Cetyltrimethylammonium Bromide-Coated Fe3O4 MagneticNanoparticles for Analysis of 15 Trace Polycyclic AromaticHydrocarbons in Aquatic Environments by Ultraperformance, LiquidChromatography With Fluorescence DetectionHao Wang,†,‡,⊥ Xiaoli Zhao,*,†,⊥ Wei Meng,*,† Peifang Wang,§ Fengchang Wu,† Zhi Tang,†

Xuejiao Han,† and John P. Giesy∥

†State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences,Beijing 100012, China‡College of Water Sciences, Beijing Normal University, Beijing 100875, China§Key Laboratory of Integrated Regulation and Resource Development on Shallow Lakes, Ministry of Education, College ofEnvironment, Hohai University, Nanjing 210098, China∥Department of Veterinary Biomedical Science and Toxicology Centre, University of Saskatchewan, 44 Campus Drive, Saskatoon,Saskatchewan S7N 5B3, Canada

*S Supporting Information

ABSTRACT: Accurate determination of polycyclic aromatichydrocarbons (PAHs) in surface waters is necessary forprotection of the environment from adverse effects that canoccur at concentrations which require preconcentration to bedetected. In this study, an effective solid phase extraction(SPE) method based on cetyltrimethylammonium bromide(CTAB)-coated Fe3O4 magnetic nanoparticles (MNPs) wasdeveloped for extraction of trace quantities of PAHs fromnatural waters. An enrichment factor of 800 was achievedwithin 5 min by use of 100 mg of Fe3O4 MNPs and 50 mg ofCTAB. Compared with conventional liquid−liquid extraction(LLE), C18 SPE cartridge and some newly developedmethods, the SPE to determine bioaccessible fraction wasmore convenient, efficient, time-saving, and cost-effective. To evaluate the performance of this novel sorbent, five natural samplesincluding rainwater, river waters, wastewater, and tap water spiked with 15 PAHs were analyzed by use of ultraperformance,liquid chromatography (UPLC) with fluorescence detection (FLD). Limits of determination (LOD) of PAHs (log Kow ≥ 4.46)ranged from 0.4 to 10.3 ng/L, with mean recoveries of 87.95 ± 16.16, 85.92 ± 10.19, 82.89 ± 5.25, 78.90 ± 9.90, and 59.23 ±3.10% for rainwater, upstream and downstream river water, wastewater, and tap water, respectively. However, the effect ofdissolved organic matter (DOM) on recovery of PAHs varied among matrixes. Because of electrostatic adsorption andhydrophobicity, DOM promoted adsorption of Fe3O4 MNPs to PAHs from samples of water from the field. This result wasdifferent than the effect of DOM under laboratory conditions. Because of competitive adsorption with the site of action on thesurface of Fe3O4 MNPs for CTAB, recoveries of PAHs were inversely proportional to concentrations of Ca

2+ and Mg2+. Thisnovel sorbent based on nanomaterials was effective at removing PAHs at environmentally relevant concentrations from waterscontaining relevant concentrations of both naturally occurring organic matter and hardness metals.

Polycyclic aromatic hydrocarbons (PAHs), of which thereare thousands of possible variations in the environment,consist of two or more fused rings without heteroatoms, withsome PAHs alkyl substituted.1,2 Most PAHs are released intothe environment during leaks or spills during extraction,transport, and refinery of petroleum hydrocarbons or duringcombustion of wood biofuels and fossil fuels such as coal andpetroleum and other paths, such as cooking, burning ofdomestic wastes.2−6 Because of their ubiquitous presence,chemical stability, potential for bioaccumulation, and carcino-

genic potential, PAHs in the environment have attractedattention globally and some have been listed as prioritypollutants by the United States Environmental ProtectionAgency (U.S. EPA).7−10

Concentrations of PAHs in ground and surface waters,sediments, and the atmosphere are increasing due to activities

Received: March 20, 2015Accepted: July 8, 2015Published: July 8, 2015

Article

pubs.acs.org/ac

© 2015 American Chemical Society 7667 DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

pubs.acs.org/achttp://dx.doi.org/10.1021/acs.analchem.5b01077

of humans.11−13 There is a need to monitor PAHs, but they canoccur at concentrations ranging from pg/L to ng/L, which, dueto their propensity to be bioaccumulated, have potential tocause adverse effects, yet be less than the LOD of standardanalytical techniques. Moreover, various environmental factors,such as chemical components, physical condition, can affectperformances of pretreatment techniques.14−16 Accuratequantification of trace concentrations of PAHs in environ-mental matrixes, especially in water at environmentally andtoxicologically relevant concentrations is needed. To achievethe required LOD, samples are concentrated and separatedfrom environmental matrixes by use of methods includingliquid−liquid extraction (LLE), solid phase extraction (SPE)and solid phase microextraction (SPME) (Table S1 in theSupporting Information). Each of these methods hasadvantages as well as limitations. Some are time-consumingand relatively expensive and result in large amounts of wastesolvents.17−23 Use of a solid adsorbent based on C18 cartridges,to selectively preconcentrate PAHs from environmentalmatrixes uses lesser amounts of organic solvents than doesLLE. An alternative to these more traditional approaches is theuse of adsorbents attached to nanoparticles that can beseparated by used of a magnetic field. One such process usescetyltrimethylammonium bromide (CTAB) coated ontomagnetic nanoparticles of iron oxide (Fe3O4) (Fe3O4−CTABMNPs). This method has excellent capacity to separate PAHfrom environmental matrixes, especially water, and is lessexpensive and quicker than the traditionally used methods thatemploy C18 disks or cartridges as the solid phase.18,23 The factthat larger volumes of water can be treated without break-through or interferences makes use of nanoparticles, such asnanocarbon, C−Fe3O4 and Ag−Fe3O4 an attractive approach toobtain lesser LODs for PAHs. Some of these solid phases mightbe unsuitable for treatment of large volumes of sample andcould be time-consuming to separate with sufficient recov-eries.24,19 The superparamagnetic properties of magneticnanoparticles (MNPs) contribute to their rapid magnetizationand separation from aqueous phases by use of external,magnetic fields. When coated with appropriate functionalgroups MNPs can enrich contaminants from large volumes ofwater.25 Additionally, advantages of MNPs including Fe3O4 andγ-Fe2O3 are their convenience, biocompatibility, and econom-ical synthesis by use of chemical coprecipitation.26−28 Whileother adsorbents such as stir bars or artificial fibers werecomplicated to produce, Fe3O4−CTAB MNPs can easily besynthesized. Because MNPs are magnetic, small particles with alarge total surface area, they are effective for rapid andquantitative adsorption of PAHs and can easily be collectedinto an organic solvent by use of a magnetic field.29,30 Onceseparated from water, the organics trapped on the surface canbe extracted by use of an organic solvent. Thus, Fe3O4−CTAB

MNPs have promise as a solid phase for extraction of PAHs inwater.Enrichment of analytes by use of MNPs is improved by

modification of surfaces of MNPs by addition of functionalgroups, such as coupling agents, surfactants, or noble metals.31

