ARTICLE pubs.acs.org/est
Kinetic and Mechanistic Investigations of the Oxidation of Tramadol by Ferrate and Ozone Saskia G. Zimmermann,†,‡,§ Annekatrin Schmukat,|| Manoj Schulz,|| Jessica Benner,||,^ Urs von Gunten,*,†,‡,§ and Thomas A. Ternes*,|| € Eawag, Swiss Federal Institute of Aquatic Science and Technology, Uberlandstr. 133, 8600 D€ubendorf, Switzerland Institute of Biogeochemistry and Pollutant Dynamics, ETH Z€urich, 8092 Z€urich, Switzerland § School of Architecture, Civil and Environmental Engineering (ENAC), Ecole Polytechnique Federale de Lausanne (EPFL), 1015 Lausanne, Switzerland Federal Institute of Hydrology (BfG), Am Mainzer Tor 1, 56068 Koblenz, Germany †
)
‡
bS Supporting Information ABSTRACT: The kinetics and oxidation products (OPs) of tramadol (TRA), an opioid, were investigated for its oxidation with ferrate (Fe(VI)) and ozone (O3). The kinetics could be explained by the speciation of the tertiary amine moiety of TRA, with apparent second-order rate constants of 7.4 ((0.4) M1 s1 (Fe(VI)) and 4.2 ((0.3) 104 M1 s1 (O3) at pH 8.0, respectively. In total, six OPs of TRA were identified for both oxidants using Qq-LIT-MS, LTQ-FT-MS, GC-MS, and moiety-specific chemical reactions. In excess of oxidants, these OPs can be further transformed to unidentified OPs. Kinetics and OP identification confirmed that the lone electron pair of the amine-N is the predominant site of oxidant attack. An oxygen transfer mechanism can explain the formation of N-oxide-TRA, while a one-electron transfer may result in the formation of N-centered radical cation intermediates, which could lead to the observed N-dealkylation, and to the identified formamide and aldehyde derivatives via several intermediate steps. The proposed radical intermediate mechanism is favored for Fe(VI) leading predominantly to N-desmethyl-TRA (ca. 40%), whereas the proposed oxygen transfer prevails for O3 attack resulting in N-oxide-TRA as the main OP (ca. 90%).
’ INTRODUCTION Tramadol (TRA) (Figure 1) is a synthetic, centrally acting analgesic agent used for the relief of moderate to severe acute and chronic pain, and shows a potency ranging between weak opioids and morphine.1 Approximately 1030% of TRA is excreted unchanged via urine. In most species, the principle metabolites produced in the liver are O-desmethyltramadol (O-DES) and N-desmethyltramadol (N-DES). These primary metabolites may be further metabolized to N,N-bidesmethyltramadol (BIDES), N,N,O-tridesmethyltramadol, and N,O-desmethyltramadol. All metabolites can undergo conjugation reactions with glucuronic acid and sulfate prior to excretion via urine.2 TRA is used in both human and veterinary medicine.3 In 2004, a total of 25.3 tons were prescribed in Germany.4 Accordingly, concentrations of up to 97 μg L1 and 6 μg L1 were reported in secondary effluent and surface water, respectively, underlining the environmental occurrence of TRA.5 Currently, ozonation of secondary effluents is discussed as one of the most promising options for the mitigation of micropollutants entering the aqueous environment via discharges from municipal wastewater treatment plants (WWTPs), and can be r 2011 American Chemical Society
Figure 1. Chemical structure of tramadol (TRA), including numbering of carbon atoms.
