Chemosphere 91 (2013) 344–350

Contents lists available at SciVerse ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Complexation of trace organic contaminants with fractionated dissolvedorganic matter: Implications for mass spectrometric quantification

0045-6535/$ - see front matter � 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.chemosphere.2012.11.059

⇑ Corresponding author at: Department of Soil, Water and EnvironmentalScience, University of Arizona, 1177 E 4th St., Tucson AZ 85721, USA. Tel.: +1 520626 5635; fax: +1 520 621 1647.

E-mail addresses: selene.hernandez27@gmail.com (S.H. Ruiz), samanthi.wic@gmail.com (S. Wickramasekara), abrell@u.arizona.edu (L. Abrell), xdgao@rice.edu(X. Gao), chefetz@agri.huji.ac.il (B. Chefetz), chorover@cals.arizona.edu(J. Chorover).

Selene Hernandez Ruiz a, Samanthi Wickramasekara b, Leif Abrell b, Xiaodong Gao a, Benny Chefetz c,Jon Chorover a,b,⇑a Department of Soil, Water and Environmental Science, University of Arizona, 1177 E 4th St., Tucson AZ 85721, USAb Arizona Laboratory for Emerging Contaminants, University of Arizona, 1040 East 4th St., Tucson AZ 85721, USAc Department of Soil and Water Sciences, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot 76100, Israel

h i g h l i g h t s

" Suwannee River natural organic matter is composed dominantly of hydrophobic acids." Fluorescence of protein and fulvic constituents was quenched by reaction with PPCPs." PPCP reaction with hydrophilic acids diminished LC–MS/MS recovery.

a r t i c l e i n f o

Article history:Received 9 July 2012Received in revised form 9 November 2012Accepted 22 November 2012Available online 28 December 2012

Keywords:Suwannee River natural organic matterPPCPEDCFluorescence quenchingLC–MS/MS

a b s t r a c t

Interaction with aqueous phase dissolved organic matter (DOM) can alter the fate of trace organic con-taminants of emerging concern once they enter the water cycle. In order to probe possible DOM bindingmechanisms and their consequences for contaminant detection and quantification in natural waters, a setof laboratory experiments was conducted with aqueous solutions containing various operationally-defined ‘‘hydrophilic’’ and ‘‘hydrophobic’’ freshwater DOM fractions isolated by resin adsorption tech-niques from reference Suwannee River natural organic matter (SROM). Per unit mass of SROM carbon,hydrophobic acids (HoA) comprised the largest C fraction (0.63 ± 0.029), followed by hydrophilic-neutrals(HiN, 0.11 ± 0.01) and acids (HiA, 0.09 ± 0.017). Aqueous solutions comprising 8 mg L�1 DOC of eachSROM fraction were spiked with a concentration range (10–1000 lg L�1) of bisphenol A (BPA), carbam-azepine (CBZ), or ibuprofen (IBU) as model target compounds in 24 mM NH4HCO3 background electrolyteat pH 7.4. Contaminant interaction with the SROM fractions was probed using fluorescence spectroscopy,and effects on quantitative analysis of the target compounds were measured using direct aqueous-injection liquid chromatography tandem mass spectrometry (LC–MS/MS). Total quenching was greaterfor the hydrophilic fractions of SROM and associations were principally with protein-like and fulvicacid-like constituents. Whereas LC–MS/MS recoveries indicated relatively weak interactions with mostSROM factions, an important exception was the HiA fraction, which diminished recovery of CBZ andIBU by ca. 30% and 70%, respectively, indicating relatively strong molecular interactions.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction freshwater sources including rivers, lakes and/or groundwater

The continuous disposal of pharmaceuticals and personal careproducts (PPCPs) into wastewater, followed by their incomplete re-moval during treatment, results in their eventual release into

(Kolpin et al., 2002; Cahill et al., 2004; Sharma et al., 2009). Fordecades, it has been suggested that dissolved organic matter(DOM) plays a mediating role in the fate of these organic contam-inants (Chin et al., 1997; Leenheer and Croue, 2003; Yamamotoet al., 2004). An established conceptual framework for predictingthe impacts of organic matter on hydrophobic organic pollutantfate has led to attempts to predict the affinity of many PPCPs fororganic matter based on contaminant hydrophobicity, e.g., throughthe use of octanol–water partition coefficients (Kow) (Tolls, 2001;Pan et al., 2009). Indeed, complexation of more polar PPCPs withDOM often exceeds that predicted from log Kow values alone,likely because bond formation involves not only hydrophobic

S.H. Ruiz et al. / Chemosphere 91 (2013) 344–350 345

interactions, but also polar interactions mediated through O, N,and S containing functional groups of PPCPs and DOM (Tolls,2001; Pan et al., 2009). Thus, the diverse physico-chemical proper-ties of both PPCPs and DOM undoubtedly influence the types andstrengths of bonds formed.

Elucidating the nature of PPCP-DOM interactions is complicatedby the multi-disperse and heterogeneous nature of DOM, which iscomprised of a diverse array of biomolecules, molecular fragments,and partially humified degradation products that derive from bothplant and microbial sources (Leenheer and Croue, 2003; Leenheer,2004). Natural DOM mixtures are polydisperse, and contain con-stituents with varying levels of hydrophilicity, and a wide rangeof functional group chemistry (Leenheer, 1981; Leenheer et al.,2003). As a result, DOM is observed to undergo sorptive fraction-ation in natural, aqueous heterogeneous systems where trends inaqueous-solid phase partitioning depend on DOM and particle sur-face chemistries (Chorover and Amistadi, 2001; Zhou et al., 2001;Guo and Chorover, 2003). Sorptive fractionation on ion exchangeor hydrophobic resins is likewise employed in analytical organicgeochemistry to isolate operationally-defined pools of DOM forfurther interrogation (Leenheer, 1981; Chefetz et al., 1998). Forinstance, through LC-MS/MS infusion experiments of fractionatedDOM (Wickramasekara et al., 2012) found that the hydrophilicacid fraction produced the greatest matrix suppression of severalPPCPs.

In the present work designed to study DOM-PPCP interactions,Suwannee River natural organic matter (SROM, an InternationalHumic Substances Society standard reference material) was frac-tionated (Leenheer, 1981; Chefetz et al., 1998) into operationally-defined hydrophobic (Ho) and hydrophilic (Hi) acids (A), bases(B) and neutrals (N). Interactions of the resulting DOM fractionswith model PPCPs were then probed at environmentally relevantpH and ionic strength. Two analytical techniques were used forcomplementary purposes: fluorescence spectroscopy was used toelucidate the molecular components of DOM fractions whose fluo-rescence is quenched by interaction with PPCPs, and direct aque-ous injection tandem mass spectrometry was used to assess therelative strength of potential bond formation, particularly with re-spect to its effect on analyte chromatographic separation andquantification of contaminant analytes. In a prior study of unfrac-tionated DOM solutions, we found that despite reproduciblequenching of DOM fluorescence principally in protein-like and ful-vic acid-like regions, PPCP recoveries by LC–MS/MS were essen-tially unimpacted (Hernandez-Ruiz et al., 2012). Hence, the samePPCPs were selected for use in the current study to probe interac-tions with DOM constituents in more detail for direct comparisonto that prior work: bisphenol-A (BPA), carbamazepine (CBZ), andibuprofen (IBU) (Table S1). All of these contaminants have been ob-served to survive secondary wastewater treatment, and to accom-pany treated wastewater discharges into the environment (Snyderet al., 2003; Dickenson et al., 2011).

2. Materials and methods

Suwannee River NOM standard (1R101N, reverse osmosis isola-tion) was purchased from the International Humic SubstancesSociety (IHSS, http://www.humicsubstances.org/sources.html). Allother materials were ACS reagent grade or better and water usedfor experiments was ultrapure treated (18 MX and UV).

2.1. SROM fractionation

SROM was fractionated into hydrophilic (Hi) and hydrophobic(Ho) acids (A), neutrals (N) and bases (B) with resin-based ap-proach. The XAD-8, Amberlyst 15 and Amberlyst 21 resins were

purchased from Sigma–Aldrich (St. Louis, MO), cleaned and pre-conditioned according to Allard et al. (1991). The fractionation pro-cedure was a slight modification of published methods (Leenheer,1981; Chefetz et al., 1998). See Supplementary materials fordetails.

2.2. Infrared spectroscopy

Subsamples of freeze-dried SROM fractions were dissolved inultrapure water for collection of Fourier transform infrared (FTIR)spectra. The GRAMS/AI software Thermo Electron Corp. (Milford,MA) was used for peak deconvolution to determine the positionand area of bond vibrations. See Supplementary materials fordetails.

2.3. Incubation of PPCPs with SROM fractions

Stock solutions for BPA, CBZ, and IBU were prepared at10.0 mg L�1 in nanopure water with 0.01% (v/v) MeOH co-solventfor 48 h. Substock solutions were then prepared at 20, 40, 400,2000 lg L�1 in an LC–MS/MS compatible 24 mM volatile NH4HCO3

buffer titrated with 0.06 M HCl to pH 7.4 for CBZ and IBU experi-ments. The limits of detection (signal to noise ratio > 3) were50 lg L�1 for BPA, 0.15 lg L�1 for CBZ and 0.16 lg L�1 for IBU.Therefore, for BPA, only the 400 and 2000 lg L�1 solutions wereused in the LC–MS/MS measurements. SROM fractions were dis-solved in the same buffer at 16 mg L�1 DOC stock concentration.

PPCP-SROM fraction incubations were performed in triplicateby adding 2 mL of each DOC stock (16 mg L�1) to 2 mL of substockPPCP solution to give a final volume of 4 mL. Negative controlscomprising SROM fractions alone (no PPCPs), and positive controlswith PPCPs at their respective concentrations but with no SROM,were included. The 4 mL bottles with no headspace were thenincubated for 24 h in an end-over-end shaker. A 3 mL aliquotwas then transferred to a quartz cuvette for fluorescence spectros-copy and the 1 mL was transferred to autosampler vial for LC–MS/MS analysis.

2.4. Fluorescence measurements

To elucidate the DOM-PPCP interactions, excitation-emissionmatrices (EEMs) of each SROM fraction were collected in the ab-sence and presence of PPCP analytes. EEMs were obtained with aJobin Yvon Horiba/Spex Fluoromax-4 fluorometer (Edison, NJ)equipped with a xenon lamp as excitation source. The 3.0 mL sub-sample for each treatment was placed in a square quartz cuvettecell (light path 10 mm � 10 mm) for EEM collection. The signal toreference detector ratio was collected and EEMs were producedusing the FluorEssence software with excitation and emissionwavelength ranges of 200–450 nm (5 nm slit) and 250–650 nm(2 nm slit), respectively, both at 5 nm increments. Quenching ofDOM fluorescence by PPCP interaction was calculated by subtract-ing the EEM for DOM fraction-PPCP treatment from that of the cor-responding DOM fraction alone. The contribution of the PPCPmolecules themselves to total fluorescence in DOM-PPCP systemswas negligible relative to that of bulk DOM, such that quenchingEEMs represent accurately the diminished fluorescence of a givenDOM fraction in the presence of PPCP. Moreover, no inner filter-ing effects were evident during trial experiments with DOM(Hernandez-Ruiz et al., 2012).

The fluorescence intensity was integrated beneath all five EEMregions previously characterized as (I) ‘‘tyrosine-like’’, (II) ‘‘trypto-phan-like’’, (III) ‘‘fulvic acid-like’’, (IV) ‘‘microbial byproduct-like’’,and (V) ‘‘humic acid-like’’ based on a large reference sample set(Chen et al., 2003). The fluorescence volume (U) beneath all

346 S.H. Ruiz et al. / Chemosphere 91 (2013) 344–350

regions of the sample fluorescence (bulk SROM or fractions) EEMsurface was calculated as:

U ¼ RExREmIðkExkEmÞDkExDkEm ð1Þ

where DkEx is the excitation wavelength interval, DkEm is the emis-sion wavelength interval and I (DkEx and Dkem) is the fluorescenceintensity >104 at each excitation–emission wavelength pair. Theintensity cut-off of 104 was applied to remove background noisein the SROM samples (e.g., apparent fluorescence at kEm < kEx

(Lakowicz, 1999). For the resulting quenching EEMs [SROM-(SROM + PPCP)] the quenching volume is:

UQ ¼ RExREmIðkExkEmÞDkExDkEm ð2Þ

2.5. PPCPs spike and LC–MS/MS recovery

The recovery of spiked PPCPs following incubation was accom-plished using a tandem mass spectrometer (LC–MS/MS) equippedwith an Acquity Ultra Performance Liquid Chromatograph and tri-ple quadrupole Quattro Premier XE mass spectrometer, with asample organizer from Waters Corporation (Milford, MA). All treat-ments and calibration standards were injected to the UPLC columnin a 24 mM NH4HCO3 buffer and pH of 7.4 in the presence and ab-sence of DOM. Analytes were separated following 5 lL direct injec-tion onto a reverse phase column maintained at 40 �C with agradient mobile phase from 80:20% (v/v) water:acetonitrile at flowrate 0.3 ml min�1. Recovery was quantified using multiple reactionmonitoring (MRM). Fragmentation transitions, collision energies(CEs), capillary voltages (CVs), cone voltages (CNVs), and desolva-tion temperatures (DTs) are provided in Table S1. Mass lynx soft-ware from Waters Corp. (Milford, MA) was used to foridentification and quantification of analytes. An ANOVA/Tukey’sstatistical test (95% CI) was used in Origin 8.5 (Northampton,MA) to assess variance in recoveries from all treatments.

250 300 350 400 450 500 550 600 650200

250

300

350

400

450

VIV

IIIII

Emission (nm)

Exci

tatio

n (n

m)

I

(A) HiA

IV

III

200

250

300

350

400

450

VIV

IIIII

Exci

tatio

n (n

m)

I

(D) HoA

IV

III

250 300 350 400 450 500 550 600 650

Emission (nm)

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450

Exci

tatio

n (n

m)

250 300 350 400

Emiss

250 300 350 400

Emiss

Fig. 1. Difference (quenching) EEM of DOM fractions by BPA [(EEM of DOM fraction) – (Eand 8 mg L�1 of DOC and BPA at 200 and 1000 lg L�1.

3. Results

3.1. Characterization of SROM hydrophilic and hydrophobic fractions

Quantitative mass balance indicates that, per unit SROM carbon,HoA is the most prevalent fraction (0.63 ± 0.029), followed by HiN(0.11 ± 0.006), HiA (0.09 ± 0.0017), HoN (0.06 ± 0.012), HoB (0.02 ±0.002) and HiB (0.02 ± 0.007) with a cumulative standard deviationof 0.073 and a total DOC recovery of 93%; (Wickramasekara et al.,2012) observed similar fractionation percentages. FTIR spectra andfluorescence EEMs are unique for each of the obtained fractions(Fig. S1), and consistent with distinct mechanisms of sorptive frac-tionation (see Supplementary materials for details).

3.2. Fluorescence quenching

Quenching EEMs for each SROM fraction – PPCP pair followed asimilar trend irrespective of the contaminant concentration andthus a representative EEM per SROM fraction-PPCPs set is pre-sented (Figs. 1–3).

The addition of BPA to all SROM fractions resulted in reproduc-ible quenching in the low excitation region (200–250 nm) of thematrix throughout the entire emission range corresponding to re-gions I, II, and III (Fig. 1A–F). However, additional quenching of re-gion V in HoA suggests that aromatic moieties are also involved inmolecular interactions. For both basic fractions (HiB and HoB)additional quenching is observed in region IV, attributed to carbo-hydrates (Her et al., 2003) and aromatic amino acids such as tryp-tophan and tyrosine (Chen et al., 2003; Her et al., 2003; Hudsonet al., 2007; Henderson et al., 2009). Fluorescence quenching byCBZ was apparent in regions I, II and III for all SROM fractions(Fig. 2A–F). However, HoA again displayed prominent quenchingin region V (Fig. 2D) and HoB in region IV and V (Fig. 2E). IBUquenched DOM fluorescence in regions I, II and III (Fig. 3A–F) for

V

III

(B) HiB

VIV

IIIIII

(C) HiN

V

III

(E) HoB

VIV

IIIIII

(F) HoN

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450

Exci

tatio

n (n

m)

450 500 550 600 650

ion (nm)

450 500 550 600 650

ion (nm)

250 300 350 400 450 500 550 600 650

Emission (nm)

250 300 350 400 450 500 550 600 650

Emission (nm)

EM of DOM fraction + PPCP)]. All treatments conducted in 24 mM NH4HCO3, pH 7.4,

250 300 350 400 450 500 550 600 650200

250

300

350

400

450

VIV

IIIII

Emission (nm)

Exci

tatio

n (n

m)

I

(A) HiA

VIV

IIIIII

(B) HiB

VIV

IIIIII

(C) HiN

VIV

IIIIII

(D) HoA

VIV

IIIIII

(E) HoB

VIV

IIIIII

(F) HoN

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450

Exci

tatio

n (n

m)

250 300 350 400 450 500 550 600 650

Emission (nm)250 300 350 400 450 500 550 600 650

Emission (nm)

250 300 350 400 450 500 550 600 650

Emission (nm)250 300 350 400 450 500 550 600 650

Emission (nm)250 300 350 400 450 500 550 600 650

Emission (nm)

Fig. 2. Difference (quenching) EEM of DOM fractions by CBZ [(EEM of DOM fraction) – (EEM of DOM fraction + PPCP)]. All treatments conducted in 24 mM NH4HCO3, pH 7.4,and 8 mg L�1 of DOC and CBZ at 10, 20, 100 and 1000 lg L�1.

200

250

300

350

400

450

VIV

IIIII

Exci

tatio

n (n

m)

I

(A) HiA

VIV

IIIIII

(B) HiB

VIV

IIIIII

(C) HiN

VIV

IIIIII

(D) HoA

VIV

IIIIII

(E) HoB

VIV

IIIIII

(F) HoN

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450

Exci

tatio

n (n

m)

200

250

300

350

400

450Ex

cita

tion

(nm

)

200

250

300

350

400

450

Exci

tatio

n (n

m)

250 300 350 400 450 500 550 600 650

Emission (nm)

250 300 350 400 450 500 550 600 650

Emission (nm)

250 300 350 400 450 500 550 600 650

Emission (nm)

250 300 350 400 450 500 550 600 650

Emission (nm)

250 300 350 400 450 500 550 600 650

Emission (nm)

250 300 350 400 450 500 550 600 650

Emission (nm)

Fig. 3. Difference (quenching) EEM of DOM fractions by IBU [(EEM of DOM fraction) – (EEM of DOM fraction + PPCP)]. All treatments conducted in 24 mM NH4HCO3, pH 7.4,and 8 mg L�1 of DOC and IBU at 10, 20, 100 and 1000 lg L�1.

S.H. Ruiz et al. / Chemosphere 91 (2013) 344–350 347

all fractions, except for HiN, for which minimal quenching was ob-served. However, as with CBZ and BPA, addition of IBU alsoquenched the HoA fraction in region V.

The total normalized quenching plot (normalized to its corre-sponding bulk SROM or fraction) for bisphenol-A shows that the

HiN, HIA and HiB fractions were quenched to a greater degreeacross treatment concentrations (Fig. S3A). Similarly CBZ quenchedthe HiB, HiN and HoB fraction to a greater extent the other frac-tions (Fig. S3B). IBU quenched the HiA, followed by the bulk SROMand the HiB fraction (Fig. S3C).

348 S.H. Ruiz et al. / Chemosphere 91 (2013) 344–350

3.3. LC–MS/MS recovery of PPCPs post SROM fractions exposure

Negative controls, (no PPCP added) produced zero percentrecoveries for the three contaminants tested, thus target analyteswere undetected in the SROM fractions themselves. All BPA-SROMfractions (200–1000 lg L�1) yielded no statistically significant dif-ference in LC–MS/MS recovery relative to respective positive con-trols (Fig. 4A). For CBZ, the HiA fraction yielded a significantreduction in recovery (ca. 30%) at all spiked CBZ concentrations.The HoN produced an apparent increase in CBZ recovery (>100%)relative to positive controls at the 10, 20, 100 and 1000 lg L�1

(Fig. 4B). The recovery of IBU in presence of most SROM fractionsdid not differ significantly from the corresponding positive con-trols, with the exception again being the HiA fraction, which pro-duced recoveries of about 28% at all spiked concentrations(Fig. 4C).

4. Discussion

4.1. PPCP quenching of DOM fluorescence

Quenching is the result of aqueous phase PPCP-DOM complex-ation that might be mediated by one or more of several mecha-nisms of intermolecular association, including hydrophobic,H-bonding, electrostatic, and van der Waals interactions. Mecha-

020406080

100120140

Perc

ent c

a. R

ecov

ery

HoHiNHiBHiA

020406080

100120140

Perc

ent c

a. R

ecov

ery

HoHiNHiBHiA

020406080

100120140

Perc

ent c

a. R

ecov

ery

HoHiNHiBHiA

(A)

(B)

(C)

Fig. 4. LC–MS/MS percent ca. recovery of spiked model contaminants [(A) BPA, (B) CBZN = 3 in 24 mM H4HCO3 of pH 7.4 buffer and 8.0 mg L�1 of DOC.

nisms of interaction leading to the quenching behavior observed(Figs. 1–3) are, therefore, likely dependent on both PPCP and SROMfraction chemistry and structure. At pH 7.4, IBU is anionic and,based on logDOW values, it exhibits the lowest hydrophobicity ofthe three compounds because of its charge. Conversely, neutralBPA is the most hydrophobic and CBZ is intermediate in hydropho-bicity. Hence, based on charge and logDOW characteristics, wehypothesized that BPA and CBZ would exhibit strongest interactionwith Ho DOM fractions via van der Waals and/or hydrophobicinteractions, whereas IBU was expected to interact with chargedHi DOM fractions either via anion exchange (e.g., on protonatedamide or amino groups) or via cation bridging interaction (e.g.,with dissociated DOM carboxyl groups).

Quenching of DOM fluorescence was observed to occur innearly all PPCP-SROM fraction pairs (Figs. 1–3). The greatest totalquenching was measured (Fig. S3) for the BPA-HiN, CBZ-HiB, andthe IBU-HiA and the lowest quenching was observed for BPA-HoN, CBZ-HoN, and IBU-HiN which is qualitatively inconsistentwith the above hypotheses. Indeed, the trends in the extent ofquenching indicate the importance of the hydrophilic SROM frac-tions with basic and acidic functional groups for interactions withneutral and acidic organic contaminants.

A secondary hypothesis was that the hydrophobic DOM frac-tions would be preferentially quenched in the ‘‘humic-acid like’’ re-gions of the EEMs, which represent the more aromatic moieties,

+CSROMHoNHoBA

+CSROMHoNHoBA

+CSROMHoNHoBA

and (C) IBU] in absence (+C is the positive control) and presence of DOM fractions.

S.H. Ruiz et al. / Chemosphere 91 (2013) 344–350 349

particularly by neutral PPCPs (BPA and CBZ). Results revealed that,for HoA, quenching in region V was indeed consistently observed(Figs. 1–3). In fact, there were significant differences in quenchingpatterns between HoA and all other fractions that showed predom-inant quenching in the fulvic-acid-like and protein-like regions ofthe EEMs, and even anionic IBU gave rise to significant quenchingof the ‘‘humic-acid-like’’ region of HoA (Fig. 3).

Peptides and proteins (e.g., including amino acids tryptophanand tyrosine) may be zwitterionic at pH 7.4, whereas polyphenolsand polysaccharides bear net negative charge due to carboxylgroup dissociation. Aliphatic molecules exhibiting a lower degreeof carboxylation should exhibit lower net negative charge, as theycomprise a greater prevalence of neutral aliphatic and aromaticmoieties that are suggested to drive the associations with neutralsurrogate PPCPs via hydrophobic and/or van der Waals interactions(Chefetz et al., 2006; Polubesova et al., 2007; Pan et al., 2008).

The interactions between the SROM fractions and BPA that gaverise to quenching are likely due to van der Waals forces (Sun et al.,2007), p–p interactions between aromatic moieties and/or hydro-gen-bonding. Similarly, CBZ could participate in p–p interactionsbetween the stilbene CBZ moiety (Bai et al., 2008) and the aromat-ics in fulvic acid-like and protein-like fluorophores, in addition tohydrogen bonding. In contrast, IBU produced significant quenchingof all fractions except HiN in the protein-like and the fulvic acidregions. Since the carboxylic group of IBU is deprotonated at theexperimental pH, anion exchange and/or cation bridging couldbe mediating the interactions among the carboxylic groups of theSROM fractions and IBU in addition to the hydrogen bonding, p–p and hydrophobic interactions and/or van der Waals associationsbetween the aromatic regions of the protein-like and fulvic acid-like groups.

Despite most fractions being predominantly quenched in theprotein-like and fulvic acid-like regions of the EEMs, the additionalquenching of HoA in the humic acid region is of particular impor-tance being that this fraction accounts for about 63% of the totalDOC contribution. The humic-acid region quenching by CBZ is con-sistent with prior observations (Bai et al., 2008). Unlike that study,however, the present experiment probed the low excitation (200–250 nm) range as well, that showed significant response.

4.2. Effects of PPCPs-DOM interaction on analyte quantification

Liquid chromatography tandem mass spectrometry has beenused previously to probe complex formation among biologicalmolecules. The bonding strength and stability of proteins and li-gands with other organic molecules has been assessed by monitor-ing their MS signal intensity relative to a control (Loo, 2000; Danielet al., 2002; Bolbach, 2005). For instance, proteins and other bio-molecules can interact weakly via hydrophobic interactions and/or van der Waals associations in addition to hydrogen bonding,but then be separated chromatographically to become labile inthe gas phase during MS analysis, whereas stronger interactions,such as electrostatic bonding between an opposite-charged organicspecies or stable cation bridging between carboxylic acids can pro-duce a signal reduction in comparison to a control (Loo, 2000;Schug and Lindner, 2005).

The near 100% LC–MS/MS recoveries of BPA, CBZ, and IBU in thepresence of most fractions (Fig. 4) suggests that the majority ofinteractions giving rise to fluorescence quenching are relativelyweak. An important exception was discovered, however, by reac-tion of HiA with CBZ and IBU, where recoveries decreased to ca.70% and 30% of the values obtained in controls, respectively(Fig. 4). The chemistry of HiA, which differs significantly from thatother fractions, is evidently responsible for this effect. The HiA EEMhad the strongest peak in the ‘‘fulvic acid-like region’’ (Fig. S2),which is consistent with intense vibrations deriving from carbox-

ylic and sugar-ring vibrations as observed by FTIR. These DOM con-stituents interact strongly with more polar target analytes (Navonet al., 2011) and, according to the present study, prevent theirquantitative recovery via mass spectrometry. Based on the fluores-cence spectroscopy data, plausible mechanisms of IBU-HiA interac-tion include cation bridging, whereas CBZ-HiA interaction couldalso include hydrogen bonding at carbonyl or amino functionalitiesand/or p–p interactions with the ring groups of the CBZ. Likewise,(Maoz and Chefetz, 2010) reported that HiA produced the highestcomplexation coefficient for CBZ at pH 7, indicating that this frac-tion may play an important role in environmental fate of CBZ andIBU. It is noteworthy that HiA comprises a relatively low (ca. 9%)contribution to the total SROM carbon mass. Thus, although ithas clear potential to strongly interact with ionized PPCPs, its im-pact on PPCP bonding to SROM in surface waters will depend on itsprevalence relative to other fractions. Moreover, the near 100%recovery of all three contaminants in presence of bulk SROM(Hernandez-Ruiz et al., 2012) suggests that when the HiA fractionis in its native bulk SROM mixture it has a diminished impact onrecovery from the bulk, either because of its lower fractional abun-dance in that case, or because of its aggregate formation (andhence blocking of reactive sites) in the presence of other SROMcomponents. A comparable effect was observed by Wickramaseka-ra et al. (2012), who showed via MS/MS infusion experiments thatthe HiA fraction produced the most significant decrease in signalMS signal intensity, quantified as a negative matrix effect.

5. Conclusions

The HiA fraction of SROM gave rise to reduced mass spectro-metric recovery of spiked CBZ and IBU irrespective of contaminantconcentration, suggesting the importance of carboxylic acid func-tional groups in DOM-PPCP molecular interactions. Thus, chemicalinteractions of polar contaminants with DOM components of acidichydrophilic character might preclude the accurate quantitativeassessment of similar organic contaminants. Fluorescence spec-troscopy proved to be a complementary, efficient and sensitiveprobe of PPCP interaction with freshwater SROM fractions. PPCPsinteracted mostly with protein-like and fulvic acid-like moieties.Probable SROM interactions for all three model contaminants in-clude hydrophobic interactions in addition to hydrogen bonding,p–p interactions and/or cation bridging.

Acknowledgments

Research support was provided by the Binational AgriculturalResearch and Development (BARD) fund, Grant # IS-3822-06, theWater Research Foundation (Award #4269) and the University ofArizona Water Sustainability Program. The comments and viewsdetailed herein may be necessarily reflecting the views of theWater Research Foundation, its officers, directors, affiliates, oragents. Analyses in the ALEC were supported by NSF Grant CBET-0722579.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.chemosphere.2012.11.059.

References

Allard, B., Boren, H., Grimvall, A., 1991. Humic substances in the aquatic andterrestrial environment. In: Proceedings of an International Symposium,Linkoping, Sweden, August 21–23, 1989. Springer-Verlag, Berlin; New York.

350 S.H. Ruiz et al. / Chemosphere 91 (2013) 344–350

Bai, Y., Wu, F., Liu, C., Guo, J., Fu, P., Li, W., Xing, B., 2008. Interaction betweencarbamazepine and humic substances: a fluorescence spectroscopy study.Environ. Toxicol. Chem. 27, 95–102.

Bolbach, G., 2005. Matrix-assisted laser desorption/ionization analysis of non-covalentcomplexes: fundamentals and applications. Curr. Pharm. Des. 11, 2535–2557.

Cahill, J.D., Furlong, E.T., Burkhardt, M.R., Kolpin, D., Anderson, L.G., 2004.Determination of pharmaceutical compounds in surface- and ground-watersamples by solid-phase extraction and high-performance liquid chromatography-electrospray ionization mass spectrometry. J Chromatogr., A 1041, 171–180.

Chefetz, B., Hadar, Y., Chen, Y., 1998. Dissolved organic carbon fractions formedduring composting of municipal solid waste: properties and significance. ActaHydrochim. Hydrobiol. 26, 172–179.

Chefetz, B., Illani, T., Schulz, E., Chorover, J., 2006. Wastewater dissolved organicmatter: characteristics and sorptive capabilities. Water Sci. Technol. 53, 51–57.

Chen, W., Westerhoff, P., Leenheer, J.A., Booksh, K., 2003. Fluorescence excitation –emission matrix regional integration to quantify spectra for dissolved organicmatter. Environ. Sci. Technol. 37, 5701–5710.

Chin, Y.P., Aiken, G.R., Danielsen, K.M., 1997. Binding of pyrene to aquatic andcommercial humic substances: the role of molecular weight and aromaticity.Environ. Sci. Technol. 31, 1630–1635.

Chorover, J., Amistadi, M.K., 2001. Reaction of forest floor organic matter at goethite,birnessite and smectite surfaces. Geochim. Cosmochim. Acta. 65, 95–109.

Daniel, J.M., Friess, S.D., Rajagopalan, S., Wendt, S., Zenobi, R., 2002. Quantitativedetermination of noncovalent binding interactions using soft ionization massspectrometry. Int. J. Mass Spectrom. 216, 1–27.

Dickenson, E.R.V., Snyder, S.A., Sedlak, D.L., Drewes, J.E., 2011. Indicator compoundsfor assessment of wastewater effluent contributions to flow and water quality.Water Res. 45, 1199–1212.

Guo, M.X., Chorover, J., 2003. Transport and fractionation of dissolved organicmatter in soil columns. Soil Sci. 168, 108–118.

Henderson, R.K., Baker, A., Murphy, K.R., Hamblya, A., Stuetz, R.M., Khan, S.J., 2009.Fluorescence as a potential monitoring tool for recycled water systems: areview. Water Res. 43, 863–881.

Her, N., Amy, G., McKnight, D., Sohn, J., Yoon, Y.M., 2003. Characterization of DOM asa function of MW by fluorescence EEM and HPLC-SEC using UVA, DOC, andfluorescence detection. Water Res. 37, 4295–4303.

