Article pubs.acs.org/JAFC
Alternative Eluent Composition for LC-MS Analysis of Perfluoroalkyl Acids in Raw Fish Samples Tõiv Haljasorg,† Jaan Saame,† Karin Kipper,*,† Anu Teearu,†,‡ Koit Herodes,† Mari Reinik,‡ and Ivo Leito† †
Institute of Chemistry, University of Tartu, 14a Ravila Street, 50411 Tartu, Estonia Estonian Health Board, Tartu Laboratory, PK 272, 50002 Tartu, Estonia
‡
ABSTRACT: A wide range of anthropogenic pollutants that possess serious environmental and health risks are known. One type of these harmful substances that have become a focus of interest during the past decade are perfluoroalkyl acids (PFAAs), which are extensively used in industry for different purposes. Due to the harmful effects that these compounds might cause in living organisms, EFSA and EU CONTAM panel have issued a monitoring program for PFAAs in foodstuffs. This has given rise to intense research dedicated to the analysis of PFAAs over the past few years. This work focuses on chromatographic analysis of three PFAAs in fish. The analytes, perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), and perfluorooctanesulfonic acid (PFOS), are commonly associated with the production of fluoropolymers. Fluorinated alcohols are used as eluent components, and their possible advantages as eluent modifiers in LC-MS analysis of PFAAs, alternative retention mechanism and enhanced ionization efficiency, are examined. The analyzed fish samples originating from Estonian fresh and marine waters had low contents of PFAAs. KEYWORDS: perfluoroalkyl-substituted acids, PFOA, PFOS, PFNA, LC-MS, food contaminants
■
INTRODUCTION Perfluorinated compounds (PFCs) are known for their high thermal and chemical stabilities and also for their water and oil repellent nature, which make them valuable for many industrial and commercial purposes. For example, polytetrafluoroethylene (PTFE)-coated nonstick cookware, stain- and soil-resistant upholstery and textiles, and waterproof apparel (Gore-Tex). Some of these compounds, notably the perfluoroalkyl acids (PFAAs), are used as surfactants, for example, perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS), during the production of fluoropolymers, which in turn are used as resistant coatings on different surfaces. At the same time some of the PFAAs are degradation products of the aforementioned polymers.1 As a result of their high chemical stability, the PFAAs tend to accumulate in the environment and living organisms. Many PFAAs are hepatotoxic2 and carcinogenic.3,4 It is therefore imperative to monitor their levels in the environment and food as well as the exposure of living organisms to PFAAs. There is currently no legislation in force limiting the levels of perfluorinated organic substances such as PFOS or PFOA in food or feed within the European Union. In 2008, The Panel of Contaminants in the Food Chain (CONTAM) of the European Food Safety Agency (EFSA) performed a risk assessment of PFOS and PFOA and established a tolerable daily intake (TDI) of 150 ng/kg body weight for PFOS and 1500 ng/kg body weight for PFOA.5 The occurrence of PFCs in food and dietary exposure was assessed in 2012 by EFSA. On the basis of the available data, exceedance of TDIs was found to be highly unlikely.1 Numerous analytical methods have been developed for the quantitation of PFAAs in the environment and food with a focus on liquid chromatography coupled with tandem mass © 2014 American Chemical Society
spectrometry (LC-MS/MS) or gas chromatography−mass spectrometry (GC-MS) methods.6,7 GC-based methods require analyte derivatization;8 therefore, LC-based methods are simpler and usually preferred. Different sample preparation techniques have been used, such as pressurized liquid extraction9,10 and solid-phase extraction (SPE) both offline and inline11,12 with numerous solvents and additional sample concentrating methods. Although LC-MS/MS is one of the most selective techniques, the LC separation is still important to achieve accurate results, because electrospray ionization is prone to matrix effects; components of a sample that are coeluted with the analyte may affect the ionization process of compounds of interest.13 Therefore, it would be ideal if the signal of an individual analyte would not be affected by coeluting compound. PFAAs, both carboxylic and sulfonic, are strong acids and consequently are anionic at any realistic mobile phase pH. PFAAs are hydrophobic and lipophobic. The interaction of the perfluoroalkyl groups with reversed phase stationary phases is not well understood, and the retention mechanism of these acids is unclear. They also undergo extensive aggregation even at very low concentration,14,15 leading to micelle formation at about a concentration of 10−3 M.16 In many cases separation of linear PFAAs with the same perfluoroalkyl chain length has been problematic.10,17−20 In particular, satisfactory liquid chromatographic separation of PFOS and perfluorononanoic acid (PFNA) (both have C8 Received: Revised: Accepted: Published: 5259