Ionic surfactants can attach homogeneously onto chargedsurfaces of MNPs by chemical self-assembly and due to theirhydrophilic groups, form hemimicelles, mixed hemimicelles oradmicelles.32 The mixed hemimicelles promote effectiveadsorption of PAHs by hydrophobic interaction with hydro-carbon moieties (Figure 1).33−35 MNPs as a substratum forsorbents successfully avoid time-consuming, blocking problemsduring conventional SPE and, relative to LLE, also reduces theamount of organic solvent used.Those PAHs, which have been designated as priority

pollutants by the U.S. EPA, were quantified by use ofultraperformance liquid chromatography in tandem withfluoresence detection (UPLC-FLD), which can convinientlyquantify all 15 PAHs within 30 min, while maintainingsufficient sensitivity, to attain LODs equivalent to the mostcommonly used analytical procedures. To our knowledge, thisis the first report of utilization of MNPs for preconcentration oftrace concentrations of PAHs from natural water.The objective of the present study was to develop a rapid,

simple, cost-effective SPE procedure using Fe3O4 MNPscoupled with UPLC-FLD for quantification of trace concen-trations of the 15 priority PAHs, designated by U.S. EPA, inwater. Several key factors that could influence recoveries andaccuracies and precision of determination of concentrations ofPAHs isolated from natural waters, such as pH, breakthroughvolume, type, and amounts of solvents used to elute analytesfrom the solid phase were determined. Effects of DOM, such asfulvic acid (FA) and humic acid (HA), and ions including Ca2+

and Mg2+ were investigated. Finally, the method was validatedby application to five environmental waters.

■ EXPERIMENTAL SECTIONReagents and Chemicals. The standard solution contain-

ing Naphthalene (Nap), Acenaphthylene (Ace), Fluorene(Flo), Phenanthrene (Phe), Anthracene (Ant), Fluoranthene(Fla), Pyrene (Pyr), Chrysene (Chr), Benzo(a)anthracene(Baa), Benzo(b)fluoranthene (Bbf), Benzo(k)fluoranthene(Bkf), Benzo(a)pyrene (Bap), Dibenzo(a,h)anthracene(DahA), Indeno(1,2,3-cd)pyrene (Icdp), and Benzo(ghi)-perylene (BghiP) (2000 mg/L) was purchased from Sigma-Aldrich (St. Louis. MO) and diluted to 1 mg/L as the stocksolution for use in spiking waters. Samples were kept in thedark at 4 °C until used. Acetonitrile (ACN), Dichloromethane(DCM), and Acetone (DMK) were HPLC grade, andpurchased from Fisher Scientific Corporation (Fair Lawn,NJ). Acetic Acid (AcOH, A.R. grade) and hydrochloric acid

Figure 1. Schematic representation of mechanism of adsorption of PAHs by Fe3O4−CTAB MNPs.

Analytical Chemistry Article

DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

7668

http://dx.doi.org/10.1021/acs.analchem.5b01077

(A.R. grade) were purchased from Xilong Chemical Corpo-ration (Guangdong, China).Cetyltrimethylammonium bromide (CTAB, A.R. grade), (1-

hexadecy) pyridinium chloride monohydrate (CPC, A.R.grade), ferric chloride (FeCl3·4H2O, A.R. grade), ferrouschloride (FeCl2·6H2O, A. R. grade), NaOH (sodium hydroxide,A.R. grade) were purchased from Sino-pharm ChemicalReagent Co., Ltd. (Beijing, China). Fulvic Acid (NordicAquatic Fulvic Acid Reference 1R105F) and Humic Acid(Leonardite Humic Acid Standard 1S104H) were purchasedfrom the International Humic Substances Society (Colorado).Synthetic, experimental ultrapure water were made from aMillipore Integral 5 water purification system (Merck,Germany).The Multi N/C3100 TOC (Analytikjena, Germany) analyzer

was employed to determine the concentration of DOM insamples, and concentrations of Ca2+, Mg2+ in samples weredetermined by use of an Ion Chromatography System 1000(Dionex Co.).Collection of Samples. Samples of surface waters

investigated included two samples of river water collectedfrom the upstream stagnant pool (low-speed flow) anddownstream reach (high-speed flow) of the Qing River(Chinese, Qinghe), one sample of rainwater (August, 2014),one sample of wastewater collected from the Qing Riverwastewater treatment plant (Haidian district, Beijing), and onesample of tap water sample from our laboratory (Chaoyangdistrict, Beijing). The total volume of each sample was 10 L andwas collected with wide-mouth jars after being cleaned withchromic acid and ultrapure water. Collected samples wereimmediately filtered through a 0.45 μm glass fiber filter(combusted at 450 °C for 4 h) combined with a filtrationdevice to remove suspended solids and stored at 4 °C. Allsamples were analyzed within 5 days.SPE Procedure and Sample Analysis. Fe3O4 MNPs were

synthesized by coprecipitation by use of a previously describedmethod.33 A 5 mL aliquant of Fe3O4 MNPs (20 mg/mL) and10 mL of CTAB (5 mg/mL) were added to 800 mL of water,either a synthetic or natural sample, and pH adjusted to 10.0

and then sonicated for 1 min. After standing for 10 min on anNd−Fe−B magnet, Fe3O4 MNPs coated with CTAB wereisolated from solution and the supernatant decanted.Preconcentrated PAHs associated with CTAB-coated Fe3O4MNPs were eluted with 2 mL of ACN solution mixed with 5%acetic acid (AcOH) (v/v) for 5 times. The eluent containingPAHs was dried under a stream of nitrogen at 45 °C anddiluted to 1 mL with ACN.Ultra performance liquid chromatography coupled with

fluorescence detection (UPLC-FLD, Waters, Massachusetts)was employed to separate, identify, and quantify individualPAHs. A CORTECS C18 column (100 mm × 2.1 mm i.d., withparticle diameter of 1.6 μm, Waters, Massachusetts) was usedto separate 15 EPA PAHs. The mobile phases were ACN andultrapure water at a flow rate of 0.4 mL/min, with an injectionvolume of 2 μL. The mobile phase was an ACN/water gradientprogram (45% ACN at start, 9.0 min hold, 15.0 min lineargradient to 60%, 19.0 min linear gradient to 67%, 22.8 minlinear gradient to 77%, 26.5 min linear gradient to 100%, 28.0min linear gradient to 45%). Excitation wavelengths were 221,289, 252, 234, 265, and 300 nm, and emission wavelengthswere 337, 322, 377, 448, 390, and 412 nm for 0−5.5, 5.5−9.0,9.0−12.0, 12.0−15.0, 15.0−18.0, 18.0−28.0 min, respectively.Calculations for quantification of PAHs were accomplished byuse of Waters Power 2.0 software. Limits of detection (LOD)for 15 PAHs were determined as being 3 times the signal-noiseratio. PAHs were quantified by use of an external standardcurve with a linear working range of 0.1−400 ng/L. Theanalytical parameters of the proposed method for PAHs areshown (Table 1).

