MS

Talanta 80 (2009) 947–953

Contents lists available at ScienceDirect

Talanta journal homepage: www.elsevier.com/locate/talanta

Optimization of multiple reaction monitoring mode for the trace analysis of veterinary sulfonamides by LC–MS/MS a ´ Anna Białk-Bielinska , Jolanta Kumirska a , Richard Palavinskas b , Piotr Stepnowski a,∗ a b

Department of Environmental Analysis, Faculty of Chemistry, University of Gda´ nsk, ul. Sobieskiego 18, 80-952 Gda´ nsk, Poland Federal Institute for Risk Assessment BfR, Thielallee 88-92, 14195 Berlin, Germany

a r t i c l e

i n f o

Article history: Received 19 April 2009 Received in revised form 14 August 2009 Accepted 17 August 2009 Available online 22 August 2009 Keywords: Veterinary drugs Sulfonamides Environment LC–MS/MS analysis MRM Soil

a b s t r a c t One of the oldest groups of veterinary chemotherapeutic agents, sulfonamides have been widely used for more than 50 years, thanks to their low cost and their broad spectrum of activity in preventing or treating bacterial infections. Nowadays, those compounds are regularly detected in a wide variety of environmental samples, including natural waters, sediments and soils. Since the environmental concentrations of sulfonamides are usually very low and their occurrence multicomponental, their determination in these matrices still pose significant analytical problems. The present paper describes the optimization of ESIMS/MS parameters and the chromatographic separation of 12 sulfonamides commonly used in veterinary medicine. The methodology developed in this study, unlike many others, satisfied the requirements of EU Commission Decision 2002/657/EC, which defines the criteria for both screening and confirmatory methods with respect to drug residues on the basis of identification points. Each MRM transition was tested not only for the qualitative but also for the quantitative analysis of sulfonamides. The method was validated for its analytical performance parameters and applied to the determination of those compounds in soil samples. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Scientists have devoted much attention to the occurrence of pharmaceuticals in the environment. Consequently, large numbers of reports have been published on traces of drugs detected in many different environmental samples, e.g., fresh-, sea-, and wastewaters, soils, sediments and biota [1]. Because of their polar nature, however, a number of drugs are not significantly adsorbed onto soils and therefore easily reach groundwater, the usual source of drinking water [2]. At present, some 3000 different active substances are used in human and veterinary medicine [3]. Drugs commonly detected in the environment include antibiotics, nonsteroidal anti-inflammatory drugs, beta-blockers, anti-epileptics, blood-lipid lowering agents, antidepressants, hormones and antihistamines [1–5]. Antibiotics are of particular concern: continuous exposure to them, even to low concentrations, can induce bacterial resistance. This constitutes pose a serious threat to public health, as fewer and fewer infections can be treated with the existing suite of antibiotics [5]. The term “antibiotic” is normally reserved for a diverse range of compounds, both natural and semi-synthetic, that possess antibacterial activity [6]. In addition to the treatment of human diseases, those drugs are used extensively in veterinary

∗ Corresponding author. Tel.: +48 58 5235448; fax: +48 58 5235 454. E-mail address: [email protected] (P. Stepnowski). 0039-9140/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2009.08.023

medicine as growth promoters and to improve feed efficiency [7]. Playing a crucial role in modern agriculture and the livestock industry, their use has been on the rise in many developing countries [4]. As antibiotics do not undergo complete metabolism in animal organisms [4–5], a large proportion of them is excreted in unchanged form in feces and urine. Both antibiotics and their metabolites are released either directly into the environment by grazing animals or indirectly during the application of manure/slurry as fertilizer. Contaminated manure is considered to be the main source from which antibiotics and their corresponding resistant genes enter different environmental compartments, food and plants [8]. Moreover, once in the aquatic environment, antibiotics have been found to break through classical wastewater treatment facilities, thereby presenting a threat to surface and groundwater [9–10]. It has also been found that the conventional drinking water treatment processes of coagulation, flocculation and sedimentation do not completely remove these compounds, so they are detected in finished drinking waters in various concentrations [11–15]. Pharmaceuticals enter the environment in low or very low concentrations; this, together with their diverse physico-chemical properties and the complexity of environmental matrices, makes their determination difficult. It is therefore necessary to develop sensitive, selective and easily reproducible methods for tracing these chemicals in different environmental materials. Since in most

948

A. Białk-Bieli´ nska et al. / Talanta 80 (2009) 947–953

Fig. 1. The chemical structures of the sulfonamides under investigation [32,44].