February 11, 2014 May 20, 2014 May 20, 2014 May 20, 2014 dx.doi.org/10.1021/jf5007243 | J. Agric. Food Chem. 2014, 62, 5259−5268
Journal of Agricultural and Food Chemistry
Article
The focus of this work is the LC-ESI-MS analysis of PFOA, PFNA, and PFOS in raw fish samples using a different approach for chromatographic separation and signal improvement: replacement of common eluent buffers (ammonium acetate) with a polyfluorinated alcohol−ammonia buffer.34,36
perfluoroalkyl chains) has not always been achieved. For difficult separations unusual stationary phases or mobile phase components have been used and in many cases successfully. Using fluorinated stationary phases21−23 or methylpiperidine as eluent modifier,24 separation of PFNA and PFOS has been achieved successfully. In addition, a number of publications show the separation of PFNA and PFOS with C18 stationary phases from different manufacturers; however, variations in eluent compositions and stationary phases make it difficult to systematically analyze the retention mechanisms of those compounds on the basis of literature data.25,26 From a mass spectrometric point of view PFOS displays some other inconvenient aspects when compared with perfluorinated carboxylic acids of similar chain length. In the MS/MS regime it has a relatively low fragmentation yield and, as a consequence, a low S/N ratio. In the case of PFOS an MRM “transition” 499 → 499 (i.e., monitoring the nonfragmented parent) has been proposed.27 However, the problem of accuracy remains, because in this case it has also been shown that certain steroidic bile acids present in fish samples have the same nominal m/z ratio of 499 with a similar fragmentation pattern and a retention time close to that of the PFOS and therefore may strongly affect the peak area of PFOS.22,27 The selection of eluent additives in LC-MS is limited by the volatility requirement. Fluorinated alcohols are volatile compounds of weak to medium acidity and are neutral in protonated form. These properties make them a potentially promising class of compounds to be used as weak acids for preparing buffers of pH value above 7 for use in LC-MS. 1,1,1,3,3,3-Hexafluoro-2-propanol (HFIP, pKa = 9.3)24 has been used as an additive to the LC mobile phase in several studies.28−35 The suitability of fluoroalcohols, HFIP and 1,1,1,3,3,3-hexafluoro-2-methyl-2-propanol (HFTB, pKa = 9.6)36, as buffer additives at concentration levels from 1 to 10 mM has been studied, and application for separation of antibiotics−fluoroquinolones and sulfonamides is demonstrated.37,38 It turned out that mobile phases with perfluorinated alcohols provide retention mechanisms that differ from the ones traditionally observed with regular eluent modifiers for C18 stationary phases.38 Alternatively, a promising group of stationary phases that have been successfully applied in pharmaceutical and drug analysis are fluorinated (pentafluorophenyl and pentafluorophenylpropyl) stationary phases.39−41 Fluorinated stationary phases are associated with various interaction mechanisms improving retention and chromatographic resolution. Most similar to common stationary phases are perfluorinated alkyl chains, for example, C8F17, used for their alternative retention in several applications for separating polar molecules,42,43 especially halogenated analytes,44−47 but also for aromatic polycyclic hydrocarbons due to the interactions between πelectrons of analytes and C−F dipole of the stationary phase.46,48 The performance of fluorinated columns has improved in recent years with respect to their pH stability and column lifetime. Additionally, stationary phase bleeding is reduced, allowing the use of these stationary phases also in MS applications. The retention of all analytes has been reported to be overall lower on the fluorinated stationary phase than on C18.47 Therefore, the retention of PFCs using a fluorinated stationary phase and fluoroalcohols with a regular C 18 stationary phase is of interest.