■ RESULTS AND DISCUSSIONCharacterization of Fe3O4−CTAB MNPs. Fe3O4 MNPs

were characterized by use of transmission electronic micros-copy (TEM) (Hitachi, Japan) at 80 kV. Particles were generallyuniform with a diameter of approximately 10 nm (Figure 2).Hysteresis was not observed, and the largest saturationmagnetism of Fe3O4−CTAB MNPs was 58.7 emu/g,

25

Table 1. Analytical Parameters of the Proposed Method

PAHs abbreviationwater solubility

(g/m3)logKow

arange of concn

(ng/L)slope ± SD [mV L

/ng]correlation

coefficient (r2)LODb

(ng/L)RSD (%)(n = 3)

naphthalene Nap 30.2 3.45 0.1−400 20.6acenaphthene Ace 3.93 4.22 0.1−400 (0.96 ± 0.21) × 103 0.782 0.7 18.0fluorene Flo 1.9 4.38 0.1−400 (1.91 ± 0.19) × 104 0.888 3.9 10.2phenanthrene Phe 1.18 4.46 0.1−400 (7.84 ± 0.06) × 103 0.992 10.3 3.9anthracene Ant 0.076 4.54 0.1−400 (1.25 ± 0.03) × 104 0.999 0.5 6.9fluoranthene Fla 0.26 5.2 0.1−400 (4.50 ± 0.04) × 103 1.000 3.5 4.4pyrene Pyr 0.135 5.3 0.1−400 (1.63 ± 0.02) × 103 1.000 6.4 2.4chrysene Chr 0.0019 5.61 0.1−400 (5.09 ± 0.12) × 104 0.998 1.0 2.4benzo[b]fluoranthene

Bbf 0.014 5.78 0.1−400 (4.49 ± 0.15) × 104 0.993 0.4 5.2

benzo(a)anthracene

Baa 0.011 5.91 0.1−400 (3.07 ± 0.12) × 104 0.996 1.7 4.3

benzo(k)fluoranthene

Bkf 0.008 6.2 0.1−400 (3.58 ± 0.11) × 105 0.995 1.2 7.5

benzo[a]pyrene Bap 0.0038 6.35 0.1−400 (2.09 ± 0.11) × 105 0.995 1.7 6.0dibenz[a,h]anthracene

DahA 0.0005 6.75 0.1−400 (9.63 ± 0.20) × 104 0.990 3.3 4.2

indeno[1,2,3-cd]pyrene

IcdP 0.0005 6.51 0.1−400 (1.18 ± 0.06) × 105 0.996 3.5 4.3

benzo[ghi]perylene BghiP 0.0003 6.9 0.1−400 (1.20 ± 0.02) × 103 0.999 2.3 8.2aWater solubilities and log Kow of 15 PAHs are quoted from Huckins et al.

57 bDetection limits were calculated by using S/N = 3.

Analytical Chemistry Article

DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

7669

http://dx.doi.org/10.1021/acs.analchem.5b01077

indicating their superparamagnetism excellence for rapidseparation.

Effect of Solution pH. pH is a key factor affectingadsorption of PAHs by Fe3O4−CTAB MNPs and in this studyrecoveries of PAHs were directly proportional to pH (Figure3a), reaching a maximum at pH of approximately 10.0.

Surfaces of Fe3O4 MNPs are negatively charged when the pHwas greater than the pH where the zeta potential of MNPs = 0,which is defined as the point of zero charge (PZC).25,36

Cationic surfactants can attach to surfaces of nanoparticles bystrong electrostatic attraction to form hydrophobic hemi-micelle, which creates a hydrophobic interaction with organicpollutants, such as PAHs (Figure 1). Octanol−water partitioncoefficients (Kow) of the 15 targeted PAHs were directlyproportional to molecular mass (Table 1), such that adsorptionof PAHs was inversely proportional to their solubilities inwater. However, recoveries of Nap, Ace, and Flo (log Kow ≤4.46) were slightly less than those of other PAHs studied. Thisresult might be due to their greater solubilities in water andgreater volatility. For PAHs with log Kow values greater than4.46, recoveries of greater than 80% were observed.

Effects of Amounts of Fe3O4 MNPs and Surfactant.CTAB was employed as the surface modifier at a ratio 1:2 (w/w) of CTAB and Fe3O4 MNPs, compared with CPC, and thedetailed information on this (Figure S1 in the SupportingInformation) and the effect of sample volume on recovery ofPAHs (Figure S2 in the Supporting Information) are providedin the Supporting Information. In order to reduce consumptionof adsorbents, the effect of amount of Fe3O4 MNPs and CTABon recoveries of PAHs was determined. Recoveries of 15 PAHsreached maxima separately as a function of the amount ofFe3O4−CTAB MNPs added (Figure 3b). However, PAHs withgreater Kow reached the maxima faster than those with lesserKow. This result might be due to stronger affinities of Fe3O4−CTAB MNPs for chemicals with greater hydrophobicity thanthose with lesser Kow. The optimal amount of adsorbents usedwas the mean of the additive amounts of adsorbents for thegreatest recovery of each PAH. On the basis of this analysis,

Figure 2. Transmission electron microscopy (TEM) image of Fe3O4MNPs.

Figure 3. Recoveries of PAHs as functions of Fe3O4 MNPs as pH (a), ratio of 2:1 to CTAB (b), standing time (c), and 4 kinds of eluents (d) inbatch mode. Sample volume, 800 mL; volume of ACN, 10 mL. (a, c, and d) 100 mg of Fe3O4 MNPs; surfactant, 50 mg of CTAB; (b) pH 10.0.

Analytical Chemistry Article

DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

7670

http://dx.doi.org/10.1021/acs.analchem.5b01077

100 mg of Fe3O4 and 50 mg of CTAB were chosen as theoptimal amounts to use. Thus, in this study, amounts ofadsorbents were optimized to be more efficient and lesswasteful than is possible in studies using cartridges.Optimization of Standing Time. Duration of separation is

a key factor for pretreatment methods. Shorter paths ofadsorption, which result in faster equilibrium are positivecharacteristics of nanoadsorbents. A duration of approximately5 min was determined to be sufficient to obtain maximumenrichment of the 15 PAHs studied (Figure 3c). This was aclear advantage compared with conventional pretreatmentmethods, such as LLE and C18 SPE cartridge, which hadtimes to maxima of 72−1 080 and 125−333 min, respec-tively.17,18,37,38 Some other methods of preconcentrationrequire a minimum of 30 min to reach their maxima. Adetailed comparison is shown in Table S1 in the SupportingInformation.Optimization of Desorption Conditions. PAHs were

eluted by mass action and destruction of hemimicelles withorganic solvents. ACN and DMK were used separately to elutePAHs; 5% DCM was added to increase their capacity, and 5%AcOH was also added to destruct the mixed hemimicellesformed by CTAB under alkaline conditions to promotedesorption of PAHs. ACN was better for eluting PAHs thanwas acetone (Figure 3d). This might have been due to greatersolubility of CTAB in ACN than acetone. In this study, 10 mL(2 mL for 5 times) ACN was sufficient to ensure sufficientrecoveries of PAHs.Effects of Fulvic Acid and Humic Acid. DOM is complex

and comprises a variety of organic substances including FA,HA, carbohydrates, sugars, amino acids, proteins, inorganicions, among others.39 Since FA and HA are the maincomponents of DOM in the environment40 and carry a varietyof functional groups, they can interfere with adsorption ofPAHs by Fe3O4−CTAB MNPs, mainly through electrostaticinteraction or hydrophobic interaction due to their differentconcentrations.41

