Chemosphere 86 (2012) 130–137

Contents lists available at SciVerse ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Sunlight-induced degradation of soil-adsorbed veterinary antimicrobialsMarbofloxacin and Enrofloxacin

Michela Sturini ⇑, Andrea Speltini, Federica Maraschi, Antonella Profumo ⇑, Luca Pretali, Elisa Fasani,Angelo AlbiniDepartment of Chemistry, via Taramelli 12, 27100 Pavia, Italy

a r t i c l e i n f o a b s t r a c t

Article history:Received 20 June 2011Received in revised form 26 July 2011Accepted 30 September 2011Available online 1 November 2011

Keywords:AntimicrobialDepollution pathFluoroquinolonePhotodegradationPhotoproductSoil

0045-6535/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.chemosphere.2011.09.053

⇑ Corresponding authors. Tel.: +39 0382 987581; faE-mail addresses: michela.sturini@unipv.it (M.

unipv.it (A. Profumo).

Marbofloxacin (MAR) and Enrofloxacin (ENR), two largely employed veterinary Fluoroquinolones (FQs),were found to be present at the micrograms per kilogram level in agricultural soils of South Lombardy(Italy) several months after manuring. Distribution coefficients (Kd) from sorption experiments indicateda strong binding to the soil. Soil samples fortified with environmentally significant FQs amounts(0.5 mg kg�1) were exposed to solar light that promoted extensive degradation (80%) of both drugs in60–150 h. Thus, photochemistry could be considered a significant depollution path in the soil, althoughit was two orders of magnitudes slower than in aqueous solution and a fraction of the drug (ca. 20%)remained unaffected. For MAR the photoprocess was the same as in solution, and involved cleavage ofthe tetrahydrooxadiazine ring. On the contrary, with ENR only some of the photoproducts determinedin water (those arising from a stepwise oxidation of the piperazine side chain) were observed. Substitu-tion of the 6-fluoro by a hydroxyl group and reduction did not occur in the soil, supporting the previouscontention that such processes required polar solvation of FQs. Consistently with this rationalization, theirradiation of thin layers of solid drugs led to essentially the same products distribution as in the soil.From the environmental point of view it is important to notice that photodegradation mainly affectsthe side-chains, while the fluoroquinolone ring, to which the biological effect is associated, is conservedup to the later stages of the degradation.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Fluoroquinolones (FQs) are antibacterial agents employed ashuman and veterinary medicines, especially in animal breeding,due to their broad activity spectrum against Gram bacteria andtheir good oral intake properties. These molecules undergo onlya partial metabolism in the body and they are largely excreted inthe pharmacological active form (Reemtsma and Jeckel, 2006;Sukul et al., 2009). This fact, coupled with the incomplete removalby wastewater treatment plants (WWTPs) (Golet et al., 2002a;Lindberg et al., 2005), leads to a continuous introduction into theenvironment that overcomes the transformation/removal rates.Thus, FQs are defined as ‘‘pseudo-persistent’’ contaminants (Huet al., 2010). Their occurrence has been documented in naturalwaters (Sturini et al., 2009; Speltini et al., 2010) and soils (Goletet al., 2002b; Morales-Muñoz et al., 2004; Prat et al., 2006;Martínez-Carballo et al., 2007; Ötker Uslu et al., 2008; Sturiniet al., 2010b), at maximum concentrations in the range of micro-grams per liter and milligrams per kilogram, respectively. These

ll rights reserved.

x: +39 0382 528544.Sturini), antonella.profumo@

drugs, initially present in water bodies, rapidly move to the soilcompartment, due to strong adsorption on minerals and organicmatter (Tolls, 2001). Additionally, the common practice of recy-cling manure from food-producing animal husbandries and sew-age sludge from WWTPs as fertilizers favors the accumulation ofthese antimicrobials in soils (Turiel et al., 2007).

FQs can be therefore considered ubiquitous contaminants ofemerging importance, although no tolerable value has as yet beenestablished for the different environmental compartments(Thiele-Bruhn, 2003). However in 1996, and more recently in the2008 revised form, the EMEA (European Agency for the Evaluationof Medicinal products) guideline fixed a threshold value of0.1 mg kg�1 for residual veterinary pharmaceuticals in soils and0.1 lg L�1 in groundwater (EMEA, 1997, 2008).

The presence of FQs in the environment may pose serious threatsto the ecosystem and human health. Indeed, a matter of concern istheir ability to stimulate bacterial resistance (Kümmerer, 2004),which makes natural-contaminated soils a terrain of resistant bac-teria potentially transferable to other bacteria living in groundwa-ter, drinking water or plants and finally to humans (Kümmerer,2004; Pruden et al., 2006). Furthermore, FQs genotoxicity has beenproven in wastewaters from hospitals (Hartmann et al., 1998),

M. Sturini et al. / Chemosphere 86 (2012) 130–137 131

whereas the potential ecotoxicity of photoproducts is still discussed(Prabhakaran et al., 2009; Vasconcelos et al., 2009).

