Multi-class method for biomonitoring of hair samples using gas chromatography-mass spectrometry

1 Multi-class method for biomonitoring of hair samples using gas chromatography - 1 mass spectrometry 2 3 Julia Martína, Monika Möderb*, Alexander Gaudlc, Esteban Alonsoa, Thorsten 4 Reemtsmab 5 6 7 a Department of Analytical Chemistry, Escuela Politécnica Superior, University of 8 Seville, C/ Virgen de África 7, 41011 Seville, Spain 9 b Helmholtz Centre for Environmental Research - UFZ, Department of Analytical 10 Chemistry, Permoserstrasse 15, 04318 Leipzig, Germany 11 c Institute of Laboratory Medicine, University of Leipzig, 04103 Leipzig, Germany 12 13 14 *corresponding author: 15 16 Monika Möder, PhD 17 Helmholtz Centre for Environmental Research - UFZ, 18 Department of Analytical Chemistry, 19 Permoserstrasse 15, 20 04318 Leipzig, 21 Germany 22 23 e-mail: [email protected] 24 Tel: ++49 (0341) 235 1413 25 Fax: ++49 (0341) 235 450822 26 27 28 29 30 31 32 33 34 2 Abstract 35 Currently, non-invasive biomonitoring of human exposure to organic pollutants bases 36 upon the analysis mainly of urine and human breast milk. While mostly persistent 37 organic pollutants are the center of interest, the aim of our study was to develop a 38 method for the determination of different chemical classes of emerging pollutants 39 (organophosphorus flame retardants, plastic additives such as phthalates, bisphenol A, 40 insecticides, antimicrobials, preservatives and musk fragrances) in hair by gas 41 chromatography-mass spectrometry. The preferred sample preparation included 42 hydrolysis of the hair with trifluoroacetic acid in methanol followed by a liquid–liquid 43 extraction using hexane/ethyl acetate. The validated method is characterized by 44 recoveries higher than 77 % for most analytes, relative standard deviations below 16 % 45 and limits of detection between 2 pg mg-1 (HHCB) and 292 pg mg-1 (propylparaben) 46 using 50 mg of dry hair. After respective blank corrections, bis-(2.ethylhexyl)phthalate 47 (DEHP) and the musk fragrance HHCB were the predominant compounds determined 48 in all hair samples at concentrations between 32 and 59 ng mg-1 and 0.8 – 13 ng mg-1, 49 respectively. The bactericide triclosan and the insect repellent N,N-diethyl-3-50 methylbenzamide (DEET) were detected in selected hair samples at 2 and 0.8 ng mg-1, 51 respectively. 52 53 54 Keywords: Emerging pollutants; Hair analysis; Gas chromatography-mass 55 spectrometry; Human biomonitoring 56 57 3 1 Introduction 58 Humans are exposed permanently to a large variety of chemicals present in their indoor 59 and outdoor environment, in food or beverages, or in clothes, personal care products etc. 60 Numerous organic pollutants are known to impose potential health risks on human and 61 the ecosystem due to their bioaccumulative and toxic character such as polychlorinated 62 dioxins and –furanes (PCDD/F), -biphenyls (PCBs), and polycyclic aromatic 63 hydrocarbons (PAHs). Other wide spread applied chemicals such as UV filters, 64 preservatives and antimicrobials, musk fragrances, and insect repellents used in personal 65 care products as well as industrial chemicals (e.g. perfluorinated compounds (PFCs), 66 phthalate esters, flame retardants (FRs) or alkyl phenols) evolved as potential hazards 67 due to their biological effects found in laboratory experiments, environmental and 68 epidemiological studies [1-3]. Along different routes, these pollutants can be 69 incorporated in human tissue and may increase potential health risks. 70 Human biomonitoring allows assessing human exposure to chemicals via working 71 environment or daily uptake from environment and diet. For epidemiological studies, 72 non-invasive samples such as urine, breast milk, saliva and hair are favoured over blood 73 and plasma because their sampling procedures are often connected with ethical and 74 practical problems [4-8]. Particularly, hair analysis attained a suitable biomonitoring 75 tool due to the easiness of sampling and sample storage that do not require any 76 restricted measures as the presence of medical staff, adapted settings, or refrigerated 77 conditions (9, 10). Furthermore, hair analysis enables a retrospective estimation of 78 chronic and past exposure which is required in drug, doping and forensic studies. 