Chemosphere 69 (2007) 554560 www.elsevier.

com/locate/chemosphere

Application of nano-FIA-Direct-EI-MS to determine diethylene glycol in produced formation water discharges and seawater samples
Achille Cappiello a,*, Giorgio Famiglini a, Pierangela Palma a, Elisabetta Pierini a, Helga Trufelli a, Chiara Maggi b, Loredana Manfra b, Michela Mannozzi b
a

` Istituto di Scienze Chimiche F. Bruner, Universita di Urbino, Piazza Rinascimento 6, 61029 Urbino, Italy b I.C.R.A.M. Central Institute for Marine Research, Via di Casalotti 300, 00166 Roma, Italy Received 17 January 2007; received in revised form 13 March 2007; accepted 14 March 2007 Available online 11 May 2007

Abstract Diethylene glycol (DEG) is extensively used on oshore gas platforms to prevent the hydrate formation during the gaswater separation process and to inhibit corrosion events. This chemical might enter in the marine environment via the produced formation water (PFW) discharge. In this study, a new approach was applied to the investigation of the DEG content in PFW discharges and seawater samples from four gas installation platforms in the Adriatic Sea (Italy). The method includes an o-line solid-phase extraction/preconcentration technique, followed by a nanoscale ow injection/direct-electron ionization (EI) mass spectrometric analysis. Direct-EI is a novel and miniaturized interface for directly coupling a liquid chromatograph with an electron ionization mass spectrometer. The capability to acquire EI spectra, and to operate in selected ion monitoring mode during actual sample analyses, allowed a precise quantication of DEG with a method limit of detection of 31 lg/l. In addition, a careful evaluation of the matrix eect showed that, as opposed to electrospray ionization, the response of the Direct-EI interface was not aected by sample interferences. 2007 Elsevier Ltd. All rights reserved.
Keywords: Diethylene glycol; Gas installation platforms; Produced formation water discharge; Seawater; Nano-FIA-Direct-EI-MS; ESI-MS/MS

1. Introduction During the oshore production processes, large volumes of aqueous waste are produced along with oil and gas. This produced formation water (PFW) is a by-product of the gas extraction operation and may include: naturally-occurring water layer present in oil and gas reservoirs (formation water), water that has been injected into the reservoirs to force the oil to the surface and any chemicals added during the production/treatment process. Once at the surface, the PFW is separated from the hydrocarbons, treated to reduce the remaining hydrocarbons and the suspended matter and discharged into the sea.

Corresponding author. Tel.: +39 0722 303344; fax: +39 0722 303311. E-mail address: acappiello@uniurb.it (A. Cappiello).

Diethylene glycol (DEG) is a widely used industrial chemical with a potential for human exposure (Lewis, 1993; Osterberg and See, 2003; Ballantyne and Snellings, 2005; Ferrari and Giannuzzi, 2005), which can be found in PFW discharges. In fact, this chemical, together with ethylene glycol (EG) and triethylene glycol (TEG), is commonly used in the dehydration processes of natural gas (Katz and Lee, 1990). Under normal conditions, natural gas is saturated with water vapor; this is undesirable because it reduces the eciency of the extraction process. In fact, when the watersaturated natural gas reaches its dew point, water molecules can combine with methane, ethane, propane, and other hydrocarbons to form crystalline-structured solid hydrates. Furthermore, when the gas is compressed or cooled, water vapor is converted to a liquid or solid phase. Liquid water can induce corrosion and reduce gas transmission eciency, and ice can plug valves, ttings, and

