- Email: [email protected]
Hydroxyl radical induced degradation of ibuprofen
Science of the Total Environment 447 (2013) 286–292
Contents lists available at SciVerse ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Hydroxyl radical induced degradation of ibuprofen Erzsébet Illés a, b,⁎, Erzsébet Takács b, András Dombi a, Krisztina Gajda-Schrantz a, c, d, Gergely Rácz b, Katalin Gonter b, László Wojnárovits b a
Institute of Chemistry, Research Group of Environmental Chemistry, University of Szeged, Szeged, Hungary Institute of Isotopes, Centre for Energy Research, Hungarian Academy of Sciences, Budapest, Hungary Department of Inorganic and Analytical Chemistry, University of Szeged, Szeged, Hungary d EMPA, Laboratory for High Performance Ceramics, Duebendorf, Switzerland b c
H I G H L I G H T S
G R A P H I C A L
A B S T R A C T
► In hydroxyl radical attack on the ring mainly hydroxylated products form ► The hydrated electron attacks the carboxyl group. ► Oxidative conditions are more effective in ibuprofen decomposition than reductive. ► Ecotoxicity of ibuprofen solution first increases then decreases with irradiation.
a r t i c l e
i n f o
Article history: Received 11 September 2012 Received in revised form 18 December 2012 Accepted 1 January 2013 Available online 11 February 2013 Keywords: Ibuprofen Advanced oxidation processes Radiolysis Hydroxyl radical Hydrated electron
a b s t r a c t Pulse radiolysis experiments were used to characterize the intermediates formed from ibuprofen during electron beam irradiation in a solution of 0.1 mmol dm −3. For end product characterization 60Co γ-irradiation was used and the samples were evaluated either by taking their UV–vis spectra or by HPLC with UV or MS detection. The reactions of •OH resulted in hydroxycyclohexadienyl type radical intermediates. The intermediates produced in further reactions hydroxylated the derivatives of ibuprofen as final products. The hydrated electron attacked the carboxyl group. Ibuprofen degradation is more efficient under oxidative conditions than under reductive conditions. The ecotoxicity of the solution was monitored by Daphnia magna standard microbiotest and Vibrio fischeri luminescent bacteria test. The toxic effect of the aerated ibuprofen solution first increased upon irradiation indicating a higher toxicity of the first degradation products, then decreased with increasing absorbed dose. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Abbreviations: UV–vis, ultraviolet–visible; HPLC, high-performance liquid chromatography; MS, mass spectrometry; IBP, ibuprofen; EC50, effective concentration for 50% luminescence reduction; AOP, advanced oxidation processes. ⁎ Corresponding author at: Institute of Isotopes, Centre for Energy Research, Hungarian Academy of Sciences, Budapest, Hungary. Tel./fax: +36 62 544 338. E-mail addresses: [email protected], [email protected] (E. Illés). 0048-9697/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scitotenv.2013.01.007
Ibuprofen (IBP) ((RS)-2-(4-(2-methylpropyl)phenyl)propanoic acid, C13H18O2) is used for the relief of symptoms of arthritis, primary dysmenorrhoea, fever, and as an analgesic, especially to heal inflammatory disease(s). Ibuprofen is known to have an antiplatelet effect,
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
287
though it is relatively mild and short-lived when compared to that of acetylsalicylic acid or other better known antiplatelet drugs. It is most often sold with trade names Nurofen, Advil and Motrin. IBP is a moderately toxic compound, its − log EC50 value (EC50, effective concentration for 50% luminescence reduction) according to the Microtox luminescent bacteria test is reported to be 3.85 (Escher et al., 2005). The solubility of IBP in water is limited to about 1 × 10−4 mol dm −3 (Zwiener and Frimmel, 2000). IBP shows little absorbance in the wavelength region of the solar spectrum, its photodegradation in surface waters is limited and its biodegradation is also slow (Packer et al., 2003). IBP is regularly detected in the effluents of wastewater treatment plants and in surface waters at ng–μg dm −3 level. Although this concentration level is very low, IBP and its probably more harmful metabolites – together with other drugs – may present a potential hazard for human health and also for the aquatic ecosystem. The potential application of several advanced oxidation processes (AOP) for the degradation of IBP in dilute aqueous solutions has already been studied, using electro-Fenton, UVA photoelectro-Fenton and solar photoelectro-Fenton (Méndez-Arriaga et al., 2008; Skoumal et al., 2009), as well as photocatalysis and sonophotocatalysis (Madhavan et al., 2010; Méndez-Arriaga et al., 2010), ozonation (Huber et al., 2003), radiolytic (Zheng et al., 2011) and Ferrate(VI) processes (Sharma and Mishra, 2006). In AOP's the degradation is due mainly to reactions with hydroxyl radical. Hydroxylated IBP molecules were detected among the first degradation products by using mass spectrometry. This hydroxylation was assumed to take place on the side chains in the tertiary positions, suggesting •OH attack on the side chains (Skoumal et al., 2009; Madhavan et al., 2010; Méndez-Arriaga et al., 2010; Zheng et al., 2011). This suggestion is in contradiction with the findings on many other substituted aromatic molecules, where the main target of radical attack was the aromatic ring (Illés et al., 2012). For compounds with low biodegradability the photogenerated •OH in surface waters around the 10 − 14 mol dm − 3 concentration level is suggested to play an important role in their degradation (Packer et al., 2003; Jones et al., 2009). Studies on the •OH + IBP reaction may help to understand the fate of the compound in the environment. High energy ionizing radiation treatment is also regarded as AOP. Radiolytic studies on IBP were previously carried out by measuring end products (Zheng et al., 2011). Jones (2007) used pulse radiolysis to determine the rate coefficients of hydroxyl radical and hydrated electron reactions with IBP. In the present study the kinetics and mechanism of •OH reaction with IBP will be investigated by transient and final-product techniques. We use irradiation of aqueous systems as the cleanest source of hydroxyl radicals. These investigations in general may help to establish AOP technologies for IBP removal, and – more specifically – they can contribute to develop an irradiation technology for this purpose.
used on an Agilent 1200 system equipped with an XB-C18 Phenomenex Kinetex (100×2.1 mm, particle size 2.6 μm) capillary column coupled with on-line mass spectrometer. In this latter case, the flow rate was 0.3 cm3 min−1 and the mobile phases were acetonitrile and 0.1% aqueous acetic acid with the gradient: 0 min 15%, 15 min 50%, and 18 min 50% acetonitrile. For detection and identification of the parent compound and the degradation products a diode array detector at 220 and 260 nm and a tandem mass spectrometer were used. Mass spectrometric experiments were performed in the negative ion mode using an Agilent 6410 triple quadrupole mass spectrometer equipped with an ESI source. The following MS parameters were used for identification: drying gas N2 (350 °C, 10 dm3 min−1); nebulizer pressure with 1.7 bar, capillary voltage with 3500 V and fragmentor voltage with 80 V. The toxicity tests of the target compound and the degradation products formed were carried out using Vibrio fischeri bacteria (Luminescent bacteria test LCK 480) according to the DIN/EN/ISO standard no. 11348-2 by a HACH-LANGE GmbH LUMIStox 300 apparatus. The pH of samples was set to 7. The inhibition of the light emission in the presence of the sample was determined against a non-toxic control solution. The second toxicity test was performed on Daphnia magna zooplankton (DAPHTOXKIT F™). This organism originates from the second tropic level. Standard microbiotests were applied according to the ISO standard no. 6341, 1996. TOXKIT tests were suited for toxicity of all chemicals in aquatic and terrestrial environment. The mortality of the species was followed during 48 h incubation with 4 dilutions of the treated IBP solutions. Pulse radiolysis investigations were carried out using 800 ns pulses of accelerated electrons, and an optical detection system with a 1 cm cell pass-length (Földiák et al., 1988). The pH values were adjusted at 4.4 and 8.5, respectively. At pH 4.4 the concentrations of protonated and deprotonated forms of IBP are equal (pKa), at pH 8.5 we investigated the reaction of deprotonated form. The results obtained at pH 8.5 were rather similar to those obtained at pH 4.4. The results at pH 4.4 will be detailed in the following sections.