79 In the context of assessing environmental exposure of humans to chemicals, hair 80 analysis has been less frequently considered as biomonitoring tool so far because the 81 substances determined in hair may reflect other metabolic pathways than those detected 82 in urine, human milk or saliva. 83 Analytical methods for hair analysis have been developed with focus on defined 84 pollutant classes such as organochlorine pesticides, brominated flame retardants 85 (BFRs), PFCs, PAHs, PCBs, PCDD/Fs and illicit drugs [9,14]. Recent reports on the 86 analysis of five metabolites of bis-(2-ethylhexyl) phthalate (DEHP) [15] and of 87 bisphenol A [16] in human hair indicated the suitability of hair samples for monitoring 88 the exposure of humans towards the ubiquitously present plastic additives. 89 Common sample preparation in hair analysis starts with washing and cutting or 90 pulverizing the hair specimen. The next step is an acidic or alkaline treatment at 91 4 elevated temperature (40 °C-80 °C) that destroys the keratin structure of the hair. 92 Subsequently the reaction mixture is extracted by liquid-liquid extraction using hexane, 93 dichloromethane or ethyl acetate as reported for the determination of BFRs, heterocyclic 94 aromatic amines and illicit drugs in hair [17-19]. Other extraction procedures preferred 95 Soxhlet extraction [20] or hollow fiber solid phase microextraction (HF-SPME) [21]. In 96 another study, ultrasonic extraction with methanol was chosen for hair extraction with 97 the aim to assess the human exposure to organophosphorous pesticides [22]. In case of 98 multicomponent analysis of pesticides, SPE was favoured over liquid-liquid extraction 99 whereas different sorbent types were required to obtain best recoveries for the pesticides 100 with different properties [23]. 101 Depending on the polarity of the target analytes, their analysis has been performed 102 either by gas chromatography-mass spectrometry (GC-MS) or liquid chromatography-103 tandem mass spectrometry (LC-MS-MS) [15]. 104 An overview on established methodologies for hair analysis is given in the electronic 105 supplement material as Table S1. 106 Although suspected of interfering with human health, pollutants such as musk 107 fragrances, preservatives or antimicrobials have been less considered in hair analysis. 108 While analysis of blood or urine samples utilizes commonly multi-class methods, 109 appropriate approaches for biomonitoring by hair are missed. In order to complement 110 data on pollutants in different biological materials, also for hair analysis highly sensitive 111 multi-class methods are required [9, 24]. 112 The aim of our investigations was to develop a multi-class method for the determination 113 of fourteen emerging pollutants in hair by gas chromatography-mass spectrometry. The 114 target compounds belonging to several chemical classes (phthalates, phenolic 115 compounds, polycyclic musks, organophosphates, aromatic amide) are applied 116 commonly as insecticides, antimicrobials, preservatives, flame retardants and 117 fragrances. The new established protocol can be included in e.g. epidemiological 118 investigations, workplace monitoring or health care studies. 119 120 121 122 123 5 2 Experimental 124 2.1 Chemicals 125 The substances investigated are listed in Table 1. BPA, ethyl (EtP) and n-propyl (n-PrP) 126 esters of 4-hydroxybenzoic acid, 4-n-nonyphenol (4n-NP) as well as triclosan (TCS) 127 were obtained from Sigma-Aldrich (Steinheim, Germany); Galaxolide (HHCB) and 128 Tonalide (AHTN) were purchased from Promochem (Wesel, Germany); N,N-diethyl-3-129 methyl-benzamide (DEET), tri-n-butyl phosphate (TBP), triphenyl phosphate (TPP), 130 tris(2-chloro-propyl) phosphate (TCPP), dimethyl phthalate (DMP), diethyl phthalate 131 (DEP), di-n-butyl phthalate (DBP) and DEHP were purchased from Dr. Ehrenstorfer 132 GmbH (Augsburg, Germany); and BPA-d16 (BPA-d16) and di-n-butyl phthalate-d4 133 (DBP-d4) were supplied by Cambridge Isotope Laboratories,Inc. (Andover, MA, USA). 