Hernandez-Ruiz, S., Abrell, L., Wickramasekara, S., Chefetz, B., Chorover, J., 2012.Quantifying PPCP interaction with dissolved organic matter in aqueoussolution: combined use of fluorescence quenching and tandem massspectrometry. Water Res. 46, 943–954.

Hudson, N., Baker, A., Reynolds, D., 2007. Fluorescence analysis of dissolved organicmatter in natural, waste and polluted waters – a review. River Res. Appl. 23,631–649.

Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M., Zaugg, S.D., Barber, L.B.,Buxton, H.T., 2002. Pharmaceuticals, hormones, and other organic wastewatercontaminants in US streams, 1999–2000: a national reconnaissance. Environ.Sci. Technol. 36, 1202–1211.

Lakowicz, J.R., 1999. Principles of Fluorescence Spectroscopy. Kluwer, New York.Leenheer, J.A., 1981. Comprehensive approach to preparative isolation and

fractionation of dissolved organic-carbon from natural-waters andwastewaters. Environ. Sci. Technol. 15, 578–587.

Leenheer, J.A., 2004. Comprehensive assessment of precursors, diagenesis, andreactivity to water treatment of dissolved and colloidal organic matter. WaterSci. Technol.: Water Supply 4, 1–9.

Leenheer, J.A., Croue, J.P., 2003. Characterizing aquatic dissolved organic matter.Environ. Sci. Technol. 37, 18A–26A.

Leenheer, J.A., Wershaw, R.L., Brown, G.K., Reddy, M.M., 2003. Characterization anddiagenesis of strong-acid carboxyl groups in humic substances. Appl. Geochem.18, 471–482.

Loo, J.A., 2000. Electrospray ionization mass spectrometry: a technology forstudying noncovalent macromolecular complexes. Int. J. Mass Spectrom. 200,175–186.

Maoz, A., Chefetz, B., 2010. Sorption of the pharmaceuticals carbamazepine andnaproxen to dissolved organic matter: role of structural fractions. Water Res.44, 981–989.

Navon, R., Hernandez-Ruiz, S., Chorover, J., Chefetz, B., 2011. Interactions ofcarbamazepine in soil: effects of dissolved organic matter. J. Environ. Qual. 40,942–948.

Pan, B., Ning, P., Xing, B., 2008. Part IV – sorption of hydorphobic organiccontaminants. Environ. Sci. Pollut. Res. 15, 564.

Pan, B., Ning, P., Xing, B., 2009. Part V – sorption of pharmaceuticals and personalcare products. Environ. Sci. Pollut. Res. 16, 106–116.

Polubesova, T., Sherman-Nakache, M., Chefetz, B., 2007. Binding of pyrene tohydrophobic fractions of dissolved organic matter: effect of polyvalent metalcomplexation. Environ. Sci. Technol. 41, 5389–5394.

Schug, K.A., Lindner, W., 2005. Noncovalent binding between guanidinium andanionic groups:   focus on biological- and synthetic-based arginine/guanidinium interactions with phosph[on]ate and sulf[on]ate residues. Chem.Rev. 105, 67–114.

Sharma, V.K., Anquandah, G.A.K., Yngard, R.A., Kim, H., Fekete, J., Bouzek, K., Ray,A.K., Golovko, D., 2009. Nonylphenol, octylphenol, and bisphenol-A in theaquatic environment: a review on occurrence, fate, and treatment. J. Environ.Sci. Health Part A – Toxic/Hazard. Sub. Environ. Eng. 44, 423–442.

Snyder, S.A., Westerhoff, P., Yoon, Y., Sedlak, D.L., 2003. Pharmaceuticals, personalcare products, and endocrine disruptors in water: implications for the waterindustry. Environ. Eng. Sci. 20, 449–469.

Sun, W.L., Ni, J.R., Xu, N., Sun, L.Y., 2007. Fluorescence of sediment humic substanceand its effect on the sorption of selected endocrine disruptors. Chemosphere 66,700–707.

Tolls, J., 2001. Sorption of veterinary pharmaceuticals in soils: a review. Environ. Sci.Technol. 35, 3397–3406.

Wickramasekara, S., Hernández-Ruiz, S., Abrell, L., Arnold, R., Chorover, J., 2012.Natural dissolved organic matter affects electrospray ionization duringanalysis of emerging contaminants by mass spectrometry. Anal. Chim. Acta.717, 77–84.

Yamamoto, H., Liljestrand, H.M., Shimizu, Y., 2004. Effects of dissolved organicmatter surrogates on the partitioning of 17 beta-estradiol and p-nonylphenolbetween synthetic membrane vesicles and water. Environ. Sci. Technol. 38,2351–2358.

Zhou, Q., Maurice, P., Cabaniss, S., 2001. Size fractionation upon adsorption of fulvicacid on goethite: equilibrium and kinetic studies. Geochim. Cosmochim. Acta.65, 803–812.