The stability of the fluoroquinolone ring causes the resistance ofthese molecules to any chemical reagent, and other functionalgroups present as substituents generally react before the aromaticnucleus is degraded. Therefore, the course of degradation in theenvironment must be monitored in detail and it must be estab-lished whether the products of partial degradation are themselvestoxic. Photodegradation is actually a major removal pathway inwater systems (Freccero et al., 2008; Sturini et al., 2010a), thoughthe structure, residual antibacterial activity and persistence of thephotoproducts are still under examination.

On the other hand, the photodegradation in solid matrices assoils, where FQs show enhanced persistence than in water, hasnot been studied in detail as yet. Indeed, the strong binding to soilsand sediments has been previously accounted for a substantialinhibition of FQs photodegradation (Holten Lützhøft et al., 2000).

For these reasons, in the present research the binding to soil andthe photochemical decay of Marbofloxacin (MAR) and Enrofloxacin(ENR), whose characteristic UV–Visible spectra are reported inFig. 1, have been studied under these conditions.

These drugs have been previously studied in our laboratory(Sturini et al., 2009, 2010a, 2010b) and they were chosen, the firstas an example of the new generation FQs, so that clarifying itsbehavior could be particularly interesting, and the second becauseit is among the most largely used FQs in veterinary medicine. Theirbehavior in the soil was studied through sorption experiments. Thetwo drugs were then determined in soils collected in the agricul-tural area between Pavia and Milan (South Lombardy plain, Italy),where these are the two most employed antibiotics, in a samplingcampaign undertaken several months after manuring and at differ-ent depths in the sub-soil. This allowed to draw a general picture ofboth persistence and mobility of these drugs. Photodegradationexperiments were carried out on soil samples exposed to naturalsunlight for a fixed time and then submitted to microwave-assistedextraction (MAE) followed by HPLC with fluorimetric detector (FD).The degradation of both drugs was accompanied by the formationof a number of byproducts, which were identified by HPLC electro-spray tandem mass spectrometry (ESI-MS/MS). Additional experi-ments on thin layers of solid FQs were carried out by exposure ofthe samples to a solar simulator.

2. Experimental section

2.1. Reagents and materials

All the chemicals employed were reagent grade or higher inquality and were used without any further purification. MAR and

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

250 300 350 400 450

ENR

MAR

Fig. 1. Characteristic UV–Visible spectra of 10�4 M MAR (solid line) and ENR(dotted line) aqueous solutions.

ENR were supplied by Fluka (Sigma–Aldrich), ACN (HPLC gradientgrade) by VWR, H3PO4 (85% w/w) by Carlo Erba and ultra-purewater (resistivity 18.2 MX cm�1 at 25 �C) was produced in labora-tory by means of a Millipore Milli-Q system. FQs stock solutions of300 mg L�1 were prepared in chloroform or methanol and stored inthe dark at 4 �C for a maximum of 3 months. Working solutions of6 mg L�1 in 25 mM H3PO4 were renewed weekly. HexahydrateMg(NO3)2 (97% w/w) and ammonia solution (30% w/w), employedfor FQs extraction from soil, were purchased from Sigma–Aldrichand Carlo Erba. All the laboratory operations were conductedunder red light.

2.2. Analytical determination

The HPLC system consists of a pump Series 200 (Perkin Elmer)equipped with vacuum degasser and a programmable fluorescencedetector (FD). The FD excitation/emission wavelengths selectedwere 297/507 nm for MAR and 280/450 nm for ENR. After an equil-ibration period of 10 min, 50 lL of each sample were injected intoa 250 � 4.6 mm, 5 lm Ascentis RP-Amide (Supelco) coupled with asimilar guard-column. The mobile phase was 25 mM H3PO4–ACN(85:15) for 30 min, followed by a 1 min-linear gradient to 100%ACN. After washing for 5 min, the initial conditions were reestab-lished by a 1 min-linear gradient, at a flow rate of 1 mL min�1.

The HPLC-ESI-MS/MS for films analysis was performed by anAgilent 1100 HPLC with a Gemini C18 (250 � 4.6 mm, 5 lm) col-umn, maintained at 30 �C. A gradient was used for the mobilephase (solvent A: HCOOH 0.5% v/v in ultrapure water; solvent B:ACN) as follows: 15% B until 10 min, 20% B from 10 to 12 minand 0% B until 1 min, 60% B from 1 to 50 min, for ENR and MAR,respectively. The flow rate was 1.2 mL min�1 and the injection vol-ume was 5 lL. The MS/MS-system consisted of a linear trap Ther-mo LXQ. ESI experiments were carried out in positive-ion modeunder the following constant instrumental conditions: source volt-age of 4.5 kV, capillary voltage of 20 V, capillary temperature of275 �C and normalized-collision energy 35. Soil extracts, obtainedby a highly selective extraction method (Sturini et al., 2010b), wereanalyzed by HPLC-ESI-MS/MS (Thermo Finnigan) equipped with aDiscovery™ BIO Wide Pore C18 column (Supelco) 5 lm,150 � 2.1 mm. The elution was performed using 0.1% v/v HCOOHin ultrapure water (solvent A) and 0.1% v/v HCOOH in ACN (solventB) at a flow rate of 0.2 mL min�1. The elution started with 87% sol-vent A for 10 min, followed by a linear gradient to 85% A in 5 minand then isocratic elution with 85% A for 5 min. Injection volumewas 100 lL. LC–MS/MS data were obtained using an LCQ ADVMAX ion trap mass spectrometer equipped with an ESI ion sourceand controlled by Xcalibur software 1.3. ESI experiments were car-ried out in positive-ion mode under the following constant instru-mental conditions: source voltage of 5.0 kV, capillary voltage of46 V, capillary temperature of 210 �C, tube lengths voltage of55 V and normalized-collision energy 35.

