Direct analysis in real time accurate mass spectrometry determination of bisphenol A in thermal printing paper

1 Direct analysis in real time accurate mass spectrometry determination of bisphenol A in thermal printing paper G. Castro, I. Rodríguez*, M. Ramil, R. Cela Department of Analytical Chemistry, Nutrition and Food Sciences. Institute for Research and Food Analysis (IIAA). Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain. *Corresponding author: Isaac Rodríguez e-mail: [email protected] Edición del artículo publicado en: Talanta 205 (2019) 120086 https://doi.org/10.1016/j.talanta.2019.06.086 2 Abstract Contact with thermal printing paper is a relevant source of dermal exposure to unbonded bisphenol A (BPA). In order to limit this exposure route, the European Union has introduced a drastic reduction in the maximum allowed concentration of BPA in thermal paper produced after beginning of year 2020. This study investigates the suitability of direct analysis in real time (DART), combined with accurate mass spectrometry, as a faster alternative to chromatography-based methods for the quantitative determination of BPA, and three analogues species, in receipts and tickets usually printed on thermal paper. The ionization efficiency of these compounds is evaluated under different conditions, and the effect of instrumental parameters of the DART source in the observed responses is discussed. The yield of the DART desorption-ionization process was greatly improved when compounds are previously converted into their acetyl derivatives; thereafter, the temperature of electronically excited helium atoms was the most relevant of the evaluated instrumental parameters. Under optimized conditions, the reported method provided recoveries in the range from 90 to 110 %, a limit of quantification of 0.004% (w:w), well below the maximum concentration established after 2020 for BPA (0.02%, w:w), and permitted to perform duplicate determinations of each sample extract with a response time around 1 min. The accuracy of BPA levels found in non-spiked samples was confirmed using GC-EI-MS as reference technique. BPA was systematically noticed in the processed samples with concentrations ranging from 0.005% to more than 6%. Keywords: bisphenol A; thermal printing paper; ambient ionization; direct analysis in realtime; mass spectrometry. *Corresponding author: Isaac Rodríguez e-mail: [email protected] 3 1. Introduction Bisphenol A (BPA) is a high production volume chemical employed as monomer in the preparation of polymeric materials and epoxy resins. Moreover, as free compound, it is used in the elaboration of thermal printing paper. In this latter application, BPA, together with a thermochromic dye in presence of an organic solvent, is combined with a base layer of paper. In contact with a hot surface, and/or upon pressure, BPA reacts with the dye to develop an image on the paper surface [1]. This printing technology has a low cost, it does not require ink, and it is extensively employed in tickets, receipts and labels stuck on retail market products. BPA is an endocrine disrupting chemical promoting a wide range of health outcomes in animals and humans [2,3]. Direct contact with thermal paper tickets is recognized as a relevant source of exposure to this compound [4]. In regards, increased urinary levels of BPA have been reported for operators of thermal paper manufacturing companies and female cashiers [5,6]. Moreover, dermal exposure has proved to turn in higher proportions of unconjugated BPA in the systemic circulation than dietary intake of the same species [7]. In order to reduce this exposure route, the EU has limited the concentration of BPA to a maximum of 0.02 % (equivalent to 0.2 mg g-1) for thermal printing paper commercialized after 2020 [8]. Simultaneously, alternative safer colour developers are under evaluation [9]. Despite these facts, recent surveys have revealed that BPA remains as the developer most often employed in thermal printing paper produced all over the world. Concentrations reported in these studies are