Determination of Diclofenac in Water Samples from Wastewater Treatment Plants by Large Volume Injection-Gas Chromatography–Mass Spectrometry

Diclofenac, a non-steroidal anti-inflammatory drug (NSAID), was analyzed in samples from two wastewater treatment plants (WWTPs) in Spain. For this process, a new analytical method using large volume injection in gas chromatography-mass spectrometry (LVI-GC–MS) via a through oven transfer adsorption desorption (TOTAD) interface was developed. Diclofenac was detected within incoming water (in the effluent after a decantation step), but not in outgoing water, demonstrating the efficiency of treatment of the WWTP.

Pharmaceuticals and personal care compounds are widely used by modern society. However, they are also a source of environmental impact and increasing concern. Non-steroidal anti-inflammatory drugs (NSAIDs) are well known drugs consumed in very large quantities; for example, in the treatment of mild to moderate pain. Moreover, some of them are available as nonprescription pharmaceuticals, and their physical and chemical properties make them easily transported into hydrologic systems, where their effects on human health and aquatic ecosystems are mostly unknown. Some studies have suggested that NSAIDs may be toxic. For example, prolonged exposure to diclofenac gives rise to its bioaccumulation and toxic effects in fish (1). Nevertheless, data about the toxicity of metabolites in humans are scarce and inconclusive, which explains why their occurrence in the environment has attracted great interest among researchers (2–5). Their main source in the aquatic environment is excretion and disposal via wastewater. After administration, pharmaceuticals can be excreted, primarily via urine and feces. Conventional wastewater treatment plants (WWTPs) are designed to remove pathogens and coliforms, and to reduce loads of carbon, nitrogen, and phosphorus by means of different physicochemical and biological techniques. These treatment technologies produce water that satisfies current legislation on water-quality standards. Nevertheless, a diverse group of unregulated pollutants, sometimes also called emerging contaminants, need to be kept in mind. It is, therefore, necessary to obtain information about these compounds, as well as their possible transformation products. Indeed, several researchers have confirmed the presence of NSAIDs in wastewater samples (6–10). In addition, due to the need to study their environmental occurrence and fate, several analytical methods for the determination of pharmaceuticals and their metabolites in aqueous solutions have been developed (11).

Modern analytical chemistry provides analytical methods that permit the measurement of trace components at very low levels (ng/L, or even below). The usual analytical methods for the identification and quantitation of NSAIDs employ different sample preparation techniques and concentration steps, the most commonly applied technique for aqueous samples being solid-phase extraction (SPE), combined with gas chromatography-mass spectrometry (GC–MS) (12) or liquid chromatography tandem mass spectrometry (LC–MS/MS) (13,14).

An alternative technique to the concentration steps is the use of online coupling LC–GC or large volume injection (LVI) techniques in GC. Online coupling LC–GC combines the high efficiencies of LC for sample preparation with the high efficiency and sensitivity of GC. LVI allows sample injections of up to several hundred microliters. Obviously, the injection of higher volumes of sample substantially increases sensitivity or reduces the need for extract concentration steps (14). The issue with both techniques is basically the same: how to introduce into the GC system a high volume of a liquid sample, extract or eluent from LC, far above that usually injected in GC. To solve this issue, reliable and versatile interfaces are needed. LVI techniques includes on-column, programmed temperature vaporization (PTV) in solvent split mode, direct sample introduction (DSI), splitless overflow, at-column, and through oven transfer adsorption desorption (TOTAD) interface (15).

The TOTAD interface has been designed and proposed with such an aim in mind. It was developed in 1999 by Pérez and co-workers (16), is totally automatic, and is a great modification of a PTV injector. The sample is introduced by a silica capillary tube (SCT), connected together with the GC column to the liner packed with an adsorbent material, and placed inside the TOTAD interface. The analytes are retained, and the solvent is vented by the carrier gas through a waste tube (WT). Once the elimination of the solvent has been completed, the TOTAD interface is rapidly heated and the analytes are transferred to the GC column (17). It has the advantage that it allows non-polar and polar solvents to be used, including water. The TOTAD interface has been used for the online LC–GC analysis of many different analytes in complex matrices such as food (18–19), or environmental samples (20–21), as well as for large volume injection (LVI) into GC in the analysis of compounds present at trace level in different samples (22–23).

In this work, we focus on diclofenac, an NSAID. The removal efficiencies of diclofenac vary between different WWTPs, 40% being a typical reported value when considering the parent compound (5). However, Radjenovic and associates reported that the removal efficiency of diclofenac could vary between 10-80% (11). The main goal of this study was to develop a fast analytical method for the determination of diclofenac in wastewater by LVI-GC–MS via the TOTAD interface. This method was applied to evaluate the elimination of diclofenac residues during the wastewater treatment WWTP processes.