cases we are dealing with complex mixtures of several chemicals from one or several groups of pharmaceuticals, chromatographic techniques need to be applied prior to their final determination. Their relatively high polarity and low thermal stability precludes gas chromatography (GC), so liquid chromatography (LC) coupled to mass spectrometry (MS), fluorescence, ultraviolet or radioimmunoassay [16] detection is used; among these latter techniques, LC coupled to MS, especially in tandem with MS, has made impressive progress in the analysis of residual pharmaceuticals [17–18]. The vast majority of studies have focused on the optimization of mobile phase composition, instrumental conditions and validation parameters (including limit of detection (LOD) and limit of quantification (LOQ)) in order to obtain favorable analytical protocols for the pharmaceuticals of interest. However, rather more methods have been described for the determination of antibiotics in different water compartments than in soil; the latter is more challenging owing to the complexity of soil matrices, interferences, and the longer sample preparation requirements (clean-up procedure). For these reasons, only a few papers have been published on the quantification of antibiotics in solid matrices (e.g. [19]). Likewise, sulfonamides as veterinary chemotherapeutics have usually been analyzed in water samples like wastewater [20–26], drinking water [13], surface/groundwater [27–28] or liquid manure [29–31] rather than in solid samples [32–37]. However, since there are no official guidelines for the trace analysis of pharmaceuticals in environmental samples, most of the reported methodologies suffer from a lack of confirmatory methods with respect to MS. Reports on the qualitative and quantitative analysis of many different pharmaceuticals (including sulfonamides) in environmental samples using HPLC systems with UV-detection are still being published [23,25,30,36,38] this method cannot be used as a confirmatory technique, because identification of a compound in complex materials based solely on retention time is not sufficient. According to the EU Commission Decision of August 12, 2002 (2002/657/EC), also referred to as SANCO 1085/2002, determination of an analyte using LC with UV/VIS detection (single

wavelength) is not suitable on its own for use as a confirmatory method. The aim of this study was therefore to optimize MS/MS parameters using ion-trap mass spectrometry with electrospray ionization for the analysis of sulfonamides – the most common veterinary chemotherapeutics – and to select characteristic fragmentation ions of these compounds in order to work in multiple reaction monitoring (MRM) mode. The experiments were performed in accordance with the above-mentioned Commission Decision, which defines the criteria for both screening and confirmatory methods with respect to drug residues on the basis of identification points. A precursor ion and a minimum of two product ions were selected for every compound, which yields four identification points and enables confirmation of the compounds analyzed. The chromatographic separation of these compounds was optimized, and also validated using parameters such as selectivity, linearity, sensitivity, limits of detection and quantification, precision (intraday and inter-day precision) and accuracy (recovery). Finally, the method was applied to the analysis of selected sulfonamides in soil samples. 2. Materials and methods 2.1. Chemicals Standards of sulfaguanidine, sulfathiazole, sulfadiazine, sulfamethiazole, sulfamerazine, sulfamethoxypyridazine, sulfachloropyridazine, sulfisoxazole, sulfamethoxazole, sulfadimethoxine were purchased from Sigma–Aldrich (Steinheim, Germany), those of sulfadimidine and sulfapyridine from Serva (Heidelberg, Germany) (Fig. 1). Methanol (MeOH) and acetonitrile (ACN) (HPLC grade), used for the mobile phases and for the dissolution of standards, were obtained from POCH S.A. (Gliwice, Poland); the acetonitrile (LC–MS Chromasolv® ) used for the LC–MS/MS analysis was purchased from Sigma–Aldrich (Steinheim, Germany). ´ Deionized water was produced by the HYDROLAB System (Gdansk, Poland). Ammonium acetate (NH4 Ac) and acetic acid (both of ana-