0045-6535/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2007.03.026

A. Cappiello et al. / Chemosphere 69 (2007) 554560

555

gas lines. Due to their high hygroscopic properties, glycols can absorb water from the natural gas stream, a characteristic that makes them the most widely used dehydration agents in the gas industry (Katz and Lee, 1990). In a typical glycol-based dehydration device, water vapor is removed from the gas stream in a glycol-absorber. The dried gas leaves the absorber for further processing or transport. The water-laden glycols are usually regenerated (stripped of water) in a stripping column. Because of the eciency of the regeneration processes occurring in the glycol-based dehydration devices is less than 100%, DEG is introduced in the marine environment by the oshore extraction plants, through the discharge of PFW. The euents associated with DEG might be of environmental concern because: (1) both benthic and pelagic organisms might be aected by DEG exposure (RIVM, 2005); (2) due to its chemical properties, DEG can act as a co-solvent for other environmental pollutants of the produced formation water discharges (Sorensen et al., 2000); (3) little research has been performed regarding the biodegradability of DEG in seawater under relevant environmental conditions (Nyholm et al., 1992); (4) although many studies about chemical composition and toxicity of PFW have been published (Brendehaug et al., 1992; Richardson, 1996; Henderson et al., 1999; Holdway, 2002; Mariani et al., 2004), information regarding the environmental impact of nding DEG in PFW and in seawater are still lacking. For these purposes, the present work was aimed at developing an innovative method for the determination of DEG in oshore euents and seawater samples. Gas chromatography coupled to mass spectrometry (GCMS) represents the method of choice for the analysis of DEG and other glycols in food (Savchuk and Kolesov, 2005), biological (Williams et al., 2000; Peters et al., 2001; Gembus et al., 2002), and environmental matrices (Drugov and Muraveva, 1995; Bensoam et al., 1999; Szymansky et al., 2001). However, the determination of hydrophilic compounds, such as glycols, in aqueous matrices by this approach is often aected by some disadvantages: preconcentration techniques such as solid phase extraction (SPE), useful for the extraction of hydrophobic compounds, usually fail because of too low recoveries due to the nal elution step in an organic solvent; GCMS assay often requires derivatization procedures which can lead to analyte losses. Due to these problems, several eorts have to be addressed on developing GCMS alternative analytical approaches for the detection of glycols. High performance liquid chromatography coupled to electrospray mass spectrometry (HPLC-ESI-MS) is now the most specic and powerful technique for the identication and quantitation of polar organic compounds in aqueous matrices. However, the sensitivity of ESI is usually reduced for glycols of low molecular weight, such as DEG, because of the occurrence of increased noise due to the presence of various adducts attributed to the mobile phase. Holcapek et al. (1999) overcome this problem through a sample derivatization with benzoyl chloride. Another drawback

associated with the ESI-MS detection may be represented by the occurrence of matrix eects related to sample interferences, when dealing with complex samples (Matuszewsky et al., 1998; Choi et al., 2001; Reemtsma, 2001; Niessen et al., 2006). In this study the identication and quantication of DEG in seawater matrices was achieved exploiting an innovative analytical approach based on a quadrupole mass spectrometer coupled with a direct-electron ionization (EI) interface, a novel device for directly coupling a liquid chromatograph with an electron ionization ion source mass spectrometer (Cappiello et al., 2001, 2002, 2003). The samples were introduced into the Direct-EI interface by ow injection analysis (FIA) at nano-ow rates. No column separation is done with this sample introduction method, allowing rapid analyses. Direct-EI takes advantage of the very low operating ow rate typical of the nano-HPLC, allowing the registration of high quality EI spectra. The interface can be placed on a pre-existing GC/MS system with a simple and reversible modication, converting it into an HPLC-EI-MS instrument. Direct-EI is not inuenced by mobile phase composition, analyte polarity and matrix; this is a clear advantage over atmospheric pressure ionization (API) based interfaces such as ESI, where a matrix eect is often observed when dealing with complex mixtures. Using this approach, we were able to detect DEG in marine water samples at a concentration P31 lg/l. An o-line solid-phase extraction procedure was employed for a suitable sample preparation and pre-concentration. The specicity needed for compound identication was reached by selected ion monitoring (SIM) analysis of three characteristic ions. In addition, the performance of the nano-FIADirect-EI approach for the characterization of DEG in seawater was compared with that of FIA-ESI-MS. In these experiments a strong signal suppression and matrix eect were observed when using electrospray ionization. 2. Materials and methods 2.1. Sampling The sampling was carried out in summer of 2005 and 2006 during ICRAM oceanographic cruises. The PFWs were sampled on four gas platforms (A, B, C and D) located in Adriatic sea; then surface seawater samples were taken at increasing distance from the PFW diusor (0 munder the diusor, 5 m, 10 m, 15 m, 20 m, 25 m). The water was collected in dark glass bottles to inhibit photochemical activity and saturated with mercury cloride to inhibit bacterial activity. Then the samples were refrigerated to 4 C until the chemical analysis. 2.2. Reagents All solvents were HPLC grade from VWR International (Milan, Italy). Reagent grade water was obtained from a