up to 3 orders of magnitude higher than the limit established for year 2020 [10–12]. The determination of BPA in thermal printing paper is a two-step procedure. First, the compound is released from paper samples with polar organic solvents, such as methanol [10,13] or acetonitrile [11]. Thereafter, liquid chromatography (LC), followed by UV [12], fluorescence [14] or mass spectrometry (MS) detection [15], is employed as analytical technique. Alternatively, BPA can be also determined by gas chromatography (GC)-MS 4 [16]. In this case, the detectability of the compound is greatly improved after derivatization, i.e. acetylation with acetic anhydride [17]. The above approaches provide LOQs low enough to satisfy the future regulation reducing the maximum concentration of BPA in thermal paper. However, the availability of faster determination procedures, not limited by the duration of the chromatographic separation step (from 10 to 30 min depending on the characteristics of the chromatographic column), is a matter of interest, particularly considering that control programs will involve the analysis of thousands of samples. Ambient ionization sources followed by tandem (MS/MS), or high resolution (HR) mass spectrometry, have been proposed for the fast control of a limited number of analytes in different matrices, avoiding a chromatographic separation step. One of the most successful direct ionization sources is DART (used as the acronym of direct analysis in real time). DART can be classified as a kind of atmospheric pressure chemical ionization source. Thus, compounds are first volatilized and then ionized, mostly through charge transfer reactions with reactive ions and, in a minor extent, through Penning ionization. Reactive ions are produced by interaction of gases existing in the atmosphere of the laboratory (from permanent gases to water and solvent vapours) and electronically excited helium atoms generated in the DART source. Additionally to compounds ionization, formation of adducts, between neutral molecules of the analyte and ionized gases, such as ammonia, has been also reported [18–21]. Several applications of DART-MS have been developed in the areas of food, forensic and manufactured goods analysis [19,22,23]. From these studies, it can be concluded that DART ionization is particularly suitable for the ionization of low molecular weight and thermally stable molecules. To the best of our knowledge, previous applications of DART to BPA determination are limited. They have been focussed in the qualitative identification of BPA in epoxy coatings used in food packing [24], and in the evaluation 5 of its relative ionization efficiency comparing DART with other ambient desorption ionization sources [25]. Herein, we investigate the suitability of DART-MS, based on a time-of-flight MS system, for the quantitative determination of BPA, and related bisphenol species, in the extracts obtained from thermal printing receipts and tickets. The ionization efficiency of selected species in the DART source, as free and acetylated compounds, is evaluated and the instrumental conditions affecting the performance of the overall procedure are discussed. The applicability of the method is assessed with analysis of thermal paper samples containing different concentrations of BPA, using GC with electron ionization (EI) MS detection as reference technique for accuracy assessment. 2. Material and methods 2.1. Standards and solvents Standards of BPA (99 %), bisphenol B (BPB), bisphenol E (BPE), bisphenol F (BPF) and bisphenol A diacetate 98% (BPADA) were supplied by Sigma-Aldrich (Milwakee, WI, USA). Deuterated BPA (BPA-d6), deuterium atoms bonded to aliphatic carbons, was obtained from Toronto Research Chemicals (North York, ON, Canada). BPA-d6 was used as internal surrogate (IS) during extraction of thermal printing paper samples, compounds acetylation and determination. Chemical structures and CAS numbers of bisphenol species are compiled in Table 1. Individual stock standards of the above compounds were prepared in methanol. Further dilutions and mixtures were made in the same solvent and stored at -20 ºC. Methanol (MeOH) (HPLC grade) was purchased by Merck (Darmstadt, Germany). Acetic anhydride, toluene and isooctane were obtained from Sigma-Aldrich. 