2.1. Materials and methods

Properties of QS, DS and NET are given in Table 1, and their molecular structures are shown in the left of Fig.1. Quinestrol (99%，SIGMA, USA), norethindrone (98%, Toronto Research Chemicals Inc.), and dienestrol (96.1%, Laboratories of Dr. Ehrenstorfer, Germany), and the organic solvents of pyridine, hexanes, acetonitrile, isopropyl alcohol and methanol applied for HPLC, were purchased from Shanghai ANPEL Scientific Instrument Co., Ltd. China. The solutions were prepared with ultrapure water (18.2 MΩ cm) produced by a Milli-Q water purification system (Millipore, USA).

Fig. 1.Full-screen View

The experimental processes.

2.2. Irradiation

Estrogens solutions at an initial concentration of 2 mg/L (10 mL) were prepared in glass dish (Φ70 mm×35 mm) with plugs, at initial pH values of 6.0–7.5 without adjustment. The samples were irradiated to 0.5–6 kGy in a ⁶⁰Co source in Shanghai Institute of Applied Physics, Chinese Academy of Sciences. All samples were filtered with 0.22 mm micro-filtration membrane before analysis. All experiments were performed at ambient temperature. The analysis processes are shown in Fig .1.

2.3. Toxicity tests

The toxicity was determined by rotifer test. Four rotifers were added to 8 cm³ of test solution in Φ16 mm×150 mm disposable glass test tubes. The tubes, at 25°C, were placed on a rotator turning at 10–15 rph. After two days, the total numbers of rotifers were counted in each tube. Each sample had three parallel samples. From the data, the rate of population growth (r) was calculated by r = (ln N_t – ln N₀)/2, where N_t is the number of female rotifers in the tube after 2 d and N₀ is the initial number of rotifers in each tube.

2.4. Analytical methods

The QS, DS and NET samples were quantified with an HPLC-1200 (Agilent, USA), equipped with a C₁₈ column (4.6 mm×250 mm×5 μm, XDB-18, Eclipse, USA) reverse-phase column and a UV detector of 220, 245 and 230 nm in detection wavelength. The column temperature was 30°C and the injection volume was 10 μL. All mobile phases were filtered using a Millipore 0.45 μm filter and were degassed by ultrasonication before use. The mobile phase of QS and NET is a mixture of degassed acetonitrile and water at pH 7(80:20, V/V). The mobile phase of DS is a mixture of degassed methanol and water at pH 7(66:34, V/V). The flow rate was 0.8 cm^3/min. The retention time of QS, DS and NET was 4.74, 5.24 and 2.45min, respectively.

Solid phase micro-extraction with GC/MS was used to analyze the byproducts. The irradiated samples of 10 mL were added to the 15 mL vials containing a magnetic stir rotor. The vials, sealed with Teflon septum, were placed on the magnetic stirring stage. They were kept at 50°C for 20 min and extracted at 200 rpm/min. After extraction, the fiber was removed from the vial and inserted into the injection port of GC/MS system for thermal desorption and analysis. Then, the fiber was removed from the injection port. Before and after the sample desorption, the fiber was inserted into the injection port of GC/MS system for blank sample analysis. Extraction of each sample was repeated three times. GC/MS analyses were performed on a 7890A gas chromatography instrument (Agilent Technologies, Palo Alto, CA, USA), equipped with a mass spectrometer detector MSD 5975C and 7693 automatic liquid sampler. The separation of QS, DS, NET and the irradiation products was carried out on a DB-5MS fused silica capillary column (30 mm×0.25 mm i.d. ×0.25 μm, J&W Scientific, Agilent, USA.) with the following oven temperature program: initial temperature of 40°C for 1 min, heated to 180°C at a rate of 8°C/min, to 310°C at a rate of 15°C/min, and kept for 10 min. Helium was used as the carrier gas, at the flow rate of 0.8 dm³/min. Injector temperature was maintained at 250°C, and the injection volume was 10 μdm³. The mass spectrometer was operated at a selected ion monitoring (SIM) mode for QS (m/z 364), DS (m/z 266), NET (m/z 294), and scan mode m/z in the range of 50 to 450 for the byproducts, respectively.