556

A. Cappiello et al. / Chemosphere 69 (2007) 554560

milli-Q water purication system (Millipore, Bedford, MA, USA). Diethylene glycol (CAS number 111-46-6, MW 106) was purchased from Fluka (Milan, Italy). A stock DEG solution was prepared in deionized water at a concentration of 10 000 mg/l. 2.3. Sample preparation

ing m/z 69, 219, and 502. Considering the molecular weight of DEG, the repeller potential was optimized for the m/z 219. No mobile phase was allowed into the ion source during calibration. The dwell times during SIM analysis were adjusted in order to obtain 0.5 cycles/s and a mean of 10 acquisition for each peak. 2.5. Electrospray mass spectrometry

The extraction procedure was carried out with 2 ml SPE cartridges packed with 200 mg of ISOLUTE ENV+ stationary phase (International Sorbent Technology, Glamorgan, UK). The cartridges were conditioned with 15 ml of acetone, followed by 10 ml of deionized water. Water samples (100 ml) were forced through the traps at a ow rate of 5 ml/min. Before extraction, the cartridges were rinsed with 10 ml of deionized water and dried under vacuum for 45 min. After that, 8 ml of acetone were allowed to pass through the traps to elute DEG. The extracts were evaporated to a volume of 200 ll, under a gentle stream of nitrogen. Afterward, 200 ll of deionized water were added and the extract was nally evaporated to reach a nal volume of 200 ll. 2.4. Direct-EI interface Direct-EI interface was mounted on a Hewlett-Packard 5989A quadrupole mass spectrometer (Palo Alto, CA, USA). In the Direct-EI interface, the whole interfacing process occurs in the electron ionization source of the mass spectrometer. The analytes, together with the mobile phase, are converted into an aerosol under high-vacuum conditions and solvent vapors are quickly eliminated prior to the ionization. The nebulizer is connected to the injector via a thermally insulated 30 lm i.d. fused silica capillary tubing (Polymicro Technologies, Phoenix, AZ, USA), which passes from an atmospheric pressure region to high-vacuum one. A target surface inside the ion source acts as hot vaporization surface. Most of the solvent vapors are eliminated through a specially designed opening. A thorough description of the characteristics of the Direct-EI interface is reported elsewhere (Cappiello et al., 2002, 2003). In the present work, solutions of standards and samples were introduced into the Direct-EI interface by FIA at a ow rate of 300 nl/min, using an injection loop of 60 nl. The ow rate of 300 nl/min permitted the generation of proper aerosol, without interferences due to chemical ionization processes. The mobile phase was composed of 100% deionized water. The ow rate of 300 nl/min was obtained with a laboratory-made splitter, placed between the pumping system (Kontron Instruments 422 dual-pump binary gradient HPLC system, Milan, Italy) and the injector valve (Berloni et al., 1994). The temperature of the ion source was kept at 250 C. The operating pressure was of $1.52.0 105 Torr ($2.02.7 103 Pa), measured in the manifold of the ion source at the operating ow rate. Mass spectrometer tuning and calibration were performed automatically using peruorotributylamine and monitor-