A. Białk-Bieli´ nska et al. / Talanta 80 (2009) 947–953

lytical reagent grade) were purchased from Chempur (Piekary ´ askie, ˛ Sl Poland). Standard stock solutions were prepared by dissolving each compound in methanol at a concentration of 500 ␮g ml−1 and stored at −18 ◦ C in the dark. Working solutions (1 ␮g ml−1 ) of sulfonamides were prepared by diluting the stock solutions in a mixture of ACN:H2 O (50:50, v/v) or in acetic buffer at pH 3.5 (acetic acid was added to the mixture of ACN:H2 O, 10:90, v/v, containing 1 mM NH4 Ac until pH 3.5 was reached), depending on the purpose, and stored at 4 ◦ C. During the experiments pH was measured using a CP-411 laboratory pH-meter (Elmetron-Zabrze, Poland). The pH of the solution for the analysis of selected antibiotics ensured their stability and the appropriate conditions for the electrospray ionization and chromatographic separation. Strata-X Polymeric Reversed Phase (SPE-Phenomenex Inc., Torrance, CA) cartridges with 200 mg of packing material and a 3 ml reservoir were used for sample preparation. 2.2. Soil samples Soils (fluvial meadow soil) were sampled from the region of Pomerania in northern Poland. The soil was air-dried, ground in a mortar and passed through a 2 mm sieve, then re-ground in a mortar with a small rubber pestle. Their organic carbon content was determined by loss-on-ignition. The pH of the soil (measured in 1 M KCl) was 5.3, the clay content (>0.01 mm fraction) was 94% and the organic content was 21.5%. 2.3. Optimization 2.3.1. Liquid chromatography conditions Agilent 1200 Series LC system (Agilent Technologies, Inc., Santa Clara, USA) and ChemStation for LC systems software were used for the chromatographic separation of the sulfonamides. For the LC–MS/MS analysis Brucker Daltonics Data Analysis software was utilized. The separation was performed on Gemini C18 and Gemini C6 columns (both 150 mm × 4.6 mm, 5 ␮m pore size) (Phenomenex Inc., Torrance, CA) with the direct infusion of the eluent from the LC to the mass spectrometer. To optimize the chromatographic separation, a serial of preliminary experiments were performed to test different mobile phases. Both acetonitrile and methanol mixed with water were tested as mobile phases. Additionally the flow rate of the eluent, injection volume and the column temperature as well as the pH of a mobile phase consisting of ammonium acetate and acetic acid was also optimized. 2.3.2. ESI-MS/MS parameters All mass spectrometric measurements were performed on an HCT Ultra ion trap mass spectrometer (Brucker Daltonics, Bremen, Germany) equipped with an electrospray ionization (ESI) source. All the compounds were tested in both positive and negative ion modes. MS–MS parameters were optimized in continuous flow mode using an integrated syringe pump at a flow rate of 5 ␮l min−1 of standard solutions. EsquireControl software was employed to optimize spectrometric conditions such as the nebulizer gas and dry gas flow rates and the dry gas temperature. The nebulizer gas and dry gas was nitrogen, and helium (99.999%) was used as the collision gas in the ion trap. The parameters were optimized manually, separately for each compound. After the best conditions for isolating the precursor ion (proton adduct of the analyte) had been determined, full scan MS/MS mode was used to record product ions from the standard solution of chemotherapeutics. For each compound, the fragmentation amplitude and isolation width were also optimized manually to increase the sensitivity and selectivity of the method and to select three of the most intensive and characteristic

949

fragmentation ions for qualitative analysis and one of the highest intensity for quantitative analysis. 2.4. Extraction method Soil samples (1 g) were spiked with the solution (2.5 ml) of 12 sulfonamides of known concentrations (250 ␮g sulfonamide kg−1 soil) and then left to dry in the dark at room temperature for three days. Next, the analytes were extracted from the dry soil with 5 ml MeOH, placed in an ultrasonic bath (Polsonic, Warsaw, Poland) for 1 h, and finally shaken for another 1 h (RS 10 Control, IKA, Germany). The samples were then centrifuged (10 min, 4000 rpm; MPW-250 Centrifuge, Warsaw, Poland) and the supernatant transferred to a 10 ml flask, after which the procedure was repeated. The resulting extract was concentrated to a volume of about 1 ml. Two alternative procedures were applied in the next step. In the first one, the residual extract was transferred to a plastic tube, evaporated to dryness under a stream of nitrogen, after which 1 ml of mobile phase A was added. The sample was centrifuged (15 min, 5500 rpm; MPW-250 Centrifuge, Warsaw, Poland), transferred to an amber vial and analyzed by LC–MS/MS. In the second procedure, 250 ml water was added to the residual extract and the whole cleaned up on a Strata-X Polymeric Reversed Phase SPE-column (3 ml, 200 mg, Phenomenex Inc.). The cartridges were rinsed with 3 ml MeOH and conditioned with 6 ml H2 O:MeOH (95:5, v/v); after sample loading the cartridge was dried and then rinsed with 3 ml H2 O:ACN (95:5, v/v) to remove interferences. Sulfonamides were eluted with 6 ml of a mixture of MeOH:ACN (50:50, v/v) and allowed to dry under a stream of nitrogen. The dried extract was reconstituted in 1 ml mobile phase A and analyzed by LC–MS/MS. 3. Results and discussion 3.1. Optimization of the LC–MS/MS method According to EU Commission Decision 2002/657/EC, the presence of the residues of veterinary drugs is confirmed so long as the drug is identified both chromatographically and spectrally. If these criteria are not satisfied, a new method or a better procedure should be developed for extracting and/or separating the analytes from interferences. Nowadays, LC–MS/MS in multiple reaction monitoring (MRM) mode is the ‘state-of-art’. It is the technique of the highest sensitivity and selectivity, which for these reasons has been used in these experiments. Comparison of positive and negative ion modes shows that working in the former mode causes more high-intensity product ions to be formed. During the optimization of sulfonamides in full scan MS, the most intensive ions were [M+H]+ ; ions like [M+Na]+ or [M+NH4 ]+ were also detected but were of lower intensity. All the MS/MS parameters were selected to obtain the most intensive precursor ions for every compound [M+H]+ . The most relevant parameters were: nebulizer gas (30 psi), dry gas (10 L min−1 ), dry temperature (300 ◦ C), maximum accumulation time (200 ms), average (3 scans), smart target (50,000), target mass (m/z = 250), capillary exit (109.8 V) and skimmer (40.0 V); ultra scan was used as the mass range mode. Moreover, for every compound a precursor ion and a minimum of two product ions were selected, earning, according to Commission Decision 2002/657/EC, four identification points for the confirmation of the compound. According to Council Directive 96/23/EC, veterinary drugs such as antibacterial agents, including sulfonamides and quinolones belong to Group B in Annex I of this Directive, and for those compounds a minimum of three identification points are required [39]. The value of a low-resolution MS precursor ion is 1.0 identification points and that of each low-