134 All reference compounds were of purity >99 %. Acetone, ethyl acetate, n-hexane, 135 methanol and trifluoroacetic acid (TFA) purchased from Merck (Darmstadt, Germany) 136 were of chromatographic analysis grade. The shampoo “herbal essences-fresh balance” 137 (Clairol) used for comparative experiments was bought in the supermarket. 138 Individual stock solutions were prepared at a concentration of 1 mg mL-1 in methanol. 139 Mix standard solutions at different concentrations were prepared in methanol, when 140 used to fortify hair samples, and in ethyl acetate, when considered to evaluate the 141 performance of the GC-MS system, respectively. 142 143 2.2 Sample collection 144 Hair samples were collected from four female volunteers of general population and 145 different age from Leipzig, Germany. All persons permitted the hair sampling and its 146 use for scientific purposes. Their hair was not chemically treated (colored, curled) but 147 the use of hairstyling products was reported within a little survey on the use of cosmetic 148 products and perfumed household products such as fabric softener. 149 Hair specimens were cut from the posterior vertex region of the head, as close as 150 possible to the scalp with a length of roughly 3-5 cm. All hair samples were stored in 151 aluminum foil, at ambient temperature until further processing and analysis. Hair 152 samples were washed twice with MilliQ water (ultrasonicated for 5 min) and 153 isopropanol (5 min). The successive washing process removes endogenous substances 154 as well as adsorbed chemicals from the surface of the hair [24]. After that, hair samples 155 were cut in small pieces (2-3 mm), further dried at room temperature and wrapped in 156 aluminum foil until the analysis. All wrapped samples were stored in the dark at room 157 6 temperature until analysis. Thus, all pollutant concentrations determined are related to 158 the weight of the dry hair. 159 160 2.3 Hair analysis 161 Fifty mg of hair were incubated with 1 mL of methanol/TFA (8.5:1.5, v/v) or for 162 comparative purposes, 1 mL 2M NaOH solution at 38 °C overnight. In each case, 20 μL 163 of the internal standard mixture at 0.4 ng mg-1 were added prior heating. After cooling 164 to room temperature, 4 mL hexane/ethyl acetate (1:1, v/v) were used for liquid-liquid 165 extraction (LLE) of the hair decomposed with methanol/TFA. 166 The samples incubated with NaOH were adjusted to pH 3 with acetic acid prior to LLE 167 with hexane/ethyl acetate. For LLE, the vial was shaken for 15 min and centrifuged at 168 4500 rpm for 10 min. The supernatant formed from the miscible solvents was separated 169 and evaporated to dryness at 40 °C under a stream of nitrogen. Finally, the dry residue 170 was reconstituted in 200 μL of ethyl acetate and 1 μL of this was injected for GC-MS 171 analysis. 172 173 2.4 Gas chromatography-mass spectrometry 174 Analyses were performed at a GC-MSD instrument (Agilent Technologies, San Jose, 175 CA, U.S.A.). Analytes were separated on a HP-5MS (30m×0.25mm, 0.25 µm, Agilent 176 Technologies) column. The oven temperature was programmed as follows: start at 50 177 °C for 1 min, increase at 10 K min-1 to 280 °C and held at 280 °C for 10 min to achieve 178 a running time of 34 min. The transfer line, ion source and quadrupole analyzer 179 temperatures were maintained at 280, 230 and 150 °C, respectively. 180 Helium was used as carrier gas at constant flow conditions of 1mL min-1. The directly 181 coupled mass spectrometer determined the substances after electron impact ionization in 182 selected ion monitoring (SIM) mode. The target ions of the analytes are listed in Table 183 1. The extract of hair sample 4 and the shampoo extract were also analyzed at full scan 184 mode (mass range 50-400 u). Instrument blank was checked injecting 1 µL ethyl acetate 185 every four analysis and a standard mixture was multiply analyzed within the batch of 186 analyses. 187 188 2.5 Method validation 189 The washed and cut hair sample was spiked with the standard mixture in ethyl acetate. 190 The solvent was evaporated completely before the hair was incubated with 191 7 methanol/TFA. Method accuracy (expressed as recovery percentage) and precision (in 192 terms of relative standard deviation (RSD)) were determined from recovery experiments 193 in triplicate at high and low concentration levels in the range from 0.2-2.8 ng mg-1 194 depending on the pollutant. Recoveries were calculated by comparing the peak areas of 195 hair samples spiked before extraction and a standard solution in ethyl acetate at the 196 same concentration level. The peak areas of analytes found in the non-spiked hair were 197 used to correct the signal areas of the spiked samples. 