A ThermoFinnigan LCQ-DUO HPLC-ESI-MS system (ThermoElectron Corporation, San Jose, CA, USA), equipped with an ion trap analyzer for MS/MS analysis, was employed. Solutions of standards and samples were introduced into the electrospray interface in ow injection analysis (FIA) conditions at a ow rate of 250 ll/min and using an injection loop of 5 ll. The solvents used to force the samples into the mass spectrometer was composed of water: methanol 90:10. DEG gave a better response in positive ion mode. Positive ions were generated and transferred into the analyzer using a 4.5 kV voltage applied between the needle and the transfer capillary. Nitrogen was used to assist both nebulization and desolvation. The pressures used were 4.14 and 1.38 bar, respectively. The heated capillary potential was xed at 10 V, while the capillary temperature was set at 225 C. MS experiments were carried out using a mass range from m/z 100 to m/z 300. Ions were accumulated in the trap for 200 ms, while three microscans were summed up to generate a full scan acquisition. The sodium adduct ion of DEG (m/z 129) was chosen as precursor ion for MS/MS experiments. Tandem mass spectra were obtained by collision induced dissociation (CID), using helium as the collision gas and using a collision energy of 23%. The maximum injection time was set at 800 ms and three microscans were summed up. The same conditions were applied for selected reaction monitoring (SRM) experiments. 3. Results and discussion 3.1. Method development In the last few years many authors have studied PFW composition and its eects on marine environment (Brendehaug et al., 1992; Richardson, 1996; Henderson et al., 1999; Holdway, 2002; Mariani et al., 2004). In the present work nano-FIA-Direct-EI-MS is proposed as an innovative and sensitive approach for the determination of the additive DEG in PFW and seawater samples. Direct-EI interface allows the direct introduction of the euent into the electron ionization source of a mass spectrometer. This attribute represents a point of strength when compared to API-based interfaces. In fact, dierently from API, DirectEI is not aected by some limitations that can lead to erratic quantitative results such as solution and mobile phase, pH, electrolyte concentration, and the presence of matrix interference components. In addition Direct-EI interface generates typical, highly informative spectra and its repro-

A. Cappiello et al. / Chemosphere 69 (2007) 554560

557

ducibility allows the comparison with thousands of spectra from commercially available sources (such as NIST or Wiley libraries). In this work, standards and samples were introduced in the Direct-EI interface by ow injection analysis; this fast and simple approach was chosen because of the lack of commercially availability of nano-HPLC columns packed with a stationary phase suitable for DEG. However, the analysis of seawater sample spiked with a known amount of DEG was tempted, using a C18 nano-column, with the aim of investigating the column capability in reducing sample interferences. These experiments demonstrated that the chromatographic separation was time consuming and not as eective as sample clean up procedure. Prior to nano-FIA-Direct-EI-MS analyses, a SPE preconcentration step was performed, due to the extreme dilution of the real samples. During preliminary experiments, several SPE packing materials were tested. The polymer packing of ISOLUTE ENV+ (polystyrenehydroxydivinylbenzene) gave the best results. Recovery was calculated on 100 ml of a blank seawater sample spiked with 100 ll of a 2000 mg/l DEG standard solution. The fortied water sample was subjected to the SPE procedure and the nal extract was analyzed by nano-FIA-Direct-EI, as reported in the experimental section. The percentage of recovery, calculated on the average of ve replicates, was of 65 16.3%. 3.2. Method validation The method was validated by the following parameters: specicity, sensitivity, linearity and precision. Fig. 1 (top) shows the recorded spectrum of a 100 mg/l DEG standard solution. The acquired spectrum was compared with those in the Wiley library, and the matching quality value obtained was of 83% (Fig. 1, bottom). Based on the recorded mass spectrum, in our experiments the specicity needed for compound identication was reached by SIM analysis on the three characteristic ions at m/z 45, 75 and 76. Sensitivity was evaluated by determining the limit of detection (LOD) of the method. This is expressed as the minimum concentration of the analyte that can be detected in the actual sample; it takes into account the initial volume of the sample (100 ml), the nal volume of the SPE extract

(200 ll), the percentage of recovery of the compound (65 16.3%), the volume that is introduced into the HPLC/MS system (60 nl), and the instrumental LOD. The instrumental limit of detection is dened as the minimum detectable amount of analyte capable of generating a peak with a signal to noise ratio of 3. LOD was determined empirically in SIM mode, injecting diluted solutions of the standard. The obtained data showed that the nano-FIA-DirectEI-MS system was capable of detecting a DEG minimum amount of 600 pg. The method LOD was of 31 lg/l. Calibration experiments permitted to asses the linearity of the system for more than three orders of magnitude starting from the instrumental LOD of 600 pg (Table 1). The precision of the method was expressed as %RSD of the peak areas and was evaluated both intraday and interday. Intraday precision was evaluated by measuring 11 replicates of a 100 mg/l DEG standard solution. Interday precision was calculated injecting the same solution over one working week. The results are reported in Table 1, demonstrating an excellent repeatability.

3.3. Evaluation of the matrix eect and comparison of the performances between the Direct-EI-MS and the ESI-MS/MS detection The chemical and physical properties of DEG make it very suitable for ESI-MS analysis: this compound is readily ionized in the positive ion mode as [M+Na]+ (m/z 129). As reported previously, a FIA-ESI-MS/MS method for the detection and quantication of DEG was developed with the aim of comparing the performances of the ESI-MS/ MS detection with those of the Direct-EI-MS. Specicity, sensitivity, linearity and repeatability were carefully evaluated, as already described for the nano-FIA-DirectEI-MS method validation. The specicity needed for compound identication was reached performing SRM experiments. The collisioninduced dissociation of the sodiated molecular ion was studied to establish the most sensitive and selective transitions to be used for DEG identication and quantication in real samples. The fragmentation pattern of the sodium adduct ion was characterized by two main losses of 18 Da (H2O) and 32 Da (CH3OH), yielding to the ion

Fig. 1. Comparison between the recorded mass spectrum (top) and Wiley library (bottom) mass spectra (5 ng injected) of DEG.

558 Table 1 Method validation data Instrumental LOD (pg) Nano-FIA-DirectEI-MS FIA-ESI-MS/MS 600 1250

A. Cappiello et al. / Chemosphere 69 (2007) 554560

Method LOD 31 lg/l n.d.

Linear regression equation Y = 1009.6X 42005 Y = 28105X 28955

R2 0.9974 0.9997

Intraday repeatability (%RSD) 5 7

Interday repeatability (%RSD) 9 11

at m/z 111 and m/z 97. These transitions were chosen for the quantitative purposes of SRM experiments. As reported in Table 1, the method was linear for two orders of magnitude starting from the instrumental LOD of 1250 pg. In addition both intraday and interday repeatability, calculated injecting ve replicates of a 1 mg/l DEG standard solution, showed satisfactory results. However, the comparison between the ESI-MS/MS and the DirectEI-MS instrumental LODs demonstrated that a lower sensitivity was achieved exploiting the ESI-MS/MS detection (Table 1). This result was attributed to the occurrence of various molecular adducts attributed to the mobile phase, which can reduce the sensitivity of ESI for low molecular compounds such as DEG. As reported in the introduction, another drawback associated with ESI-MS is represented by matrix-related signal suppression or enhancement eects. These phenomena could cause serious discrepancies between signal responses obtained from standard and matrix samples, resulting in diculties in both qualitative and quantitative characterization. In this work, matrix eect was evaluated following the procedure proposed by Matuszewsky et al. (2003). These authors suggest to express matrix eect as the ratio of the mean peak area of an analyte spiked postextraction, to the mean peak area of the same analyte standard: a value of >100% indicates ionization enhancement, and a value of <100% indicates ionization suppression. Following this scheme, a seawater sample free of DEG was extracted using the SPE procedure. The nal extract was spiked in order to have a nal DEG concentration of 100 mg/l (total injected amount of 5 ng), for the nanoFIA-Direct-EI-MS detection, and of 1 mg/l (total injected amount of 5 ng) for the FIA-ESI-MS/MS analysis. Matrix eect was calculated by measuring ve replicates as follows: ME % A=B 100 1