2.2. Samples and sample preparation 6 Thermal printing paper samples (cash receipts and tickets) were collected during January and February of 2019 in different establishments from Spain and England. After reception, each sample was folded using aluminium paper, and stored individually, at room temperature, until analysis (c.a. 2-3 weeks). Before extraction, samples were cut in 4 mm diameter circled pieces with a stainless steel punch [13]. Extraction was carried out using 50 mg of circled pieces from each sample (ticket or receipt) spiked with the IS. MeOH (10 mL) was employed as extraction solvent. The process was carried out in closed glass vessels, for 10 min, under sonication [10,13]. After centrifugation, an aliquot of the obtained supernatant (0.1 mL unless otherwise stated) was acetylated as described elsewhere [26]. In brief, derivatization was carried out mixing the methanolic extract, or standard solution, with 8 mL of a K2HPO3·3H2O 0.3 M (aqueous solution), using 50 µL of acetic anhydride as derivatization reagent, and 2 mL of isooctane to recover the acetylated derivatives. Tubes were capped, shaken manually for 5 min and centrifuged (2500 rpm for 5 min) to facilitate the separation between aqueous and organic phases. The fraction of isooctane was transfered into a 2 mL autosampler vial, ready to analysis either using DART-TOF-MS, or GC-EI-MS employed as reference technique. 2.3. Equipment and determination conditions DART-QTOF-MS The DART-QTOF-MS system consisted of a DART-SVP ion source (IonSense Inc. Saugus, MA, USA, model number SVPS-200), equipped with a linear rail and Quick Strip transmission sample cards, with a 12-possition frame of stainless steel. The DART source was coupled to a QTOF-MS, Agilent 6520 model acquired from Agilent Technologies (Wilmington, DE, USA), through the commercial Vapur chamber, which reduces the entrance of helium and nitrogen in the high vacuum region of the MS instrument. A flow rate of helium, 2.5 L min-1, was used during compounds desorption 7 and ionization. In the standby mode, helium was replaced by nitrogen. During method development, DART was operated in positive and negative modes for native and acetylated bisphenol compounds ionization, applying a grid voltage of 400 V for positive and 350 V for negative mode. Under final conditions, compounds were determined as acetylated species using the positive-ion DART mode. The temperature of the source was set at 400 ºC and the speed of the linear rail fixed at 0.2 mm s-1. Standards and sample extracts (from 1 to 4 µL) were deposited in the stainless-steel mesh of cards using a 10 µL volume syringe. After solvent evaporation (c.a. 5 min), the card was loaded in the linear rail for analysis. The first position of each card was used to record the background spectrum corresponding to ambient ions generated in the DART source. This spectrum was subtracted to those obtained for standards and sample extracts. Samples were analysed in duplicate, leaving a non-spiked position in the card between different samples to prevent cross-contamination problems. Thus, 4 different samples can be processed in duplicate using the same card. The QTOF instrument operated in high resolution (4 GHz) mode. Under final working conditions, analytes (as acetylated derivatives) were quantified in positive ionization mode applying a capillary voltage of 1000 V. The fragmentor voltage was set at 130 V. Accurate mass data were recorded in the range of m/z values from 50 to 1700, at a rate of 1 spectra s-1 (13700 scans are accumulated in each spectrum). The identity of acetylated compounds was confirmed with their MS/MS spectra recorded in the autoMS/MS mode. Selection of precursor ions was restricted to base peaks (ions) in the positive-ion DART spectra of acetylated compounds. Two product ion spectra were recorded by precursor, when