Concentrations of total organic carbon (TOC) in aquaticenvironments ranges for 20−100 mg/L, depending on soiltypes in the watershed, climate, and hydrologic conditions,39

thus concentrations of FA and HA considered in this studyranged from 0 to 120 mg/L and their effects on recoveries of

PAHs were assessed separately. FA and HA had similar effectson recoveries of PAHs by Fe3O4−CTAB MNPs (Figure 4a,b).The effect of DOM on recovery of PAHs could be divided intothree stages: the recovery of PAHs initially declined with theaddition of FA for 0−40 mg/L (HA for 0−50 mg/L), thenincreased with the addition of FA for 40−80 mg/L (HA for50−100 mg/L), and slightly decreased at the concentration ofFA for 80 mg/L (HA for 100 mg/L). The stages observed: (1)competitive adsorption of DOM with Fe3O4 MNPs to bindCTAB and PAHs, resulted in lesser recoveries of PAHs. DOMis generally electronegative at pH 10.0, because there are morenegative charged functional groups than positive groups onDOM.42−44 However, the majority of added DOM partitionedinto the aqueous phase, adsorbed CTAB to form hemimicellestructure by electrostatic interactions and had a competitionwith Fe3O4−CTAB MNPs for adsorbing PAHs by thehydrophobic effect;36,43 and meanwhile, less DOM adsorbedon the surface of Fe3O4−CTAB MNPs and competed withPAHs for adsorption sites.45−47 The above both effects ofDOM resulted in the decrease of recoveries of PAHs byFe3O4−CTAB MNP. Thus, PAHs adsorbed by DOM orDOM-CTAB would not have been extracted by Fe3O4 MNPsdue to the electrostatic repulsion between DOM or DOM-CTAB and MNPs (Figure 5a,b). These combinations of pH-dependent phenomena resulted in recoveries of PAHs beinginversely proportional to concentration of DOM until aconcentration of about 40 mg/L for FA and 50 mg/L forHA. (2) Recoveries of PAHs were increasing until theconcentration of FA reached about 80 mg/L or for HA 100mg/L. This phenomenon has been less reported. And at thisstage, the newly added FA and HA also adsorbed the CTABand PAHs, which might reduce electrostatic repulsion betweenthe DOM complex and Fe3O4−CTAB MNPs. As a result,DOM complexes would be adsorbed onto surfaces of Fe3O4−CTAB MNPs, and some polymers such as flocculation48,49

were gradually formed with addition of FA and HA due toelectrostatic and hydrophobic interactions. It has been reportedof removal of DOM from water with bentonite andbenzyltrimethylammonium bromide by the flocculation reac-tion (Figure 5c).50 (3) When more than 80 mg/L FA and 100mg/L HA was added to the solution, newly added DOM wouldalso compete with previously added DOM and Fe3O4−CTAB

Figure 4. Recoveries of PAHs as functions of FA (a) and HA (b) in batch mode. Amount of metal oxide: 100 mg of Fe3O4 MNPs. Surfactant, 50 mgCTAB; pH, 10.0; sample volume, 800 mL. Volume of ACN, 10 mL.

Analytical Chemistry Article

DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

7671

http://dx.doi.org/10.1021/acs.analchem.5b01077

MNPs for adsorbing PAHs. One possible mechanism is thatFe3O4−CTAB MNPs and DOM or DOM-CTAB would bothbe more electronegative because the competitive adsorption ofthe new added DOM to CTAB and PAHs from previouslyformed polymers. Therefore, electrostatic repulsion betweenDOM complexes and Fe3O4−CTAB MNPs was recovered dueto electrostatic repulsion regenerated by their electronegativity,which decreased adsorption of PAHs by Fe3O4−CTAB MNPs(Figure 5d).Method Parameters. Calibration curves were run for 15

PAH in the range of 0.1−400 ng/L. Coefficients ofdetermination (r2) for PAHs (log Kow ≥ 4.46) were all greaterthan 0.99, and LODs were calculated by using 3 times thesignal-to-noise and ranges from 0.4 to 10.3 ng/L, whichindicated suitability of MNPs for preconcentration of neutral,

hydrophobic organic pollutants, such as PAHs. However, lesserrecoveries of Nap, Ace, and Flo were likely due to their greatervolatilities and solubilities in water. Therefore, coefficients ofdetermination (r2) for Ace and Flo were 0.78 and 0.88, but anadequate standard curve could not be obtained for Nap.

Analyses of Environmental Water Samples. Reprodu-cibility of recoveries of PAHs by Fe3O4−CTAB MNPs wasinvestigated by spiking known amounts of the standard mixtureof PAHs into samples of rainwater, two samples of river water,wastewater, and tap water. Figure 6 shows the chromatogramsof PAHs in the water samples of the Qing river by using UPLC-FLD. Effects of physical-chemical factors were also investigated.Total concentrations of 15 PAHs in the samples for rainwater,upstream and downstream river water, wastewater, and tapwater were 924.3 ± 80.71, 1206.99 ± 89.20, 1669.91 ± 148.65,2232.47 ± 38.12, and 305.54 ± 25.07 ng/L, respectively.Recoveries of PAHs were about 90, 80, 70 and 60% inrainwater, both samples of river water, wastewater, and tapwater, respectively (Table 2).Concentrations of Ca2+, Mg2+, and DOM in natural water

affected performance of Fe3O4−CTAB MNPs on extraction ofPAHs. There could be two aspects of their interactions: (1)Fe3O4 MNPs were electronegative at pH 10.0, and acompetitive adsorption of metal ions with CTAB on surfacesof Fe3O4 MNPs prevented formation of mixed hemimicellesand resulted in poorer recoveries of PAHs51,52 and (2) becauseof strong complexation between Ca2+ and Mg2+ and somefunctional groups (carboxyl, phenolic hydroxyl) ofDOM,42,46−48 which was also due in part to lesser adsorptionof CTAB to DOM. Thus, presence of DOM would indirectlyreduce the effect of metal irons on adsorption activity of MNPsto PAHs, as well as increasing development of mixedhemimicelles, which increased adsorption of PAHs byFe3O4−CTAB MNPs.53,54Concentrations of TOC in five environmental waters are

23.22 ± 0.62, 8.41 ± 0.32 and 3.58 ± 0.44 mg/L for wastewaterand upstream and downstream river water, respectively, with noDOC detected in tap water or rainwater (Table S2 in theSupporting Information). Recoveries of PAHs from rainwaterwere greater due to lesser concentrations of Ca2+, Mg2+, and

Figure 5. Schematic representation of interactions among CTAB,DOM, PAHs, Fe3O4 MNPs (a) from adsorption of CTAB onto Fe3O4MNPs, (b) from adsorption of CTAB onto Fe3O4 MNPs and lessDOM separately, (c) formed from DOM complex and Fe3O4 MNPsby the bridging effect of CTAB, and (d) decrease recovery of PAHs bysorption supersaturation of Fe3O4 MNPs to more DOM additives.

Figure 6. Solid-phase extraction/UPLC-FLD chromatograms: (a) Qing river water sample and (b) Qing river water sample spiked with 60 ng/L ofPAHs.