not possible to calculate the method LOD using the FIAESI-MS/MS approach (Table 1). 3.4. Analysis of real samples Direct-EI-MS detection allowed the identication of DEG in the actual samples on the basis of its characteristic ion signals (Fig. 1). The concentration of this analyte was calculated by comparing peak areas to those obtained from the calibration curve, and correcting them according to the percentage of recovery. Considering the high simplicity of the FIA diagrams, the peak areas were calculated from the TIC proles in order to prevent unexpected variations of the single ion currents due to instrumental uctuations. The results obtained are briey summarized in Table 2. DEG was found in all the investigated PFW samples (samples 1, 8, 15, 22), thus demonstrating a reduced eciency of

where A indicates the mean peak area for the extract spiked after the SPE procedure, and B corresponds to the mean peak area obtained for a standard solution having the same concentration of the extract. Our results demonstrate that DEG detection in seawater samples was aected by a strong signal suppression using the FIA-ESI-MS/MS approach (ME = 16%) (Fig. 2a). On the contrary, the Direct-EI-MS detection allows the determination of DEG in the investigated extracts without any interference from the other components present in the sample (ME = 105%) (Fig. 2b). Due to the strong matrix eect it was

Fig. 2. Total ion current diagrams for the evaluation of the matrix eect: (a) FIA-ESI-MS/MS; (b) FIA-Direct-EI-MS. Injections: (1) 1 mg/l spiked seawater extract; (2) 1 mg/l DEG standard solution; (3) 100 mg/l spiked seawater extract; (4) 100 mg/l DEG standard solution.

A. Cappiello et al. / Chemosphere 69 (2007) 554560 Table 2 DEG concentration in the investigated samples Sample number Sample description Sampling points Discharge length (m) A Platform 1 2 3 4 5 6 7 B Platform 8 9 10 11 12 13 14 C Platform 15 16 17 18 19 20 21 D Platform 22 23 24 25 26 27 28 PFW Seawater Seawater Seawater Seawater Seawater Seawater On platform Under diusor 5 10 15 20 25 Depth (m) On platform 0 0 0 0 0 0 1.56 0.730 0.490 0.350 0.360 0.390 0.300

559

DEG concentration (mg/l)

PFW Seawater Seawater Seawater Seawater Seawater Seawater

On platform Under diusor 5 10 15 20 25

On platform 0 0 0 0 0 0

2.04 0.640 0.270 0.390 0.620 0.600 0.590

PFW Seawater Seawater Seawater Seawater Seawater Seawater

On platform Under diusor 5 10 15 20 25

On platform 0 0 0 0 0 0

9.60 1.60 1.64 1.40 1.82 1.52 1.57

PFW Seawater Seawater Seawater Seawater Seawater Seawater

On platform Under diusor 5 10 15 20 25

On platform 0 0 0 0 0 0

13.0 1.33 2.31 2.16 1.51 1.26 1.72

the DEG gas platforms regeneration devices. The analyses on the seawater samples, collected at dierent distance from the PFW outlets, showed that DEG can be found as far as 25 m distance from the PFW diuser. However, the dispersion of this analyte around the investigated oshore platforms was quite dierent. For example, the results regarding the sampling area around the platforms A and B revealed the highest DEG concentrations in the sampling stations located under the water discharge pipes (samples 2 and 9). On the other hand, the nal results relative to the sampling area around the platforms C and D revealed that moderately high levels were found, respectively, at a 15 and 5 m distance from the platforms (sample 19 and 24). These data clearly suggest that several factors, depending from both hydrographic proles and seawater chemicalphysical properties, can inuence the dispersion of DEG in the marine environment. In addition, the DEG determination in seawater samples permitted to observe the presence of a PFW component as far as 25 m from the diusor; these results indicate that when PFW is discharged into the sea, its plume dissipates in some meters.