successfully operated at the full-scale level in municipal WWTPs as a tertiary treatment.6,7 Ferrate (Fe(VI)) is a promising new oxidant for wastewater treatment achieving both oxidation of a broad range of micropollutants as well as precipitation of phosphate from wastewater.810 The use of Fe(VI) as an oxidant in the treatment of bromide-containing waters has the additional advantage of not producing bromate,11 a potentially carcinogenic Received: September 23, 2011 Accepted: November 15, 2011 Revised: November 14, 2011 Published: December 21, 2011 876
dx.doi.org/10.1021/es203348q | Environ. Sci. Technol. 2012, 46, 876–884
Environmental Science & Technology
ARTICLE
by-product, which is formed during ozonation.12 Research to implement Fe(VI) application in wastewater treatment is currently carried out, and mainly focuses on the development of on-site Fe(VI) production techniques by electrochemical methods.13 Oxidative treatment does not typically result in full mineralization but in transformation of the target compounds. For a comprehensive environmental assessment of an oxidative technology especially with regard to ecotoxicological concerns, it is crucial to elucidate the transformation pathways of micropollutants. Mechanistic information on ozonation products formed from micropollutants containing reactive functional groups such as olefins,14 phenols15 or amines1619 is already available, but much less is known about oxidation products (OPs) and mechanisms for Fe(VI) reactions. Studies for olefinic carbamazepine20 as well as the phenolics triclosan21 and bisphenol A22 are available, however, limited information is available for amine-containing micropollutants. In general, Fe(VI) reacts with inorganic and organic compounds via one-electron transfer, hydrogen abstraction or oxygen transfer.11,23 An example of a hydrogen abstraction mechanism is the oxidation of phenol by Fe(VI) producing a phenoxy radical and Fe(V), finally resulting in the formation of quinone.23,24 Oxygen transfer has been proposed for the oxidation of AsO33 to AsO43.25 Some studies investigating the oxidation of the N-containing compounds glycine and nitrilotriacetic acid by Fe(VI) showed cleavage of N-C bonds.26,27 Nitrobenzene derived from oxygen transfer as well as coupling products were identified as Fe(VI) OPs from aniline,28 and in analogy, nitroso- and nitro-OPs from sulfamethoxazole attack.29 In the present paper, the kinetics and OP formation of TRA, containing an aliphatic tertiary amine moiety, are investigated for Fe(VI) and O3. Reaction mechanisms are postulated based on kinetic data, identified OPs, mass balances in buffered solutions, and available information from literature.
to fully decay. In total, this procedure was repeated 14 times to reach sufficient Fe(VI) exposure. The pH was controlled after each step and if necessary, readjusted to 8.5 by the addition of H3PO4 (0.1 - 1.0 M). The solution was kept at 4 C until complete Fe(III)(hydr)oxide precipitation. The unfiltered supernatant was used for product identification. An O3 stock solution was added to the reaction solution containing TRA (100 μM) and phosphate buffer (50 mM) at pH 8.0, resulting in O3:TRA ratios of 1:5, 1:1, 5:1, and 10:1. To exclude the influence of hydroxyl radicals during OPs formation, 100 mM tert-butanol was spiked as a radical scavenger. Preconcentration of OPs formed by ozone was subsequently achieved by freeze-drying 1 L of the 10:1 reaction solution, and redissolving it in deionized water. Separation by HPLC-UV/FLD-Fraction Collector. An Agilent 1100 system (Agilent Technologies, Santa Clara, CA) was used to separate the mixtures of OPs. HPLC conditions are described in SI Text S3 and Table S2. Fractions of the eluate were subsequently collected by a fraction collector (Advantec SF-2120 Super Fraction Collector, Techlab GmbH, Erkerode, Germany). Fe(VI) OPs fractions were concentrated in brown glass bottles heated to 40 C in a water bath under a gentle N2 stream. Determination of Molecular Weights and Fragmentation by Mass Spectrometry. Molecular weights of TRA OPs produced from Fe(VI) or O3 were determined by Q1 scans (501000 m/z) of a hybrid triple quadrupole-linear ion trap mass spectrometer (Qq-LIT-MS, Applied Biosystems Sciex 4000 Q Trap (Langen, Germany)). Structural elucidation of TRA OPs was based on their MS2 and MS3 fragmentation pathways obtained from product