Bioconcentration of Xenobiotics in Trout Bile: A

Proposed Monitoring Aid for Some Waterborne Chemicals

Abstract. A technique is proposed for the monitoring of certain xenobiotic pollu-tants in suspect aquatic environments by fish bile analysis. Bile removed from rain-bow trout Salmo gairdneri! exposed to nine diferent radioactive compounds in vivacontained concentrations of radioactivity greater than those in the surrounding wa-ter. Bile-to-water radioactivity ratios as high as 10,000: 1 were found after 24-hourexposures. The results of these experiments suggest that analysis of bile of wild orcaged fish from a suspect site may be useful as a qualitative moniton'ng aid for cer-tain types ofxenobioticsin water.

Studies of rainbow trout in this labora-tory have established that several foreigncompounds can be conjugated withglucuronic acid and excreted into bile inhigh concentrations �!. More recently,the results of other investigators have in-dicated that other fish species are able toconjugate certain phenols, such as pen-tachlorophenol, with sulfate �!. AI-

Table 1, Biliary concentration of various xenobiatics by rainbow trout Satmo gairdneri!. Expo-sures were made at 12'C for 24 hours. Water hardness was 134 parts per million, measured bythe CaCO, method, and pH was 7.2. Radioactivities are expressed as disintegrations per minute dprn! per milliliter', each value of the 24-hour bile radioactivity is the mean of a minimum of fiveanima!s from at least two separate exposures. Abbreviation: UL, uniformly labeled,

C pecan Radioactivity� pm/ml!

in H,O H,O Bile mg/ther! � hours! �4 hours!

Ratio Me- bile "C!/ tabo- H,O "C! lites

Compound

0.05 3,010 30,500,000 10,100 1

0,5 1,070 265,000

0,005 310 796,0000.25 1,030 975,000

0.005 305 127,0000. 1 4,070 21,$00,000

0. 5 3,640 39,000

0.1 180 22,500

0.5 2,020 2,150,000 1,064 1

2',5 Dichloro-4'-nitrosalicylanilide Bayer 73; chlorosalicylic acid;ring-UL-"C!

Di-2-ethylhexylphthalate DEHP;carboxyl-"C!

Methylnaphthalene ring-UL-"C!1-Naphthy!-/V-methylcarbamate

carbaryl; naphthyl-1-"C!Naphthalene nng-UL-"C!Pentachlorophenol PCP;

ring-UL "C!2,5,2',5'-Tetrachlorobiphenyl

TCB; ring-UL-"C!1,1,1- Trichloro-2,2-bis p-chlo-

rophenyl!ethane p st '-DDT;ring-UL-"C!

3-Trifiuoromethy�-nitrophenol TFM; ring-UL-"C!

though the biliary concentration of a vari-ety of organic anions including con-jugates of foreign compounds is thoughtto occur through a specific transportmechanism in mammals �!, very fewstudies have dealt with this process infish �!. We have recently reported onthe biliary concentration of several xeno-biotic substances in rainbow trout �!.

247 5?

2,570 ?947 3

414 25,360 2?

11 2?

124 1

This report deals with the biliary concen-tration of a structurally diverse group ofchemical compounds and indicates thatthe sampling of bile may be of potentialuse as an aid in monitoring water qualityor as a diagnostic tool in the investiga-tion of chemically related fish kills.

Rainbow trout usually 10 g of biomassper liter! were placed in a glass tank thatcontained 50 liters of dechlorinated wa-ter pH 7.2! and the '"C-labeled com-pound or compounds. The tanks wereaerated and kept at 12'C and the systemwas allowed to remain undisturbed for 24hours. The concentrations of the com-pounds in the exposures were below thelevel of acute toxicity for the times in-dicated and had no observable effects onthe fish during the 24-hour exposures.The concentrations were chosen for con-venience in metabolite detection ratherthan to simulate environmental levels.The amount of "C in the tank water wasdetermined by counting suitable portionsin 15 ml of ACS scintillation mixture Amersham/Searle! in a model 6872 Searle Analytic! liquid scintillationcounter. After exposure, the fish werekilled by cervical dislocation and the bilewas collected by ga!!bladder puncture.Portions of crude bile were then placedin the scintillation mixture for counting,and the remainder was pooled for metab-olite identification. The pooled bile wasdiluted with water and passed over a bed� by 15 cm! of XAD-2 resin in a glasscolumn and washed with two bed-vol-umes of distified water. The radioactivematerials were eluted from the columnswith three bed-volumes of methanol. Themethanol was then concentrated to 30 ml.Thin-layer chromatography was per-formed on 0,25-mm silica gel plates. Theplates were scanned for radioactivity byscraping segments � by 2 cm! of silicagel from them from the origin to the sol-vent front and counting the gel in ACSscintillation mixture.