2.3. Sorption experiments and sample extraction

As suggested by OECD Test Guideline 106 (OECD, 1997), FQsbatch sorption experiments were performed in 0.01 M CaCl2, atconcentrations in the range 2-10 lg mL�1.

MAR and ENR were quantitatively extracted from soil samplesaccording to the green MAE method recently developed in thislaboratory (Sturini et al., 2010b) and selective for this class ofcompounds. The acidified extracts from manured soils were pre-concentrated by solid-phase extraction (SPE) on WAX–HLBcartridges (Sturini et al., 2009, 2010b), whereas SPE was avoidedin the case of the fortified samples used for the photodegradationexperiments, in reason of the larger – though realistic – spikes.

132 M. Sturini et al. / Chemosphere 86 (2012) 130–137

2.4. Field samples

A sampling campaign was carried out on fields near cattle andswine farms (South Lombardy plain), known to employ regularlyFQs as a therapeutic treatment. Soil samples were collected at dif-ferent depths (0–5 and 10–15 cm) by means of a steel cylinder,several months after fertilization. Samples were dried at roomtemperature, sieved (2 mm), stored in the dark and analyzed forMAR and ENR within 2 weeks.

A sample of agricultural soil, collected in Ferrera Erbognone(South Lombardy plain), was used for the experiments below re-ported because of its negligible MAR and ENR content, less thanMDLs (Sturini et al., 2010b), and its physico-chemical properties.The characteristics of sampled soils are listed in Table 1.

2.5. Irradiation experiments

Agricultural soil samples (1 g) from Ferrera Erbognone were for-tified with MAR and ENR (0.5 mg kg�1). After spiking, the soil wasmanually mixed for about 15 min to assure a homogeneous distri-bution of the analytes in the whole sample and stored in the darkovernight to promote FQs sorption equilibrium to the matrix sites,and solvent evaporation. Afterwards, samples were homoge-neously dispersed on glass capsulae Petri (£ 8.5 cm, depth 2 cm)to obtain a monolayer (thickness below 1–2 mm). These were ex-posed to sunlight outdoor (9.00 am–5.00 pm) during the summer(June–September), at temperatures ranging from 25 to 35 �C. Thesolar power ranged from 170 to 470 W m�2 (in the visible range)and from 8 to 30 W m�2 (in the UV), respectively. At the same time,unfiltered river water samples (pH 8.1, conductivity 385 lS cm�1,DOC 1.23 mg L�1), enriched with 50 lg L�1 of each drug, werephotolyzed in an open glass container (20 mm depth, exposed sur-face 280 � 200 mm). The flux was measured by means of a HD9221 (Delta OHM) (450–950 nm) and of Multimeter (CO.FO.ME.-GRA) (295–400 nm) pyranometers. At regular intervals each oneof the irradiated soil samples was extracted by a Mg(II) solution,a convenient procedure because, as previously reported (Sturiniet al., 2010b), the cation forms high-stability constants chelateswith FQs and byproducts. As for river water, aliquots (1 mL) of eachsample (500 mL) were withdrawn and immediately injected in theHPLC–FD system, after 0.45 lm filtration. All experiments wereperformed in triplicate (n = 3).

Further irradiation experiments were carried out on FQs thinlayer films obtained by evaporation of the solutions (0.1 mL of34 lg mL�1 corresponding to 3 lg) deposited on glass capsulae

Table 1Physico-chemical characteristics of the soils sampled.

Linarolo Belgioioso FerreraErbognone

pH in H2O 6.5 6.5 6.4pH in KCl 5.6 6.0 5.5Total organic carbon (%) 1.4 1.8 1.1Cation exchange capacity

(meqv/100 g)14 18 13

Composition of soilparticles (%)

Sand 44.8 43.3 46.0Clay 11.6 8.2 12.6Silt 43.0 48.5 41.4

Mineralogicalcomposition (%)

Serpentine 7 n.d. 5Chlorite 13 10 8Mica 16 19 24Quartz 23 27 28Feldspar 12 20 8Plagioclase 19 17 22Calcite n.d. n.d. n.d.Amphibole 10 7 5

n.d.: not detectable.