These DEG determinations together with the information about physicalchemical parameters of water column have supplied the necessary data to a dispersion model to study the PFW discharges (Manfra, 2006). 4. Conclusions From the results obtained, it can be concluded that the combination between an o-line SPE with the nano-FIADirect-EI detection proved to be an ecient, precise, rapid and sensitive approach for the detection of DEG in seawater samples. The method fullled analytical validation criteria: sensitivity and linearity were relevant to environmental water analysis, and a precise and accurate measurement of DEG in actual samples was achieved. Speed of analysis is another important point of strength, which can address a future application of the method for routine analyses. In fact, the specicity achieved in SIM allowed to use FIA as sample introduction method, avoiding a timeconsuming chromatographic separation. In addition, the Direct-EI-MS detection allowed to overcome the matrix

560

A. Cappiello et al. / Chemosphere 69 (2007) 554560 Holdway, D.A., 2002. The acute and chronic eects of wastes associated with oshore oil and gas production on temperature and tropical marine ecological processes. Mar. Pollut. Bull. 44, 185203. Katz, D.L., Lee, R.L., 1990. Natural Gas Engineering-Production and Storage. McGraw-Hill, New York, pp. 617626. Lewis, R.J., 1993. Hawleys Condensed Chemical Dictionary. Van Nostrand Reinhold Company, New York, 391 p. Manfra, L., 2006. Dispersione in mare delle acque di produzione e valutazione ecotossicologica degli eetti indotti. Ph.D. Thesis, Uni` versita degli Studi di Napoli Federico II, 137 p. Mariani, L., Manfra, L., Maggi, C., Savorelli, F., Di Mento, R., Cicero, A.M., 2004. Produced formation waters: a preliminary study on chemical characterization and acute toxicity by using sh larve Dicentrarchus labrax L., 1758. Fresen. Environ. Bull. 13 (12), 1427 1432. Matuszewsky, B.K., Constanzer, M.L., Chavez-Eng, C.M., 1998. Matrix eect in quantitative LC/MS/MS analyses of biological uids: a method for determination of nasteride in human plasma at pictogram per millilitre concentrations. Anal. Chem. 70, 882889. Matuszewsky, B.K., Constanzer, M.L., Chavez-Eng, C.M., 2003. Strategies for the assessment of matrix eect in quantitative bioanaytical methods based on HPLC-MS/MS. Anal. Chem. 75, 30193030. Niessen, W.M.A., Manini, P., Andreoli, R., 2006. Matrix eects in quantitative pesticide analysis using liquid chromatographymass spectrometry. Mass Spectrom. Rev. 25, 881899. Nyholm, N., Damborg, A., Lindgaard-Joergensen, P., 1992. A comparative study of test methods for assessment of the biodegradability of chemicals in seawater screening tests and simulation tests. Ecotox. Environ. Safe. 23 (2), 173190. Osterberg, R.E., See, N.A., 2003. Toxicity of excipients a food and drug administration perspective. Int. J. Toxicol. 22 (5), 377380. Peters, F.T., Paul, L.D., Kraemer, T., 2001. Validated gas chromatographicmass spectrometric assay for determination of the antifreezes ethylene glycol and diethylene glycol in human plasma after microwave-assisted pivalylation. J. Chromatogr. B 754 (2), 401409. Reemtsma, T., 2001. The use of liquid chromatography-atmospheric pressure ionization-mass spectrometry in water analysis Part II: Obstacles. Trends Anal. Chem. 20 (10), 533542. Richardson, M., 1996. Environmental Xenobiotics. Taylor and Francis Publishers, London. RIVM Report 601501016/2005. Verbruggen, E.M.J., Traas, T.P., Fleuren, R.H.L.J., Ciarelli, S., Posthumus, R., Vos, J.H., Scheepmaker, J.W.A., van Vlaardingen, P.L.A., 2005. Environmental Risk Limits for alcohols, glycols, and some other relatively soluble and/or volatile compounds 1. Ecotoxicological evaluation. Savchuk, S.A., Kolesov, G.M., 2005. Chromatographic techniques in the quality control of cognacs and cognac spirits. J. Anal. Chem. 60 (8), 752771. Sorensen, J.A., Gallagher, J.R., Hawthorne, S.B., Aulich, T.R., 2000. AAD Document Control, National Energy Technology Laboratory (US Department of Energy); JV TASK3 Gas Industry Groundwater Research Program Final Report AAD. Szymansky, A., Wyrwas, B., Szymanowska, M., Lukaszawsky, Z., 2001. Determination of short-chained poly(ethylene glycols) and ethylene glycol in environmental samples. Water Res. 35 (15), 35993604. Williams, R.H., Shah, S.M., Maggiore, J.A., Erickson, T.B., 2000. Simultaneous detection and Quantitation of diethylene glycol, ethylene glycol, and the toxic alcohols in serum using capillary column gas chromatography. J. Anal. Toxicol. 24 (7), 621626.

