, the preferred reaction pathway is the reaction (1–2a) owing to the lowest free energy (9.05 kcal mol−1). IM2 continuously reacts with ·OH to form water and IM9, which isomerized to IM10 in water. The free energy barrier for the isomerization with water (11.01 kcal mol−1) is lower than without water (35.96 kcal mol−1), and the reaction rate of the former (7.26 × 104 s−1) is also higher than the latter (2.91 × 10–5 s−1), so water plays the catalytic role in the isomerization. Due to the lower free energy barrier, IM10 can be converted via the TS1-2d to IM11, which further reacts with ·OH to form IM8 (2-Hydroxypyrazine) detected by QTOF and IM6. Another possible pathway to form IM8 is the pathways (1–1b and 1–1c), the radical reaction between IM1 and ·OH can generate carbamic acid and IM7 converted further to IM8. IM3 and IM4, as the active intermediates, react rapidly with ·OH via the reaction pathways (1–3b and 1–4b) to form ultimately IM12 found by QTOF.

Scheme 1
scheme 1

Proposed the degradation processes of PZA by EBI

Full size image

In addition, H· as reductive species may attack PZA through the reaction pathways (2–1a)–(2–6a) whose free energy barriers are 10.42, 8.53, 9.91, 9.92, 8.36 and 8.92 kcal mol−1, respectively. Pathway (2–7a) and (2–8a) are hardly achieved due to the higher free energy barriers of 15.57 and 15.07 kcal mol−1 and the lower reaction rate constants of 7.24 × 101 (M s)−1 and 1.50 × 104 (M s)−1. There is another pathway to generate IM15, hydrogen ions combined with the charged complex (IM22) formed by electron transfer of PZA and hydrated electrons. The instability of the radical intermediates generated by pathways (2–1a)–(2–6a) decides that they are highly susceptible to ·OH or H·. The calculations show that these intermediates tend to react with ·OH to form PZA and water. Moreover, the leaving of the amide group is hardly occurred due to the endergonic process by 4.13 kcal mol−1 and the lower reaction rate constant (1.21 × 101 s−1). Therefore, it cannot produce pyrazine in treating PZA solution by EBI.

Based on the analysis of degradation products and quantum chemical simulations, a possible degradation pathway is shown in Scheme 2. IM8 and IM 12 are the primary stable intermediate. IM12 is converted to IM23 under further attack of ·OH, which IM23 can undergo the break of C–C bond via transition state TS4b to generate IM24 and ·CONH2. Similar to the isomerization of IM7, IM24 can also be isomerized to IM26. Likewise, 2-Hydroxypyrazine can be further oxidized via the reactions (5a–5b) to form IM26. IM26 undergoes the break of the six-membered ring by ·OH to produce IM28. The calculations show that the hydrogen ionization reaction in the carboxyl group of IM28 is less favorable due to the endergonic process and the decrease in pH in the radiolytic degradation of the organic molecule. In fact, the occurrence of the reaction (5e) is more thermodynamically favorable. In the presence of H·, the C-N single bond of IM29 breaks to form IM30 found by QTOF and IM6. The addition reaction is then carried out in the C-N double bond of IM30, thus forms to IM32, which is further oxidized to IM34 through the reactions (5i and 5j). IM34 can be decomposed to IM35, ·CHO and water with the action of ·OH due to the instability of two hydroxyl groups on the same carbon atom. ·OH is combined with ·CHO to produce formic acid in the experiment, and IM35 to produce water and IM37. The calculations of IM37 show that the hydrogen ionization reaction in the carboxyl group and the reaction between IM37 and ·OH are hard to carry out due to the endergonic process. The analysis of QTOF proves the abundant presence of IM37. It is found that IM37 tends to react with hydrated electrons in electron exchange. The free energy barrier of this reaction is 7.53 kcal mol−1, which is lower than ·OH or H·. Further, the generated charged complex reacts rapidly with the hydrated proton to finally form the carboxyl group and NH2COOH. The carboxyl group can combine with ·OH to produce carbonic acid. It is reported that NH2COOH can be decomposed to NH3 and CO2 in aqueous [50]. The calculation of NH2COOH also shows that it can undergo the hydrogen transfer reaction to produce NH3 and CO2.

Scheme 2
scheme 2

Proposed the possible degradation processes of the main intermediates by EBI

Full size image

Conclusions

The present work showed that EBI was efficient to degrade PZA. The results showed that electron beam irradiation was highly effective in the removal of PZA from water. Some inorganic anions and fulvic acid exhibited inhibition of the PZA degradation in this study. The inhibition effects of inorganic anions were CO32− > NO3− > HCO3− while Cl− and SO42− have negligible effects on the degradation of PZA. K2S2O8 and lower concentrations of H2O2, such as 1 and 3 mM, could effectively promote the removal of PZA. However, 5 mM H2O2 inhibited the degradation at lower doses, and promoted the degradation at higher doses. In addition, TOC removal efficiency increased from 23.8% to 41.0% and 27.6% with the additions of 5 mM K2S2O8 and H2O2, respectively. Thus, K2S2O8 was the better additive to enhance the mineralization of PZA. The experiments of radical scavengers showed that ·OH was the main reactive species in PZA degradation. Moreover, the degradation mechanism of PZA by EBI was suggested by combining LCMS, IC and quantum chemiscal calculations. It was confirmed that both ·OH and H· played a significant role in the degradation of intermediates and the mineralization of PZA.