Petri (£ 8.5 cm, depth 2 cm) in a solar simulator (SolarBox 1500e, CO.FO.ME.GRA), set at a power factor 250 W m�2 andequipped with a UV outdoor filter of IR treated soda lime glass.At predetermined intervals the irradiated films were redissolvedin 10 mL 25 mM H3PO4 and injected in the HPLC–FD system asabove.

2.6. Identification of photoproducts

Experiments were carried out on soil samples (1 g), fortifiedwith 102 mg kg�1 of each FQ, under natural sunlight. In the thinfilm experiments, 1.5 mL of 300 lg mL�1 stock solutions, corre-sponding to 0.5 mg of each FQ, were deposited on glass capsulaePetri and irradiated in a solar simulator. The soil extracts and theredissolved films were then analyzed by HPLC-ESI-MS/MS.

3. Results and discussion

3.1. Sorption studies

As it is well known, Kd is strongly dependent on the type of ma-trix (Nowara et al., 1997). The sorption coefficient experimentallydetermined for ENR was 3.4(3) L g�1 (n = 3), in accordance withvalues reported in literature (0.26–6.31 L g�1) (Nowara et al.,1997). The adsorption of MAR, for which no previous data wereavailable, turned out to be of the same order of magnitude asENR, viz. Kd 2.5(1) L g�1 (n = 3). The organic carbon-normalizedsorption coefficient Koc, the preferred measure of sorption in envi-ronmental risk assessment (Tolls, 2001), could be obtained from Kd

by normalizing to the organic carbon content and was315(20) L g�1 for ENR and 234(13) L g�1 (n = 3) for MAR.

3.2. Actual samples

Analysis of different farm soil samples proved the presence ofMAR and ENR in the Lombardy plain at concentration levels inthe micrograms per kilogram range, confirming their persistencealso after several months from fertilization. Results are reportedin Table 2. It can be seen that the FQ concentration decreases withincreasing soil depth, similarly to what recently observed on ENR(Ötker Uslu et al., 2008; Profumo et al., 2009), Ciprofloxacin andNorfloxacin (Golet et al., 2003).

3.3. Photodegradation in soil

It has been generally assumed that, contrary to what happens inaqueous solution, light has a minor effect on the degradation ofantibiotics in a solid matrix than in water, or when spread as con-taminants in sludge and slurries (Picò and Andreu, 2007; Chenxiet al., 2008). However, Lai and Lin recently proved that natural illu-mination plays a key role in the degradation of Oxolinic acid and

Table 2FQs concentrations (lg kg�1) determined in field soil samples (n = 3) at severalmonths after fertilization and different depths.

Sampled site Depth (cm) MAR ENR

Belgioioso 0–5 43 (6) 51 (7)10–15 <MQLa 29 (5)

Linarolo 0–5 15 (7) 50 (11)10–15 <MQLa 25 (6)

Torre d’Isola 0–5 24 (2) 23 (3)10–15 <MQLa 10 (2)

a Sturini et al. (2010b).

M. Sturini et al. / Chemosphere 86 (2012) 130–137 133

Flumequine in pond sediment slurry under specific light intensity(Lai and Lin, 2009).

In the present investigation, irradiation under natural sunlightwas proven to contribute substantially to FQs photodegradationin agricultural soil spiked at realistic concentration (0.5 mg kg�1).Fig. 2 shows the photodegradation profiles obtained by solar irra-diation of soil (1 g) and river water (500 mL) samples, fortifiedrespectively with 0.5 mg kg�1 and 50 lg L�1 of each antibiotic. Insoil, the degradation profiles of MAR and ENR are similar, with aslow decrease that reaches a plateau corresponding to 20% of theinitial amount (0.5 mg kg�1) after about 50 h irradiation. The pho-toreactivity is certainly affected by several parameters, such as soilchemical composition, pH, and solar light power, as already ob-served in aqueous systems (Andreozzi et al., 2003; Belden et al.,2007; Sturini et al., 2010a). Possibilities are hydrogen abstractioneither by some excited state of matrix constituents or by hydroxylradicals arising from components of the soil (Stangroom et al.,1998). However, it was shown that photogenerated hydroxyl rad-icals are readily scavenged within a few microseconds by dissolvedorganic carbon (DOC) in solution. All these considerations can bereadily apply to the solid matrix, where light adsorption is re-stricted to the first layer of the matrix (7–8 mm), further loweringthe quantity of light available. Diffusion of eventually photogener-ated radicals would also be more difficult and scavenging phenom-ena should occur faster due to the nature of the solid matrix(reduced mobility and higher concentrations of NOM and otherscavenger than in solution). Irradiation experiments have beendeliberately carried out under natural sunlight during the entireday, in order to better simulate realistic conditions. Under thesame experimental conditions, about 70% of the ENR and MAR ini-tial amount (0.5 mg kg�1 in agricultural soil and 50 lg L�1 in riverwater) was still present in soil after more than 10 h exposure,while the same decrease required about 5 min in river water, witha rate of 10�4 lg min�1 and 10�2 lg min�1, respectively. Goodreproducibility was observed, RSD < 5% in water and RSD � 10%in soil. These results indicate that the persistence of FQs is mark-edly higher in solid matrix than in natural waters.