2. Experimental
yields also OH radicals, the reacting radicals and their yields are: hydroxyl radical 0.56 μmol J − 1, hydrogen atom 0.06 μmol J − 1. The intensity of the characteristic aromatic absorption band in the UV spectrum of IBP in the 250–300 nm region is very weak (Fig. 1). When the solution is irradiated with a dose of 0.25 kGy, the intensity of the band around 265 nm becomes higher, indicating the formation of changed aromatic molecules. In the radiolysis of many aromatic molecules hydroxylation was observed at the beginning of the transformations. This hydroxylation may take place in the side chains (Zheng et al., 2011) and also in the ring. Substitution in the side chain does not change the absorption spectrum at ~ 265 nm. A simple hydroxylation in the ring of IBP can result in two isomers. As the inset shows, a dose of 0.25 kGy results in transformation of more than 60% of the initial molecules to new molecules. The secondary degradation of the primary products may have already started. (The G-values of
IBP and the chemicals for pH setting were purchased from Spectrum-3D or Carlo Erba. In end product experiments the irradiation was carried out by using a 60Co γ-irradiation facility with 5 kGy h−1 dose rate. The pH values were set by HCl and NaOH. These solutes at low concentration do not influence considerably the radiolytic reactions of organic molecules. All experiments were carried out at room temperature. The samples were evaluated either by taking their UV–vis spectra using a JASCO 550 UV–vis spectrophotometer with a 1 cm cell, or by HPLC separation. For chromatographic separation two different systems were used. The Agilent 1100 HPLC system consists of a C18 Kinetex 2.6 μm 100A, 100 × 4.6 mm, Phenomenex column where the elution was performed using a 50:50 mixture of 1% aqueous acetic acid solution and acetonitrile (flow rate of 0.9 cm3 min−1). Linear gradient elution was
3. Results and discussion 3.1. Hydroxyl radical reactions − Hydroxyl radical ( •OH, 0.28), hydrated electron (eaq , 0.28) and hy• drogen atom (H , 0.06) are the main products of water radiolysis. All of them can be classified as “reactive intermediates”. The values in brackets are the yields, the so-called G-values in μmol J −1 units (Spinks and Woods, 1990). Fig. 1 depicts the UV absorption spectra of 1 × 10 −4 mol dm −3 N2O saturated IBP solutions before and after irradiation. When the solution is saturated with N2O, the transformation
eaq
−
•
−
þ N2 O þ H2 O → OH þ OH þ N2
ð1Þ
288
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
Fig. 1. UV absorption spectrum of 1 × 10−4 mol dm−3 IBP solution saturated with N2O (•OH reaction), pH~4.4. Inset: relative concentration of IBP in irradiated solution obtained by HPLC separation of the components and integrating the area under the chromatographic peak.
IBP degradation are calculated as 0.25, 0.18, 0.12 μmol J −1 at doses 0.25, 0.5 and 0.75 kGy, respectively.) With the increase of absorbed dose the band at 250–300 nm disappears, indicating the disintegration of the aromatic structure. A characterless absorption band remains here; it is probably due to the absorbance of a large number of degradation products. In the transient absorption spectrum of IBP, obtained in pulse radiolysis experiments, three bands are observed (Fig. 2). Two bands appear at 285 and 315 nm, respectively, with comparable intensities. The third one is a much weaker band at 385 nm. The band at 315 nm decays in about 1 ms, the decay of the 285 nm band is somewhat slower, while the intensity of the 385 nm band does not change considerably on the 2 ms timescale of the measurement. The different time behaviors of these bands may suggest different intermediates. It should be noted, however, that stable final products may also contribute to the observed absorbance below 300 nm. • OH radical reacts with IBP with a rate coefficient of ~ 7 × 10 9 mol − 1 dm 3 s − 1 (Huber et al., 2003; Packer et al., 2003). The molecule offers three special places for the attack: attachment to the aromatic ring forming in the first step hydroxycyclohexadienyl radical
Fig. 2. Transient absorption spectra obtained in 1×10−4 mol dm−3 IBP solution saturated with N2O (•OH reaction), pH~4.4, 16 Gy/pulse.
(two isomers), and hydrogen atom abstraction from the tertiary position of the carboxyl group bearing side chain, or hydrogen atom abstraction from the secondary position of the other side chain. In the latter two reactions, benzyl type radical is expected. Cyclohexadienyl type radicals typically have wide absorption bands in the 300–350 nm range with molar absorption coefficients of 3000–5000 mol − 1 dm 3 cm − 1. The benzyl type radicals have a very intense absorption peak around 260–275 nm with an absorption coefficient of 10,000–16,000 mol − 1 dm 3 cm − 1 as well as one or two weaker bands at 300–320 nm with molar absorption coefficients of 2000–4000 mol − 1 dm 3 cm − 1. The band at 385 nm is probably due to phenoxyl radical which may have been formed in the secondary reactions. Since phenoxyl radicals have molar absorption coefficients of several thousand mol − 1 dm 3 cm − 1 around 400 nm, the phenoxyl radical yield is very low. The bands at lower wavelengths probably belong to both hydroxycyclohexadienyl and benzyl type radicals. Due to the relatively low intensity of absorbance around 285 nm (molar absorption coefficient ~ 3000 mol − 1 dm 3 cm − 1), we assume that the hydroxycyclohexadienyl type radical dominates. This suggestion is in agreement with the pulse radiolysis results on large number of alkyl-substituted benzenes: less than 20% of •OH reacts with alkylbenzenes in hydrogen atom abstraction, the majority of the radicals add to the ring forming hydroxycyclohexadienyl radical intermediates (Sehested et al., 1975). The formation of products hydroxylated in the ring also supports this idea. − 3.2. Reactions under reducing conditions (eaq + H •)
The reactions of hydrated electron are studied in solutions containing tert-butanol. In such solutions the hydroxyl radical is converted to rather inactive •CH2(CH3)2COH radical in the reaction (Spinks and Woods, 1990): •
•
OH þ ðCH3 Þ3 COH → H2 O þ CH2 ðCH3 Þ2 COH:
ð2Þ
Since the reaction of H • is slow: •
•
H þ ðCH3 Þ3 COH → H2 þ CH2 ðCH3 Þ2 COH;
ð3Þ
H • may also react with IBP. As Fig. 3 shows, there is a gradual decrease of the UV absorption intensities during degradation. The shape of
Fig. 3. UV absorption spectrum of 1 × 10−4 mol dm−3 IBP solution saturated with N2, − containing 0.5 mol dm−3 tert-butanol (eaq reaction), pH ~4.4. Inset: relative concentration of IBP in irradiated solution obtained by HPLC separation of the components by integrating area under the chromatographic peak.
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
289
− the spectrum changes just slightly during eaq /H • reactions: the band around 265 nm is shifted slightly to longer wavelengths. From the absorption spectra and the integrated chromatographic peaks of IBP, it is obvious that the degradation is less efficient under reducing conditions (G ≈ 0.03 μmol J −1) than under oxidative conditions in solutions saturated with N2O, or with air (see later). The rate coefficient of hydrated electron reaction with IBP was found here to be 8.5 × 10 9 mol − 1 dm 3 s − 1. A value of 8.9×109 mol−1 dm3 s−1 was published in the literature (Jones, 2007). The rate coefficient is about one third of the rate coefficient found for ketoprofen 2.6 × 10 10 mol − 1 dm 3 s − 1. In the case of ketoprofen the hydrated electron attack takes place on the carbonyl oxygen (Illés et al., 2012), whereas the carboxyl group of IPB seems to be the main target. Electron attachment is probably followed by an immediate protonation.