Methods

Sampling and Studied WWTP

Albacete is in the southeastern part of Spain, in the autonomous community of Castilla-La Mancha. Albacete WWTP treats the wastewater from the city of Albacete, as well as surrounding villages with a total population of approximately 315,000 inhabitants. The WWTP treats a daily average of 40,000 cubic meters of water, 33,000 cubic meters coming from the sewage system, and 7,000 cubic meters from industrial wastewater.

The Albacete plant has three treatment lines: the water line, the sludge line, and the gas line. Grab sampling was performed at three different sampling points within the plant: influent (incoming), water after secondary decantation, and effluent (outgoing). Other samples were collected from the WWTP at a small village close to Albacete. This plant treats the wastewater from approximately 600 people. The village plant has two treatment lines: water line and sludge line. The samples were taken in two different points, incoming and outgoing. The samples were collected in June of 2016 on three different days, and were individually analyzed. They were filtered twice through a 0.22 µm nylon filter (Chromatography Research Supplies, Inc., Louisville, Kentucky), and diluted to 30% with acetonitrile. Samples were stored at 4 °C, in the dark. The standard was prepared by dissolving diclofenac in water:acetonitrile (70:30, v/v), and used to optimize the analytical conditions. Different concentrations, ranging from 5 ng/mL to 500 ng/mL, were prepared. Diclofenac was purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany), and water and acetonitrile were used to dilute and to inject the aqueous samples into the TOTAD interface were HPLC grade (Labscan). Standard samples were filtered once through a 0.22 µm nylon filter.

Instrumentation

A Konik 5000B gas chromatograph, equipped with a TOTAD interface, was coupled to a mass spectrometer (MS) (Konik MS Q12). The TOTAD interface was used to inject a large volume of aqueous samples into the GC. For large volume sampling, a manual injection valve (model 7125, Rheodyne, California), provided with a 100 µL loop, was used. A ternary pump (Konik model 550) was used to inject the large volume of aqueous samples into the TOTAD interface. Data acquisition and processing were performed with Konikrom Plus and MS Control (Konik, Sant Cugat Del Vallés, Barcelona, Spain) software.

TOTAD Operation Mode

The operating mode of the TOTAD interface has been extensively described elsewhere (16). The TOTAD interface and GC oven temperature were stabilized at 125 °C and 60 °C, respectively; a helium flow enters the liner through both the oven side (b) and the opposite side (a) (Figure 1) during the transfer step at 300 mL/min; the elimination time of the remaining solvent was 1 min. The TOTAD interface was heated to 325 °C, and maintained at this temperature for 5 min to achieve the thermal desorption and transfer of the analytes to the GC column impelled by the helium. The samples reached the glass liner at 0.1 mL/min. Tenax TA, 80–100 mesh (Supelco, Bellefonte, Pennsylvania) was used as adsorbent inside the liner of the TOTAD interface, holding 1 cm of Tenax TA in place by means of two plugs of glass wool. The liner prepared was then conditioned in a helium stream, increasing the temperature to 50 °C in 10 min to reach 350 °C, and then maintaining this for 1 hr.

Figure 1: Schematic of the through oven transfer adsorption desorption (TOTAD) interface during the injection step.

GC Conditions

GC separations were carried out on a DB-17MS (30 m × 0.25 mm internal diameter (i.d.), 0.25 µm) (J&W Scientific, Folsom, California). During the transfer and solvent elimination steps the oven temperature was maintained at 60 °C. During GC analysis, the helium used as carrier gas enters through A at 1 mL/min. The oven initial temperature was kept at 60 °C for 1 min, raised to 270 °C at 20 °C/min, and maintained for 10 min. The analytical conditions GC–MS were as follows: interface temperature, 270 °C, MS was performed in electron impact mode at 70 eV, ion source temperature, 110 °C, MS operating in selected-ion monitoring (SIM) mode, in order to improve the detection limits using different ions as 214, 242, and 295.