Fig. 2. Photodegradation profiles obtained by solar irradiation of soil samples (diamondpharmaceutical, respectively.

The observed FQs concentration decrease was due to photolysisonly, as no degradation was observed in ground samples stored inthe dark at room temperature for 1 month after fortification. Thesame blank experiments also excluded a role of biodegradation, afact that was justified by the high fixation rates of quinolones tothe surface or in the pores of the matrix, which is likely to preventmicrobial activity in accordance with previous suggestions (Lai andLin, 2009). Further experiments were carried out on soil samplesfortified with each of the FQs and stirred every 2 h during irradia-tion. Experimental data were not significantly different from thoseobtained from non mixed samples. This supports that the incom-plete degradation up to 150 h is not due to soil particles not ex-posed to solar light, but rather to part of the analyte differentlylocated/complexed. Furthermore, experiments were carried outon FQs films (3 lg on a 1.5 cm2 surface, for an average thicknessof 2 � 10�6 cm) deposited on glass capsulae Petri. In this case, bothantibiotics where completely degraded within 2 h in solar box.Thus, a thin layer of solid FQs reacts rapidly, excluding that theincomplete degradation in soil is due to the formation of smallcrystals that are less photolabile.

To summarize, these experiments suggest that part of the drug(ca. 20%) interacts more strongly with the matrix, or it is located inthe inner pores, not reached by light.

3.3.1. Photochemical paths: ENRThe matrix affects also the product distribution. Thus, the

photodegradation of ENR gives six products in appreciableamounts (N, P, O, Q, E, S, as shown in Figs. 3 and 4), both in the soiland in solid film. The structure was assigned on the basis of massspectra (see Table 3) and retention times, as well as by analogywith previous more extensive characterizations for the reactionin solution (Burhenne et al., 1997a, 1997b; Burhenne et al., 1999;Parshikov et al., 2000; Mella et al., 2001; Sturini et al., 2010a)(see Fig. 3).

In Fig. 5, the overlay of four FD-chromatograms obtained underthe same HPLC conditions is reported. These refer to irradiated riv-er water and soil spiked with 0.25 and 0.5 lg of ENR respectively,

s) and river water samples (circles) fortified with 0.5 mg kg�1 and 50 lg L�1 of each

134 M. Sturini et al. / Chemosphere 86 (2012) 130–137

as well as to thin layers (500 lg) from the evaporation of chloro-form and methanol solutions.

Among the photoproducts, E was identical to a previously char-acterized product from the photolysis in water (Sturini et al.,2010a) and Q was a close analogue of photoproducts obtained fromrelated FQs, (Fasani et al., 1999) and the structure was safely as-signed by comparison of the mass spectrum with literature databy Burhenne et al. (1997a), who reported also the NMR spectrum.The attribution of structure S was supported by a MW 16 amuhigher than that of ENR and by the previous identification of thecorresponding 3-keto-piperazine derivative (Sturini et al., 2010a)in water solution. As for products O and N, the mass spectra withMW differing from the starting compound by 30 and 2 amu,

Fig. 3. Products distribution in the ph

ENR

E

N

O

OH

OF

NN

HOS

N

O

OH

OF

NHN Q

F

NHN

O

h

hi

i

i

i

i

Fig. 4. Products distribution and photodeg

respectively, are in accordance with structures arising fromstepwise oxidative degradation of product Q at the piperazineside-chain. Product P had a mass spectrum in accordance with acomplete oxidative degradation of the benzene ring and the struc-ture proposed is consistent with those previously proposed byBurhenne et al. (1999).

To summarize, of the three main processes occurring in watersolution (Sturini et al., 2010a), viz. (i) oxidative degradation ofthe ethyl-piperazine side chain, (ii) fluorine solvolysis and (iii)reductive defluorination, only the first one operates in solid matrix.Side chain oxidation on the intact ENR leads to products Q, viaintermediate E, and S through parallel paths. This is in agreementto what was previously found for ENR degradation in water

otolysis of ENR in water solution.

N

O

OH

O

OO

N

O

OH

OF

NNH2 O

N

BNH

OH

OO

P

CO2i i

i

radation path of ENR in solid matrix.

Table 3ENR photodegradation products from MS/MS fragmentation.

Fragment HPLC/ESI-MS/MS

Comp. Q Comp. S Comp. O Comp. N Comp. P

m/e int. m/e int. m/e int. m/e int. m/e int.