The pH and TOC values were measured by a pH meter (Dual Star, Thermo Inc. America) and a TOC analyzer.

2.5. Quality assurance and quality control

Field blanks (n = 8) were analyzed with every batch of samples against contaminations or interferences. Sample concentrations were determined from external calibration curves prepared at concentrations ranging from 0.5 to 1000 ng /mL for QS, DS and NET. Strong coefficients of 0.9998, 0.9999 and 0.9994 for QS, DS and NET, respectively, were determined. No interferences were detected in the field blanks.

2.6. First-order kinetic model of estrogen Gamma-irradiation degradation

The following pseudo-first-order gamma-irradiation degradation kinetic model was assumed to relate changes in the total estrogen concentration in batch tests ^[25].

where C₀ and C are concentrations of the pollutant before and after irradiation, respectively; d is the dose constant in units of a reciprocal dose; and D is absorbed dose. Knowing d, the first-order rate law, Eq. (1), can be used to calculate the dose requirement (D) to achieve desired decreases in pollutant concentrations. The dose constant can be used in certain investigations into the fundamentals of radiolysis reaction mechanisms. In some cases, it can be deemed as a dose dependent pseudo first-order “rate constant”^[26].

2.7. Estrogens radiation chemical yield (G-value)

Radiation chemical yield (defined by the number of molecules formed or destroyed in solutions absorbing 100 eV of radiation energy) can be calculated using Eq. (3)^[27-29],

where, ΔR is the amount of reduced estrogens (mol/L); N_A =6.023×10²³ (molecules/mol) is Avogadro constant; D is the absorbed dose (Gy); and 6.24×10¹⁶ is the conversion coefficient from Gy to 100 eV/L.

3.1. Degradation kinetics of estrogens

Radiation-induced degradation of organic compounds was initiated by primary species produced from water radiolysis, as shown in Eq.(4). In neutral solutions, OH, e_aq⁻ and H are major primary species. In the presence of dissolved oxygen or air, the reducing species, H and e_aq⁻ can be converted into oxidizing species: perhydroxyl radicals (HO₂) and super oxideradical anions (O₂), as shown in Eqs. (5) and (6)^[30-32]. Thus, in the presence of air, OH is considered to be significant oxidizing transient for estrogens degradation. Oxidizing radicals, especially •OH radicals, attack double bonds (C-C in estrogens molecular), forming OH^- adducts. The carbon atom can be attacked by •OH radicals, leading to the cleavage of C-C bond, producing some intermediates. The OH^-adducts can easily scavenge oxygen to form the corresponding unstable peroxyl-radicals. Hydrolysis, radical–radical recombination or other reactions can take place, forming various final products ^{[33, 34]}. In the presence of air, reduction system occurs during gamma irradiation, as shown in Eqs.(7) and (9).

The estrogens in aqueous solution were irradiated to 0.5, 1.0, 3.0, 4.0 and 6.0 kGy by ⁶⁰Co γ-rays. The concentrations of QS, DS and NET varying with doses are shown in Fig.2. The estrogen concentrations decreased with increasing doses. It changed slowly after 4 kGy to reach the degradation efficiency of about 100% at 6 kGy.

Fig. 2.Full-screen View

Degradation of QS, DS and NET estrogens irradiated to different doses.

Radiation-induced degradation of estrogens in aqueous solution of different systems is shown in Fig.3. The estrogens were degraded rapidly with increasing doses. In N₂-saturated aqueous solutions with 5% isopropanol, the HO· radicals are scavenged by isopropanol. Therefore, the short-lived species reacting with estrogens is e_aq⁻. The degradation of estrogens was at reduction system. The degradation rate of estrogens increased with dose. In O₂ saturated aqueous solutions, most of the electrons and H atoms react with oxygen. Therefore, the short-lived species reacting with estrogens is HO·. The estrogens removal in O₂ saturated aqueous solutions was the highest of all conditions.

Fig. 3.Full-screen View

Degradation kinetics of QS, DS and NET estrogens irradiated in different systems.