198 The linearity of the analysis method was studied by multi-level calibration curves built 199 from the analysis of standard solutions in ethyl acetate in triplicate at 6 different 200 concentration levels. Thus, the instrumental limits of detection (LOD) and 201 quantification (LOQ) were estimated as the concentrations of the analyte corresponding 202 to a signal-to-noise ratio of 3:1 and 10:1. In cases of DEHP, DBP, TCPP and BPA, the 203 blank signals were considered for corrections. In order to calculate the method limits of 204 detection (MDL) and quantification (LOQ), the recovery data (obtained with the low 205 spiking concentration) were taken into account. 206 For quantification, the internal standards 4-n-NP, BPA-d16 , DBP-d4 were used (Table 207 1). Intra-day precision was determined using 2 hair samples spiked at 2 ng mg-1 each 208 analyte and measured in triplicate. For the determination of the inter-day precision, the 209 same procedure was repeated on 3 different days.Matrix effects (ME) on the response of 210 analytes were evaluated relating the signal areas of a spiked hair extract to a standard 211 solution at the same concentration level (100 ng mL-1 for each substance in ethyl 212 acetate, corresponds to 0.4 ng mg-1 hair). The signal areas of the analyte in the non-213 spiked extract were considered for correction of the spiked sample (Equ. 1). 214 215 𝑀𝐸 =("#$%&'(&)*&(+&#)(*,-)&.-("/#0*1)3("#$%&'(&)*&(+&#)(*,-)&.-(%4-("/#0*1) "#$%&'(&)*&(567(( Equation 1 216 217 2.6 Blank analysis and instrument performance 218 Blanks of the individual solvents (1 mL of methanol/TFA and of 4 mL hexane/ethyl 219 acetate) and of both mixtures together were measured following the sample preparation 220 protocol. Results of the blank analyses are included in Table S2 (Supplement material) 221 The blank signals e.g. of DEHP, DBP, TCPP, and BPA were subtracted in terms of the 222 ion area (in counts per second, cps) from the respective target ion signals. The signal 223 areas of the real sample extracts were corrected in the same way by the blank values. 224 8 The performance of the instrument was checked regularly by an external standard 225 contained each target analytes at 100 ng mL-1. 226 227 3 Results and discussion 228 3.1 Method optimization 229 One difficulty in hair analysis is to distinguish between the amount of pollutants 230 adsorbed to the outer surface of the hair and the pollutants inside the hair which were 231 incorporated from the human body. Thus commonly, the first step in hair analysis is a 232 cleaning procedure to remove contaminants from the surface together with biogenic 233 compounds, such as fatty acids, sphingolipids and steroids suspected to affect analysis. 234 (Table S1). 235 236 3.1.1 Washing process 237 In a preliminary experiment, the efficiency of the cleaning procedure was evaluated 238 using hair from a volunteer who has freshly washed the hair with a defined amount of 239 commercial shampoo (2 g wet weight, “herbal essences-fresh balance”, Clairol, UK). 240 After air drying, 32 mg hair were taken from the volunteer and subjected to the cleaning 241 procedure as described in 2.2. Another portion of 37 mg underwent the same cleaning 242 procedure twice. 243 In each case the target substances remained on the hair surface were extracted with 4 244 mL hexane/ethyl acetate (1:1, v/v). After evaporation and reconstitution in ethyl acetate 245 (200 µL), the target substances were analyzed. Most of the target substances were 246 detected at trace concentration below 1 ng mg-1 hair except DEHP, DBP and AHTN 247 which were detected in the range of 1 - 3 ng mg-1(Fig. 1). However, these 248 concentrations are significantly below the pollutant concentration detected later in the 249 real hair samples. This suggests that the washing procedure was strong enough to 250 remove contaminants from the hair’s surface. Furthermore, the substance