[M + 1]+ 332 100 376 100 362 100 334 100 140 100[M + 1–H2O]+ 314 35 358 25 344 20 316 35[M + 1–CO]+ 348 5 334 15 306 5[M + 1–CO2]+ 318 20[M + 1–C3H5NO]+ 305 3

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

ENR

S

E

B

Q

O

P

A

N

C

D

13

sign

al(m

V)

time (min)

Fig. 5. Overlay of four LC–FD chromatograms from irradiated river water (green), soil (blue) spiked with ENR (25 lg and 0.5 lg, respectively), and thin layer (500 lg) fromevaporation of chloroform (black) and methanol (red) solutions. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version ofthis article.)

sign

al(m

V)

time (min)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

D

ENR

SE

N

QA

C

B

Fig. 6. Chromatograms of the extracts from soils: irradiated after ENR spike (red), fortified with all the aqueous photolysis products (black). (For interpretation of thereferences to colour in this figure legend, the reader is referred to the web version of this article.)

M. Sturini et al. / Chemosphere 86 (2012) 130–137 135

solution, viz. the oxidation of the methylene groups a to the aminonitrogen, a general, if usually inefficient, process in amine photo-chemistry. Products O and N arise from Q via stepwise oxidationof the side-chain. Further degradation leads to product P and

eventually to mineralization, as previously reported for ENR andrelated FQs by Burhenne et al. (1997b). The mechanism whichleads to oxidized products has not yet been fully clarified. It isknown that both natural organic matters (NOM) and other

NO N

O

N

FOH

O

N

hNHOH

O

N

FOH

O

NNHOH

O

N

FOH

O

N

HN

OH

O

N

FHO

O

N

MAR F G

Fig. 7. Products distribution and photodegradation path of MAR in solid matrix.

136 M. Sturini et al. / Chemosphere 86 (2012) 130–137

macro-constituents of the soil can affect the photodegradation pro-cesses of FQs in aqueous systems (Andreozzi et al., 2003; Beldenet al., 2007; Sturini et al., 2010a). The role of oxygen radicalsand/or oxygen sensitization and the effect of NOM on such pathshas been previously discussed (Sturini et al., 2010a), and well ex-plains the stepwise photodegradation in path (i), which affects firstthe more nucleophilic/oxidizable moieties present (pipera-zine > benzo ring > pyridone ring). This involves radical intermedi-ates and occurs undisturbed both in solution and in the solid state.In contrast, the competitive primary photoprocesses, reductivedehalogenation (iii) and solvolysis (ii), involve ionic intermediated(see Fig. 3) that are active in solution, but inhibited in solid matrix.This is confirmed also by recovery tests from soils samples spikedwith all the aqueous photolysis products (results in Fig. 6) showingthat the absence of products D and A in irradiated soil samples isnot due to failure in the recovery process but to the effective inhi-bition of paths (i) and (ii) in this matrix.

3.3.2. Photoreaction paths: MAROn the other hand, the products formed by irradiation of MAR in

solid matrix (F and G) are the same as in water solution (Sturiniet al., 2010a), as reported in Fig. 7. The reaction follows the samepath as in solution and the products distribution seems not to beinfluenced by the medium. This is in agreement with the previousconclusion that, differently from most FQs, the photoreaction ishere homolytic with the cleavage of the weak N–N bond, whichcauses loss of an N-methylimine group to yield the observed prod-ucts. Indeed, this process takes place also in apolar organic solventsand involves the triplet state (Pretali et al., 2010).

4. Conclusions

The present study demonstrates the ubiquitary distribution ofveterinary FQs in regions where husbandry is diffuse, and theirpersistence. These drugs are mainly found adsorbed on the man-ured soil and are present at concentrations (lg kg�1) that representa serious environmental threat. Since leaching and biodegradationare of minor importance in this matrix, the periodical plowing ofagricultural fields is a potential way of exposing sorbed FQs to sun-light. Indeed, it has been demonstrated that photochemistry repre-sents a significant degradation path for these otherwise chemicallyresistant pollutants, even though the rate of degradation is two or-der of magnitude lower than in aqueous solution. Furthermore,while photoprocesses following an ionic path (those involvingthe C–F bond in ENR) do not occur in the solid or in the adsorbedstate, homolytic processes (amine moiety oxidation, fragmentationof the oxazine ring in MAR) are little affected by the medium char-acteristics. In both cases, however, it is a substituent, not the quin-olone nucleus that is affected by the initial steps. This suggests thatmicrobiological activity may be preserved in the early degradationphase and prompts studies of that issue.

Acknowledgments

This work was partially supported by FAR, Pavia University. TheAuthors are grateful to Dr. Stefania Nicolis, University of Pavia andDr. Alessandro Granata, Lab Analysis S.r.l. (Casanova Lonati, Pavia)for HPLC-ESI-MS/MS analysis.

References

Andreozzi, R., Marotta, R., Paxéus, N., 2003. Pharmaceuticals in STP effluents andtheir solar photodegradation in acquatic environment. Chemosphere 50, 1319–1330.