225

The data shown in Table 1 indicatethat the ratios of "C in bile to "C in wa-ter after the 24-hour exposures to the in-dicated compounds range from a low val-ue of 1 1 for 2,5,2',5'-tetrachlorobiphenyl TCB! to 10,000 for 2',5-dichloro-4'-ni-trosalicylanilide Bayer 73!. In most cas-es the "C in bile was associated with me-tabolites of the parent compounds andsome of these biliary metabolites havebeen characterized l, 2!. lt is evidentthat the lowest bile-to-water ratios wereassociated with compounds DDT andTCB! that have a high lipid solubility,and this may be due to a low rate of me-tabolism or conjugation related to the se-questration of these compounds by tis-sue lipids. Of great interest are the highbile-to-water ratios of the compoundsthat have comparatively lower lipid sol-ubilities, since from a monitoring pointof view, more polar compounds mayhave a low bioaccumulation potential�!. Although much attention has beengiven in the past several years to moni-toring for chemicals that tend to accumu-late in the food chain, there are few in-novations in the area of monitoring forpotentially hazardous chemicals thathave lower bioaccumulation potentials,such as phenols and certain componentsof petroleum products. A recent reporthas suggested the use of liver ben-zopyrene hydroxylase activity as a moni-tor for petroleum pollution �!. The datain Table 1 concerning naphthalene andmethylnaphthalene appear to be rele-vant, since both of these compounds areconstituents of crude oil 8!, and theappearance of metabolites of these com-pounds in bile in high concentrationssuggests the possibility of using this tech-nique as a tool in the monitoring of petro-leum pollution.

Although more work needs to be doneconcerning the qualitative and quan-

titative aspects of the biliary con-centrating system in the diverse speciesof fish and in the development of specificidentification techniques for xenobioticcompounds and their metabolites, thevalue of capitalizing on this process is ap-parent. The careful design of monitoringmethods based on bile collection from ei-

ther captured fish or caged fish placed ata suspect site may well serve to provideincreasingly needed environmental in-dices 9!,

CHARLES N. STATHAMMARK J. MELANCON, JR.

JOHN J. LECH"

Department of Pharmacology,Medical College of Wisconsin,Mi lwatrkee 53233

References and Notes

I. J. J, Lech, Toxlcol. Appl. Pharmacol. 24, 114,�973!; C. N. Statham and J. J. Lech, J. Fish.Res. Board Can. 32, SIS �975!; C. N, Statham,S. K, Pepple, J. J. Lech, Drug hferob. Dispos.3, 400�975!; M. J. Melancon, Jr., and J. J. Lech,ibid. 4, 112 �976!.

2. K. Kobayashi, H. Akitake, T. Tomiyama, Bull.Jpn. Soc. Sci. Fish. 36, 103;1970!; H. Akitakeand K. Kobayashi, ibid. 41, 321 �975!.

3. R. L, Smith, Prog. Drug Res. 9, 299 �966!; A.M. Guarino and L. S. Schanker, J. Pharmocol.Exp. Ther. 164, 387 �96$!.

4. D. A. Schmidt and L. J. Weber, J. Fish, Res.Board Cart. 30, 1301, �973!; J, B. Hunn and J.L. Allen, Annu. Rev. Pharmacol. 14, 47 �974!;Comp. Gert. Pharmacal. 6, IS �975!.

S. J, J, Lech, S. K. Pepple, C. N. Statham, Tax-i col. Appl. Phormacol. 2$, 430 �973!.

6. J. L. Hamelink, R. C, Waybrant, R. C. Bail,Trans. Am. Fish. Soc. 1$$, 207 �971!; D, R.Branson, G. E. Blau, H. C. Alexander, W. B.Neely,ibid. 104, 785�975!.

7. J. F. Payne, Sciertce 191, 945 �976!.8. R. F, Lee, R. Sauerheber, G. H. Dobbs, ttfar.

Biol. I'7, 201 �972!; S. Warner, Battelle Memo-rial Laboratories, Columbus, Ohio, unpublishedanalytical report.