pattern of the 251 shampoo extract was not reflected in the extract of the cleaned hair underlining the 252 efficient removal of e.g. shampoo components by the washing procedure. Particularly 253 the presence of AHTN has to be pointed up because this polycyclic musk compound 254 was not determined in the shampoo extract. An extraction from inner hair parts cannot 255 be excluded although for these cleaning experiments, the hair was not cut in small 256 pieces. Thus, twice cleaning can reduce the substances on the hair surface (Fig. 1) but 257 on the other site, multiple cleaning could extract already substances incorporated in the 258 9 hair. For this reason, in the protocol for the analysis of flame retardants Kucharska et al. 259 abstained completely from a cleaning step of the hair [25b]. 260 During our method development, one cleaning step consisting of twice washing with 261 water and once with isopropanol was applied. 262 263 3.1.2 Denaturation of the hair sample 264 After the cleaning step, the denaturation of the hair by acidic, alkaline or enzymatic 265 treatment [17, 18, 24] facilitates the release of the incorporated contaminants. In this 266 study, acidic hydrolysis using MeOH/TFA and basic hydrolysis by aqueous NaOH were 267 compared. Finally, the LLE extracts obtained from acidic hydrolysis were cleaner and 268 exhibited lower noise in the GC-MS analysis compared to those of the alkaline 269 hydrolysis that yielded turbid and dark extracts (Figure S1 in supplement material). 270 These findings are in agreement with literature reports [17, 26]. 271 Therefore, the hair treatment with TFA/methanol was selected for further optimization. 272 The acidic hydrolysis is also advantageous with respect to the stability of esters such as 273 phthalates and parabens. At acidic conditions, the hydrolysis half-life times of DBP and 274 DEHP are > 500 d and > 800 d, respectively [27], and also parabens are known to resist 275 acidic hydrolysis [28]. 276 277 3.1.3 Liquid-liquid extraction 278 The LLE of a reaction mixture from 50 mg hair spiked with 2 ng mg-1 of each analyte 279 was optimized. Hexane/ethyl acetate (1:1, v/v) and alternatively, dichloromethane were 280 considered as extraction solvents because their suitability to extract substances of a 281 broad range of polarity has been reported previously (e.g. cannabinol derivatives [19] 282 and PCBs [17]). Similar recoveries were obtained for most analytes, except for TBP, 283 HHCB and TCS (Fig. 2). For these three analytes, the recoveries were significantly 284 higher using hexane/ethyl acetate (82 %, 103 %, 99 %) compared to dichloromethane 285 (33 %, 68 %, 41%). Correspondingly, hexane/ethyl acetate was preferred for LLE. 286 287 3.2 Method validation 288 The method performance for each target substance was characterized in accordance to 289 the FDA Guideline [29] determining the linearity of calibration of the instrumental 290 method, the limits of detection (MDL) and quantification (LOQ) of the entire method 291 using hair as well as its accuracy and precision . The calibration curves covered 292 16 21. Ebrahimi M, Eshaghi Z, Samadi F, Hosseini M-S (2011) Ionic liquid mediated sol-480 gel sorbents for hollow fibre solid-phase microextraction of pesticide residues in water 481 and hair samples. J Chromatogr A 1218:8313– 8321 482 22. Tsatsakis AM, Barbounis MG, Kavalakis M, Kokkinakis M, Terzi I, Tzatzarakis 483 MN (2010) Determination of dialkyl phosphates in human hair for the biomonitoring of 484 exposure to organophosphate pesticides. J Chromatogr B 878:1246–1252 485 23. Duca RC, Salquebre G, Hardy, E, Appenzeller BMR (2014) Comparison of solid 486 phase- and liquid/liquid-extraction for the purification of hair extract prior to multiclass 487 pesticides analysis. J Chromatogr B 955-956:98-107 488 24. Schramm KW (2008) Hair biomonitoring of organic pollutants. Chemosphere 489 72:1103–1111 490 25.a) Kucharska A, Covaci A, Vanermen G, Voorspoels S (2014) Development of a 491 broad spectrum method for measuring flame retardants- Overcoming the challenges of 492 non-invasive human biomonitoring studies. Anal Bioanal Chem 406:6665-6675 493 25.b) Kucharska A, Covaci A, Vanermen G, Voorspoels S (2015) Non-invasive 494 biomonitoring for