■
MATERIALS AND METHODS
Instrumentation for Sample Preparation and Analysis. For primary sample homogenization a Retsch Grindomix GM 200 homogenizer was used. A Heidolph 900 DIAX homogenizer was used during the PFC extraction step, and a centrifuge Eppendorf 5430R was used during sample preparation. Chromatographic separation of the analytes was carried out on a Waters 2695 series chromatograph (Waters, Milford, MA, USA) equipped with a binary pump, degasser, autosampler, and column thermostat. Detection and quantification of analytes was carried out with a QuattroMicro triplequadrupole mass spectrometer (Micromass, Manchester, UK). Tested analytical columns were Epic FO LB (150 mm × 3.0 mm, 3 μm) and YMC Triart C18 columns (150 mm × 3 mm, 3 μm). Preliminary tests for chromatographic separation were performed with an Agilent 1290 series UHPLC (Agilent) and a J-320 mass spectrometer (Varian). Chemicals. PFOA, PFNA, potassium perfluorooctanesulfonate standards, HFIP, HFTB, and anhydrous MgSO4 were purchased from Sigma (St. Louis, MO, USA). Acetonitrile and methanol were obtained from J. T. Baker (Deventer, The Netherlands), formic acid and ammonium hydroxide were from Riedel-de-Haën, and ammonium acetate was from Fluka (Buchs, Germany). All solvents were of reagent grade or higher quality. Water was purified in-house using a Milli-Q Advantage A10 system from Millipore (Bedford, MA, USA). Standard and Buffer Solutions. Calibration solutions were prepared in methanol so that the concentrations were in the range of 0.5−500 ng/mL. Spiking solutions were prepared in acetonitrile at concentrations of 7.5, 75, and 750 ng/mL; in that case the analyte concentrations in samples were 1.0, 10, and 100 μg/kg, respectively. Solvent A (see below) was prepared by dissolving 0.965 g of HFTB in 1 L of Milli-Q water, and the pH was adjusted to 10.0 by adding a concentrated solution of ammonium hydroxide. Sample Preparation. The fish samples were purchased from local markets and delivered to the laboratory on the same day for homogenization. Samples included the most common fish species available on the Estonian market (Table 1). The initial amount of sample was typically >1 kg. Upon arrival, the fish samples were cleaned of skin, intestines, and bones. Edible parts of fish (muscle and fat tissue) were homogenized with a Retsch Grindomix GM 200 at 2000 rpm for 4 min. Homogeneous samples were stored in a freezer at −20 °C until analysis. Further sample preparation was carried out using a modified “Quick, Easy, Cheap, Effective, Rugged, and Safe” protocol (QuEChERS), essentially as described in ref 49. As a modification, the cleanup step with dispersive SPE was omitted, because there was no need for it due to the already low matrix effects. In addition, the use of HFTB/HFIP increased analyte signal intensity and reduced matrix effects. Homogenized fish tissue was weighed (7.5 g) into a 50 mL polypropylene (PP) tube, and 10 mL of water was added. The sample was then additionally homogenized during 1 min with a Heidolph Diax 900 homogenizer at 18000 rpm. Fifteen milliliters of acetonitrile was added to the tubes, and the tubes were shaken by hand for 10 s. A preweighed mixture of salts (6 g of anhydrous MgSO4 and 1.5 g of NaCl) was then added to the samples, and the tubes were shaken vigorously by hand for 1 min and for 1 min on a vortex-mixer. Samples were then centrifuged at 7000 rpm for 15 min. Twelve milliliters of clear supernatant was transferred into a 15 mL PP tube containing 1.8 g of anhydrous MgSO4. Tubes were shaken once more for 1 min by hand and on a vortex mixer and then centrifuged at 7000 rpm for 15 min. Eight milliliters of supernatant was transferred into a 15 mL glass vial and evaporated to dryness under a stream of nitrogen at 40 °C. The dry residue was dissolved in 1 mL of methanol. The final extract was filtered through a 0.45 μm regenerated cellulose syringe filter into a chromatographic vial. 5260
dx.doi.org/10.1021/jf5007243 | J. Agric. Food Chem. 2014, 62, 5259−5268
Journal of Agricultural and Food Chemistry
Article
Table 1. Analyzed Fish Samples content of PFAAs (μg/kg) sample
scientific name
pike ide European perch
Esox lucius Leuciscus Idus Perca fluviatilis
Baltic herring European whitefish vimba bream common bream European flounder European plaice Atlantic salmon
Clupea harengus membras Coregonus lavaretus Vimba vimba Abramis brama Platichthys f lesus Pleuronectes platessa Salmo salar
pike-perch
Sander lucioperca
common carp rainbow trout
Cyprinus carpio Oncorhynchus mykiss
Crucian carp African sharptooth catfish European sprat
Carassius carassius Clarias gariepinus Sprattus sprattus
place of origin
PFOA
PFNA
PFOS
river Emajõgi lake Lämmijärv Baltic Sea lake Võrtsjärv Baltic Sea Baltic Sea Baltic Sea river Emajõgi Baltic Sea Norway Norway Norway river Emajõgi lake Võrtsjärv Härjanurme fish farm (Estonia) Karilatsi fish farm (Estonia) Härjanurme fish farm (Estonia) Baltic sea Sõmerpalu fish farm (Estonia) Baltic Sea