Belden, J.B., Maul, J.D., Lydy, M.J., 2007. Partitioning and photodegradation ofciprofloxacin in aqueous systems in the presence of organic matter.Chemosphere 66, 1390–1395.

Burhenne, J., Ludwig, M., Nikoloudis, P., Spiteller, M., 1997a. Photolytic degradationof fluoroquinolones carboxylic acids in aqueous solution. Primaryphotoproducts and half-lives. Environ. Sci. Pollut. Res. 4, 10–15.

Burhenne, J., Ludwig, M., Spiteller, M., 1997b. Photolytic degradation offluoroquinolones carboxylic acids in aqueous solution. Isolation and structuralelucidation of polar photometabolites. Environ. Sci. Pollut. Res. 4, 61–67.

Burhenne, J., Ludwig, M., Spiteller, M., 1999. Polar photodegradation products ofquinolones determined by HPLC/MS/MS. Chemosphere 38, 1279–1286.

Chenxi, W., Spongberg, A.L., Witter, J.D., 2008. Determination of the persistence ofpharmaceuticals in biosolids using liquid-chromatography tandem massspectrometry. Chemosphere 73, 511–518.

EMEA, 1997. EMEA/CVMP/055/96-Final. <http://www.ema.europa.eu/pdfs/vet/regaffair/005596en.pdf>.

EMEA, 2008. EMEA/CVMP/ERA/418282/2005-rev.1consultation. <http://www.ema.europa.eu/pdfs/vet/era/41828205-Rev1.pdf>.

Fasani, E., Barberis Negra, F.F., Mella, M., Monti, S., Albini, A., 1999. Photoinduced C–F bond cleavage in some fluorinated 7-amino-4-quinolone-3-carboxylic acids. J.Org. Chem. 64, 5388–5395.

Freccero, M., Fasani, E., Mella, M., Manet, I., Albini, A., 2008. Modeling thephotochemistry of the reference phototoxic drug lomefloxacin by steady-stateand time-resolved experiments, and DFT and post-HF calculations. Chem. Eur. J.14, 653–663.

Golet, E.M., Alder, A.C., Giger, W., 2002a. Environmental exposure and riskassessment of fluoroquinolone antibacterial agents in wastewater an riverwater of the Glatt valley watershed, Switzerland. Environ. Sci. Technol. 36,3645–3651.

Golet, E.M., Strehler, A., Alder, A.C., Giger, W., 2002b. Determination offluoroquinolone antibacterial agents in sewage sludge and sludge-treated soilusing accelerated solvent extraction followed by solid-phase extraction. Anal.Chem. 74, 5455–5462.

Golet, E.M., Xifra, I., Siegrist, H., Alder, A.C., Giger, W., 2003. Environmental exposureassessment of fluoroquinolone antibacterial agents from sewage to soil.Environ. Sci. Technol. 37, 3243–3249.

Hartmann, A., Alder, A.C., Koller, T., Widmer, R.M., 1998. Identification offluoroquinolone antibiotics as the main source of umuC genotoxicity in nativehospital wastewater. Environ. Toxicol. Chem. 17, 377–382.

Holten Lützhøft, H.C., Vaes, V.H.J., Freidig, A.P., Halling-Sørensen, B., Hermens, J.L.M.,2000. Influence of pH and other modifying factors on the distribution behaviorof 4-Quinolones to solid phases and humic acids studied by ‘‘negligible-depletion’’ SPME–HPLC. Environ. Sci. Technol. 34, 4989–4994.

Hu, X., Luo, Y., Zhou, Q., 2010. Simultaneous analysis of selected typical antibioticsin manure by microwave-assisted extraction and LC–MS. Chromatographia 71,217–223.

Kümmerer, K., 2004. Resistance in the environment. J. Antimicrob. Chemother. 54,311–320.

Lai, H.T., Lin, J.J., 2009. Degradation of oxolinic acid and flumequine in aquaculturepond waters and sediments. Chemosphere 75, 462–468.

Lindberg, R.H., Wenneberg, P., Johansson, M.I., Tysklind, M., Andersson, B.A.V., 2005.Screening of human antibiotic substances and determination of weekly massflows in five sewage treatment plants in Sweden. Environ. Sci. Technol. 39,3421–3429.

M. Sturini et al. / Chemosphere 86 (2012) 130–137 137

Martínez-Carballo, E., González-Barreiro, C., Scharf, S., Gans, O., 2007.Environmental monitoring study of selected veterinary antibiotics in animalmanure and soils in Austria. Environ. Pollut. 148, 570–579.

Mella, M., Fasani, E., Albini, A., 2001. Photochemistry of 1-cyclopropyl-6-fluoro-1,4-dihydro-4-oxo-7-(piperazin-1-yl)quinoline-3-carboxylic acid (=ciprofloxacin)in aqueous solutions. Helv. Chim. Acta 84, 2508–2519.

Morales-Muñoz, S., Luque-García, J.L., Luque de Castro, M.D., 2004. Continuousmicrowave-assisted extraction coupled with derivatization and fluorimetricmonitoring for the determination of fluoroquinolone antibacterial agents fromsoil samples. J. Chromatogr. A 1059, 25–31.

Nowara, A., Burhenne, J., Spiteller, M., 1997. Binding of fluoroquinolone carboxylicacid derivatives to clay minerals. J. Agric. Food Chem. 45, 1459–1463.

Organization for Economic Cooperation and Development (OECD), 1997.Adsorption–Desorption Using a Batch Equilibrium Method. TechnicalGuideline 106.

Ötker Uslu, M., Yediler, A., Akmehmet Balcıoglu, I., Schulte-Hostede, S., 2008.Analysis and sorption behavior of fluoroquinolones in solid matrices. Water AirSoil Pollut. 190, 55–63.

Parshikov, I.A., Freeman, J.P., Lay Jr., J.O., Beger, R.D., Williams, A.J., Sutherland, J.B.,2000. Microbiological transformation of enrofloxacin by the fungus Mucorramannianus. Appl. Environ. Microbiol. 66, 2664–2667.

Picò, Y., Andreu, V., 2007. Fluoroquinolones in soil-risks and challenges. Anal.Bioanal. Chem. 387, 1287–1299.

Prabhakaran, D., Sukul, P., Lamshöft, M., Maheswari, M.A., Zühlke, S., Spiteller, M.,2009. Photolysis of difloxacin and sarafloxacin in aqueous systems.Chemosphere 77, 739–746.

Prat, M.D., Ramil, D., Compañó, R., Hernández-Arteseros, J.A., Granados, M., 2006.Determination of flumequine and oxolinic acid in sediments and soils bymicrowave-assisted extraction and liquid chromatography-fluorescence. Anal.Chim. Acta 567, 229–235.

Pretali, L., Fasani, E., Dondi, D., Mella, M., Albini, A., 2010. The unexpectedphotochemistry of marbofloxacin. Tetrahedron Lett. 51, 4696–4698.

Profumo, A., Sturini, M., Speltini, A., Maraschi, F., Pretali, L., Fasani, E., 2009.Determination of trace veterinary fluoroquinolones in soils and waters and roleof photochemistry in their degradation. In: Presented at Euroanalysis XV,Innsbruck (September 6–10).

Pruden, A., Pe, R.T., Storteboom, H., Carlson, K.H., 2006. Antibiotic resistance genesas emerging contaminants: studies in northern Colorado. Environ. Sci. Technol.40, 7445–7450.

Reemtsma, T., Jeckel, M., 2006. Organic Pollutant in the Water Cycle. Wiley-VCH,Weinheim.

Speltini, A., Sturini, M., Maraschi, F., Profumo, A., 2010. Fluoroquinolone antibioticsin environmental waters: sample preparation and determination. J. Sep. Sci. 33,1115–1131.

Stangroom, S.J., Macleod, C.L., Lester, J.N., 1998. Photosensitized transformation ofthe herbicide 4-chloro-2-methylphenoxy acetic acid (MCPA) in water. WaterRes. 32, 623–632.

Sturini, M., Speltini, A., Pretali, L., Fasani, E., Profumo, A., 2009. Solid-phaseextraction and HPLC determination of fluoroquinolones in surface waters. J. Sep.Sci. 32, 3020–3028.

Sturini, M., Speltini, A., Maraschi, F., Profumo, A., Pretali, L., Fasani, E., Albini, A.,2010a. Photochemical degradation of marbofloxacin and enrofloxacin in naturalwaters. Environ. Sci. Technol. 44, 4564–4569.

Sturini, M., Speltini, A., Maraschi, F., Rivagli, E., Profumo, A., 2010b. Solvent-freemicrowave-assisted extraction of fluoroquinolones from soil and liquidchromatography-fluorescence determination. J. Chromatogr. A 1217, 7316–7322.

Sukul, P., Lamshöft, M., Kusari, S., Zühlke, S., Spiteller, S., 2009. Metabolism andexcretion kinetics of 14C-labeled and non-labeled difloxacin in pigs after oraladministration, and antimicrobial activity of manure containing difloxacin andits metabolites. Environ. Res. 109, 225–231.

Thiele-Bruhn, S., 2003. Pharmaceutical antibiotic compounds in soils – a review. J.Plant Nutr. Soil Sci. 166, 145–167.

Tolls, J., 2001. Sorption of veterinary pharmaceuticals in soils: a review. Environ. Sci.Technol. 35, 3397–3406.

Turiel, E., Martín-Esteban, A., Tadeo, J.L., 2007. Molecular imprinting-basedseparation methods for selective analysis of fluoroquinolones in soils. J.Chromatogr. A 1172, 97–104.

Vasconcelos, T.G., Henriques, D.M., König, A., Martins, A.G., Kümmerer, K., 2009.Photo-degradation of the antimicrobial ciprofloxacin at high pH: identificationand biodegradability assessment of the primary by-products. Chemosphere 79,487–493.