ion (PI), precursor or MS3 scans using Qq-LIT-MS. The exact masses and hence sum formula of the OPs were further determined by a linear triple quadrupole ion trap Fourier transformation mass spectrometer (LTQ-FT-MS, LTQ Orbitrap Velos, Thermo Scientific, Bremen, Germany). Both MS systems were equipped with an electrospray ionization source (ESI). The optimized MS conditions for each OP were included into the multiple reaction mode (MRM) transition method, and are provided in SI Table S3. Derivatization and Subsequent GC-MS Detection. The fractions of OP 249 and OP 264a were derivatized with a variety of different agents and subsequently measured by GC-MS (SI Text S4). Purpald Test. Each OP fraction was tested for aldehydes by the Purpald test.30,31 An aliquot (500 μL) of each fraction was added to 2 mL reactant solution, consisting of 150 mg 4-amino-3hydrazino-5-mercapto-1,2,4-triazole (Purpald) dissolved in 2 mL 1 N NaOH. Glutaraldehyde was used as positive control, and Milli-Q water and the final HPLC eluent mixture were used as negative controls. pH Dependence of LC Retention Time. The presence or absence of basic or acid moieties in isolated OP fractions was checked by determining the retention times (Rt) at pH 3.5, 7.0, and 10.0 (with ammoniumacetate and methanol as eluents) in RP-LC-Qq-LIT-MS, using a 4.6 100 mm nonmodified, endcapped Chromolith C18 column (Merck, Darmstadt, Germany). Mass Balance. A mass balance experiment for TRA OPs with Fe(VI) was carried out at pH 8.0 in 100 mM phosphate buffer with [TRA]0 = 50 μM, and [Fe(VI)]0 = 500 μM. Fe(VI) was added nine times in order to keep [Fe(VI)] in excess to [TRA] until complete TRA oxidation was achieved (SI Figure S1). The pH was controlled after each Fe(VI) addition and if necessary, readjusted to pH 8.0 with H3PO4 (0.11.0 M). At proper time intervals, [Fe(VI)] was determined by the ABTS method,32 and
’ MATERIALS AND METHODS Chemicals and Reagents. All chemicals and solvents were of analytical grade (g95%). Information on authentic reference standards and oxidant preparation is given in Table S1 and Text S1 of the Supporting Information (SI). Reaction Kinetics. All kinetic experiments were carried out at 22 ( 2 C in duplicate or triplicate. Apparent second-order rate constants kapp for the reaction of TRA and N-DES with Fe(VI) were measured in excess of Fe(VI) in the pH range 6.510.5. For BIDES, N-oxide-tramadol (N-OXID) and anisole, kapp values for the reaction with Fe(VI) were measured in excess of the compound at pH 8.0. For O3, kapp values for the reaction with TRA were determined in excess of O3 in the pH range 3.07.5 by direct measurements in batch experiments, and in excess of compound in the pH range 7.58.5 using competition kinetics with cephalexin as a competitor. 20 mM tert-butanol was spiked as a radical scavenger. Competition kinetics experiments were used for determining kapp values for the reaction of TRA with hydroxyl radicals at pH 8.0. Further details are given in SI Text S2. Identification of Oxidation Products. Sample Preparation. TRA OPs from Fe(VI) and O3 treatment were produced in bench-scale experiments. A Fe(VI) stock solution was added to a 1670 μM TRA solution buffered to pH 8.5 (125 mM phosphate) to yield 200 μM Fe(VI). After complete decay of the 200 μM Fe(VI), a second aliquot of the Fe(VI) stock solution was added to yield 200 μM Fe(VI). The 200 μM Fe(VI) were again allowed 877
dx.doi.org/10.1021/es203348q |Environ. Sci. Technol. 2012, 46, 876–884
Environmental Science & Technology
ARTICLE
Table 1. Second-Order Rate Constants kapp [M1 s1] for the Reaction of Fe(VI), O3 and Hydroxyl Radicals with Tramadol as well as for the Reaction of Fe(VI) with the Tramadol Oxidation Products N-Desmethyl-Tramadol, N,N-Bidesmethyl-Tramadol, and N-Oxide-Tramadol, and the Model Compound Anisole at 22 ( 2 Ca substance tramadol (TRA)
reacting
k1 / k2 (Fe(VI)) k5
kapp
kapp
pKa
species
/ k6 (O3)
(pH 7)
(pH 8)
9.4 b
HFeO4 + PH+/HFeO4 + P
1.7 ((0.2) 101/1.2 ((0.1) 103
1.4 ((0.1) 101
7.4 ((0.4)
7.7 ((0.2) 10 /1.0 ((0.1) 10
2.2 ((0.2) 103
4.2 ((0.3) 104
+
1
O3 + PH /O3 + P •
6
6.3 ((0.2) 109 c
+
OH + PH /P N-desmethyl-tramadol (N-DES) N,N-bidesmethyl-tramadol (BIDES)
10.6 na
N-oxide-tramadol (N-OXID)
na
anisole a
b
d
HFeO4 + PH+/HFeO4 HFeO4/FeO42‑+ PH+/P HFeO4/FeO42‑+ PH+/P HFeO4/FeO42‑ + P
+P
2.4 ((0.6) 10 /7.9 ((1.6) 10 1
3
2.8 ((0.3) 101
7.6 ((0.8) 3.5 ((0.3) 101 1.3 ((0.2)