9. R. E. Train, Science 17$, 121 �972!.10. J.J.L. is a recipient of National Institute of

Environmental Health Sciences career develop-ment award ES00002. C.N.S. is a staif fellow mthe Pharmacology Research Associate Programof the National Institute of General Medica! Sci-ences, National Institutes of Health, Bethesda,Md. 20014. Supported by NIH grant ES01080and EPA grant R80-397-1010.To whom reprint requests should be addressed.

19 May 1976

226

Isolation of Xenobiotic Chemicals fromTissue Samples by Gel Permeation

Chromatography

D. W. Kuehl and E. N. Leonard

Reprinted with permission from Araattlticat Chemistry,50 l!:182-85 January 1978!. Copyright 1977, American

Chemical Society.

227

Reprinted from ANALYTICAL CHEMI8TRY, VoL 80, Page 182, January 1978Copyright 19r 7 by the American Cheinicel Society snd reprinted by permission of the copyright owner

Isolation ol Xenoblotio Chemfoaia from TiasiIe Sampieeby Gel Permeation Chromatography

Oouglae W. Kuehl' and Edward lrL LeonardU.S. Enrdronmenral ~ Agency, Enveonrnenarl ~ fabonsioiy~. 8201 Conrfgcn Bcsdsrssrrt Duluth, ~ 55804

229

Since automatic gel permeation chromatographic GPC!systems were first described for the cleanup of samples withhigh fat content l, 2!, efforts have been made to characterize�! and utilize �, 5! these systems more fully. Mulder and

Buytenhuys �! had previously reviewed applications of GPCfor the separation of a variety of orgsnics on Bio-Beads orSephadex LH-20 with various organic solvents. An excellentreview on the fundamental gel network structure, its ability

ANALYTICAL CHEMISTRY, VOL 50, NO. 1, JANUARY 1978 ~

Table I, Gel Permeatio e and Recovery DataCyclohexene/CH,Cl,

n Chromatography Retention Volum

CH,Cl,Retn vol, mL Recovery, %

160-190 100150-186 95168-199 100168-195 100144-168 90144-171 100162-178 100156-174 100158-176 89152-169 84156-174 95156-181 98162-183 94170-195 100172-197 90166-196 95170-195 94154-180 89166-192 100172-200 100147-161 100167-181 100174-2 00 75180-210 93154-176 100176-200 100168-195 99

Compound

tLuTHRI volvss sicIso vs sM

~ LUTION VOI,Vus Wl

Rgure 2. GPC ahromstograms showing the sepersffaa of Zrw mafscuevweight organic chemicals LMWO! from iipids in the upper trace andfrectionatloa ai LMWO'e ia iower trace

pramelas! tissue with enough anhydrous sodium sulfate to drysample. Extract with hexene/acetone �+1! on Soxhlet extractorfor 8 h, Evaporate solvent and dilute concentrated oil to 100mg/mL with CHsCle Inject a seriee of 5-mL aliquots on aGPC/CHzClz system, operating at 3,5 mL/min. Collect properfraction in a Kuderna-Danish apparatus, Concentrate sampleto 5 mL.

Equilibrate GPC system by recycling CHiCls/cyclohexene.Inject 5-mL sample and collect fractions.

Apparatus and Reagents. Salaents. Hexane, cyclohexane,methylene chloride, and acetone, redietilled pesticide grade Burdick and Jackson, Muekegon, Mich.!.

Gel Permeation System. Dual 25 cm x 2.5 cm glass columnsconnected in series and filled with 100-200 mesh'Bio-Rad SX-2beads. A high pressure all-Teflon sample valve Durrum Madel24089, Palo Alto, Calif.! with a 5-mL sample loop was used forsample injection, A 254-nm ultraviolet detector Verien Aero-graph! and a recorder Varian Aeragraph inadel A-25! were usedto monitor the chromatogram. All connections and sampletransfer lines were Teflon,

Gas-Liquid Chramatagraph-hfass Spectrometer, VerianAerograph Model 1700 GC equipped with flame ionization de-tector and 6-ft x '/<in. i.d. glass column packed with 3% OV-101on 80-100 mesh Gas-Chrom Q wae used for GPC recovery studies.The GLC-MS system is e Varien MAT CH-5 system, and the

230

Arociar 1254A roc! or 1 01 6HexachlorobenzensNaphthaleneHexechlorobutadienep,p-DDTo-ChloraphenolPentechloraenisole2,4,6- Tribromoanisole2,4-Dibromoanisolep-Bromosnieoleo-Bramophenol2,4-Dibromophenoi2,4-Dichiorophenol2,4,6- TrichiorophenolPentschlorophenoi3,4.DichloroanilineDipheayiamine1-Naphthylaminem-Chloroaniline2,4,6-Tribromophenol5-Bromoindoie1,2,4-Trichlorobenzenep-Cblorophenoip-BromophenoiPyrenePhenanthrene

Figure 1. GPC chromatogram showing separation capabiiigee ofmethyleae chloride-cycfohexsae mixed solvent system on a Bxx4tadSX-2 columnta fractionate on the basis of steric exclusion, and the chemicalcontributians of chromatographic affinity hae been presentedby Freeman �!. Gel permeation, gel filtration, or molecularsieve chromatography are synonymous terms for separationby eteric exclusion, i.e., differences in solute molecular size,In theory, inert gels have the ability to function as a sort ofmaes spectrometer where the degree of permeation variesinversely with solute molecular size. However, chromato-graphic affinity due principally to hydrogen bonding betweenthe solute and the gel network has been shown to have adefinite effect upon column performance 8!. The finding ofan orderly relationship between measured affinity and soluteproton-donor strength suggests a new framework for studyinghydrogen bonding, for measuring the proton-donor strengthsoF chemicals and for performing chemical separations 9!.

The objective of this work wse to develop an efficient, rapidmethod for the isolation of low molecular weight polar organicein fatty tissue for subsequent gas-liquid chromatographic GLC! � mass spectrometric MS! analysis. We describe a2-etep GPC cleanup procedure for samples with high fatcontent that uses first the steric exclusion principle and thentakes advantage of the combination of the steric exclusion andchromatographic affinity phenomenon.

EXPERIMENTALProcedure. Blend ground fathead minnow Pimephales

Retn vol, mL160-190150-188168-198170-195148-169152-178170-198172-196172-198168-202174-200174-208182-211190-218192-244186-216200-255184-214206-2S7216-246194-214224-270234-266240-284220-244252-294250-284

Recovery %1007284868757

1007598676084

10087819091836084806164

100958787

Retnvol shift

0 0 0 24 8 8

1614161818202020203030404450586060667682

~ ANALYTICAL CHEMISTRY, VOL. 50, NO. 1, JANUARY 1978

Fish Tissue by GC/MS after Sample

Fractions 5, 6, and 7

ChlorophenolIndoleBromoindoleDibromoindoleTribromoindoleDibromomethylindoleTrichloromethylindolePentachlorophenolTrichlorophenylphenolNicotinamide1,3-DiphenylpyrazolineDichlorodibenzofuran

MethylnaphthyleneC-alkylnaphthyleneC-alkylnaphthyleneBiphenylDichlorobiphenylTrichlorobiphenylTetrechlorobipheny1TrichlorobenzeneTetrachloroanisoleTribromoanieolecis-Chlordanetrans-Chlordanecis-Nonechlortrans-Nonachlorp,p'-DDE

CHzCLr; b! 10% CHsCls/90% CsHis c! 33% CHsCli/67%CsHig d! 50% CH,Cl,/50% CsH�; and e! 75% CHrCI,/25%CsH». The 50:50 mixture was the best compromise of highpercentage recoveries and compound separations. Figure Ishows an example of the resolution obtained with the 50:50mixture. The retention volume and percentage recovery datawith the 50:50 mixture end 100% CHzClz are presented inTable 1.

All compounds tested eluted atgow retention volumes innarrow bands and at approximately the same retention volumewhen 100% CHzC1, was the solvent. This system, therefore,can be used as a rapid and efficient technique for the bulkseparation of lipids from low molecular weight organicchemicals LMWO!. The unique elution of compounds fromthe mixed solvent GPC system can then be used to fractionatethe LMWOs into polar and nonpolar organics. This 2-stepprocedure has been demonstrated for an extract of fish thathad previously been exposed to a bromine chloride-disinfectedwastewater effluent �0!.

Figure 2 shows a GPC trace for the lipid/LMWO separation step 1!, and a GPC trace for fractionation of the LMWO bypolarity step 2!. After each sample was screened on the GC,fractions 3 and 4 were combined; 5, 6, and 7 were combined;and 8, 9, and 10 were combined for GLC-MS analysis. Thequalitative results are presented in Table II. These fractionscontained compound types such as phenols, anisoles, andheterocyclic aromatics dibenzofurans, dibenzothiaphenes,indoles, etc.!. Fractions 1 and 2 were analyzed separately andwere basically PCBs and chlorobenzenes.

Commercially available GPC units can be used for bothsteps of this technique. Typically a timing unit is used toprovide "waste", "collect", and "wash" cycles for repetitiveprocessing of a single sample. Step 1 can therefore becompleted, after sample loading, without operator assistance.Step 2 can also be automated by routing the effluent to afraction collector during the "collect" cycle. However, sincestep 2 is done only once for each sample, and each sample maybe slightly different, it may be advantageous to allow theoperator to decide when each fraction should be collected.

Varian MAT Spectrosystem-100 MS date system was used fordata acquisition and processing.

Standards. All standard compounds used for retention volumeand recovery studies Table I! were from the Lab Assist kit Chemservices, Inc., West Chester, Pe.!. Solutions of 1 mg/mLCHrCIr were prepared.

RESULTS AND DISCUSSIONThe isolation of nonpolar xenobiotic organics from fatty

tissue on Bio-Rad SX-2 copoly styrene-2% divinylbenzene!!with cyclohexane as the solvent is a very efficient samplepreparation technique for subsequent GLC-MS analysis �!.Recoveries of PCBs, for instance, are generally better than95%. A disadvantage of this system, however, is the lowrecovery of polar organics such as pentachlorophenol, whichwas only 10%, On this system, however, a mixture of equalquantities of corn oil, Aroclor 1254, and pentachlorophenolcan be completely separated. Excellent chromatographicresolution can therefore be obtained if low recoveries of thepolar chemicals are considered acceptable. Polar solventsystems will reduce retention volumes, reduce band broad-ening, and increase recoveries of polar compounds to ac-ceptable levels. Johnson et al. �! observed that polar solventsystems such as mixtures of toluene and ethyl acetate willresult in high percentage recoveries of both polar and nonpolarchemicals. This system, however, did not provide anychromatographic resolution of polar and nonpolar compounds.Nonpolar solvents readily elute nonpolar solutes from the gelnetwork, whereas larger quantities of solvent are required toeventually elute polar solutes at low yields and both polar andnonpolar solutes co-elute at high yields with polar solvents.

It was necessary then to develop a GPC system that wouldyield high recoveries of both polar and nonpolar chemicalsand would give good chromatographic resolution of polar andnonpolar chemicals. In addition, the solvent system had tobe highly volatile so that the more volatile chemicals isolatedfrom samples would not be lost during solvent removal.

Retention volume and recovery studies for various polarand nonpolar organics ranging from p-chlorophenol to PCBswere conducted for the following solvent systems: a! 100%

231

Table II. Compounds Identified in

Fractions 3 and 4

Cleanup by Two.Step GPC MethodFractions 8, 9, and 10

AnisoieDibromoanisoleTribromoanisoleChlorobromoanisoleTetrachloroanisolePhenolChlorophenolDibromophenolBromodichlorophenolChlorodibromophenolEthylphenolDihromocresoiDichlorobenzeneTrichlorobenzeneChloronaphtheleneDichloronaphthaleneDibenzofuranBenzthiazoleMethylbenzothiapheneDibenzothiapheneNaph thylemineMethylcarbazoleIn doleDibromoindoleTribromoindoleDibromomethylindoleTribromomethylindoleTetrebromomethylindolePentschloroanilineDibromomethylbenzothiazole

LITERATURE CITED

232

Future developtnents involving automatic GPC units for thecleanup of fatty tissue are the addition of an ezpandedfraction-collection systetn, the addition of gradient solvent.elution capabilities, and the use of a variety of detectors fora more cotnplete characterization of the GPC elution patternof a satnple.

�! R. C. Tlndkt end D. L. Staang, Anal. Chem� ii. '�68-'t773 �972!.�! D. L atagng, R. C. TkuIe, and J. L Johnson, J, Aaaoc. Olr. Ared. Q»m.,

5$, 32-38 �972!,�! K. R, &IIIII nnd J. C. Crew, J. Anode, Oly. Anal. Chem., $7, 188-�2

974!.