950

A. Białk-Bieli´ nska et al. / Talanta 80 (2009) 947–953

Table 1 Precursor ion and product ion masses selected during optimization for the MRM detection of sulfonamides. Substance

Precursor ion

Product ions

Fragmentation amplitude (V)

2.0

156 173 122

0.55 0.50 0.55

250 [M+H]+

0.9

156 184 108

0.50 0.60 0.55

Sulfadiazine

251 [M+H]+

0.8

156 174 108

0.75 0.55 0.75

Sulfamethoxazole

254 [M+H]+

2.0

156 188 147

0.65 0.65 0.50

Sulfathiazole

256 [M+H]+

2.0

156 108 92

0.55 0.60 0.65

Sulfamerazine

265 [M+H]+

2.0

Sulfisoxazole

268 [M+H]+

2.0

190 174 156 156 113 108

0.65 0.60 0.55 0.70 0.65 0.65

Sulfamethiazole

271 [M+H]+

2.0

156 108 92

0.60 0.70 0.60

Sulfamethazine

279 [M+H]+

1.9

204 124 156

0.60 0.55 0.55

Sulfamethoxypyridazine

281 [M+H]+

1.9

156 126 188

0.45 0.55 0.50

Sulfachloropyridazine

285 [M+H]+

2.0

156 108 92

0.75 0.60 0.75

Sulfadimethoxine

311 [M+H]+

2.0

156 245 218

0.60 0.60 0.65

Sulfaguanidine

+

215 [M+H]

Sulfapyridine

Isolation width

resolution MS product ion is 1.5 IPs. For these reasons, a minimum of two product ions in low-resolution MS must be monitored for the minimum number of identification points to be acquired. Note, that working in SIM or MRM mode with only one product ion in low-resolution MS does not fulfill the minimum IP requirement. Table 1 presents the results together with selected values of the optimized fragmentation amplitude for each of the transitions and the isolation width for every product ion. The fragmentation of sulfonamides has already been described [40]. For sulfonamides, the most characteristic ions described in the literature, such as m/z 156, 108 and 92, corresponding to [M−RNH2 ]+ , [M−RNH2 −SO2 ]+ , and [M−RNH2 −SO]+ , respectively, were also observed in this experiment. Additionally, fragmentation of some of those compounds produced such ions as 124, 186 and 204, which correspond to [RNH2 +2H]+ , [RNH2 +SO2 ]+ and [RNH2 +SO2 +H2 O]+ . Although complete separation is not necessary for selective MS–MS detection, it much improves detectability and reduces the ion suppression effect; this was one of the main reasons for performing the experiments [18]. Therefore, the influence of different parameters on retention time was optimized. The separation was done by reversed-phase chromatography on two types of column—C6 and C18. The C-phenyl column was excluded from the later experiments because of the very strong hydrophobic and ␲· · ·␲ interactions between the phenyl groups on the stationary phases and the specific groups in the sulfonamide structures. Retention times were therefore longer and peaks were broader.