þ
R−COOH þ eaq
−
H3 O → R−C• O− OH → R−C• ðOHÞ2 þ H2 O
ð4Þ
Under pulse radiolysis conditions the transient absorption spec− is very weak showing a maxitrum produced by the reaction of eaq mum around 340 nm (Fig. 4). This characterless spectrum belongs probably to the radical at the carboxyl group. Cyclohexadienyl type radicals produced by H• addition to the ring – with a rate coefficient of 4 × 109 mol−1 dm3 s−1 as measured here – may also contribute to the spectrum. These cyclohexadienyl radicals, as opposed to the hydroxycyclohexadienyl radicals produced in the • OH reaction, do not yield hydroxylated products. When the solution is deoxygenated by N2 bubbling, but does not − contain the tert-butanol, both •OH and eaq may react with the solute (Fig. 5). The UV spectra obtained after γ irradiation are similar to the spectra measured in N2O saturated solution, although the effect of irradiation is smaller in the solution with N2 bubbling. In the latter solution the transformations mainly take place in the reactions of • OH. The transient absorption spectrum in Fig. 6 also shows the similarities to the spectrum obtained in the solution saturated with N2O. Apparently, the cross-reaction of intermediates formed in •OH and − eaq reactions is negligible.
Fig. 5. UV absorption spectrum of 1 × 10−4 mol dm−3 IBP solution saturated with N2, − (•OH + eaq reaction), pH ~4.4. Inset: relative concentration of IBP in irradiated solution obtained by HPLC separation of the components by integrating the area under the chromatographic peak.
and H• produce O2−•/HO2• pair in the following reactions (pKa = 4.8) (Spinks and Woods, 1990): eaq •
−
þ O2 → O2 •
−•
H þ O2 → HO2 :
ð5Þ ð6Þ
From practical point of view the investigations in air saturated solu− tions are of utmost importance. When dissolved oxygen is present, eaq
In such a system •OH remains to be the main reacting intermediate, which is expected to act the same way as in the solution saturated with N2O. The reactivity of the O2−•/HO2• pair with aromatic molecules is very low (Jovanovic et al., 1995). In the UV absorption spectra of solutions containing dissolved oxygen and irradiated with doses of 0.25, 0.5 and 0.75 kGy, there is a well resolved absorption band with a maximum at 260 nm (Fig. 7). Its intensity first increases with the dose and decreases having reached a maximum. As mentioned previously, this change in absorbance is attributed to the formation and decay of products hydroxylated in the aromatic ring. Such products are formed when •OH radical reacts with the aromatic ring and the hydroxycyclohexadienyl transient radical finally transforms to
Fig. 4. Transient absorption spectra obtained in 1×10−4 mol dm−3 IBP solution saturated with N2, containing 0.5 mol dm−3 tert-butanol, pH~4.4, 18 Gy/pulse.
Fig. 6. Transient absorption spectra obtained in 1×10−4 mol dm−3 IBP solution saturated − with N2, without tert-butanol (•OH+eaq reaction), pH~4.4, 19 Gy/pulse.
3.3. Reactions in the presence of dissolved O2
290
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
Fig. 8. Total ion chromatogram (TIC) and chromatogram monitored at m/z 221, ibuprofen solution saturated with N2O, after 0.25 kGy dose. Fig. 7. UV absorption spectrum of 1 × 10−4 mol dm−3 IBP solution saturated with air (•OH + O2−•/HO2• reaction), pH ~4.4, irradiated with different doses. Inset: relative concentration of IBP in irradiated solution obtained by HPLC separation of the components and integrating the area under the chromatographic peak.
a hydroxylated molecule. The cyclohexadienyl radicals readily react with dissolved O2, forming peroxy radicals. These peroxy radicals may eliminate HO2• (Roder et al., 1990), in this elimination reaction phenol type molecule is formed:
ð7Þ This elimination reaction gives an explanation to the higher intensity of the aromatic peak around 260 nm in air saturated solution than in N2O bubbled solution in spite of the reversed order of •OH yields (0.28, and 0.56 μmol J − 1, respectively). The peroxy radicals may undergo also ring opening reaction forming linear dicarboxylic acid molecules (von Sonntag and Schuchmann, 2001; von Sonntag, 2006). The ring-opening reaction is an important step of mineralization. In air saturated solution the estimated yield (G-value, 0.25 μmol J −1) agrees practically with the yield of •OH (0.28 μmol J−1). The reason is that the radical scavenging reaction opens a new reaction channel for degradation and the intermediates cannot be transformed back to the starting molecules. 3.4. Identification of final products Due to the low solubility, low solute concentration and low absorbed doses were used in these studies. In the HPLC chromatograms large number of primary, secondary, etc. products appeared with low yields. As a result, their identification and establishing the degradation mechanism were rather complicated. Except the tert-butanol containing solutions saturated with N2, at the lowest absorbed dose, the main products had a mass number of 221, being higher by 16 units than the m/z value of IBP (205) in the negative ion mode. Thus, the main products at the lowest absorbed dose are hydroxylated molecules. In our chromatograms taken in case of solutions saturated with N2O or air, 6 peaks appear with the same 221 m/z value (Fig. 8). One of these peaks is predominant, having outstandingly large peak area in the MS detection. In the literature, as it
was mentioned in the Introduction, the side chain of the molecule in the tertiary position (Fig. 9, structure (7)) (Marco-Urrea et al., 2009; Méndez-Arriaga et al., 2010) and/or in the secondary position are suggested to be hydroxylated (Caviglioli et al., 2002; Madhavan et al., 2010; Méndez-Arriaga et al., 2010; Zheng et al., 2011). In contrast to these suggestions our γ-radiolytic experiments with diode array detection made parallel with MS detection clearly show that hydroxylation in the ring is of high importance (see above, in Section 3.1.). From the two possible variations, one of the isomers is shown in Fig. 9 (structure (8)). In addition to hydroxylation of the ring, the products observed suggest that hydroxylation of the side chains also takes place. From the large number of products in Fig. 9, those with m/z values of 279, 351 and 237 should also be mentioned. Product (10) and (11) are formed only in solutions saturated with N2 and containing also tert-butanol. These are tert-butanol adducts. Product (9) is one of the double hydroxylated derivatives of ibuprofen. Double hydroxylated products are formed when solutions saturated with N2O or air were radiolysed. Product (1) with an m/z value of 133 was also detected by Caviglioli et al. (2002) and Zheng et al. (2011). It may have been formed in •OH induced degradation. This product was not detected in N2 saturated, tert-butanol containing solutions. The literature mentioned 4-isobutylacetophenone (3) as a degradation product of IBP with an m/z value of 175 (Castell et al., 1987; Skoumal et al., 2009). In our experiments under aerobic conditions six minor products with this mass number were separated. In the chromatogram there was only one product which elutes later than the ibuprofen, with m/z value of 171. Its presumable structure could not be determined. 3.5. Ecotoxicity Ibuprofen is considered to be a moderately toxic compound. First, the acute toxicity was determined with V. fischeri liquid-dried luminescent bacteria. Based on the experimental results, as Fig. 10 shows, in spite of the considerable decrease of IBP concentration upon irradiation, the inhibition of the light emission initially increases and then decreases. This phenomenon can be explained with the formation of toxic degradation products during the radiolysis of ibuprofen. In the inset of Fig. 10, the change of the concentration of some main products can be seen. The second toxicity test was performed on D. magna zooplankton which showed that ibuprofen is not toxic in the applied concentration. Radiolysis caused some increase in the mortality, this increase was, however, statistically not relevant. A slight increase of toxicity was also reported in the gamma radiolytic investigations of Zheng et al. (2011). Due to the large number of by-products it is difficult to claim which of them was responsible for the substantial increase of toxicity. Based
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
291
Fig. 9. Ibuprofen degradation products formed during radiolysis.
on some publications (Castell et al., 1987; Miranda et al., 1991) 3 (Fig. 9) is a toxic by-product because it has a lytic activity on erythrocytes. In the literature 1,2-dihydroxy ibuprofen (one of the double hydroxylated derivatives of ibuprofen) with a molecular mass of 238 was declared to be more toxic than IBP (Marco-Urrea et al., 2009). 1-(4-Isobutylphenyl)-1-ethanol (4) was also found to have toxic effect examined by measuring the amount of cell protein and intracellular enzymatic activity of LDH and GOT in cultured fibroblasts (Castell et al., 1987; Marco-Urrea et al., 2009). Dystrophic lesions affecting the liver, kidney, and skin were caused by degradation product (1) using zebrafish (Danio rerio) as the test object (Rácz et al., 2011). The ibuprofen solution after an absorbed dose of 2 kGy still preserved its toxic effect, because of the formation of aldehydes, carboxylic acids with small molecular weight and organic peroxides coming from the degradation of the aromatic part. These compounds have light absorption at short wavelength; therefore, cannot be observed in our system.