Analytical Chemistry Article

DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

7672

http://dx.doi.org/10.1021/acs.analchem.5b01077

Table

2.Resultsof

DeterminationandRecoveriesof

Natural

Water

Samples

Spiked

with60

ng/L

ofPAHs

samples

rainwater

downstream

water

upstream

water

wastewater

tapwater

PAHs

detecteda(ng/L)

recoveryb(%

)detecteda(ng/L)

recoveryb(%

)detecteda(ng/L)

recoveryb(%

)detecteda(ng/L)

recoveryb(%

)detecteda(ng/L)

recoveryb(%

)

Nap

97.12±

6.24

45.99±

24.32

181.94

±29.36

52.86±

19.91

395.88

±12.96

86.34±

21.25

360.00

±13.74

62.64±

16.95

128.6±

14.00

19.40±

1.71

Ace

270.47

±37.39

63.53±

24.48

492.39

±17.28

91.35±

24.07

558.42

±71.68

79.69±

18.31

1177.08±

19.55

67.06±

15.25

65.01±

3.21

47.71±

3.55

Flo

329.57

±22.75

85.22±

16.11

319.28

±27.21

94.00±

21.39

421.18

±49.12

86.71±

13.50

395.01

±2.39

80.79±

11.44

69.75±

4.46

56.89±

8.04

Phe

168.75

±6.02

80.04±

7.24

92.33±

10.80

89.67±

12.92

167.82

±8.70

84.60±

10.20

178.39

±0.96

95.19±

5.05

30.41±

2.19

63.60±

0.19

Ant

18.25±

0.98

82.78±

6.35

13.63±

0.50

96.27±

6.40

11.09±

0.50

96.11±

7.32

9.16

±0.16

72.52±

1.05

2.40

±0.12

66.55±

1.91

Fla

12.54±

0.75

87.44±

9.23

10.97±

0.83

93.36±

3.03

18.16±

2.24

87.74±

7.07

18.26±

0.15

85.47±

3.35

2.69

±0.17

63.65±

0.04

Pyr

8.77

±1.11

85.10±

8.36

14.10±

0.69

83.33±

4.21

12.09±

2.57

82.46±

1.66

10.31±

0.14

88.25±

2.45

0.06

±0.08

63.25±

6.06

Chr

2.81

±0.21

92.44±

9.83

8.69

±0.16

86.82±

1.61

7.81

±0.04

81.06±

1.74

7.75

±0.02

77.83±

3.59

0.75

±0.05

63.50±

3.62

Baa

1.87

±0.21

94.80±

11.02

10.33±

0.19

84.64±

0.19

12.31±

0.09

82.21±

3.13

12.10±

0.08

76.89±

3.24

1.15

±0.03

62.94±

2.11

Bbf

2.73

±0.26

96.86±

10.53

10.69±

0.26

85.55±

0.80

9.37

±0.06

78.87±

1.12

9.61

±0.02

69.88±

3.30

0.15

±0.01

63.95±

2.58

Bkf

1.44

±0.31

97.40±

10.87

9.26

±0.13

85.68±

14.52

10.37±

0.07

77.80±

1.98

10.20±

0.01

75.21±

3.46

0.48

±0.01

62.34±

2.97

Bap

0.77

±0.45

102.99

±3.62

13.17±

0.19

79.06±

2.77

16.83±

0.11

74.15±

1.53

16.66±

0.01

72.92±

2.87

0.41

±0.00

61.02±

5.34

DahA

1.54

±0.83

101.53

±12.13

10.01±

0.40

89.93±

4.02

12.05±

0.09

80.76±

1.90

11.98±

0.03

82.17±

4.32

0.47

±0.23

65.60±

2.89

IcdP

4.25

±2.43

92.04±

12.28

13.22±

0.44

89.97±

0.58

13.23±

0.33

79.22±

2.49

12.64±

0.57

78.29±

3.79

1.98

±0.29

63.57±

3.26

BghiP

3.42

±0.77

111.04

±8.24

6.98

±0.76

86.38±

4.86

3.30

±0.09

85.66±

1.76

3.32

±0.29

98.42±

9.80

1.23

±0.22

64.50±

2.21

total

924.3±

80.71

87.95±

16.16c

1206.99±

89.22

82.89±

5.25c

1669.91±

148.65

85.92±

10.19c

2232.47±

38.12

78.90±

9.90c

305.54

±25.07

59.23±

3.10c

aMeanof

threedeterm

inations.bStandard

deviationforthreedeterm

inations.cMeanrecovery

of15

PAHs.

Analytical Chemistry Article

DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

7673

http://dx.doi.org/10.1021/acs.analchem.5b01077

DOM. The relatively greater concentrations of Ca2+ and Mg2+

in tap water reduced adsorption of PAHs due to electrostaticbinding to Fe3O4 MNPs, which are competitive for CTAB, andcould prevent formation of mixed hemimicelles.55 DOM inriver water and wastewater reduced effects of Ca2+ and Mg2+ onperformance of Fe3O4 MNPs due to potential complexation ofthe metal ions, which resulted in greater recoveries of PAHs.56

Some polymer micromolecules (PAHs-DOM-CTAB-Fe3O4)might have formed due to the hydrophobic interaction ofDOM, CTAB, and Fe3O4 MNPs.

50 Therefore, Fe3O4−CTABMNPs performed better in extracting stable hydrophobicorganic pollutions. Because of the negative and positive effectsof both DOM and metal ions, it is suggested that the mostaccurate and reproducible method employing Fe3O4−CTABMNPs would be the use of internal standards using mass-labeled PAHs.

■ CONCLUSIONSThe Fe3O4−CTAB MNPs, used in the present study, hadseveral advantages for extraction of PAHs from water. First,their relatively large specific surface area provided moreadsorption sites for PAHs and their superparamagnetism alsobenefited from rapid separation. Second, separation of PAHswas completed within 5 min, and the preconcentration processwas convenient, which greatly shortened the duration requiredfor maximal extraction. Third, relatively small amounts oforganic solvents were needed, which avoided wasting ofsolvents. Fourth, the Fe3O4−CTAB MNPs can easily besynthesized with several low-cost chemicals, which might bemore suitable to industrialization for the determination of traceorganic pollutants. Last, the lesser biotoxicity of Fe3O4 MNPsand CTAB might potentially reduce pollution of the environ-ment, compared with other nanomaterials. In conclusion,Fe3O4−CTAB MNPs as a solid adsorbent combined withUPLC-FLD presented excellent performance in analyzing tracePAHs in water.

■ ASSOCIATED CONTENT*S Supporting InformationSelection and additive amount of surfactants, selection ofsample volume and collected parameters of preconcentrationtechniques for PAHs from water, Tables S1 and S2, and FiguresS1−S3. The Supporting Information is available free of chargeon the ACS Publications website at DOI: 10.1021/acs.analchem.5b01077.

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: zhaoxiaoli_zxl@126.com. Phone: (+86)10-84931804.Fax: (+86)10-84931804.*E-mail: mengwei@craes.org.cn. Phone: (+86) 10-84915193.Fax (+86) 10-84915194.

Author Contributions⊥H.W. and X.Z. contributed equally to this work

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThe research was supported by the National Natural ScienceFoundation of China (Grants 41222026, 41130743, and21007063).