eect signal suppression, a limitation observed with the ESI-MS/MS, allowing the determination of DEG without any interference from other components present in the sample. The DEG determination in seawater samples showed that, when PFW was discharged into the sea, its plume dissipated in some meters. These data together with informations regarding water column physicalchemical parameters will be used in future publications for the development of a PFW discharge dispersion model. References
Ballantyne, B., Snellings, W.M., 2005. Developmental toxicity study with diethylene glycol dosed by gavage to CD rats and CD-1 mice. Food Chem. Toxicol. 43, 16371646. Bensoam, J., Cicolella, A., Durjardin, R., 1999. Improved extraction of glycerol ethers from water by solid-phase microextraction by carboxen polydimethylsiloxane-coated ber. Chromatographia 50 (3/4), 155 159. Berloni, A., Cappiello, A., Famiglini, G., Palma, P., 1994. Generation of split-ow micro-gradients for capillary HPLC. Chromatographia 39, 279284. Brendehaug, J., Johnsen, S., Bryne, K.H., Gjose, A.L., Eide, T.H., Aamot, E., 1992. Toxicity testing and chemical characterization of produced water a preliminary study. In: Ray, J.P., Engelhardt, F.R. (Eds.), Produced Water: Tecnological/Environmental Issues and Solutions. Plenum Press, New York, pp. 245252. Cappiello, A., Famiglini, G., Mangani, F., Palma, P., 2001. New trends in the application of electron ionization to liquid chromatographymass spectrometry interfacing. Mass Spectrom. Rev. 20, 88104. Cappiello, A., Famiglini, G., Mangani, F., Palma, P., 2002. A simple approach for coupling liquid chromatography and electron ionization mass spectrometry. J. Am. Soc. Mass Spectr. 13, 265273. Cappiello, A., Famiglini, G., Palma, P., 2003. Electron ionization for LC/ MS. Anal. Chem. 75, 497A503A. Choi, B.K., Hercules, D.M., Gusev, A.I., 2001. LC-MS/MS signal suppression eects in the analysis of pesticides in complex environmental matrices. Fresen. J. Anal. Chem. 369 (34), 370377. Drugov, Yu.S., Muraveva, G.V., 1995. Gas-chromatographic determination of diethylene glycol in ambient air. J. Anal. Chem. 50 (9), 899 902. Ferrari, L.A., Giannuzzi, L., 2005. Clinical parameters, postmortem analysis and estimation of lethal dose in victims of a massive intoxication with diethylene glycol. Forensic Sci. Int. 153, 4551. Gembus, V., Goulle, J., Lacroix, C., 2002. Determination of glycols in biological specimens by gas chromatographymass spectrometry. J. Anal. Toxicol. 26 (5), 280285. Henderson, S.B., Grigson, S.J.W., Johnson, P., Roddie, B.D., 1999. Potential impact of production chemicals on the toxicity of produced water discharges from north sea oil platforms. Mar. Pollut. Bull. 12, 11411151. Holcapek, M., Virelizier, H., Chamot-Rooke, J., Jandera, P., Moulin, C., 1999. Trace determination of glycols by HPLC with UV and electrospray ionization mass spectrometric detection. Anal. Chem. 71 (13), 22882293.