PFRs and PBDEs: new insights in analysis of human hair externally 495 exposed to selected flame retardants. Sci Total Environ 505:1062-1071 496 26. Tadeo JL, Sánchez-Brunete C, Miguel E (2009) Determination of polybrominated 497 diphenyl ethers in human hair by gas chromatography–mass spectrometry.Talanta 498 78:138–143 499 27. Lertsirisopon R, Soda S, Sei K, Ike M (2009) Abiotic degradation of four phthalic 500 acid esters in aqueous phase under natural sunlight irradiation. J Environ Sci 21: 285–501 290 502 28.!!Alves A, Kucharska A, Erratico C, Xu F, Den Hond E, Koppen G, Vanermen G, 503 Covaci A, Voorspoels S (2014) Human biomonitoring of emerging pollutants through 504 non-invasive matrices: state of the art and future potential.Anal Bioanal Chem 505 406(17):4063-88. 506 29. U.S. Department of Health and Human Services, Food and Drug Administration 507 (2015) “Analytical Procedures and Methods Validation for Drugs and Biologics”, 508 Guidance for Industry, pp.3 509 17 30. Shanmugam G, Ramawamy BR, RadhakrishnanV, Tao H (2010) GC–MS method 510 for the determination of paraben preservatives in the human breast cancerous tissue. 511 Microchem J 96: 391-396 512 31.!Król S, Namieśnik J, Zabiegała B (2014) Occurrence and levels of polybrominated 513 diphenyl ethers (PBDEs) in house dust and hair samples from Northern Poland; an 514 assessment of human exposure. Chemosphere 110:91–96!515 32. Cousins AP, Holmgren T, Remberger M (2014) Emissions of two phthalate esters 516 and BDE 209 to indoor air and their impact on urban air quality. Sci Total Environ 470-517 471: 527-535 518 33. Api AM (2001) Toxicological profile of diethyl phthalate: a vehicle for fragrance 519 and cosmetic ingredients. Food Chem Toxicol. 39(2):97-108. 520 34. Altshul L, Covaci A, Hauser R (2004) The relationship between levels of PCBs and 521 pesticides in human hair and blood: preliminary results. Environ Health Perspect 522 112:1193–1199 523 35. Covaci A, Tutudaki M, Tsatsakis AM, Schepens P (2002) Hair analysis: another 524 approach for the assessment of human exposure to selected persistent organochlorine 525 pollutants. Chemosphere 46, 413–418 526 36. Chojnacka K, Górecka H, Górecki H (2006) The effect of age,sex, smoking habit 527 and hair color on the composition of hair.Environ Toxicol Pharmacol 22:52–57 528 37. Reiner JL, Wong CM, Arcaro KF, Kannan K (2007) Synthetic musk fragrances in 529 human milk from the United States. Environ Sci Technol 41:3815–20 530 38. Hond ED, Paulussen M, Geens T, Bruckers L, Baeyens W, David F, Dumont E, 531 Loots I, Morrens B, de Bellevaux BN, Nelen V, Schoeters G, Van Larebeke N, Covaci 532 A (2013) Biomarkers of human exposure to personal care products: results from the 533 Flemish Environment and Health Study (FLEHS 2007-2011). Sci Total Environ 463-534 464:102–110 535 536 537 538 539 540 18 541 542 543 Legend of figures 544 545 Figure 1. above: The pattern of the target substances in a commercial hair shampoo 546 extract (2 g extracted with ethyl acetate), below: Efficiency of the number of cleaning 547 processes using the procedure included in the protocol 548 Figure 2. Influence of the solvent extraction on the recoveries of the procedure (n = 3), 549 2 ng mg-1 of each analyte was used for spiking the hair 550 Figure 3. SIM chromatogram of a hair sample superimposed to a standard solution (500 551 ng mL-1) and to a blank of ethyl acetate 552 553 554 Tables 555 Table 1. Studied compounds, their log Kow values and target ions used for MS detection 556 (SIM) 557 Table 2. Performance parameters of the method, Rlow: Recovery of low concentration (2 558 ng mg-1 for TCS, BPA, DEHP, EtP and PrP; and 0.2 ng mg-1 for the rest of analytes), 559 Rhigh: Recovery of high concentration (2.8 ng mg-1 for TCS, BPA, DEHP, EtP and PrP; 560 and 2 ng mg-1 for the rest of analytes).