■ REFERENCES(1) Pitts, J. N. In Particulate Polycyclic Organic Matter; NationalAcademy of Sciences: Washington D.C., 1972; pp 1−13.(2) Keyte, I. J.; Harrison, R. M.; Lammel, G. Chem. Soc. Rev. 2013,42, 9333−9391.(3) Crone, T. J.; Tolstoy, M. Science 2010, 330, 634−634.(4) Zhang, Y.; Tao, S. Atmos. Environ. 2009, 43, 812−819.(5) Khalili, N. R.; Scheff, P. A.; Holsen, T. M. Atmos. Environ. 1995,29, 533−542.(6) Nguyen, T. C.; Loganathan, P.; Nguyen, T. V.; Vigneswaran, S.;Kandasamy, J.; Slee, D.; Stevenson, G.; Naidu, R. Ecotoxicol. Environ.Saf. 2014, 104, 339−348.(7) D'Adamo, R.; Pelosi, S.; Trotta, P.; Sansone, G.Mar. Chem. 1997,56, 45−49.(8) Mayer, P.; Holmstrup, M. Environ. Sci. Technol. 2008, 42, 7516−7521.(9) U.S. Environmental Protection Agency. http://water.epa.gov/scitech/methods/cwa/pollutants.cfm. (accessed December 3, 2014).(10) Pashin, Y. V.; Bakhitova, L. M. Environ. Health Persp. 1979, 30,185−189.(11) Chen, B.; Xuan, X.; Zhu, L.; Wang, J.; Gao, Y.; Yang, K.; Shen,X.; Lou, B. Water Res. 2004, 38, 3558−3568.(12) Qin, N.; He, W.; Kong, X.; Liu, W.; He, Q.; Yang, B.; Ouyang,H.; Wang, Q.; Xu, F. Chemosphere 2013, 93, 1685−1693.(13) Wang, X.; Liu, S.; Zhao, J.; Zuo, Q.; Liu, W.; Li, B.; Tao, S.Environ. Toxicol. Chem. 2014, 33, 753−760.(14) Ballesteros Goḿez, A.; Rubio, S.; Peŕez Bendito, D. J.Chromatogr. A 2009, 1216, 530−539.(15) Belkessam, L.; Lecomte, P.; Milon, V.; Laboudigue, A.Chemosphere 2005, 58, 321−328.(16) Szolar, O. H. J.; Rost, H.; Braun, R.; Loibner, A. P. Anal. Chem.2002, 74, 2379−2385.(17) Brum, D. M.; Cassella, R. J.; Pereira Netto, A. D. Talanta 2008,74, 1392−1399.(18) Li, N.; Lee, H. K. J. Chromatogr. A 2001, 921, 255−263.(19) Du, J.; Jing, C. J. Phys. Chem. C 2011, 115, 17829−17835.(20) Azvedo, D. d. A.; Gerchon, E.; Reis, E. O. d. J. Braz. Chem. Soc.2004, 15, 292−299.(21) Ma, J.; Xiao, R.; Li, J.; Yu, J.; Zhang, Y.; Chen, L. J. Chromatogr.A 2010, 1217, 5462−5469.(22) Dias, A. N.; Simaõ, V.; Merib, J.; Carasek, E. Anal. Chim. Acta2013, 772, 33−39.(23) Brown, J. N.; Peake, B. M. Anal. Chim. Acta 2003, 486, 159−169.(24) Heidari, H.; Razmi, H.; Jouyban, A. J. Chromatogr. A 2012, 1245,1−7.(25) Zhao, X.; Cai, Y.; Wu, F.; Pan, Y.; Liao, H.; Xu, B. Microchem. J.2011, 98, 207−214.(26) Ankamwar, B.; Lai, T.; Huang, J.; Liu, R.; Hsiao, M.; Chen, C.;Hwu, Y. Nanotechnology 2010, 21, 075102.(27) Xie, J.; Chen, K.; Lee, H. Y.; Xu, C.; Hsu, A. R.; Peng, S.; Chen,X.; Sun, S. J. Am. Chem. Soc. 2008, 130, 7542−7543.(28) Zhao, X.; Cai, Y.; Wang, T.; Shi, Y.; Jiang, G. Anal. Chem. 2008,80, 9091−9096.(29) Mao, X.; Hu, B.; He, M.; Fan, W. J. Chromatogr. A 2012, 1260,16−24.(30) Mayer, P.; Vaes, W. H. J.; Wijnker, F.; Legierse, K. C. H. M.;Kraaij, R.; Tolls, J.; Hermens, J. L. M. Environ. Sci. Technol. 2000, 34,5177−5183.(31) Govan, J.; Gun'ko, Y. K. Nanomaterials 2014, 4, 222−241.(32) Ballesteros Goḿez, A.; Rubio, S. Anal. Chem. 2009, 81, 9012−9020.(33) Zhao, X.; Shi, Y.; Cai, Y.; Mou, S. Environ. Sci. Technol. 2008, 42,1201−1206.(34) Liao, M.; Chen, D. J. Mater. Chem. 2002, 12, 3654−3659.(35) Du, J.; Jing, C. J. Colloid Interface Sci. 2011, 358, 54−61.(36) Hu, J. D.; Zevi, Y.; Kou, X. M.; Xiao, J.; Wang, X. J.; Jin, Y. Sci.Total Environ. 2010, 408, 3477−3489.

Analytical Chemistry Article

DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

7674

http://pubs.acs.orghttp://pubs.acs.org/doi/abs/10.1021/acs.analchem.5b01077http://pubs.acs.org/doi/abs/10.1021/acs.analchem.5b01077mailto:zhaoxiaoli_zxl@126.commailto:mengwei@craes.org.cnhttp://water.epa.gov/scitech/methods/cwa/pollutants.cfmhttp://water.epa.gov/scitech/methods/cwa/pollutants.cfmhttp://dx.doi.org/10.1021/acs.analchem.5b01077