Results and Discussion

Method Optimization

The operation parameters of the TOTAD interface were fixed as indicated in the previous section entitled "TOTAD Operation Mode". Before GC was coupled to the MS, and to verify the complete elimination of solvent, a water resistant flame ionization detector (FID) was used. It was observed that there was no solvent peak in the GC chromatogram obtained (results not shown). A helium flow pushes the water to the Tenax, which is on the opposite side of the GC column (see Figure 1), thus preventing the introduction of water into the GC column. Only the water retained in the Tenax or the glass wool of the liner enters the GC column, which, after the remaining solvent elimination step, is a very low amount. If water were pumped to the GC column and the MS detector, the solvent background would increase but, as can be seen in Figure 2 and Figure 3, this does not happen. A standard solution of the target NSAID was analyzed and characterized by GC–MS in the SIM mode. As can be observed from the chromatogram obtained by LVI-GC–MS, in SIM mode performed with 100 µL injection volume of diclofenac in water–acetonitrile at 50 ng/mL (Figure 2a) and at 5 ng/mL (Figure 3a), there is no problem with the volume injected, because water is totally eliminated by the TOTAD interface before GC analysis. Moreover, up to 75 mL of aqueous samples can be injected without risk of damaging the MS (24). The sample is directly injected without a sample pre-treatment step other than filtration, thus avoiding the use of organic solvent and disposable cartridges used in SPE. LVI techniques allow for an increase in the sensitivity of the methods, as the analyte concentration reaching the detector is much higher compared to the injection of 1 or 2 µl (25). In this method, 100 µl of sample was directly injected. The concentration step occurs inside the TOTAD interface, where the analyte is retained, and the solvent vented. Other analytical methods require efficient sample treatment to concentrate the analytes, and to remove matrix interferences (26).

Figure 2: Chromatograms obtained by large volume injection-gas chromatography- mass spectrometry (LVI-GC-MS), using the selected ion monitoring (SIM) mode. The real water samples are from Albacete Waste Water Treatment Plant (WWTP), injected volume of100 µL. The chromatograms correspond to: (a) standard solution of diclofenac at 50 µg/L, (b) incoming water, (c) water after decantation, and (d) outgoing water. The peak identified [1] corresponds to diclofenac.

Validation Method

In order to determine the linearity of the method, 100 µL of standard solutions were injected over a wide range of concentrations from 5–500 ng/ mL. The correlation coefficient (R²) of the method was 0.992. The linearity of the whole procedure was good. The repeatability of the method was measured by injecting 100 µL at 50 ng/mL five times. The result shows a very good relative standard deviation (RSD) of the absolute peak areas, lower than 5%.

Figure 3: Chromatograms obtained by large volume injection-gas chromatography- mass spectrometry (LVI-GC-MS), in selected ion monitoring (SIM) mode. The real water samples are from village Waste Water Treatment Plant (WWTP), injected volume of 100 µL. The chromatograms correspond to: (a) standard solution of diclofenac at 5 µg/L, (b) incoming water, and (c) outgoing water. The peak identified as [1] corresponds to diclofenac.

The limit of detection (LOD) and limit of quantification (LOQ), calculated as the analyte concentration, resulted in a signal to noise (S/N) of 3 and 10, respectively, corresponding to concentration levels of 0.2 ng/mL and 0.6 ng/mL. The data are listed in Table I.

Application to Real Samples

Real WWTP samples of two different WWTPs were analyzed. Albacete WWTP uses a more thorough process based on primary, secondary, and biological treatment. The village WWTP process only features pretreatment and secondary treatment. For this reason, three samples were analyzed in Albacete: incoming, water effluent after decantation, and outgoing. In the village WWTP, only two samples were analyzed, incoming and outgoing. The analyses were carried out without sample preparation, and samples were only filtered, without any extraction or concentration step. As can be seen in Figures 2b and 3b, diclofenac was detected for both WWTPs within the incoming samples, at 0.87 ng/mL in Albacete WWTP and at a lower concentration than the LOQ within the village WWTP. Other authors have found similar concentrations using different techniques, for instance, 0.46 ng/mL (12), 0.492 to 0.996 ng/mL (13), 1.4 ng /mL (27). After the secondary decantation in Albacete, it was detected at a lower level that in incoming water (Figure 2c). In outgoing samples, diclofenac was not detected in Albacete or the village WWTP (Figures 2d and 3c). These observations indicate the efficiency of the removal processes involved. Some unidentified peaks from the matrix were also detected in incoming samples from both WWTPs. These peaks decreased as the water treatment process progressed.

Conclusions

LVI-GC coupled to a MS detector proved to be an efficient and easy method for the determination of diclofenac in dirty water samples such as the WWTP samples. The wastewater samples only have to be filtered, and need no extraction or concentration step, being directly injected into the GC–MS. The method developed presents good linearity and repeatability, and the sensitivity is as good as that obtained with other methods.

Although diclofenac was found in incoming water samples from both WWTPs, it could only be quantified in Albacete WWTP (at 0.87 ng/mL, a concentration lower than the LOQ that was detected for the village WWTP). The drug was not found in either outgoing water samples, demonstrating the efficiency of the respective treatment plants. Although this is a preliminary study, the data obtained are very promising, and further studies will be carried out including more NSAIDs and also their degradation products.

Conflict of Interest

There are no conflicts to declare.

Acknowledgments

The authors wish to express their gratitude to Marta Nieto from Aguas de Albacete WWTP for helping with the sampling and providing valuable information about WWTP. Consejería Educación y Ciencia of Junta de Comunidades de Castilla-La Mancha project SBPLY/17/180501/000377 is gratefully acknowledged. J.M Cortés thanks the CYTEMA campus for his grant.