*fortification with 100 -ng mL-1-(0.4 ng mg-1 561 hair) each component, negative value = signal suppression, positive = signal 562 enhancement 563 Table 3. Concentration of selected pollutants in four hair specimens (n=2) 564 Figure 1 0,00 0,50 1,00 1,50 2,00 2,50 3,00 3,50 4,00 ng/mg hair 1x clean 2x clean 1 10 100 1000 10000 100000 1000000 ng/ shampoo applied Shampoo extract 0 40 80 120 160 200 Recovery (%) Analyte Dichloromethane Hexene/Ethyl acetate (1:1) 15.50 16.00 16.50 17.00 17.50 18.00 18.50 4000 8000 12000 16000 20000 24000 Time (min) Abundance DBP (m/z 104) 11.50 12.00 12.50 13.00 13.50 14.00 14.50 0 10000 20000 30000 40000 50000 60000 70000 80000 90000 100000 Time (min) Abundance DMP (m/z 163) 14.00 14.20 14.40 14.60 14.80 15.00 15.20 5000 10000 15000 20000 25000 30000 35000 Time (min) Abundance DEET (m/z 119) 14.00 14.20 14.40 14.60 14.80 15.00 15.20 0 5000 10000 15000 20000 25000 30000 35000 Time (min) Abundance DEP (m/z 177) 14.40 14.60 14.80 15.00 15.20 0 1000 3000 5000 7000 9000 11000 13000 15000 17000 Time (min) Abundance TBP (m/z 211) 14.40 14.60 14.80 15.00 15.20 0 1000 3000 5000 7000 9000 11000 13000 15000 17000 Time (min) Abundance TCPP (m/z 277) 17.40 17.60 17.80 18.00 18.20 18.40 18.60 0 10000 30000 50000 70000 90000 110000 Time (min) Abundance HHCB (m/z 243) AHTN (m/z) 243 23.00 23.40 23.80 24.20 24.60 0 20000 60000 100000 140000 180000 220000 260000 300000 340000 Time (min) Abundance DEHP (m/z 167) 21.80 22.00 22.20 22.40 22.60 22.80 23.00 23.20 0 4000 8000 12000 16000 20000 24000 Time (min) Abundance TPP (m/z 326) Blank ethyl acetate Mixture standard solution (500 ng/ml) Hair sample EtP (m/z 121) 13.50 14.00 14.50 15.00 15.50 16.00 0 5000 10000 15000 20000 25000 30000 35000 40000 Time (min) Abundance PrP (m/z 121) Table 1. Studied compounds, their log Kow values and target ions used for MS detection (SIM) and gas chromatographic retention times Rt Target compound (abbreviation) CAS number Chemical class/ usage Log Kow a Target ions for SIM mode (m/z)b Rt (min) Internal Standard for quantitation Triclosan (TCS) 3380-34-5 chlorinated phenoxy phenol/ antimicrobial 4.76 218, 288, 290 21.3 4n-NP N,N-diethyl-meta- toluamide (DEET) 134-62-3 phenyl amide/ insect repellent 2.18 91, 119, 190 15.3 4n-NP Galaxolide (HHCB) 1222-05-5 polycyclic musk/ fragrance 5.9 243, 258, 213 18.7 4n-NP Tonalide (AHTN) 1506-02-1 5.2 243, 258 18.9 4n-NP Tri-n-butyl phosphate (TBP) 126-73-8 organophosphorus compound/ flame retardant 4 211, 155 16.2 DBP-d4 Triphenyl phosphate (TPP) 115-86-6 4.59 325, 326 24.0 DBP-d4 Tris(chloro-2-propyl) phosphate (TCPP) 13674-84-5 2.89 277, 279 18.1 DBP-d4 Bisphenol A (BPA) 80-05-7 phenolic compounds/ antioxidant 3.32 213, 228 21.8 BPA-d16 Dimethyl phthalate (DMP) 131-11-3 phthalates/ plasticizer 1.6 163, 77 13.7 DBP-d4 Diethyl phthalate (DEP) 84-66-2 2.42 149, 177 15.5 DBP-d4 Di-n-butyl phthalate (DBP) 84-74-2 4.5 104, 149 19.7 DBP-d4 Di-2-ethylhexyl phthalate (DEHP) 117-81-7 7.6 149, 167 25.2 DBP-d4 Ethylparaben (EtP) 120-47-8 hydroxyl-benzoic acid ester/ preservative 2.47 121, 138, 166 14.5 4n-NP n-Propylparaben (PrP) 94-13-3 3.04 121, 138, 180 15.8 4n-NP Internal standards 4n-Nonylphenol (n-NP)a 25154-52-3 5.71 107, 220 18.7 Bisphenol A d16 (BPA- d16)a 96210-87-6 224, 242 21.9 Di-n-butyl phthalate-d4 (DBP-d4)a 93952-11-5 153, 223 19.6 a: Experimental values from Database ChemSpider; b: quantifier ion bold Table 2. Performance parameters of the method, Rlow: Recovery of low concentration (2 ng mg-1 for TCS, BPA, DEHP, EtP and PrP; and 0.2 ng mg-1 for the rest of analytes), Rhigh: Recovery of high concentration (2.8 ng mg-1 for TCS, BPA, DEHP, EtP and PrP; and 2 ng mg-1 for the rest of analytes).*fortification with 100 -ng mL-1- (0.4 ng mg-1 hair) each component, negative value = signal suppression, positive = signal enhancement Compound r2 Linear range (MDL to…) (ng mg-1) Method precision (RSD %) (n=3) (2 ng mg-1 each) MDL (ng mg-1) LOQ (ng mg-1) Recovery R in % (± % RSD, n=3) Matrix effect* (%) Intra-day Inter-day Rlow Rhigh TCS 0.9948 4 2 5 0.276 0.920 99 (4) 80 (6) -5 DEET 0.9957 0.8 10 12 0.022 0.072 109 (12) 111 (14) 12 HHCB 0.9992 4 2 10 0.002 0.007 103 (13) 90 (13) 5 AHTN 0.9998 0.8 9 9 0.002 0.007 106 (3) 104 (13) 13 TBP 0.9966 0.8 2 5 0.031 0.102 77 (10) 82 (9) -5 TPP 0.9986 0.8 6 8 0.031 0.102 154 (8) 134 (2) 44 TCPP 0.9995 4 5 8 0.021 0.069 104 (1) 93 (10) 6 BPA 0.9988 4 9 13 0.041 0.137 51 (20) 72 (14) -10 DMP 0.9974 0.8 9 7 0.001 0.005 83 (3) 105 (10) -3 DEP 0.9991 4 2 6 0.019 0.062 93 (9) 110 (13) -4 DBP 0.9985 4 9 10 0.020 0.065 61 (13) 87 (13) -15 DEHP 0.9940 80 4 8 0.025 0.080 106 (14) 120 (9) 32 EtP 0.9940 4 8 12 0.149 0.498 53 (15) 32 (21) 15 PrP 