(37) Hawthorne, S. B.; Grabanski, C. B.; Martin, E.; Miller, D. J. J.Chromatogr. A 2000, 892, 421−433.(38) Chen, Y.; Zhu, L.; Zhou, R. J. Hazard. Mater. 2007, 141, 148−155.(39) Wu, F. C.; Xing, B. S. In Natural Organic Matter and ItsSignificance in the Environment; Xie, H. Y., Luo, J., Eds.; Beijing SciencePress: Beijing, China, 2009; pp 20−21.(40) Selberg, A.; Viik, M.; Ehapalu, K.; Tenno, T. J. Hydrol. 2011,400, 274−280.(41) Cho, H. H.; Choi, J.; Goltz, M. N.; Park, J. W. J. Environ. Qual.2002, 31, 275−280.(42) Avena, M. J.; Koopal, L. K. Environ. Sci. Technol. 1998, 32,2572−2577.(43) Grybos, M.; Davranche, M.; Gruau, G.; Petitjean, P.; Ped́rot, M.Geoderma 2009, 154, 13−19.(44) Zhao, X.; Li, J.; Shi, Y.; Cai, Y.; Mou, S.; Jiang, G. J. Chromatogr.A 2007, 1154, 52−59.(45) Li, J.; Werth, C. J. Environ. Sci. Technol. 2001, 35, 568−574.(46) Chin, Y. P.; Weber, W. J.; Eadie, B. J. Environ. Sci. Technol. 1990,24, 837−842.(47) Moon, J. W.; Goltz, M. N.; Ahn, K. H.; Park, J. W. J. Contam.Hydrol. 2003, 60, 307−326.(48) Kretzschmar, R.; Sticher, H.; Hesterberg, D. Soil Sci. Soc. Am. J.1997, 61, 101−108.(49) Kim, E. K.; Walker, H. W. Colloids Surf., A 2001, 194, 123−131.(50) Shen, Y. H. Environ. Technol. 2002, 23, 553−560.(51) Beckett, R.; Le, N. P. Colloids Surf. 1990, 44, 35−49.(52) Pan, B.; Qiu, M.; Wu, M.; Zhang, D.; Peng, H.; Wu, D.; Xing, B.Environ. Pollut. 2012, 161, 76−82.(53) Schlautman, M. A.; Morgan, J. J. Environ. Sci. Technol. 1993, 27,961−969.(54) Haftka, J. H.; Govers, H. J.; Parsons, J. Environ. Sci. Pollut. Res.2010, 17, 1070−1079.(55) Hayes, K. F.; Leckie, J. O. J. Colloid Interface Sci. 1987, 115,564−572.(56) Riedel, T.; Biester, H.; Dittmar, T. Environ. Sci. Technol. 2012,46, 4419−4426.(57) Huckins, J. N.; Petty, J. D.; Orazio, C. E.; Lebo, J. A.; Clark, R.C.; Gibson, V. L.; Gala, W. R.; Echols, K. R. Environ. Sci. Technol.1999, 33, 3918−3923.

Analytical Chemistry Article

DOI: 10.1021/acs.analchem.5b01077Anal. Chem. 2015, 87, 7667−7675

7675

http://dx.doi.org/10.1021/acs.analchem.5b01077

S-1

Cetyltrimethylammonium Bromide-coated Fe3O4 Magnetic 1

Nanoparticles for Rapid Analysis of 15 Trace Polycyclic Aromatic 2

Hydrocarbons in Aquatic Environments by UPLC-FLD 3

4

5

Hao Wang,†,‡,#

Xiaoli Zhao,†,‡,*

Wei Meng,‡,*

Peifang Wang,§ Fengchang Wu,

‡ Zhi Tang,

‡ Xuejiao 6

Han,‡ John P. Giesy

£ 7

‡State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research 8

Academy of Environmental Sciences, Beijing 100012, China; 9

#College of Water Sciences, Beijing Normal University, Beijing 100875, China; 10

§Key Laboratory of Integrated Regulation and Resource Development on Shallow Lakes, Ministry 11

of Education, College of Environment, Hohai University, Nanjing 210098, China; 12

£Department of Veterinary Biomedical Science and Toxicology Centre, University of 13

Saskatchewan, 44 Campus Drive, Saskatoon, SK, Canada. 14

15

*Corresponding Authors: zhaoxiaoli_zxl@126.com, mengwei@craes.org.cn 16

Tel.: (+86)10-84931804; Fax: (+86)10-84931804; 17

Author Contributions: †These authors contributed equally to this work 18

This supporting information contains 2 Tables and 3 Figures. This document contains 10 pages, 19

including this cover page.20

S-2

■ EXPRIMENTS 21

Ultra Performance Liquid Chromatography Tandem Fluorescence Detector. Samples were 22

analyzed at 45 °C. Mobile phases WERE ACN and ultrapure water (UP Water) at a flow rate of 23

0.4 mL/min, with an injection volume of 2 μL. Calculations for quantification of PAHs were 24

accomplished by use of Waters Power 2.0 software. Limits of detection for 15 PAHs were 25

calculated with 3 times the ratio of signal-noise (S/N). PAHs were quantified by use of an external 26

standard curve with coefficient of determination (r2) are all > 0.78 with a linear working range of 0, 27

0.1, 1, 10, 100, 200 and 400 ng/L. 28

■ RESULTS AND DISCUSSION 29

Selection of type and Amounts of Surfactants. CTAB and CPC were chosen as surface 30

modifiers for Fe3O4 MNPs. CPC contains a pyridine moiety, which can contribute to adsorption of 31

PAHs by π-π interaction of the aromatic portions of the two molecules. However, Greater 32

recoveries of PAHs were obtained by using CTAB than CPC (Figure S1a). Fluorescence 33

characteristics of pyridine interfered with quantification of some PAHs, such as Bkf and Bap. For 34

these reasons, CTAB was selected as the more appropriate surface amendment for Fe3O4 MNPs. 35

Recoveries of PAHs were directly proportional to the amount of CTAB added, with a 36

maximum recovery at 50 mg CTAB in the presence of 100 mg Fe3O4 MNPs (Figure S1b). 37

Addition of CTAB to Fe3O4 MNPs at pH 10.0, resulted in CTAB on surfaces of nanoparticles 38

attaching through their cationic ends by chemical self-assembly to form mixed hemi-micelles, 39

which resulted in greater concentrations of PAHs adsorbing to MNPs. However, when the amount 40

of CTAB was greater than 50 mg, hydrophobic interaction between the hydrophobic tails of CTAB 41

molecules occurred, instead of interactions between CTAB and PAHs. This then decreased 42

S-3

enrichment of PAHs from water by iron oxide nanoparticles. Therefore, to optimize subsequent 43

experiments CTAB was employed as the surface modifier at a ratio 1:2 (w/w) of CTAB and Fe3O4 44

MNPs. 45

Selection of Sample Volume. Breakthrough volume is a major parameter in preconcentration 46

of samples by use of SPE. A range of volume of samples ranging from 200 to 2000 mL was tested. 47

Recoveries of PAHs were inversely proportional to volume of the sample extracted by a fixed 48

amount of adsorbent. Due to their greater solubilities in water, the effect was greater for Nap, Ace, 49

and Flo. An optimal sample volume of 800 mL was selected for further studies (Figure S2). 50

Comparison between C18 cartridges and Fe3O4 MNPs. In this study, two kinds of commonly 51

used C18 cartridges, Supelclean LC-18 (PA., USA) and Waters Sep-Pak Vac C18 (Massachusetts, 52

USA), were employed to extract PAHs in downstream water of the Qing river (Beijing, China). 53

The experimental procedure of enrichment of PAHs by using both of C18 cartridges were as 54

follows: Samples of water were collected on May 25, 2015, and immediately filtered through glass 55

fiber filters (pore diameter 0.45μm), 800 mL water sample with and without 100 ng/L PAHs were 56

used, and three replicates were conducted for each SPE technique; Cartridges were activated with 57

5 mL methanol and 5 mL ultrapure water orderly before PAHs enrichment process, and then water 58

samples were filtered with a flow velocity of 5 mL/min, and later, 5 mL dichloromethane and 10 59

mL acetone were employed for eluting PAHs from cartridges. 60

Based on experimental results shown in Figure S3, recoveries of PAHs were generally similar 61

for the three methods. However, PAHs extraction process by Fe3O4 MNPs was complete in 0.5 62

hour, while more than 12 hours was spent by two kinds of C18 cartridges, even than all samples of 63

water had been filtered by 0.45 μm glass fiber filters. By comparing among methods of 64

javascript:void(0);