Both acetonitrile and methanol were tested as mobile phases, as were the pH and different salt contents. It was demonstrated that acetonitrile and water were preferable for sulfonamide separation. Moreover, it was necessary to work under acidic conditions for two reasons: first, to achieve better separation, and second, to improve ionization in the mass spectrometer. In the HPLC system, trifluoroacetic acid is usually used for this purpose; but it causes ion suppression so is not recommended for LC–MS/MS analysis. This was why an acetic buffer of different pH was tested. Generally, it was found that the lower the pH, the higher the retention time and the better the separation. This is because most veterinary sulfonamides have at least two nitrogen-containing functional groups (Fig. 1). The amine attached to the aromatic ring-referred to as N4 – is protonated at pH < 2.5, whereas the one attached to the sulfurreferred as N1 – is deprotonated at pH > 5.5–7.0. Hence, most of these compounds are positively charged under acidic conditions, negatively charged under alkaline conditions, and neutral between pH 2.5–6.0 [32]. Additionally, very weakly basic or acidic sites may also be present in the side R-group, which can also play a crucial role in retention on the stationary phase. As observed by Botsoglou et al. [41], the retention time of the investigated sulfonamides increased slightly in the pH range 2.5–3.5, the greatest retention being obtained for sulfonamides bearing an additional methyl group on the R-side-chain. This suggests that the R-side-chain plays a crucial role in the hydrophobic interactions of the compounds with the stationary phase [41], which was also observed in the

A. Białk-Bieli´ nska et al. / Talanta 80 (2009) 947–953

951

Fig. 2. Chromatograms of the optimized method: (A) (HPLC–UV) and (B) (MRM-LC–MS/MS) chromatograms of the standard mixture (1 ␮g ml−1 ) of 12 sulfonamides: 1, sulfaguanidine; 2, sulfadiazine; 3, sulfathiazole; 4, sulfapyridine; 5, sulfamerazine; 6, sulfadimidine; 7, sulfamethiazole; 8, sulfamethoxypyridazine; 9, sulfachloropyridazine; 10, sulfamethoxazole; 11, sulfisoxazole; 12, sulfadimethoxine; C18 -column, detection wavelength 270 nm, 300 ␮l min−1 , injection volume 50 ␮l, mobile phase A – H2 O:ACN (90:10, v/v, 1 mM NH4 Ac, HAc at pH 3.5), mobile phase B – 100% ACN; gradient program: 0–64% B for 32 min, temperature 25 ◦ C); (C) (MRM-LC–MS/MS) chromatogram of sulfonamides extracted from soil (procedure 1 – without any clean-up) spiked with the standard solution (250 ␮g kg−1 ); (D) blank sample of soil.

present experiments. The selection of pH 3.5 for the mobile phase and the solution of the standard compounds was based on the pKa of the sulfonamides (Fig. 1). Since it was below the lowest value of pKa , a pH of 3.5 ensured that the sulfonamides were in their neutral form, their retention time increased and the baseline separation was better; all these observations were confirmed by Haller et al. [32]. Moreover, mobile phase A consisted of 1 mM ammonium acetate, which is the right concentration for ionization, because the higher the salt content, the greater the loss of MS sensitivity during

measurements. Additionally, the flow rate in the 150–300 ␮l min−1 range as well as the injection volume in the range 10–50 ␮l and the temperature in the 25–45 ◦ C range were also optimized. Increasing the temperature did not significantly improve sulfonamide separation but did diminish the intensity of chromatographic peaks. A temperature of 25 ◦ C was therefore the best option. In summary, the antibiotics were separated on a C18 -column at a temperature of 25 ◦ C, a wavelength of 270 nm, an injection volume of 50 ␮l, and a flow rate of 300 ␮l min−1 . Mobile phase A was a mixture of

952

A. Białk-Bieli´ nska et al. / Talanta 80 (2009) 947–953

Table 2 Validation parameters of the optimized method (n is the number of replicates). Transitions

LOD (␮g L−1 ), n = 10

LOQ (␮g L−1 ), n = 10

215 > 156 250 > 156 251 > 156 254 > 156 256 > 156 265 > 156 268 > 156 271 > 156 279 > 204 281 > 156 285 > 156 311 > 156