detected in the accurate MS function. After 0.5 min, selection of precursor ions was re-activated in order to record product ion spectra from compounds in the following spot. Collision energies for MS/MS determination were 15 eV (BPA and BPB) and 10 eV (BPE and BPF). Recalibration of the m/z axis in the TOF MS analyser was performed using background ions with m/z values of: 135.1016, 8 152.1281 and 391.2860. The first two values correspond to the [M+H]+ and [M+NH4]+ ions associated to diethylene glycol monoethyl ether (DEGMEE) [27]; the latter one corresponds to the [M+H]+ ion of bis(ethylhexyl) phthalate. Both species are recognized as ubiquitous in indoor environments and provided signals with enough intensity for continuous re-calibration of the TOF MS analyser. According to the design of the card holder mounted in the linear rail module (see Fig. S1), the ionization flow of excited helium atoms is blocked between consecutive samples (spots); thus, recalibration of the TOF instrument is only possible during ionization of spots, but not when the linear rail is moving from spot to spot. GC-EI-MS GC-EI-MS was used as reference technique. The employed system was an Agilent 7890A gas chromatograph, equipped with an autosampler (Agilent 7693 model) and connected to a quadrupole mass spectrometer (Agilent MS 5975 C model). The MS analyser was operated in the single ion monitoring mode (SIM), selecting two characteristic ions per compound. Separations were carried out in a HP-5MS capillary column (30 m x 250 µm x 0.25 µm) acquired from Agilent and operated at a constant carrier gas flow of 1.2 mL min-1 (He, 99.999%). The temperature of the GC oven was as follows: 70 ºC (1 min) rate at 10 ºC min-1 to 280 ºC (5 min). The electron impact source (EI) and the quadrupole mass analyser were set at 230 ºC and 150 ºC, respectively. Standards and sample extracts (2 µL volume), as acetylated species, were injected in splitless mode with the injector temperature at 280 ºC. The transfer line between the GC and the MS was also set at 280 ºC. Retention times and m/z ratios corresponding to quantification and qualification ions of each bisphenol are given as supplementary information, Table S1. 2.4. Recoveries assessment and samples quantification 9 The yield of the sample preparation procedure was investigated with samples of thermal printing paper, containing low levels of BPA, spiked with target compounds at two different levels: 0.02 and 0.2 %. Concentrations in thermal paper extracts were determined by comparison with a set of calibration standards, acetylated under same conditions as sample extracts, and containing the equivalent concentration of BPA-d6. Responses for calibration standards and extracts for spiked and non-spiked samples were obtained by DART-TOF-MS, using a mass window of 25 ppm around the m/z values for the [M+NH4+] species in the positive-ion DART spectra of acetylated compounds. Confirmation of target compounds in non-spiked samples required mass errors below 20 ppm for at least one of the two product ions observed in the MS/MS spectra recorded using the autoMS/MS function. Concentrations in non-spiked samples were also calculated by comparison with responses obtained for acetylated standards. Quite often, the concentrations of BPA existing in thermal printing paper samples overpassed the linear response range of the DART-TOF-MS system, and also that for the GC-EI-MS. In that case, the primary methanolic extract obtained from thermal paper samples was diluted before acetylation. Using GC-EI-MS determination, the concentrations in sample extracts were also established by comparison with calibration curves obtained for acetylated species, using BPA-d6 as internal standard. 3. Results and discussion 3.1. DART-MS spectra of bisphenol compounds For a given compound, the yield of DART ionization depends on a balance among mass and/or charge transfer process, formation of adducts with reactive ions generated from ambient species, and thermal stability during volatilization from the stainless steel mesh in the Quick Strip cards. 16 4. Conclusions For the first time, DART QTOF-MS is proposed for the quantitative determination of four bisphenol-type compounds in samples of thermal printing paper. Acetylation of target compounds turned a key parameter to improve the efficiency of the desorption-ionization process at the DART source. Combination with QTOF-MS guarantees the selectivity of compounds determination, whereas product ion scan spectra permitted to verify the identity of responses in the chronograms corresponding to non-spiked samples. Under final working conditions, DART-MS achieved LOQs in the low ng per mL level, with a linear response range covering more than two-orders of magnitude. Four samples, with the corresponding blank positions, can be processed in duplicate with the same Quick Strip card within an analysis time of a few minutes. Recoveries obtained for spiked samples and concentrations measured in non-spiked samples of receipts and tickets were in agreement with those obtained by GC-MS. BPA was ubiquitous in the processed samples. In most cases, the concentrations of this compound are well above the maximum permitted levels by the EU regulation to be implemented in year 2020. Acknowledgements This study was supported by Xunta de Galicia, Spanish Government and EU (grant numbers GRC-ED431C 2017/36 and CTQ2015-68660-P). Conflict of interest: none. 17 Table 1. Names, abbreviations, structures and CAS numbers of bisphenol compounds studied in this work. Compound Abbreviated name Structure Formula Monoisotopic mass CAS Bisphenol A BPA C15H16O2 228.1150 80-05-7 Bisphenol A diacetate BPADA C19H20O4 312.1362 10192-62-8 Bisphenol B BPB C16H18O2 242.1307 77-40-7 Bisphenol E BPE C14H14O2 214.0994 2081-08-5 Bisphenol F BPF C13H12O2 200.0837 620-92-8 Bisphenol A dimethyl-d6 BPA-d6 C15D6H10O2 234.1527 86588-58-1 18 Table 2. Performance of the DART-TOF-MS system (without considering sample preparation) for acetylated bisphenol compounds determination. Compound [M+NH4]+ Linearity (10- 2000 ng mL-1, n= 8 levels) Repeatability (RSDs %; n=5) Reproducibility (RSDs %; n=3) LOQs (ng mL-1) R2 (no IS correction) R2 (after IS correction) BPA 330.1705 0.9702 0.9990 8% 5% 8 BPB 344.1862 0.9726 0.9981 11% 5% 5 BPE 316.1549 0.9905 0.9977 11% 4% 8 BPF 302.1392 0.9746 0.9988 10% 5% 6 BPA-d 6 336.2082 - - - - - 19 Table 3. Summary of recoveries (%), with RSDs, for spiked samples of a thermal printing paper sample, n=4 replicate extractions and determinations. Compound Addition level 0.02% Addition level 0.2% DART-TOF-MS GC-EI-MS DART-TOF-MS GC-EI-MS BPA 107 ± 5 106 ± 10 96 ± 6 97 ± 4 BPB 96 ± 4 117 ± 10 96 ± 6 112 ± 3 BPE 105 ± 6 108 ± 13 98 ± 9 99 ± 1 BPF 107 ± 8 102 ± 12 103 ± 8 105 ± 7 20 Table 4. Concentrations (%) of BPA in receipts and tickets. Average values for duplicate determinations. Code Type Concentration (%) 1 Supermarket 4.4 2 Supermarket 1.7 3 Supermarket * 6.3 4 Fruit shop 1.9 5 Food * 0.014 6 Sport shop 0.005 7 Fashion shop 4.0 8 Fashion shop 0.016 9 Fashion shop 2.3 10 Fashion shop 0.014 11 Pound shop 3.0 12 Toy shop 2.3 13 Library 3.2 14 Pharmacy 1.9 15 Cinema ticket 2.4 16 Train ticket 0.007 17 Train ticket * 0.006 *Samples from United Kingdom 21 Captions to figures: Fig. 1. DART-MS spectra of BPA under positive (A) and negative (B) ionization modes. C, positive-ion DART spectra of BPADA. D, E and F chronograms for the base peak in above MS spectra corresponding to a 10 µg m L-1 standard. G, product ion scan spectrum of BPADA. Fig. 2. Positive-ion DART spectra (A to C) and product ion scan spectra (D to F) of BPB, BPE and BPF. Fig. 3. A, effect of DART helium gas temperature in the normalized responses for [M+NH4]+ ions of analytes, n=2. B, responses as function of the volume of solution deposited in the Quick Strip card, normalized values (n=3) to those obtained for 3 µL. Fig. 4. A, DART-MS chronogram obtained under final working conditions for BPA in non-spiked and spiked (addition level 0.02%) pieces obtained from the same receipt. Duplicate determinations. B, product ion scan spectrum of BPA in the spiked sample. Fig. 5. Correlation plot between concentrations measured by GC-EI-MS and DART-MS for BPA in receipts and tickets. Values as percentage. 22 Fig. 1. DART-MS spectra of BPA under positive (A) and negative (B) ionization modes. C, positive-ion DART spectra of BPADA. D, E and F chronograms for the base peak in above MS spectra corresponding to a 10 µg m L-1 standard. G, product ion scan spectrum of BPADA 5 x10 0 0.4 0.8 1.2 1.6 2 2.4 2.8 3.2 Scan (rt: 2.326-2.499 min, 12 scans) Frag=130.0V BPADA_pos.d Subtract 330.1704 Counts vs. Mass-to-Charge (m/z) 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500 C 19 H 20 O 4 [M+NH 4 ] + Calc. 330.1705 Error: -0.3 ppm 3 x10 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75 3EIC(227.1078) Scan Frag=130.0V BPA_neg.d Counts vs. Acquisition Time (min) 1.5 1.6 1.7 1.8 1.9 22.1 2.2 2.3 2.4 2.5 3 x10 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 C 15 H 16 O 2 [M-H] - Calc. 227.1078 Error: -2.2 ppm Scan (rt: 2.082-2.240 min, 20 scans) Frag= 130.0V BPA_neg 227.1073 89.1429 258.0956 77.0230 121.0277 186.0565 139.0036 Counts vs. Mass-to-Charge (m/z) 80 100 120 140 160 180 200 220 240 260 5 x10 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 EIC(330.1704) Scan Frag=130.0V BPADA_pos.d Counts vs. Acquisition Time (min) 2.15 2.2 2.25 2.3 2.35 2.4 2.45 2.5 2.55 2.6 2.65 2.7 2.75 2.8 2.85 2.9 2.95 3 4 x10 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Scan (rt: 1.517-1.733 min, 14 scans) Frag= 130.0V BPA_pos.d 135.0855 72.0697 90.0809 152.1354 166.1508 198.1489 234.1707 270.1950 Counts vs. Mass-to-Charge (m/z) 60 80 100 120 140 160 180 200 220 240 260 280 300 A B C C 9 H 11 O Calc. 135. 0804 Error: 37 ppm 4 x10 0 1 2 3 4 5 6 7 EIC(135.0805) Scan Frag=130.0V BPA_pos.d Counts vs. Acquisition Time (min) 0.9 11.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 D E F 4 x10 0 0.5 1 1.5 2 2.5 Product Ion (rt: 2.34-2.47 min, 8scans) Frag=130.0V CI[email protected] (330.1713[z=1] -> **) BPADA … 135.0792 177.0916 330.1702 Counts vs. Mass-to-Charge (m/z) 60 80 100 120 140 160 180 200 220 240 260 280 300 320 C 11 H 13 O 2 Calc. 177.0910 Error: 3.4 ppm C 9 H 11 O Calc. 135.0804 Error: -8.9 ppm G 23 Fig. 2. Positive-ion DART spectra (A to C) and product ion scan spectra (D to F) of BPB, BPE and BPF. 5 x10 0 0.5 1 1.5 2 2.5 3 3.5 +Scan (rt: 2.810-2.965 min, 15 scans) Frag=130.0V BPB_ace.d Subtract 344.1885 Counts vs. Mass-to-Charge (m/z) 100 150 200 250 300 350 400 450 500 550 600 BPB C 20 H 22 O 4 [M+NH 4 ] + Cal. 344.1862 Error: 6.8 ppm 5 x10 0 1 2 3 4 5 6 +Scan (rt: 4.818-4.929 min, 11 scans) Frag=130.0V BPE_ac.d Subtract 316.1580 Counts vs. Mass-to-Charge (m/z) 100 150 200 250 300 350 400 450 500 550 600 5 x10 0 1 2 3 4 5 6 7 8 +Scan (rt: 4.019-4.142 min, 12 scans) Frag=130.0V BPF_ac.d Subtract 302.1418 Counts vs. Mass-to-Charge (m/z) 100 150 200 250 300 350 400 450 500 550 600 BPE C 18 H 18 O 4 [M+NH 4 ] + Cal. 316.1542 Error: 11.7 ppm BPF C 17 H 16 O 4 [M+NH 4 ] + Cal. 302.1392 Error: 8.6 ppm 3 x10 0 1 2 3 4 5 +Product Ion (rt: 2.484 min) Frag=130.0V [email protected] (344.1862[z=1] -> **) MSMS_BPE … 149.0947 191.1074 344.1882 Counts vs. Mass-to-Charge (m/z) 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 3 x10 0 1 2 3 4 -NH 3 +Product Ion (rt: 2.481 min) Frag=130.0V [email protected] (316.1542[z=1] -> **) MSMS_BPE … 121.0665 316.1531 163.0755 299.1272 Counts vs. Mass-to-Charge (m/z) 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 3 x10 0 0.2 0.4 0.6 0.8 1 1.2 +Product Ion (rt: 2.433 min) Frag=130.0V [email protected] (302.1386[z=1] -> **)MSMS_BPF … 107.0500 149.0601 302.1369 72.0830 285.1559 199.1631129.1070 Counts vs. Mass-to-Charge (m/z) 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 C 7 H 7 O Cal. 107.0491 Error: 8.4 ppm C 9 H 9 O 2 Cal. 149.0597 Error: 2.7 ppm O O O O C 10 H 13 O Cal. 149.0961 Error: -9.4 ppm C 12 H 15 O 2 Cal. 191.1067 Error: 3.7 ppm O O HO OO O O HO C 8 H 9 O Cal. 121.0648 Error: 14.0 ppm C 10 H 11 O 2 Cal. 163.0754 Error: 0.6 ppm O O A B C D E F O O O O OH O O 24 Fig. 3. A, effect of DART helium gas temperature in the normalized responses for [M+NH4]+ ions of analytes, n=2. B, responses as function of the volume of solution deposited in the Quick Strip card, normalized values (n=3) to those obtained for 3 µL. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% BPA BPB BPE BPF 250 ºC 300 ºC 350 ºC 400 ºC 0% 20% 40% 60% 80% 100% 120% BPA BPB BPE BPF 1 µL 2 µL 3 µL 4 µL A B 25 Fig. 4. A, DART-MS chronogram obtained under final working conditions for BPA in non-spiked and spiked (addition level 0.02%) pieces obtained from the same receipt. Duplicate determinations. B, product ion scan spectrum of BPA in the spiked sample. A 5 x10 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 EIC(330.1705) Scan Frag=130.0V AUTOMSMS_AdCODE17.d Smooth Counts vs. Acquisition Time (min) 55.5 66.5 77.5 88.5 99.5 10 non-spiked sample Added concentration 0.02% B 3 x10 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 Product Ion (rt: 8.9-9.11 min, 2 scans) Frag=130.0V CI[email protected] (330.1705[z=1] -> **) … 135.0797 177.0895 330.1672 Counts vs. Mass-to-Charge (m/z) 80 100 120 140 160 180 200 220 240 260 280 300 320 340 Calc. 135.0804 Error: -5.2 ppm Calc. 177.0910 Error: -8.4 ppm Calc. 330.1705 Error: -10 ppm 32 Table S1: Summary of retention times, quantification and qualification ions, and instrumental performance of the GC-EI-MS system for determination of bisphenol compounds as acetylated derivatives. Compound Retention time (min) Quantification ion (m/z) Qualification ion (m/z, relative abundance) aR2 (10- 2000 ng mL-1, n=8 levels) a,bRepeatability (RSDs %; n=5) LOQs (ng mL -1 ) BPA 18.94 213 228(26) 0.9996 4% 2 BPB 19.62 255 297 (65) 0.9987 5% 5 BPE 18.49 214 256 (54) 0.9925 9% 8 BPF 18.07 200 242 (29) 0.9914 13% 5 BPA-d6 18.93 216 234 (26) - - - aValues after IS correction bData for a 100 ng mL-1 standard 33 Table S2. Experimental m/z ratios and mass errors for the [M+NH4]+ ions of acetylated compounds in the series of calibration standards. [M+NH4] + 10 ng mL -1 25 ng mL -1 50 ng mL -1 100 ng mL -1 Compound Cal. m/z Exp. m/z Error (ppm) Exp. m/z Error (ppm) Exp. m/z Error (ppm) Exp. m/z Error (ppm) BPA 330.1705 330.1721 4.8 330.1714 2.7 330.1703 -0.6 330.1702 -0.9 BPB 344.1862 344.1881 5.5 344.1891 8.4 344.1859 -0.9 344.1861 -0.3 BPE 316.1549 316.1587 12.0 316.156 3.5 316.1553 1.3 316.1548 -0.3 BPF 302.1392 302.1426 11.3 302.1424 10.6 302.1409 5.6 302.14 2.6 Table S2. Cont. [M+NH4 + ] 250 ng mL -1 500 ng mL -1 1000 ng mL -1 2000 ng mL -1 Compound Cal.m/z Exp. m/z Error (ppm) Exp. m/z Error (ppm) Exp. m/z Error (ppm) Exp. m/z Error (ppm) BPA 330.1705 330.1703 -0.6 330.171 1.5 330.1711 1.8 330.171 1.5 BPB 344.1862 344.1861 -0.3 344.1869 2.0 344.1873 3.2 344.1869 2.0 BPE 316.1549 316.1551 0.6 316.1559 3.2 316.1557 2.5 316.1556 2.2 BPF 302.1392 302.14 2.6 302.1403 3.6 302.1403 3.6 302.1404 4.0 In bold, errors above 10 ppm. 34 Fig. S1.A, Picture of the DART source connected to QTOF-MS instrument. B, detail of the 12-possitions Quick Strip card and the card metallic holder mounted on the linear rail module. A B 35 Fig. S2. Chronograms obtained for acetylated bisphenols in the series of calibration standards from 10 to 2000 ng mL-1. Extraction window 25 ppm. 4 x10 0 1 2 3 4 5 6 7 8 9 + EIC(330.1705) Scan Frag=130.0V Calibrado_bisfenoles_4GHz. d Counts vs. Acquisition Time (min) 0.5 11.5 22.5 33.5 44.5 55.5 66.5 77.5 88.5 99.5 10 0.40 .6 0.8 11 .21 .4 1 .6 1 .8 22 .2 2.4 2 .6 2 .8 33.2 10 25 50 100 250 500 1000 2000 Blank Blank 10 25 5 x10 0 0.5 1 1.5 2 + EIC(316.1549) Scan Frag=130.0V Calibrado_bisfenoles_4GHz.d Counts vs. Acquisition Time (min) 0.5 11.5 22.5 33.5 44.5 55.5 66.5 77.5 88.5 99.5 10 5 x10 0 0.2 0.4 0.6 0.8 1 1.2 1.4 + EIC(302.1392) Scan Frag=130.0V Calibrado_bisfenoles_4GHz.d 1 Counts vs. Acquisition Time (min) 0.5 11.5 22.5 33.5 44.5 55.5 66.5 77.5 88.5 99.5 10 5 x10 0 0.2 0.4 0.6 0.8 1 1.2 1.4 + EIC(344.1862) Scan Frag=130.0V Calibrado_bisfenoles_4GHz.d 1 Counts vs. Acquisition Time (min) 0.5 11.5 22.5 33.5 44.5 55.5 66.5 77.5 88.5 99.5 10 1000 2000 500 250 100 50 25 10 1000 2000 500 250 100 50 25 10 1000 2000 500 250 100 50 25 10 BPA BPB BPE BPF