S-4

pretreatment given in the literature and the method developed during this study (Table S1), the 65

advantages of Fe3O4 MNPs was demonstrated. 66

S-5

Table S1. Parameters for pre-concentration techniques for PAHs from water 67

Method PAHs adsorbent sample volume

(mL)

Volume of consumed

Organic solvent (mL)

Time

(min)

the detective

limit (ng/L) analyzer reference

LLE 16 hexane 300 77 72 0.033 – 0.13 HPLC1-FLD

(S1)

SPE 16 C18 column 1000 10 125 1 – 5 GC2-MS

3

(S2)

SPE 15 C18 column 1000 15 333 NP HPLC-FLD (S3)

SPE 16 C18 column 1000 10 167 1 GC-FID4

(S4)

SPE 12 C18 column 1000 90 333 –500 0.01 – 5.00 HPLC-UV5

(S5)

SPE 10 C18 column 200 14 30 0.5 –25.0 GC-MS (S6)

SPE 10 C18 column 1000 13 356 10.0 – 166.9 HPLC-UV (S7)

SPE 16 C18 disk 2500 115 30~45 0.01 – 11.3 HPLC-UV-FLD (S8)

SPE 16 nano carbon 500 15 145 2.0 – 8.5 GC-MS (S9)

SPE 8 C14- Fe3O4 NPs 350 3 17 0.1 – 0.25 HPLC-FLD (S10)

SPE 15 Stir bar 10 0.15 60 0.2 – 2.0 HPLC-FLD (S11)

SPME 12 Fiber 25 0 60 0.03 GC-SIM6-MS

(S12)

SPME 16 PDMS fiber 10 0 90 30 – 590 GC-MS (S13)

SPME 5 C-Fe3O4/C NPs 25 2 25 0.7 – 49.6 HPLC-FLD (S14)

SPE 15 Fe3O4-CTAB NPs 800 10 30 0.13 –20.6 UPLC-FLD Our study

SPE 15 C18 column7 800 15 >720 – UPLC-FLD Our study

NP: No Reported; 1: high-performance liquid chromatography; 2: Gas Chromatography; 3: Mass Spectrometer; 4: Flame ionization detector; 5: ultraviolet detector; 68

6: selected ion monitoring; 7: the detail were presented in Figure S3 69

S-6

Table S2. Parameters of field samples. 70

Sample type TOC (mg/L) Ca2+

(mg/L) Mg2+

(mg/L)

Sewage water 23.22±0.62 117±1.4 55±1.4

upstream water 8.41±0.32 67±1.4 19±2.8

downstream water 3.58±0.44 95.00±4.2 41±4.2

tap water NDa 67±4.2 27±4.2

rainwater ND 3±1.4 1±1.4

a: no detected

S-7

0 20 40 60 80 100 120

10

20

30

40

50

60

70

0 20 40 60 80 100 120 140 160

0

20

40

60

80

100

ba

CPC

CTAB

aver

age

reco

ver

y (

%)

Surfactants concentration (mg/100 mg Fe3O

4 MNPs)

Rec

over

y (

%)

Concentration of CTAB per 100 mg Fe3O

4 MNPs

Nap

Ace

Flo

Phe

Ant

Fla

Pyr

Chr

Baa

Bbf

Bkf

Bap

DahA

IcdP

BghiP

71

Figure S1. Recoveries of 15 PAHs as functions of several surfactants (a) and additive amount of CTAB (b), during batch mode. Amount of metal 72

oxide: 100 mg of Fe3O4 MNPs. pH:10.0, sample volume: 800 mL. Volume of ACN: 10 mL. 73

S-8

400 800 1200 1600 20000

20

40

60

80

100R

eco

ver

y (

%)

Volume of Samples (mL)

Nap

Ace

Flo

Phe

Ant

Fla

Pyr

Chr

Baa

Bbf

Bkf

Bap

DahA

IcdP

BghiP

74

Figure S2. Effects of sample volume on recoveries of PAHs 75

S-9

Nap A

ce Flo

Phe

Ant Fl

aPy

rChr

Baa B

bfBkf

Bap

Dah

AIc

dP

Bgh

iP

0

20

40

60

80

100

Rec

over

y (

%)

PAHs

Supelclean LC-18 cartridge

Waters Sep-Pak Vac C18 cartridge

Fe3O

4 MNPs

76

Figure S3. The recoveries of PAHs in river water sample by Supelclean LC-18 77

Cartridge, Waters Sep-Pak Vac C18 Cartridge and Fe3O4 MNPs 78

S-10

References 79

(S1) Brum, D. M.; Cassella, R. J.; Pereira Netto, A. D. Talanta 2008, 74, 1392-1399. 80

(S2) Li, N.; Lee, H. K. Journal of Chromatography A 2001, 921, 255-263. 81

(S3) Chen, Y.; Zhu, L.; Zhou, R. Journal of Hazardous Materials 2007, 141, 148-155. 82

(S4) Zhou, J. L.; Maskaoui, K. Environmental Pollution 2003, 121, 269-281. 83

(S5) Kabziński, A.; Cyran, J.; Juszczak, R. Polish Journal of Environmental Studies 2002, 11, 84

695-706. 85

(S6) Elaine, G. J. Braz. Chem. Soc. 2004, 15, 292-299. 86

(S7) Moja, S. J.; Mtunzi, F.; Madlanga, X. Journal of Environmental Science and Health, Part A 87

2013, 48, 847-854. 88

(S8) Brown, J. N.; Peake, B. M. Analytica Chimica Acta 2003, 486, 159-169. 89

(S9) Kicinski, H. G.; Adamek, S.; Kettrup, A. Chromatographia 1989, 28, 203-208. 90

(S10) Ma, J.; Xiao, R.; Li, J.; Yu, J.; Zhang, Y.; Chen, L. Journal of Chromatography A 2010, 91

1217, 5462-5469. 92

(S11) Popp, P.; Bauer, C.; Wennrich, L. Analytica Chimica Acta 2001, 436, 1-9. 93

(S12) Dias, A. N.; Simão, V.; Merib, J.; Carasek, E. Analytica Chimica Acta 2013, 772, 33-39. 94

(S13) Doong, R.; Chang, S.; Sun, Y. Journal of Chromatography A 2000, 879, 177-188. 95

(S14) Heidari, H.; Razmi, H.; Jouyban, A. Journal of Chromatography A 2012, 1245, 1-7. 96

wang et alwang suppinfo

undefined: containing relevant concentrations of both naturally occurring organic matter and hardness metals: sediments and the atmosphere are increasing due to activities: CPC: CTAB: Offundefined_2: undefined_3: undefined_4: undefined_5: undefined_6: undefined_7: undefined_8: undefined_9: undefined_10: Nap: Flo: Phe: Pyr: Chr: Baa: Bbf: DahA: undefined_11: undefined_12: undefined_13: undefined_14: BghiP: 0: 0_2: undefined_15: undefined_16: undefined_17: undefined_18: undefined_19: Flo_2: Phe_2: Fla: Pyr_2: Chr_2: Baa_2: Bbf_2: DahA_2: BghiP_2: undefined_20: 400: