Article pubs.acs.org/est
Cite This: Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Identification of Photoinitiators, Including Novel Phosphine Oxides, and Their Transformation Products in Food Packaging Materials and Indoor Dust in Canada Runzeng Liu* and Scott A. Mabury
Environ. Sci. Technol. Downloaded from pubs.acs.org by UNIV OF LOUISIANA AT LAFAYETTE on 04/04/19. For personal use only.
Department of Chemistry, University of Toronto, 80 Saint George Street, Toronto, M5S 3H6, Ontario, Canada S Supporting Information *
ABSTRACT: Although photopolymerization is generally considered a green technology, the contamination of foodstuffs by photoinitiators (PIs), an essential component of photopolymerization systems, has recently attracted notice. Despite this interest, little attention has been paid to PI contamination in the environment. To date, only one study, performed in China, has reported the occurrence of PIs in the environment. In the present study, the occurrence of 25 PI additives with discrete molecular structures was investigated in food packaging materials and indoor dust. The PIs studied here include benzophenones (BZPs), thioxanthones (TXs), amine co-initiators (ACIs), and novel phosphine oxides (POs). Twenty-four PIs were detected in food packaging materials. Total concentrations of PIs (∑PIs) ranged between 122 and 44 113 ng/g, with a geometric mean (GM) of 3375 ng/g. The photodegradation of PIs in food packaging materials was investigated for the first time, and the halflives of PIs in these materials were found to range from 32 to 289 h. These 24 PIs were also detected in indoor dust samples (GM of ∑PIs = 1483 ng/g). The relative abundances of different PIs were found to vary between the packaging materials and the indoor dust, which is attributed in part to the different stabilities of different PIs under simulated sunlight. Using standards synthesized in our lab, four TX transformation products (GM: 34.8 ng/g) were also detected in indoor dust. The concentrations of the transformation products were higher than the concentrations of the parent chemicals in indoor dust. Thus, further studies exploring human exposure to TXs should include these transformation products to avoid underestimation. This is the first report of PIs and relevant transformation products in the indoor environment in North America.
■
laminobenzoate (EDMAB) were also present in foodstuffs.5,6 The contamination of foodstuffs was attributed to PI migration from UV-curable inks and varnishes, which are widely used in food packaging materials.6 Toxicity studies showed that 2-ITX had antiestrogenic and antiandrogenic properties in yeast-based estrogen and androgen bioassays. 7,8 Furthermore, 2-ethylhexyl-4(dimethylamino)benzoate (EHDAB) was reported to have estrogenic and androgenic activities using in vitro assays,9 and MK was found to be potentially carcinogenic.10 Because of human health concerns surrounding the presence of PIs in foodstuffs, techniques such as the application of an impermeable barrier between the foodstuffs and the packaging materials were implemented to prevent the migration of PIs.11 Furthermore, allowable amounts of PI residue in packaged food were legislated in Europe.12 For instance, a specific migration threshold of 0.05 mg/kg was set for 2-ITX.13
INTRODUCTION The last several decades have witnessed the rapid transfer of photopolymerization techniques from academic research to industrial applications.1 Indeed, photopolymerization, which is characterized by low energy requirements and temperatures of operation, and no release of volatile organic compounds, is among the most efficient technologies available for the rapid production of polymeric materials.1 Photopolymerization is mainly used in light-sensitive, UV-curable materials such as ink, coatings, and adhesives.2 Photoinitiators (PIs) are indispensable to photopolymerization systems. They are synthetic chemicals that are widely used to initiate polymerization during the optical absorption process by generating active substances such as free radicals.3 Reports of contamination of foodstuffs by PIs in recent years have challenged the perception of photopolymerization as a green technology. PIs first attracted the attention of the public when high concentrations of 2-isopropylthioxanthone (2-ITX) were found in infant formula and drinks in Europe.4 Subsequent studies showed that several other PI analogues such as benzophenone (BP), 4,4′-bis(dimethylamino)benzophenone (Michler’s ketone, MK), and ethyl-4-dimethy© XXXX American Chemical Society
Received: January 3, 2019 Revised: March 12, 2019 Accepted: March 26, 2019
A
DOI: 10.1021/acs.est.9b00045 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 1. Structures of the novel phosphine oxide photoinitiators.
samples were wrapped in aluminum foil and sealed in polypropylene bags to prevent cross-contamination. The packaging materials investigated in this study were taken from packages of dry foodstuffs including staple foods (e.g.; rice, noodles), dry vegetables, snacks (e.g., cookies, tarts, pie, and candy), tea, and sugar. The outside surfaces of the packaging materials were covered with colorful inks, but these surfaces do not contact the food directly. In addition to the cardboard packaging materials, inner plastic bags were used to package the foods. There is no direct contact between the foods and cardboard packaging materials. These food packaging materials collected from local markets are assumed to be representative of the packaging materials the local residents are bringing into their homes. Dust samples were collected from offices (n = 62, one sample per office) and rooms in residential houses (n = 30, 1−2 samples per house) located in Toronto between December 2017 and February 2018. Dust samples from each room were collected using a vacuum cleaner with cotton pads (4 × 4 in., 118 × 60 threads per square inch, VWR, Mississauga, Canada) inserted on the tube extender. More details on dust collection were described in our previous study.19 Field blanks were prepared by collecting Na2SO4 using the same procedure and vacuum cleaner. Photodegradation Experiments. In order to investigate the stability of PIs, samples of the food packaging material measuring 2 × 4 cm were put in the sunlight simulator (24 °C), and the exposure time was varied (0, 2, 6, 12, 24, 48, and 72 h). Control experiments were conducted by keeping food packaging materials in the dark, also at 24 °C, for the same amounts of time. The Suntest CPS sunlight simulator (Atlas, Mt Prospect, IL), which is equipped with a xenon lamp that emits light across the actinic spectrum (290−800 nm), was used to initiate the photodegradation of the food packaging materials. The sunlight simulator was operated with the internal cooling fan on and was chilled by a water/ethylene glycol recirculator (PolyScience, Niles, IL). The internal temperature of the chamber can be adjusted by setting a designed temperature of the water/ethylene glycol recirculator, which was linked to the bottom of the chamber. The internal temperature was set to 24 °C during the experiments. Spectroradiometry was conducted on the sunlight simulator to confirm its spectral characteristics and assess the intensity of the light to which the samples were subjected. The emission spectrum of the solar simulator was measured with a BlackComet C-50 spectroradiometer (Stellar Net Inc.). The measured irradiance was 533 W/m2. A first-order exponential decay model was used to describe the photodegradation of PIs in food packaging materials in the sunlight simulator, as shown in eq 1,
Despite these reports on the potential toxicity of PIs (e.g., endocrine disrupting effects and carcinogenicity),7−10 few studies have investigated PI contamination in environmental matrices. Environmental contamination caused by PIs has only been reported once, in China, where three kinds of PIs, including benzophenones (BZPs), thioxanthones (TXs), and amine co-initiators (ACIs), were repeatedly detected in indoor dust and sewage sludge at concentrations ranging from partper-billion to part-per-million levels.14 Phosphine oxides (POs), such as 2,4,6-trimethylbenzoyldiphenylphosphine oxide (TPO), are also widely used as PIs in photopolymerization, due to their high quantum yield of free radicals and ability to cure thick films.15 Recent toxicity studies have shown that the EC50 of TPO is 10 times lower than that of EDMAB, meaning that TPO is 10 times more cytotoxic than EDMAB, a well-known PI analogue in the ACI class.16 In a study testing the toxicity of resin-based composites, TPO was found to be significantly more cytotoxic and genotoxic than the other constituents of the composites, such as camphorquinone and N,N-dimethylaminoethyl methacrylate.17 Despite the widespread use of POs in various products and their reported potential for toxicity, their presence as contaminants in the environment has not been reported. Somewhat surprising, 18 PIs were detected in human sera from U.S. donors in our recent study,18 with concentrations of PIs from 423 to 2870 pg/mL, indicating significant human exposure to PIs in North America. Unfortunately, no data are available concerning human exposure to these chemicals in the indoor environment in North America. In the present study, the presence, concentration, and composition profiles of four kinds of PIs, including BZPs, TXs, ACIs, and POs, were investigated in the indoor environment. Samples, including food packaging materials and indoor dust were collected in Toronto, Canada. The stability of the detected PIs in the food packaging materials was investigated using a sunlight simulator, and the relevant transformation products were explored to provide a first impression of the environmental fates of these emerging organic contaminants.
■
MATERIALS AND METHODS Materials. The names, abbreviations, and structures of the analytes and other relevant information are shown in Figure 1 and in the Supporting Information (Table S1 and Figure S1). Native standards, isotopically labeled standards [benzophenone-d10 (BP-d10), 2-isopropyl-thioxanthone-d7 (2-ITX-d7), triphenyl phosphate-d15 (TPHP-d15)], and other materials used in this study are shown in the Supporting Information. Four transformation products of 2-ITX and 2,4-diethylthioxanthone (DETX) were synthesized in our lab. Their purities were demonstrated to be >95%. Details were shown in our previous study.18 Sample Collection. Cardboard food packaging samples (n = 33) were obtained from local markets in Toronto. All
Ct = C0 × e−kt
(1)
where C0 is the initial PI concentration in food packaging material, Ct is the remaining PI concentration at time t, t is the B
DOI: 10.1021/acs.est.9b00045 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology Table 1. Descriptive Statistics of the Measured PI Concentrations (ng/g) in Food Packaging Materials (n = 33)
average proportion (%) compounds
GM
Benzophenones (BZPs) BP 1571 benzil 7.97 PI-184 202 PI-651 114 4-MBP 32.9 MBB 158 PBZ 22.2 EAQ 1.15 ∑BZPs 2469 Amine Co-Initiators (ACIs) MK 36.6 MEK 85.6 EAB 2.12 EDMAB 58.4 EHDAB 5.42 PI-369 44.3 PI-907 23.9 DABP 0.62 DMAB 4.87 ∑ACIs 391 Thioxanthones (TXs) TX 0.46 2-ITX 70.9 DETX 18.7 2-Cl-TX