0.9971 4 7 16 0.292 0.972 104 (8) 110 (12) 19 Table 3. Concentration of selected pollutants in four hair specimens (n=2) Compound Concentration (ng mg-1) Sample 1 Sample 2 Sample 3 Sample 4 TCS <MDL <MDL <MDL <MDL <MDL <MDL 1.84 1.87 DEET 0.14 0.13 0.07* 0.07* 0.19 0.22 0.79 0.73 HHCB 4.00 4.20 0.82 0.84 11.95 12.7 2.68 2.67 AHTN 0.23 0.29 0.12 0.12 1.16 1.26 0.45 0.51 <MDL: below method limit of detection; * below LOQ (see Tab. 2) TBP <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL TPP 0.78 0.91 0.14 0.10 0.21 0.20 0.78 0.67 TCPP 1.00 1.09 0.19 0.18 0.48 0.57 1.70 1.55 BPA <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL DMP 0.13 0.13 0.06 0.07 0.02 0.02 0.36 0.36 DEP 2.41 2.40 0.49 0.44 0.06* 0.06* 15.31 15.42 DBP 6.25 6.48 5.23 5.16 1.04 1.49 7.08 6.74 DEHP 42.91 46.31 34.96 32.52 43.21 58.78 55.04 46.33 EtP 1.83 1.98 0.81 0.84 <MDL <MDL <MDL <MDL PrP 1.42 1.52 0.42* 0.40* <MDL <MDL <MDL <MDL S7 Figure S1. Total ion chromatograms of GC-MS(SIM) analysis of hair extracts (non-spiked) a) treated with MeOH/TFA and extracted with ethyl acetate and b) treated with NaOH solution, then adjusted to pH 3 with acetic acid and extracted with ethyl acetate. Table S2. Blanks of the solvents used for the individual sample preparation steps and of the complete method without hair (n=2). Values are given as amount per injection. (n.d. = not detected) MeOH/TFA hexane/ethylacetate Complete method pg/µL (±%RSD) pg/µL (±%RSD) pg/µL (±%RSD) TCS n.d. 1 (6) n.d. DEET n.d. n.d. n.d. HHCB 1 (8) 1 (5) 1(6) AHTN 1 (6) n.d. n.d. TBP n.d. n.d. n.d. TPP n.d. 1(1) 1(3) TCPP 6 (2) 6 (1) 7(2) BPA n.d. 2 (10) 2(8) DMP n.d. n.d. n.d. DEP 1 (4) 1 (4) 2 (4) DBP 11 (5) 8 (6) 13(6) DEHP 14 (5) 13 (5) 16(3) EtP n.d. n.d. n.d. PrP n.d. n.d. n.d. S8 S9 References for SI Behrooz RD, Barghi M, Bahramifar N, Esmaili-Sari A (2012) Chemosphere 86:235–241 Chen H, Xiang P, Shen M (2014) J Forensic Leg Med 22:62–67 Cuong LP, Evgen’ev MI, Gumerov FM (2012) J Supercritical Fluids 61:86–91 Eshaghi Z, Ebrahimi M, Hosseini M-S (2011) J Chromatogr A 1218:3400–3406 Grova N, Salquèbre G, Appenzeller BMR (2013) Anal Bioanal Chem 405:8897–8911 John C, Ghosh P, Varshney KM, Kaur S, Shukla SK, Satyanarayana S (2014) J Liquid Chromatogr Related Technol 37:1917–1928 Kataoka H, Inoue T, Saito K, Kato H, Masuda K (2013) Anal Chim Acta 786:54–60 Kavalakis MP, Tzatzarakis MN, Theodoropoulou EP, Barbounis EG, Tsakalof AK, Tsatsakis AM (2013) Chemosphere 93:2612-2620 Kim J, Ji D, Kang S, Park M, Yang W, Kim E, Choi H, Lee S (2014b) J Pharm Biomed Anal 89:99– 105 Li J, Guo F, Wang Y, Zhang J, Zhong Y, Zhao Y, Wu Y (2013) Environ Int 53:47–52 Liang B, Liu X, Hou J, Liang G, Gong W, Xu D, Zhang L (2014) Environ Pollut 185:10–15 Lu D, Feng C, Lin Y, Wang D, Ip HSS, Qiu X, Wang G, She J (2014) Chemosphere 114:327–336 Ma J, Cheng J, Wang W, Kunisue T, Wu M, Kannan K (2011) J Haz Mat 186:1966–1971 Man YB, Chan JKY, ShengWang H, Wu SC, HungWong M (2014) Environ Int 65:73–82 Mercadante R, Polledri E, Giavini E, Menegola E, Bertazzi PA, Fustinoni S (2012) Toxicol Letters 210:169–173 Montesano C, Johansen SS, Nielsen MKK (2014) J Pharm Biomed Anal 88:295–306 Perez F, Llorca M, Farré M, Barceló D (2012) Anal Bioanal Chem 402:2369–2378 S10 Posecion Jr N, Ostrea Jr E, Bielawski D, Corrion M, Seagraves J, Jin Y (2006) Chromatographia 64:681– 687 Salquèbre G, Schummer C, Millet M, Briand O, Appenzeller BMR (2012) Anal Chim Acta 710:65–74 Schummer C, Appenzeller BMR, Millet M, Wennig R (2009) J Chromatogr A 1216: 6012–6019 Schummer C, Salquèbre G, Briand O, Millet M, Appenzeller BMR (2012) Toxicol Letters 210:203–210 Sporkert F, Pragst F (2000) J Chromatogr B 746:255–264 Toriba A, Kuramae Y, Chetiyanukornkul T, Kizu R, Makino T, Nakazawa H, Hayakawa K (2003) Biomed Chromatogr 17:126–132 Tsatsakis AM, Tzatzarakis MN, Tutudaki M (2008a) Forensic Sci Int 176:67–71 Tsatsakis AM, Tzatzarakis MN, Tutudaki M, Babatsikou F, Alegakis AK, Koutis C (2008b) Hum Exp Toxicol 27:933–940 Wielgomas B, Czarnowski W, Jansen EHJM (2012) Chemosphere 89: 975–981 Zhao G, Wang Z, Dong MH, Rao K, Luo J, Wang D, Zha J, Huang S, Xu Y, Ma M (2008) Sci Total Environ 397:46–57 Zheng J, Luo X-J, Yuan J-G, Wang J, Wang Y-T, Chen S-J, Mai B-X, Yang Z-Y (2011) Environ Pollut 159:3706–3713