3.2. G values

According to Eq. (3), the G values of estrogens (DS, QS and NET, 2 mg/L each) at different doses in oxygen gas are shown in Table 2. The G values decreased with the increasing doses. This may due to the competitive reactions of radicals between the parent compounds and degradation products, or the recombination of radical–radical (·OH, e_aq⁻ and H·), so, the radical concentrations for reaction with estrogens were reduced. The trend of radiation chemical yield is consistent with published findings for radiolytic degradation of malathion and lindane ^[35].

3.3. Removal of TOC and r values

TOC values of the radiolytic products in O₂ atmosphere were examined to evaluate the effect of mineralization of DS, QS and NET estrogens at different doses (Fig. 4a), with their initial concentration of 2 mg/L. The TOC at 0.5 kGy was the highest, decreasing with increasing doses. Below 3 kGy, TOC values decreased slowly, while above 3 kGy, it decreased quickly. This is due to that irradiation products of DS, QS and NET remained in the solutions without being mineralized at lower absorbed doses. The order of TOC is QS > NET>DS.

Fig. 4.Full-screen View

TOC (a) and r values of QS, DS and NET irradiated to different doses in O₂ atmosphere.

It was reported that parent estrogens and some of the intermediate products of pollutants can disturb the normal physiological response and lead to sex steroid imbalance^[36]. Hence, effect of the intermediate products of pollutants should be tested before discharging them into the ecosystem. In the irradiated solutions, the negative effect decreased with increasing doses (Fig.4b). It is obvious that even at the start of irradiation the r value of the reactive intermediates decreased. The order of r value is QS >NET> DS. We think that the rate of degradation was affected by the amount of benzene contained in the estrogens. Quinestrol can be completely mineralized under γ-irradiaition. The population of QS is low.

3.4. Toxicity analysis

Rotifers can be used as biomarkers for the presence of endocrine disrupting chemicals in the environment ^{[22, 23]}。 Thus, we attempted to explore the toxicity of QS, DS and NET with rotifers before and after their irradiation.

AS shown in Fig. 5, toxicity of the three irradiated estrogens declined with increasing doses, decreased rapidly by 1 kGy irradiation, and changed slowly from 1 kGy to 6 kGy, especially from 4 kGy to 6 kGy. The order of toxicity values of the irradiated QS, DS and NET is DS > NET > QS. These indicate that the estrogens in the environment can affect the normal biological reproduction, which were connected with other study about estrogens or endocrine disruptors ^[22-24].

Fig. 5.Full-screen View

Toxicity of the irradiated NET, QS and DS.

3.5. Estrogens derivatives from γ- irradiation process

In order to identify the QS, DS and NET derivatives, samples irradiated to 1 kGy were analyzed by GC-MS. The derivatives and their mass spectrums are shown in Fig. 6.

Fig. 6.Full-screen View

TIC of estrogens from GC-MS a-d: the main m/z after gamma irradiation. (a) Quinestrol (b) Dinestrol (c) Norethindrone

It was observed that an attack of hydroxyl radicals on the propanoic acid and isobutyl substituents of QS, DS and NET structures result in the formation of products such as phthalic acid, 1,2-Benzenedicarboxylic acid and benzoic acid. In Fig. 6 the peaks at m/z=133 and m/z=151 indicate benzoic acid and m/z=149 shows the formation of phthalic acid and 1,2-benzenedicarboxylic acid. In Fig. 6(a), the peak at m/z=149 is from phthalic acid, the peak at m/z=44 is from carbon dioxide, and molecular weight 364 is quinestrol. In Fig.6(c), the peak at m/z =44 is carbon dioxide, the peak at m/z =129 is benzoic acid and m/z= 296 decreased 2H from norethindrone ^[37]. Tao Tang et al. ^[10] reported the formation of products estrone, organic acid and carbon dioxide, in their study of UV degradation of QS. Dai K et al. ^[38] argued that the hydroxylation process could be the first step of the degradation, followed by a second step of demethylation or decarboxylation with other different byproducts with smaller m/z values ^[3]. Tao Tang et al. and Rajib Roy Chowdhury et al.^{[10, 39]} reported the formation of products benzeneacetic acid and phenylacetic acid by the UV/sonolysis degradation of quinestrol. However, Huber et al. and Kurisu et al. ^{[40, 41]} suggested that the final photodegradation products were acid derivatives.