Maybe some hydrogen peroxides are also present in the solution after irradiation. These compounds have acute toxicity, however, under natural conditions they degrade easily. Due to the formation of toxic by-products, it is clear, that the absorbed dose during the treatment should be higher than needed for complete degradation of IBP since the toxic by-products should also be decomposed. Our experiments showed that the toxicity of the solutions could be strongly reduced with prolonged irradiation. 4. Conclusion The results of this work show that the degradation of initial IBP molecules is more efficient under oxidative conditions than under reducing conditions. The UV spectra obtained after γ irradiation in N2 bubbled solution are similar to the spectra measured in N2O saturated solution, although the effect of irradiation is smaller. Large number of primary and secondary products was formed with low yields during radiolysis. Except for the solutions saturated with N2 and containing also tert-butanol, where the hydrated electron reacts with the solute, the main products are hydroxylated molecules at the lowest absorbed dose. Due to the formation of toxic by-products, the absorbed dose should be higher during the treatment than needed for complete degradation of IBP. Our experiments showed that with prolonged irradiation the ecotoxic effect of the treated solutions could be strongly reduced. Acknowledgment The authors thank the Hungarian Science Foundation (OTKA, No. CK 80154 and NK 105802), the Swiss-Hungarian project (No. SH7/2/14) and the International Atomic Energy Agency (Contract No. 16485) for the support.
Fig. 10. Changes of fluorescence inhibition (I, %) in Vibrio fischeri toxicity test, and the relative concentration of unaltered IBP molecules as a function of dose, measured in 1×10−4 mol dm−3, IBP solution saturated with air (•OH+O2−•/HO2• reaction). The inset shows the absorbance changes of products determined with diode array detection after HPLC separation.
References Castell JV, Gomez-L. MJ, Miranda MA, Morera IM. Photolytic degradation of ibuprofen. Toxicity of the isolated photoproducts on fibroblasts and erythrocytes. Photochem Photobiol 1987;46:991–6.
292
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
Caviglioli G, Valeria P, Brunella P, Sergio C, Attilia A, Gaetano B. Identification of degradation products of ibuprofen arising from oxidative and thermal treatments. J Pharm Biomed Anal 2002;30:499–509. Escher BI, Bramaz N, Maurer M, Richter M, Sutter D, Kanel CV, et al. Screening test battery for pharmaceuticals in urine and wastewater. Environ Toxicol Chem 2005;24:750–8. Földiák G, Hargittai P, Kaszanyiczki L, Wojnárovits L. A computer-controlled pulse-radiolysis laboratory. J Radioanal Nucl Chem 1988;125:19–28. Huber MM, Canonica S, Park G-Y, von Gunten U. Oxidation of pharmaceuticals during ozonation and advanced oxidation processes. Environ Sci Technol 2003;37:1016–24. Illés E, Takács E, Dombi A, Gajda-Schrantz K, Gonter K, Wojnárovits L. Radiation induced degradation of ketoprofen in dilute aqueous solution. Radiat Phys Chem 2012;81: 1479–83. Jones GK. Applications of radiation chemistry to understand the fate and transport of emerging pollutants of concern in coastal waters. Ph.D. dissertation, North Caroline State University, Raleigh, North Caroline, 2007. http://repository.lib.ncsu.edu/ir/ bitstream/1840.16/4346/1/etd.pdf. Jones KG, Cooper WJ, Mezyk SP. Bimolecular rate constant determination for the reaction of hydroxyl radicals with domoic and kainic acid in aqueous solution. Environ Sci Technol 2009;43:6764–8. Jovanovic SV, Hara Y, Steenken S, Simic MG. Antioxidant potential of gallocatechins. A pulse radiolysis and laser photolysis study. J Am Chem Soc 1995;117:9881–8. Madhavan J, Grieser F, Ashokkumar M. Combined advanced oxidation processes for the synergistic degradation of ibuprofen in aqueous environments. J Hazard Mater 2010;178:202–8. Marco-Urrea E, Pereztrujillo M, Vicent T, Caminal G. Ability of white-rot fungi to remove selected pharmaceuticals and identification of degradation products of ibuprofen by Trametes versicolor. Chemosphere 2009;74:765–72. Méndez-Arriaga F, Torres-Palma RA, Pétrier C, Esplugas S, Gimenez J, Pulgarin C. Ultrasonic treatment of water contaminated with ibuprofen. Water Res 2008;42:4243–8. Méndez-Arriaga F, Esplugas S, Giménez J. Degradation of the emerging contaminant ibuprofen in water by photo-Fenton. Water Res 2010;44:589–95.
Miranda MA, Morera I, Vargas F, Gómez-Lechón MJ, Castell JV. In vitro assessment of the phototoxicity of anti-inflammatory 2-arylpropionic acids. Toxicol In Vitro 1991;5: 451–5. Packer JL, Werner JJ, Latch DE, McNeill K, Arnold WA. Photochemical fate of pharmaceuticals in the environment: naproxen, diclofenac, clofibric acid, and ibuprofen. Aquat Sci 2003;65:342–51. Rácz G, Csenki Z, Kovács R, Hegyi Á, Baska F, Sujbert L, et al. Subacute toxicity assessment of water disinfection byproducts on Zebrafish. Pathol Oncol Res 2011;18: 579–84. Roder M, Wojnárovits L, Földiák G. Pulse radiolysis of aqueous solutions of aromatic hydrocarbons in the presence of oxygen. Radiat Phys Chem 1990;36:175–6. Sehested K, Corfitzen H, Christensen HC, Hart EJ. Rates of reaction of O−, OH and H with methylated benzenes in aqueous solution. Optical spectra of radicals. J Phys Chem 1975;79:310–5. Sharma VK, Mishra SK. Ferrate(VI) oxidation of ibuprofen: a kinetic study. Environ Chem Lett 2006;3:182–5. Skoumal M, Rodríguez RM, Cabot PL, Centellas F, Garrido JA, Arias C, et al. Electro-Fenton, UVA photoelectro-Fenton and solar photoelectro-Fenton degradation of the drug ibuprofen in acid aqueous medium using platinum and boron-doped diamond anodes. Electrochim Acta 2009;54:2077–85. Spinks JWT, Woods RJ. Introduction to radiation chemistry. 3rd ed. New York: Wiley-Interscience; 1990. von Sonntag C. Free-radical-induced DNA damage and its repair: a chemical perspective. ed. Heidelberg: Springer; 2006. von Sonntag C, Schuchmann H-P. The chemistry behind the application of ionizing radiation in water-pollution abatement. Stud Phys Theor Chem 2001;87:657–70. Zheng BG, Zheng Z, Zhang JB, Luo XZ, Wang JQ, Liu Q, et al. Degradation of the emerging contaminant ibuprofen in aqueous solution by gamma irradiation. Desalination 2011;276:379–85. Zwiener C, Frimmel FH. Oxidative treatment of pharmaceuticals in water. Water Res 2000;34:1881–5.
Contents lists available at SciVerse ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Hydroxyl radical induced degradation of ibuprofen Erzsébet Illés a, b,⁎, Erzsébet Takács b, András Dombi a, Krisztina Gajda-Schrantz a, c, d, Gergely Rácz b, Katalin Gonter b, László Wojnárovits b a
Institute of Chemistry, Research Group of Environmental Chemistry, University of Szeged, Szeged, Hungary Institute of Isotopes, Centre for Energy Research, Hungarian Academy of Sciences, Budapest, Hungary Department of Inorganic and Analytical Chemistry, University of Szeged, Szeged, Hungary d EMPA, Laboratory for High Performance Ceramics, Duebendorf, Switzerland b c
H I G H L I G H T S
G R A P H I C A L
A B S T R A C T
► In hydroxyl radical attack on the ring mainly hydroxylated products form ► The hydrated electron attacks the carboxyl group. ► Oxidative conditions are more effective in ibuprofen decomposition than reductive. ► Ecotoxicity of ibuprofen solution first increases then decreases with irradiation.
a r t i c l e
i n f o
Article history: Received 11 September 2012 Received in revised form 18 December 2012 Accepted 1 January 2013 Available online 11 February 2013 Keywords: Ibuprofen Advanced oxidation processes Radiolysis Hydroxyl radical Hydrated electron
a b s t r a c t Pulse radiolysis experiments were used to characterize the intermediates formed from ibuprofen during electron beam irradiation in a solution of 0.1 mmol dm −3. For end product characterization 60Co γ-irradiation was used and the samples were evaluated either by taking their UV–vis spectra or by HPLC with UV or MS detection. The reactions of •OH resulted in hydroxycyclohexadienyl type radical intermediates. The intermediates produced in further reactions hydroxylated the derivatives of ibuprofen as final products. The hydrated electron attacked the carboxyl group. Ibuprofen degradation is more efficient under oxidative conditions than under reductive conditions. The ecotoxicity of the solution was monitored by Daphnia magna standard microbiotest and Vibrio fischeri luminescent bacteria test. The toxic effect of the aerated ibuprofen solution first increased upon irradiation indicating a higher toxicity of the first degradation products, then decreased with increasing absorbed dose. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Abbreviations: UV–vis, ultraviolet–visible; HPLC, high-performance liquid chromatography; MS, mass spectrometry; IBP, ibuprofen; EC50, effective concentration for 50% luminescence reduction; AOP, advanced oxidation processes. ⁎ Corresponding author at: Institute of Isotopes, Centre for Energy Research, Hungarian Academy of Sciences, Budapest, Hungary. Tel./fax: +36 62 544 338. E-mail addresses: [email protected], [email protected] (E. Illés). 0048-9697/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scitotenv.2013.01.007
Ibuprofen (IBP) ((RS)-2-(4-(2-methylpropyl)phenyl)propanoic acid, C13H18O2) is used for the relief of symptoms of arthritis, primary dysmenorrhoea, fever, and as an analgesic, especially to heal inflammatory disease(s). Ibuprofen is known to have an antiplatelet effect,
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
287
though it is relatively mild and short-lived when compared to that of acetylsalicylic acid or other better known antiplatelet drugs. It is most often sold with trade names Nurofen, Advil and Motrin. IBP is a moderately toxic compound, its − log EC50 value (EC50, effective concentration for 50% luminescence reduction) according to the Microtox luminescent bacteria test is reported to be 3.85 (Escher et al., 2005). The solubility of IBP in water is limited to about 1 × 10−4 mol dm −3 (Zwiener and Frimmel, 2000). IBP shows little absorbance in the wavelength region of the solar spectrum, its photodegradation in surface waters is limited and its biodegradation is also slow (Packer et al., 2003). IBP is regularly detected in the effluents of wastewater treatment plants and in surface waters at ng–μg dm −3 level. Although this concentration level is very low, IBP and its probably more harmful metabolites – together with other drugs – may present a potential hazard for human health and also for the aquatic ecosystem. The potential application of several advanced oxidation processes (AOP) for the degradation of IBP in dilute aqueous solutions has already been studied, using electro-Fenton, UVA photoelectro-Fenton and solar photoelectro-Fenton (Méndez-Arriaga et al., 2008; Skoumal et al., 2009), as well as photocatalysis and sonophotocatalysis (Madhavan et al., 2010; Méndez-Arriaga et al., 2010), ozonation (Huber et al., 2003), radiolytic (Zheng et al., 2011) and Ferrate(VI) processes (Sharma and Mishra, 2006). In AOP's the degradation is due mainly to reactions with hydroxyl radical. Hydroxylated IBP molecules were detected among the first degradation products by using mass spectrometry. This hydroxylation was assumed to take place on the side chains in the tertiary positions, suggesting •OH attack on the side chains (Skoumal et al., 2009; Madhavan et al., 2010; Méndez-Arriaga et al., 2010; Zheng et al., 2011). This suggestion is in contradiction with the findings on many other substituted aromatic molecules, where the main target of radical attack was the aromatic ring (Illés et al., 2012). For compounds with low biodegradability the photogenerated •OH in surface waters around the 10 − 14 mol dm − 3 concentration level is suggested to play an important role in their degradation (Packer et al., 2003; Jones et al., 2009). Studies on the •OH + IBP reaction may help to understand the fate of the compound in the environment. High energy ionizing radiation treatment is also regarded as AOP. Radiolytic studies on IBP were previously carried out by measuring end products (Zheng et al., 2011). Jones (2007) used pulse radiolysis to determine the rate coefficients of hydroxyl radical and hydrated electron reactions with IBP. In the present study the kinetics and mechanism of •OH reaction with IBP will be investigated by transient and final-product techniques. We use irradiation of aqueous systems as the cleanest source of hydroxyl radicals. These investigations in general may help to establish AOP technologies for IBP removal, and – more specifically – they can contribute to develop an irradiation technology for this purpose.
used on an Agilent 1200 system equipped with an XB-C18 Phenomenex Kinetex (100×2.1 mm, particle size 2.6 μm) capillary column coupled with on-line mass spectrometer. In this latter case, the flow rate was 0.3 cm3 min−1 and the mobile phases were acetonitrile and 0.1% aqueous acetic acid with the gradient: 0 min 15%, 15 min 50%, and 18 min 50% acetonitrile. For detection and identification of the parent compound and the degradation products a diode array detector at 220 and 260 nm and a tandem mass spectrometer were used. Mass spectrometric experiments were performed in the negative ion mode using an Agilent 6410 triple quadrupole mass spectrometer equipped with an ESI source. The following MS parameters were used for identification: drying gas N2 (350 °C, 10 dm3 min−1); nebulizer pressure with 1.7 bar, capillary voltage with 3500 V and fragmentor voltage with 80 V. The toxicity tests of the target compound and the degradation products formed were carried out using Vibrio fischeri bacteria (Luminescent bacteria test LCK 480) according to the DIN/EN/ISO standard no. 11348-2 by a HACH-LANGE GmbH LUMIStox 300 apparatus. The pH of samples was set to 7. The inhibition of the light emission in the presence of the sample was determined against a non-toxic control solution. The second toxicity test was performed on Daphnia magna zooplankton (DAPHTOXKIT F™). This organism originates from the second tropic level. Standard microbiotests were applied according to the ISO standard no. 6341, 1996. TOXKIT tests were suited for toxicity of all chemicals in aquatic and terrestrial environment. The mortality of the species was followed during 48 h incubation with 4 dilutions of the treated IBP solutions. Pulse radiolysis investigations were carried out using 800 ns pulses of accelerated electrons, and an optical detection system with a 1 cm cell pass-length (Földiák et al., 1988). The pH values were adjusted at 4.4 and 8.5, respectively. At pH 4.4 the concentrations of protonated and deprotonated forms of IBP are equal (pKa), at pH 8.5 we investigated the reaction of deprotonated form. The results obtained at pH 8.5 were rather similar to those obtained at pH 4.4. The results at pH 4.4 will be detailed in the following sections.
2. Experimental
yields also OH radicals, the reacting radicals and their yields are: hydroxyl radical 0.56 μmol J − 1, hydrogen atom 0.06 μmol J − 1. The intensity of the characteristic aromatic absorption band in the UV spectrum of IBP in the 250–300 nm region is very weak (Fig. 1). When the solution is irradiated with a dose of 0.25 kGy, the intensity of the band around 265 nm becomes higher, indicating the formation of changed aromatic molecules. In the radiolysis of many aromatic molecules hydroxylation was observed at the beginning of the transformations. This hydroxylation may take place in the side chains (Zheng et al., 2011) and also in the ring. Substitution in the side chain does not change the absorption spectrum at ~ 265 nm. A simple hydroxylation in the ring of IBP can result in two isomers. As the inset shows, a dose of 0.25 kGy results in transformation of more than 60% of the initial molecules to new molecules. The secondary degradation of the primary products may have already started. (The G-values of
IBP and the chemicals for pH setting were purchased from Spectrum-3D or Carlo Erba. In end product experiments the irradiation was carried out by using a 60Co γ-irradiation facility with 5 kGy h−1 dose rate. The pH values were set by HCl and NaOH. These solutes at low concentration do not influence considerably the radiolytic reactions of organic molecules. All experiments were carried out at room temperature. The samples were evaluated either by taking their UV–vis spectra using a JASCO 550 UV–vis spectrophotometer with a 1 cm cell, or by HPLC separation. For chromatographic separation two different systems were used. The Agilent 1100 HPLC system consists of a C18 Kinetex 2.6 μm 100A, 100 × 4.6 mm, Phenomenex column where the elution was performed using a 50:50 mixture of 1% aqueous acetic acid solution and acetonitrile (flow rate of 0.9 cm3 min−1). Linear gradient elution was
3. Results and discussion 3.1. Hydroxyl radical reactions − Hydroxyl radical ( •OH, 0.28), hydrated electron (eaq , 0.28) and hy• drogen atom (H , 0.06) are the main products of water radiolysis. All of them can be classified as “reactive intermediates”. The values in brackets are the yields, the so-called G-values in μmol J −1 units (Spinks and Woods, 1990). Fig. 1 depicts the UV absorption spectra of 1 × 10 −4 mol dm −3 N2O saturated IBP solutions before and after irradiation. When the solution is saturated with N2O, the transformation
eaq
−
•
−
þ N2 O þ H2 O → OH þ OH þ N2
ð1Þ
288
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
Fig. 1. UV absorption spectrum of 1 × 10−4 mol dm−3 IBP solution saturated with N2O (•OH reaction), pH~4.4. Inset: relative concentration of IBP in irradiated solution obtained by HPLC separation of the components and integrating the area under the chromatographic peak.
IBP degradation are calculated as 0.25, 0.18, 0.12 μmol J −1 at doses 0.25, 0.5 and 0.75 kGy, respectively.) With the increase of absorbed dose the band at 250–300 nm disappears, indicating the disintegration of the aromatic structure. A characterless absorption band remains here; it is probably due to the absorbance of a large number of degradation products. In the transient absorption spectrum of IBP, obtained in pulse radiolysis experiments, three bands are observed (Fig. 2). Two bands appear at 285 and 315 nm, respectively, with comparable intensities. The third one is a much weaker band at 385 nm. The band at 315 nm decays in about 1 ms, the decay of the 285 nm band is somewhat slower, while the intensity of the 385 nm band does not change considerably on the 2 ms timescale of the measurement. The different time behaviors of these bands may suggest different intermediates. It should be noted, however, that stable final products may also contribute to the observed absorbance below 300 nm. • OH radical reacts with IBP with a rate coefficient of ~ 7 × 10 9 mol − 1 dm 3 s − 1 (Huber et al., 2003; Packer et al., 2003). The molecule offers three special places for the attack: attachment to the aromatic ring forming in the first step hydroxycyclohexadienyl radical
Fig. 2. Transient absorption spectra obtained in 1×10−4 mol dm−3 IBP solution saturated with N2O (•OH reaction), pH~4.4, 16 Gy/pulse.
(two isomers), and hydrogen atom abstraction from the tertiary position of the carboxyl group bearing side chain, or hydrogen atom abstraction from the secondary position of the other side chain. In the latter two reactions, benzyl type radical is expected. Cyclohexadienyl type radicals typically have wide absorption bands in the 300–350 nm range with molar absorption coefficients of 3000–5000 mol − 1 dm 3 cm − 1. The benzyl type radicals have a very intense absorption peak around 260–275 nm with an absorption coefficient of 10,000–16,000 mol − 1 dm 3 cm − 1 as well as one or two weaker bands at 300–320 nm with molar absorption coefficients of 2000–4000 mol − 1 dm 3 cm − 1. The band at 385 nm is probably due to phenoxyl radical which may have been formed in the secondary reactions. Since phenoxyl radicals have molar absorption coefficients of several thousand mol − 1 dm 3 cm − 1 around 400 nm, the phenoxyl radical yield is very low. The bands at lower wavelengths probably belong to both hydroxycyclohexadienyl and benzyl type radicals. Due to the relatively low intensity of absorbance around 285 nm (molar absorption coefficient ~ 3000 mol − 1 dm 3 cm − 1), we assume that the hydroxycyclohexadienyl type radical dominates. This suggestion is in agreement with the pulse radiolysis results on large number of alkyl-substituted benzenes: less than 20% of •OH reacts with alkylbenzenes in hydrogen atom abstraction, the majority of the radicals add to the ring forming hydroxycyclohexadienyl radical intermediates (Sehested et al., 1975). The formation of products hydroxylated in the ring also supports this idea. − 3.2. Reactions under reducing conditions (eaq + H •)
The reactions of hydrated electron are studied in solutions containing tert-butanol. In such solutions the hydroxyl radical is converted to rather inactive •CH2(CH3)2COH radical in the reaction (Spinks and Woods, 1990): •
•
OH þ ðCH3 Þ3 COH → H2 O þ CH2 ðCH3 Þ2 COH:
ð2Þ
Since the reaction of H • is slow: •
•
H þ ðCH3 Þ3 COH → H2 þ CH2 ðCH3 Þ2 COH;
ð3Þ
H • may also react with IBP. As Fig. 3 shows, there is a gradual decrease of the UV absorption intensities during degradation. The shape of
Fig. 3. UV absorption spectrum of 1 × 10−4 mol dm−3 IBP solution saturated with N2, − containing 0.5 mol dm−3 tert-butanol (eaq reaction), pH ~4.4. Inset: relative concentration of IBP in irradiated solution obtained by HPLC separation of the components by integrating area under the chromatographic peak.
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
289
− the spectrum changes just slightly during eaq /H • reactions: the band around 265 nm is shifted slightly to longer wavelengths. From the absorption spectra and the integrated chromatographic peaks of IBP, it is obvious that the degradation is less efficient under reducing conditions (G ≈ 0.03 μmol J −1) than under oxidative conditions in solutions saturated with N2O, or with air (see later). The rate coefficient of hydrated electron reaction with IBP was found here to be 8.5 × 10 9 mol − 1 dm 3 s − 1. A value of 8.9×109 mol−1 dm3 s−1 was published in the literature (Jones, 2007). The rate coefficient is about one third of the rate coefficient found for ketoprofen 2.6 × 10 10 mol − 1 dm 3 s − 1. In the case of ketoprofen the hydrated electron attack takes place on the carbonyl oxygen (Illés et al., 2012), whereas the carboxyl group of IPB seems to be the main target. Electron attachment is probably followed by an immediate protonation.
þ
R−COOH þ eaq
−
H3 O → R−C• O− OH → R−C• ðOHÞ2 þ H2 O
ð4Þ
Under pulse radiolysis conditions the transient absorption spec− is very weak showing a maxitrum produced by the reaction of eaq mum around 340 nm (Fig. 4). This characterless spectrum belongs probably to the radical at the carboxyl group. Cyclohexadienyl type radicals produced by H• addition to the ring – with a rate coefficient of 4 × 109 mol−1 dm3 s−1 as measured here – may also contribute to the spectrum. These cyclohexadienyl radicals, as opposed to the hydroxycyclohexadienyl radicals produced in the • OH reaction, do not yield hydroxylated products. When the solution is deoxygenated by N2 bubbling, but does not − contain the tert-butanol, both •OH and eaq may react with the solute (Fig. 5). The UV spectra obtained after γ irradiation are similar to the spectra measured in N2O saturated solution, although the effect of irradiation is smaller in the solution with N2 bubbling. In the latter solution the transformations mainly take place in the reactions of • OH. The transient absorption spectrum in Fig. 6 also shows the similarities to the spectrum obtained in the solution saturated with N2O. Apparently, the cross-reaction of intermediates formed in •OH and − eaq reactions is negligible.
Fig. 5. UV absorption spectrum of 1 × 10−4 mol dm−3 IBP solution saturated with N2, − (•OH + eaq reaction), pH ~4.4. Inset: relative concentration of IBP in irradiated solution obtained by HPLC separation of the components by integrating the area under the chromatographic peak.
and H• produce O2−•/HO2• pair in the following reactions (pKa = 4.8) (Spinks and Woods, 1990): eaq •
−
þ O2 → O2 •
−•
H þ O2 → HO2 :
ð5Þ ð6Þ
From practical point of view the investigations in air saturated solu− tions are of utmost importance. When dissolved oxygen is present, eaq
In such a system •OH remains to be the main reacting intermediate, which is expected to act the same way as in the solution saturated with N2O. The reactivity of the O2−•/HO2• pair with aromatic molecules is very low (Jovanovic et al., 1995). In the UV absorption spectra of solutions containing dissolved oxygen and irradiated with doses of 0.25, 0.5 and 0.75 kGy, there is a well resolved absorption band with a maximum at 260 nm (Fig. 7). Its intensity first increases with the dose and decreases having reached a maximum. As mentioned previously, this change in absorbance is attributed to the formation and decay of products hydroxylated in the aromatic ring. Such products are formed when •OH radical reacts with the aromatic ring and the hydroxycyclohexadienyl transient radical finally transforms to
Fig. 4. Transient absorption spectra obtained in 1×10−4 mol dm−3 IBP solution saturated with N2, containing 0.5 mol dm−3 tert-butanol, pH~4.4, 18 Gy/pulse.
Fig. 6. Transient absorption spectra obtained in 1×10−4 mol dm−3 IBP solution saturated − with N2, without tert-butanol (•OH+eaq reaction), pH~4.4, 19 Gy/pulse.
3.3. Reactions in the presence of dissolved O2
290
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
Fig. 8. Total ion chromatogram (TIC) and chromatogram monitored at m/z 221, ibuprofen solution saturated with N2O, after 0.25 kGy dose. Fig. 7. UV absorption spectrum of 1 × 10−4 mol dm−3 IBP solution saturated with air (•OH + O2−•/HO2• reaction), pH ~4.4, irradiated with different doses. Inset: relative concentration of IBP in irradiated solution obtained by HPLC separation of the components and integrating the area under the chromatographic peak.
a hydroxylated molecule. The cyclohexadienyl radicals readily react with dissolved O2, forming peroxy radicals. These peroxy radicals may eliminate HO2• (Roder et al., 1990), in this elimination reaction phenol type molecule is formed:
ð7Þ This elimination reaction gives an explanation to the higher intensity of the aromatic peak around 260 nm in air saturated solution than in N2O bubbled solution in spite of the reversed order of •OH yields (0.28, and 0.56 μmol J − 1, respectively). The peroxy radicals may undergo also ring opening reaction forming linear dicarboxylic acid molecules (von Sonntag and Schuchmann, 2001; von Sonntag, 2006). The ring-opening reaction is an important step of mineralization. In air saturated solution the estimated yield (G-value, 0.25 μmol J −1) agrees practically with the yield of •OH (0.28 μmol J−1). The reason is that the radical scavenging reaction opens a new reaction channel for degradation and the intermediates cannot be transformed back to the starting molecules. 3.4. Identification of final products Due to the low solubility, low solute concentration and low absorbed doses were used in these studies. In the HPLC chromatograms large number of primary, secondary, etc. products appeared with low yields. As a result, their identification and establishing the degradation mechanism were rather complicated. Except the tert-butanol containing solutions saturated with N2, at the lowest absorbed dose, the main products had a mass number of 221, being higher by 16 units than the m/z value of IBP (205) in the negative ion mode. Thus, the main products at the lowest absorbed dose are hydroxylated molecules. In our chromatograms taken in case of solutions saturated with N2O or air, 6 peaks appear with the same 221 m/z value (Fig. 8). One of these peaks is predominant, having outstandingly large peak area in the MS detection. In the literature, as it
was mentioned in the Introduction, the side chain of the molecule in the tertiary position (Fig. 9, structure (7)) (Marco-Urrea et al., 2009; Méndez-Arriaga et al., 2010) and/or in the secondary position are suggested to be hydroxylated (Caviglioli et al., 2002; Madhavan et al., 2010; Méndez-Arriaga et al., 2010; Zheng et al., 2011). In contrast to these suggestions our γ-radiolytic experiments with diode array detection made parallel with MS detection clearly show that hydroxylation in the ring is of high importance (see above, in Section 3.1.). From the two possible variations, one of the isomers is shown in Fig. 9 (structure (8)). In addition to hydroxylation of the ring, the products observed suggest that hydroxylation of the side chains also takes place. From the large number of products in Fig. 9, those with m/z values of 279, 351 and 237 should also be mentioned. Product (10) and (11) are formed only in solutions saturated with N2 and containing also tert-butanol. These are tert-butanol adducts. Product (9) is one of the double hydroxylated derivatives of ibuprofen. Double hydroxylated products are formed when solutions saturated with N2O or air were radiolysed. Product (1) with an m/z value of 133 was also detected by Caviglioli et al. (2002) and Zheng et al. (2011). It may have been formed in •OH induced degradation. This product was not detected in N2 saturated, tert-butanol containing solutions. The literature mentioned 4-isobutylacetophenone (3) as a degradation product of IBP with an m/z value of 175 (Castell et al., 1987; Skoumal et al., 2009). In our experiments under aerobic conditions six minor products with this mass number were separated. In the chromatogram there was only one product which elutes later than the ibuprofen, with m/z value of 171. Its presumable structure could not be determined. 3.5. Ecotoxicity Ibuprofen is considered to be a moderately toxic compound. First, the acute toxicity was determined with V. fischeri liquid-dried luminescent bacteria. Based on the experimental results, as Fig. 10 shows, in spite of the considerable decrease of IBP concentration upon irradiation, the inhibition of the light emission initially increases and then decreases. This phenomenon can be explained with the formation of toxic degradation products during the radiolysis of ibuprofen. In the inset of Fig. 10, the change of the concentration of some main products can be seen. The second toxicity test was performed on D. magna zooplankton which showed that ibuprofen is not toxic in the applied concentration. Radiolysis caused some increase in the mortality, this increase was, however, statistically not relevant. A slight increase of toxicity was also reported in the gamma radiolytic investigations of Zheng et al. (2011). Due to the large number of by-products it is difficult to claim which of them was responsible for the substantial increase of toxicity. Based
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
291
Fig. 9. Ibuprofen degradation products formed during radiolysis.
on some publications (Castell et al., 1987; Miranda et al., 1991) 3 (Fig. 9) is a toxic by-product because it has a lytic activity on erythrocytes. In the literature 1,2-dihydroxy ibuprofen (one of the double hydroxylated derivatives of ibuprofen) with a molecular mass of 238 was declared to be more toxic than IBP (Marco-Urrea et al., 2009). 1-(4-Isobutylphenyl)-1-ethanol (4) was also found to have toxic effect examined by measuring the amount of cell protein and intracellular enzymatic activity of LDH and GOT in cultured fibroblasts (Castell et al., 1987; Marco-Urrea et al., 2009). Dystrophic lesions affecting the liver, kidney, and skin were caused by degradation product (1) using zebrafish (Danio rerio) as the test object (Rácz et al., 2011). The ibuprofen solution after an absorbed dose of 2 kGy still preserved its toxic effect, because of the formation of aldehydes, carboxylic acids with small molecular weight and organic peroxides coming from the degradation of the aromatic part. These compounds have light absorption at short wavelength; therefore, cannot be observed in our system.
Maybe some hydrogen peroxides are also present in the solution after irradiation. These compounds have acute toxicity, however, under natural conditions they degrade easily. Due to the formation of toxic by-products, it is clear, that the absorbed dose during the treatment should be higher than needed for complete degradation of IBP since the toxic by-products should also be decomposed. Our experiments showed that the toxicity of the solutions could be strongly reduced with prolonged irradiation. 4. Conclusion The results of this work show that the degradation of initial IBP molecules is more efficient under oxidative conditions than under reducing conditions. The UV spectra obtained after γ irradiation in N2 bubbled solution are similar to the spectra measured in N2O saturated solution, although the effect of irradiation is smaller. Large number of primary and secondary products was formed with low yields during radiolysis. Except for the solutions saturated with N2 and containing also tert-butanol, where the hydrated electron reacts with the solute, the main products are hydroxylated molecules at the lowest absorbed dose. Due to the formation of toxic by-products, the absorbed dose should be higher during the treatment than needed for complete degradation of IBP. Our experiments showed that with prolonged irradiation the ecotoxic effect of the treated solutions could be strongly reduced. Acknowledgment The authors thank the Hungarian Science Foundation (OTKA, No. CK 80154 and NK 105802), the Swiss-Hungarian project (No. SH7/2/14) and the International Atomic Energy Agency (Contract No. 16485) for the support.
Fig. 10. Changes of fluorescence inhibition (I, %) in Vibrio fischeri toxicity test, and the relative concentration of unaltered IBP molecules as a function of dose, measured in 1×10−4 mol dm−3, IBP solution saturated with air (•OH+O2−•/HO2• reaction). The inset shows the absorbance changes of products determined with diode array detection after HPLC separation.
References Castell JV, Gomez-L. MJ, Miranda MA, Morera IM. Photolytic degradation of ibuprofen. Toxicity of the isolated photoproducts on fibroblasts and erythrocytes. Photochem Photobiol 1987;46:991–6.
292
E. Illés et al. / Science of the Total Environment 447 (2013) 286–292
Caviglioli G, Valeria P, Brunella P, Sergio C, Attilia A, Gaetano B. Identification of degradation products of ibuprofen arising from oxidative and thermal treatments. J Pharm Biomed Anal 2002;30:499–509. Escher BI, Bramaz N, Maurer M, Richter M, Sutter D, Kanel CV, et al. Screening test battery for pharmaceuticals in urine and wastewater. Environ Toxicol Chem 2005;24:750–8. Földiák G, Hargittai P, Kaszanyiczki L, Wojnárovits L. A computer-controlled pulse-radiolysis laboratory. J Radioanal Nucl Chem 1988;125:19–28. Huber MM, Canonica S, Park G-Y, von Gunten U. Oxidation of pharmaceuticals during ozonation and advanced oxidation processes. Environ Sci Technol 2003;37:1016–24. Illés E, Takács E, Dombi A, Gajda-Schrantz K, Gonter K, Wojnárovits L. Radiation induced degradation of ketoprofen in dilute aqueous solution. Radiat Phys Chem 2012;81: 1479–83. Jones GK. Applications of radiation chemistry to understand the fate and transport of emerging pollutants of concern in coastal waters. Ph.D. dissertation, North Caroline State University, Raleigh, North Caroline, 2007. http://repository.lib.ncsu.edu/ir/ bitstream/1840.16/4346/1/etd.pdf. Jones KG, Cooper WJ, Mezyk SP. Bimolecular rate constant determination for the reaction of hydroxyl radicals with domoic and kainic acid in aqueous solution. Environ Sci Technol 2009;43:6764–8. Jovanovic SV, Hara Y, Steenken S, Simic MG. Antioxidant potential of gallocatechins. A pulse radiolysis and laser photolysis study. J Am Chem Soc 1995;117:9881–8. Madhavan J, Grieser F, Ashokkumar M. Combined advanced oxidation processes for the synergistic degradation of ibuprofen in aqueous environments. J Hazard Mater 2010;178:202–8. Marco-Urrea E, Pereztrujillo M, Vicent T, Caminal G. Ability of white-rot fungi to remove selected pharmaceuticals and identification of degradation products of ibuprofen by Trametes versicolor. Chemosphere 2009;74:765–72. Méndez-Arriaga F, Torres-Palma RA, Pétrier C, Esplugas S, Gimenez J, Pulgarin C. Ultrasonic treatment of water contaminated with ibuprofen. Water Res 2008;42:4243–8. Méndez-Arriaga F, Esplugas S, Giménez J. Degradation of the emerging contaminant ibuprofen in water by photo-Fenton. Water Res 2010;44:589–95.
Miranda MA, Morera I, Vargas F, Gómez-Lechón MJ, Castell JV. In vitro assessment of the phototoxicity of anti-inflammatory 2-arylpropionic acids. Toxicol In Vitro 1991;5: 451–5. Packer JL, Werner JJ, Latch DE, McNeill K, Arnold WA. Photochemical fate of pharmaceuticals in the environment: naproxen, diclofenac, clofibric acid, and ibuprofen. Aquat Sci 2003;65:342–51. Rácz G, Csenki Z, Kovács R, Hegyi Á, Baska F, Sujbert L, et al. Subacute toxicity assessment of water disinfection byproducts on Zebrafish. Pathol Oncol Res 2011;18: 579–84. Roder M, Wojnárovits L, Földiák G. Pulse radiolysis of aqueous solutions of aromatic hydrocarbons in the presence of oxygen. Radiat Phys Chem 1990;36:175–6. Sehested K, Corfitzen H, Christensen HC, Hart EJ. Rates of reaction of O−, OH and H with methylated benzenes in aqueous solution. Optical spectra of radicals. J Phys Chem 1975;79:310–5. Sharma VK, Mishra SK. Ferrate(VI) oxidation of ibuprofen: a kinetic study. Environ Chem Lett 2006;3:182–5. Skoumal M, Rodríguez RM, Cabot PL, Centellas F, Garrido JA, Arias C, et al. Electro-Fenton, UVA photoelectro-Fenton and solar photoelectro-Fenton degradation of the drug ibuprofen in acid aqueous medium using platinum and boron-doped diamond anodes. Electrochim Acta 2009;54:2077–85. Spinks JWT, Woods RJ. Introduction to radiation chemistry. 3rd ed. New York: Wiley-Interscience; 1990. von Sonntag C. Free-radical-induced DNA damage and its repair: a chemical perspective. ed. Heidelberg: Springer; 2006. von Sonntag C, Schuchmann H-P. The chemistry behind the application of ionizing radiation in water-pollution abatement. Stud Phys Theor Chem 2001;87:657–70. Zheng BG, Zheng Z, Zhang JB, Luo XZ, Wang JQ, Liu Q, et al. Degradation of the emerging contaminant ibuprofen in aqueous solution by gamma irradiation. Desalination 2011;276:379–85. Zwiener C, Frimmel FH. Oxidative treatment of pharmaceuticals in water. Water Res 2000;34:1881–5.