Multi-contaminant analysis of organophosphate and halogenated flame retardants in food matrices using ultrasonication and vacuum assisted extraction, multi-stage cleanup and gas chromatography–mass spectrometry

Highlights

• First specific method for multi-class flame retardants in food matrices.

• New extraction technique based on QuEChERS with a high extraction yield.

• Multi-step clean-up for removal of lipids and pigments from food samples.

• Method LOQs (wet weight) were in the pg/g range for BFRs and ng/g range for PFRs.

Abstract

A multi-residue analytical method was developed for the determination of a range of flame retardants (FRs), including polybrominated diphenyl ethers (PBDEs), emerging halogenated FRs (EFRs) and organophosphate FRs (PFRs), in food matrices. An ultrasonication and vacuum assisted extraction (UVAE), followed by a multi-stage clean-up procedure, enabled the removal of up to 1 g of lipid from 2.5 g of freeze–dried food samples and significantly reduce matrix effects. UVAE achieves a waste factor (WF) of about 10%, while the WFs of classical QuEChERS methods range usually between 50 and 90%. The low WF of UVAE leads to a dramatic improvement in the sensitivity along with saving up to 90% of spiking (internal) standards. Moreover, a two-stage clean-up on Florisil and aminopropyl silica was introduced after UVAE, for an efficient removal of pigments and residual lipids, which led to cleaner extracts than normally achieved by dispersive solid phase extraction (d-SPE). In this way, the extracts could be concentrated to low volumes, e.g. <100 μL and the equivalent matrix concentrations were up to 100 g ww/mL. The final analysis of PFRs was performed on GC–EI-MS, while PBDEs and EFRs were measured by GC–ECNI-MS. Validation tests were performed with three food matrices (lean beef, whole chicken egg and salmon filet), obtaining acceptable recoveries (66–135%) with good repeatability (RSD 1–24%, mean 7%). Method LOQs ranged between 0.008 and 0.04 ng/g dw for PBDEs, between 0.08 and 0.20 ng/g dw for EFRs, and between 1.4 and 3.6 ng/g dw for PFRs. The method was further applied to eight types of food samples (including meat, eggs, fish, and seafood) with lipid contents ranging from 0.1 to 22%. Various FRs were detected above MLOQ levels, demonstrating the wide-range applicability of our method. To the best of our knowledge, this is the first method reported for simultaneous analysis of brominated and organophosphate FRs in food matrices.

Introduction

Flame retardants (FRs), such as polybrominated diphenyl ethers (PBDEs), emerging halogenated flame retardants (EFRs) and organophosphate flame retardants (PFRs), have been used as additives in furniture, electronics, foams, building materials, textiles, etc to reduce the risk of fire spreading [1], [2], [3], [4]. In the past decade, the widely used PBDEs were gradually banned worldwide, due to increasing evidence of their toxicity, persistency and other health and environment concerns [1], [5], [6]. These bans and restrictions have led to the increased usage of EFRs and PFRs as alternatives [1], [2], [7]. As representative EFRs, dechlorane plus (DP), 1,2-bis(2,4,6-tribromophenoxy)ethane (BTBPE), and decabromodiphenyl ethane (DBDPE) are used as replacements of Octa- and Deca-BDE, respectively, while 2-ethylhexyl-2,3,4,5-tetrabromobenzoate (EH-TBB) and bis(2-ethylhexyl)-3,4,5,6-tetrabromo-phthalate (BEH-TEBP) were used as alternatives for Penta-BDE [1], [3]. PFRs, such as tris(1,3-dichloropropyl) phosphate (TDCIPP), triphenyl phosphate (TPHP) and tris(1-chloro-2-propyl) phosphate (TCIPP) are also used as substitutes for different PBDEs. Some PFRs, like TPHP and tris-(butoxyethyl)-phosphate (TBOEP), are also used as plasticizers [2].

FRs have been ubiquitously found in the environment, such as in air, dust, soil and sediment [1], [2], [6], [7], [8], [9]. Some studies have detected EFRs in biota and food matrices [10], [11], [12]. Lipids and food pigments are two major interferences, which could introduce high signal background, lead to bad shape of the chromatographic peaks or even damage the column if the extract is not properly cleaned. PFRs and some EFRs, such as EH-TBB and BEH-TEBP, have seldom been investigated in biota or food matrices due to their non-persistent chemical character and the complex matrix effect from food samples. PFRs, BEH-TEBP and EH-TBB cannot resist strong acid or base clean-up, which is a traditional method for purifying persistent organic pollutants (POPs). Zheng et al. [10] used gel permeation chromatography (GPC) to remove lipids from chicken egg samples for FRs analysis, but this method was lengthy and consumed large amounts of solvents (usually >100 mL), while a further clean-up step was still required.

Diet is an important pathway of human exposure to contaminants [13]. Contamination of food may result not only from the environment, but also during food processing or storage. Many studies have reported the presence of PBDEs in food [10], [14], [15], but very few studies have investigated the presence of PFRs in food, such as eggs, or in aquatic biota [15], [16], [17], [18]. Recently, PFRs were even detected in human breast milk in Asian mothers [19]. Labunska et al. [14] recently analyzed food samples from Chinese e-waste facilities and found EH-TBB and BEH-TEBP in the food produced near these e-waste sites. These studies raised the serious concern of EFRs and PFRs as food contaminants, which may further be linked to issues related to public health. Since PFRs have similar physical-chemical properties with PBDEs [2] and thus will tend to be present in the same samples, it is necessary to develop an universal method for several classes of FRs, including EFRs and PFRs, in food.

Recently, QuEChERS (quick, easy, cheap, effective, rugged and safe) methods have been reported for the analysis of organic pollutant residues in food matrices, due to their simplicity and good lipid removal [20], [21], [22], [23]. Such method typically uses acetonitrile to minimize co-extraction of lipids. However, it requires a further clean-up step to ensure lipid removal, which otherwise could significantly affect the chromatography. Dispersive solid-phase extraction (d-SPE) is usually applied, using sorbent like Z-SEP, PSA or C18 [21], [24]. Although a large amount of sample (usually 10 g) is processed by a QuEChERS method, only a small fraction is actually present in the final extract (equivalent of 0.5–1 g, waste factor (WF) > 90%) [21]. In this case, the amounts of lipids, pigments and other interfering compounds in extract can be removed by a single simple clean-up step. However, the method's sensitivity has to be sacrificed. Some studies [20], [21] reported the analysis of FRs and other organic pollutants in food using QuEChERS methods, but neither have they achieved a good clean-up for samples rich in lipids and pigments, nor has one specifically focused on the analysis of EFRs and PFRs in food matrices.

To fill this research gap, we have developed a new method using ultrasonication and vacuum assisted extraction (UVAE) followed by a further clean-up and fractionation procedure, for the analysis of different groups of FRs in lipid- and pigment-rich foods. To the best of our knowledge, this method achieves an efficient removal of lipids and other interferences, while reaching a better sensitivity and selectivity of FRs than other published methods. It is also the first study on analyzing PFRs in different food matrices, other than fish.