1.05 2.08 2.63 2.74 0.94 0.52 1.26 1.49 0.70 0.44 2.61 0.31

2.10 4.16 5.26 5.48 1.88 1.03 2.52 2.99 1.40 0.88 5.22 0.62

Intra-day precision RSD (%), n = 6 3.61–9.87 2.46–12.89 5.28–13.54 3.47–12.01 2.22–5.91 2.27–7.68 11.97–12.59 1.87–5.56 3.71–6.37 2.81–8.21 1.86–12.67 1.30–4.92

H2 O:ACN (90:10, v/v, 1 mM NH4 Ac/AcH, pH 3.5) and mobile phase B was 100% ACN; both were selected in the gradient program. Elution began with 100% of mobile phase A, which was reduced to 36% within 32 min. Complete separation of 10 of the 12 sulfonamides was possible within less than 30 min (Fig. 2A and B), which is a satisfactory result compared to the methods described in the literature [23,25,27,30], where, for example, five sulfonamides were separated in 16 min or eight in 35 min. 3.2. Validation of the analytical method 3.2.1. Linearity Linearity is a measure of the method’s sensitivity: the greater the slope of the calibration curve, the more sensitive the method [42]. To determine the linearity of the method a minimum of five different concentrations were analyzed with six replicates each. The calibration curves obtained for both the quantification and the confirmation MRMs were linear for all compounds over a wide range of concentrations from 2.5 to 1000 ␮g L−1 with a correlation coefficient (R2 ) > 0.99. However, after taking into consideration the expected concentration of sulfonamides in environmental samples, and to improve the accuracy of the method, it was decided to work with the lowest concentrations. Only one MRM transition for the fragmentation ion of the highest intensity was selected for the quantification (Table 2), which resulted in lower LODs and LOQs. All calibration curves displayed good linearity with R2 from 0.9984 to 1.000. 3.2.2. Limits of detection and quantification The limits of detection (LODs) of the analytical methods developed in this study for all sulfonamides were evaluated by the measurement of 10 independent blank water samples fortified at the lowest acceptable concentration: LODs were equal to three times standard deviation of the measured signals [43]. The limits of quantification (LOQ) were calculated as twice the LOD values. The limits of detection (LODs) ranged from 0.31 to 2.74 ␮g L−1 (Table 2). The method thus enables sulfonamides to be determined at very low levels. 3.2.3. Precision and accuracy The precision of the method (expressed as RSD values) was determined in the intra-day and inter-day tests. The intra-day precision of the method was determined by calculating the relative standard deviation (RSD) for the repeated measurements (n = 6). The accuracy of the method was determined by assessing the agreement between the measured and known concentration of analysed samples. All the results are presented in the Table 2. To determine the inter-day precision of the method, three selected concentrations (5, 10 and 50 ␮g L−1 ), with the exception of sulfadiazine, sulfathiazole and sulfisoxazole (10, 50 and 500 ␮g L−1 ) were analyzed by the same person under the same conditions on

Inter-day precision RSD (%), n = 3 14.27–14.74 14.78–17.54 2.95–17.73 4.76–8.52 8.16–14.13 6.99–9.92 8.03–14.34 18.24–28.00 9.14–13.88 6.60–8.78 1.84–11.91 5.24–8.74

Accuracy (%), n = 6 80.07–113.89 97.13–119.77 98.07–115.98 96.75–113.19 87.98–108.41 82.19–109.93 80.18–90.02 113.08–125.59 85.16–113.79 76.18–100.73 101.20–108.51 86.33–115.80