References

(1) J. Schwaiger, H. Ferling, U. Mallow and H. Wintermayr, Aquat. Toxicol. 68, 141–150 (2004).

(2) M. Villar Navarro, M. Ramos Payan, R. Fernández-Torres, M.A. Bello-López, M. Callejón and M.A. Guiráum-Pérez, Electrophoresis 32, 2107–2113 (2011).

(3) A.I. Olives, V. González-Ruiz and M.A. Martín, Anti-Inflamm. Anti-Allergy Agents Med. Chem. 11, 65–95 (2012).

(4) E. Sagristà, E. Larsson, M. Ezoddin, M. Hidalgo, V. Salvadó and J.Å. Jönsson, J. Chromatogr. A 1217, 6153–6158 (2010).

(5) E. Larsson, A. Rabayah and J.Å. Jönsson, J. Environ. Prot. 4, 946–955 (2013).

(6) L.H.M.L.M. Santos, M. Gros, S. Rodriguez-Mozaz, C. Delerue-Matos, A. Pena, D. Barceló and M.C.B.S.M. Montenegro, Sci. Total Environ. 461–462, 302–316 (2013).

(7) M. Lavén, T. Alsberg, Y. Yu, M. Adolfsson-Erici and H. Sun, J. Chromatrogr. A 1216, 49-62 (2009).

(8) J.B. Quintana, R. Rodil, S. Muniategui-Lorenzo, P. López-Mahía and D. Prada-Rodríguez, J. Chromatrogr. A 1174, 27-39 (2007).

(9) A. Togola and H. Budzinski, J. Chromatrogr. A, 1177, 150-158 (2008).

(10) J. Wu, X. Qian, Z. Yang and L. Zhang, J. Chromatrogr. A, 1217, 1471-1475 (2010).

(11) J. Radjenovic, M. Petrovic and D. Barceló, TrAC, Trends Anal. Chem. 26, 1132-1144 (2007).

(12) N. Migowska, M. Caban, P. Stepnowski and J. Kumirska, Sci. Total Environ. 441, 77–88 (2012).

(13) N. Gilart, R. M. Marce, F. Borrull and N. Fontanals, J. Sep. Sci. 35, 875–882 (2012).

(14) J.A. Rivera-Jaimes, C. Postigo, R.M. Mendoza-Alemán, J. Aceña, D. Barceló, M. López de Alda, Sci. Total Environ. 613–614, 1263-1274 (2018).

(15) E. Hoh and K. Mastovska, J. Chromatrogr. A 1186, 2-15 (2008).

(16) M. Pérez, J. Alario, A. Vázquez and J. Villén, J. Microcolumn Sep. 11, 582–589 (1999).

(17) G. Purcaro, S. Moret, L. Conte, J. Chromatrogr. A 1255, 100–111 (2012).

(18) J.M. Cortés, R.M. Toledano, J. Villén and A. Vázquez J. Agric. Food Chem. 56, 5544–5549 (2008).

(19) A. Aragón, J.M. Cortés, R.M. Toledano, J. Villén and A. Vázquez, J. Chromatrogr. A 1218, 4960–4965 (2011).

(20) M. Pérez, J. Alario, A. Vázquez and J. Villén, Anal. Chem. 72, 846–852 (2000).

(21) J. M. Cortés, J. C. Andini, R. M. Toledano, C. Quintero, J. Villén and A. Vázquez, Int. J. Environ. Anal. Chem. 93, 461–471 (2013).

(22) R. M. Toledano, J. M. Cortés, J. C. Andini, J. Villén and A. Vázquez, J. Chromatrogr. A 1217, 4738–4742 (2010).

(23) J. M. Cortés, R. Sánchez, E.M. Díaz-Plaza, . Villén and A. Vázquez, J. Agric. Food Chem. 54, 1997–2002 (2006).

(24) A. Aragón, R.M. Toledano, A. Vázquez, J. Villén and J.M. Cortés, Talanta 139, 1–5 (2015).

(25) S.H.G. Brondi, F.C. Spoljaric, F.M. Lanças, J. Sep. Sci. 28, 2243–2246 (2005).

(26) K.M. Dimpe, P.N. Nomngongo, Trends in Anal. Chem. 82, 199-207 (2016).

(27) M. Becerra-Herrera, L. Honda and P. Richter, J. Chromatrogr. A 1423, 96–103 (2015).

álvaro Aragón, Rosa M. Toledano, and Jesús Villén are with the Botanical Institute of the Botanic Garden University Campus in Albacete. Spain. Ana Vázquez and José M. Cortés are with the Universidad de Castilla-La Mancha, in Albacete. Spain.Direct correspondence to: alvaroarse@gmail.com