six consecutive days with three replicates on each day. Instrumental repeatability, expressed as RSD, was in general <20%. According to the U.S. EPA recommendations, the method is considered precise when RSD is less than or equal to 20% (Table 2). 3.2.4. Selectivity Selectivity is the degree to which a method can quantify the analyte accurately in the presence of interferents. It is particularly important to check interferents that are likely, on chemical principles, to respond to the test [42]. To determine the selectivity of the method the blank sample of the soil extract was analyzed (Fig. 2C and D). None of the analyzed compounds was detected. Therefore, presented method is characterized by good selectivity. 3.3. Applicability of the method The method’s applicability was assessed by an analysis of 12 sulfonamides in soil samples. A very quick, simple and cost-effective method without any sample clean-up was implemented for this analysis using LC–MS/MS in MRM mode. The great advantage of this mode is its high selectivity and sensitivity when working with very complex matrices, which was also demonstrated in this experiment. Two different procedures in three replicates were performed to compare the results: one without any sample preparation and the other with sample clean-up using Solid Phase Extraction (see Section 2.4). The recoveries of the 12 sulfonamides from the spiked soil samples were calculated for both methods: they ranged from 8.78 ± 0.79% (sulfisoxazole) to 96.66 ± 3.55% (sulfadiazine) for the first procedure (six compounds had the recovery level over 50%, three between 20–50% and three lower than 20%), and from 5.59 ± 0.47% (sulfisoxazole) to 61.39 ± 5.59% (sulfadiazine) for the second procedure (three compounds had the recovery level over 50%, three between 20–50% and six lower than 20%) (the uncertainty is expressed as the standard deviation). The only matrix effect to be observed in the samples prepared without any cleanup was with sulfaguanidine, the recovery values of which were extremely high (293.65 ± 27.72%); however, after the extract had been cleaned up on an SPE-column, the recovery was 7.43 ± 0.75%. Generally, the recovery results show that there is no significant difference between the two different sample preparation steps, from which it may be inferred that this methodology can be used to determine sulfonamides in soil samples without prior clean-up. These experiments are still at a preliminary stage, however, so further investigations into the extraction procedure are necessary to improve recoveries from soil samples and to validate this analytical procedure. In summary, the optimized method for determining veterinary sulfonamides with LC–MS/MS in MRM mode, unlike many other methods, satisfied both the criteria laid down in the Commission Decision of August 12, 2002 (2002/657/EC) and the validation requirements. In the literature, many different meth-

A. Białk-Bieli´ nska et al. / Talanta 80 (2009) 947–953

ods are described for determining these compounds, which usually use LC–MS/MS full scan mode, SIM mode or only one MRM transition [26,27,31–33,35]. According to the Commission Decision, this is not sufficient to confirm the presence of these compounds. The methodology developed in the present work is characterized by good validation parameters, low LODs and LOQs, as well as good precision, repeatability and applicability to real samples. The fact that no preconcentration step was necessary is also worthy of note. 4. Conclusion The optimized chromatographic separation and MS/MS parameters for the analysis of common veterinary sulfonamide antibiotics by LC–MS/MS was done in order to deliver a systematic analytical methodology for their determination in both environmental and biological samples. The experiments were performed using ion trap mass spectrometry with electrospray ionization in order to work in multiple reaction monitoring (MRM) mode. Four identification points for every compound (a precursor ion and a minimum of two daughter ions) were selected according to the existing criteria for both screening and confirmatory methods for drug residues based on the identification points proposed by EU Commission Decision 2002/657/EC. Each MRM transition was tested not only for the qualitative but also for quantitative analysis of sulfonamides, which is a huge advantage due to the fact, that the quantification of the sulfonamides in a very complex matrices can be done not only based on one-selected transition. If there are some interferences, it is always possible to use another transition for the quantification purposes. Unlike other methodologies, this one satisfied these criteria and is applicable to the qualitative and quantitative analysis of sulfonamide residues in environmental and biological matrices. Acknowledgement Financial support was provided by the Polish Ministry of Research and Higher Education under grants DS 8200-4-0085-9 and BW 8000-5-0357-9. References [1] [2] [3] [4] [5] [6]

A. Nikolaou, S. Meric, D. Fatta, Anal. Bioanal. Chem. 387 (2007) 1225–1234. S.K. Khetan, T.J. Collins, Chem. Rev. 107 (2007) 2319–2364. S.D. Richardson, Anal. Chem. 78 (2006) 4021–4046. A.K. Sarmah, M.T. Meyer, A.B.A. Boxall, Chemosphere 65 (2006) 725–759. N. Kemper, Ecol. Indicat. 8 (2008) 1–13. I. Kanfer, M.F. Skinner, R.B. Walker, J. Chromatogr. A 812 (1998) 255–286.

953

[7] J.B. Addison, Residue Rev. 92 (1984) 1–28. [8] M. Grote, C. Schwake-Anduschus, R. Michel, H. Stevens, W. Heyser, G. Langenkämper, T. Betsche, M. Freitag, Landbauforschung Völkenrode 1 (2007) 25–32. [9] C.S. McArdell, E. Molnar, M.J.F. Suter, W. Giger, Environ. Sci. Technol. 37 (2003) 5479–5486. [10] M.S. Díaz-Cruz, D. Barceló, Anal. Bioanal. Chem. 386 (2006) 973–985. [11] T. Heberer, J. Hydrol. 266 (2002) 175–189. [12] W.C. Lin, H.C. Chen, W.H. Ding, J. Chromatogr. A 1065 (2005) 279–285. [13] Z. Ye, H.S. Weinberg, M.T. Meyer, Anal. Chem. 79 (2007) 1135–1144. [14] A.J. Watkinson, E.J. Murby, S.D. Costanzo, Water Res. 41 (2007) 4164–4176. [15] A. Gulkowska, H.W. Leung, M.K. So, S. Taniyasu, N. Yamashita, L.W.Y. Yeug, B.J. Richardson, A.P. Lei, J.P. Giesy, P.K.S. Lam, Water Res. 42 (2008) 395–403. [16] W.L. Shelver, N.W. Shappell, M. Franek, F.R. Rubio, J. Agric. Food Chem. 56 (2008) 6609–6615. ˜ [17] F. Hernández, J.V. Sancho, M. Ibánez, C. Guerrero, Trends Anal. Chem. 26 (2007) 466–485. [18] S.D. Richardson, Anal. Chem. 80 (2008) 4373–4402. ´ D. [19] S.C. Kim, K. Carlson, Analysis of antibiotics in solid samples, in: M. Petrovic, Barceló (Eds.), Comprehensive Analytical Chemistry, vol. 50, 2007, pp. 95–131. [20] C. Hartig, T. Storm, M. Jekel, J. Chromatogr. A 854 (1999) 163–173. [21] S. Yang, J. Cha, K. Carlson, J. Chromatogr. A 1097 (2005) 40–53. [22] V.K. Balakrishnan, K.A. Terry, J. Toito, J. Chromatogr. A 1131 (2006) 1–10. [23] J.D. Li, Y.Q. Cai, Y.L. Shi, S.F. Mou, G.B. Jiang, J. Chromatogr. A 1139 (2007) 178–184. [24] E.L. McClure, C.S. Wong, J. Chromatogr. A 1169 (2007) 53–62. [25] C. Lin, S. Huang, Anal. Chim. Acta 612 (2008) 37–43. [26] C. Hong, H. JianYing, W. LeZheng, S. Bing, Chin. Sci. Bull. 53 (2008) 514–520. [27] M.E. Lindsey, M. Meyer, E.M. Thurman, Anal. Chem. 73 (2001) 4640–4646. [28] A.L. Batt, D.S. Aga, Anal. Chem. 77 (2005) 2940–2947. [29] A.M. Jacobsen, B. Halling-Sørensen, Anal. Bioanal. Chem. 384 (2006) 1164–1174. [30] N.T. Malintan, M.A. Mohd, J. Chromatogr. A 1127 (2006) 154–160. [31] W. Ben, Z. Qiang, C. Adams, H. Zhang, L. Chen, J. Chromatogr. A 1202 (2008) 173–180. [32] M.Y. Haller, S.R. Müller, C.S. McArdell, A.C. Alder, M.J.F. Suter, J. Chromatogr. A 952 (2002) 111–120. [33] A.M. Jacobsen, B. Halling-Sørensen, F. Ingerslev, S.H. Hansen, J. Chromatogr. A 1038 (2004) 157–170. [34] A. Göbel, A. Thomsen, C.S. McArdell, A.C. Alder, W. Giger, N. Thei␤, D. Löffler, T.A. Ternes, J. Chromatogr. A 1085 (2005) 179–189. [35] E. Martínez-Carballo, C. González Barreiro, S. Scharf, O. Gans, Environ. Pollut. 148 (2007) 570–579. [36] M. de Liguoro, C. Poltronieri, F. Capolongo, C. Montesissa, Chemosphere 68 (2007) 671–676. [37] A. Nieto, F. Borrull, R.M. Marcé, E. Pocurull, J. Chromatogr. A 1174 (2007) 125–131. ˜ ˜ [38] J.J. Berzas Nevado, G. Castaneda Penalvo, F.J. Guzmán Bernard, Anal. Chim. Acta 442 (2001) 241–248. [39] G. Balizs, A. Hewitt, Anal. Chim. Acta 492 (2003) 105–131. [40] A. Gentili, D. Perret, S. Marchese, Trends Anal. Chem. 24 (2005) 704–733. [41] N.A. Botsoglou, D.J. Fletouris, E.J. Simeonidou, I.E. Psomas, Chromatographia 46 (1997) 477–482. [42] M. Thompson, S. Ellison, R. Wood, Pure Appl. Chem. 74 (2002) 835–855. [43] I. Taverniers, M. de Loose, E.V. Bockstaele, Trends Anal. Chem. 23 (2004) 535–551. [44] W.M.A. Niessen, J. Chromatogr. A 812 (1998) 53–75.