ANALYTICAL Ct%MISTRY, VCL. 60, NO. 1, JANUARY 1$7$ ~

i! Q. D. Vetth, D. W. Kutdd, and J. Roaend»I, J. Aaacc. OI7. Anal. Chem.,$$, 1-6 �976!,

�! L, D. Johnson, R, H. Walla. J. P. Venery, and F. 6. Knlaer, J. Aaacc.Oty. Anal. Q»m., 5$, tro-179 I97$!,

�! J. L Mtddar and F. A. ~ J. Chmmakkr., $1, 469-477 �970!.�! D. H. Freeman, J. Chrom»togr�dcl., 11. 176-180 �973!. 8! D. H. Freeman and R, M. AIOdm. J, Chroma»gr., gcl., 12, 730-736

tgyd!, 9! D. A. Fraen»n, R. M, Ange!en, D. P. Enagonb, and W. My, Anal. Chem.,

d$, 7~TO '19'73!.�0! H. L. opperman, D, W. KueM, and 6, 6, gaea, Proceedlnga ol 9»

~ On EnrtrOnnant» Impaot OI Water Cttkntnadcn., Oak RklgaHeIonal ~, Oak Rtdge. Tenn., Octet»r 22-2i, 1975, pp 327-3a6.

Rg!cggvgn for review August 26, 1977. Accepted October 17,1977.

Organic Compounds in the Delaware River

L. S. Sheldon and R. A. Hites

Reprinted with permission from Znoironmental Scienceond Technology, 1R�0!:1188-94 October 1978!. Copy'-

right 1978, American Chemical 8ociety.

233

Reprinted from ENVIRONMENT SCIENCE gt TECHNOLOGY, Vol. 12, Page 1188, October 1978Copyright I 1978 by the American Chemical Society and reprinted by permission of the copyright owner

Organic Campounds in the Delaware River

cancer is very high in the areas surrounding the DelawareRiver �0, 11!.

Based on these considerations, it is logical to ask if there isa correlation among cancer incidence, orgsnics in the drinkingwater, and organics in the Delaware River. This is obviouslya very complex and difficult question, but the answer mustbegin with a complete study of the organic compounds in theriver itself. Our study includes compound identification andquantitation as well as a preliminary assessment of theirsources, The analytical techniques used in this study includedvapor stripping of volatile organic compounds, liquid chro-matographic LC! fractionation, high-resolution gas chro-matography, computerized gas chromatographic mass spec-trometry GC/MS! in both the electron impact EI! andchemical ionization CI! modes, and high-resolution massspectrometry HRMS!.

River System. The Delaware River is a 350-mile-longwater-way rising in central New York State, running throughheavily industrialized areas of New Jersey, Pennsylvania, andDelaware to the Delaware Bay, Water flow in the lower thirdof the river from Trenton, N.J. river mile 132!, to the Bay, isdominated by tidal action; tidal volumes around Philadelphiaare at least an order of magnitude greater than the down-stream river flow �2!. In general, any effluent discharged intothe river will travel approximately 16 miles during one tidalcycle � 7 miles upstream during high tide, and 8 miles down-stream with ebb flaw �2!. This tidal action is important forthe movement of industrial wastewaters discharged into theriver both by dispersing the effluent particularly in the up-stream direction! and by prolonging residence times. Addi-tionally, during periods of normal flow the lower 55 miles ofthe river is an estuary characterized by salinity gradientswhich limit both domestic and industrie! usage �3!. Back-ground data provided by the Delaware River Basin Commis-sion show that the ratio of municipal to industrial dischargeis approximately 2 to 1 8!, Information provided on effluentsource locations 8! aided in selecting representative samplingsites,

~ Nearly 100 compounds were identified in Delaware Riverwater samples taken in August 1976 and March 1977 betweenMarcus Hook, Pa. river mile 78!, and Trenton, N.J. rivernule 132!. Extraction with CHzClq, liquid chromatographiccleanup, and gas chromatographic mass spectrometry wereused for compound separation and identification. The ob-served compounds included natural products, municipalwastes, and industrial contaminates. The latter were of threetypes: those found in industrialized urban areas with no par-ticular production source such as aromatic hydrocarbons andphenols; those commonly used in manufacturing processessuch as plasticizers and industrial solvents; and those specificto a single source and traceable to that source such as a seriesof chlorinated aliphatic and aromatic polyethylene glycolswhich were specific to a plant in the Philadelphia area.

Thirty million people in the United States drink watertaken from rivers �!. In many cases, these rivers also receivewastewaters from surrounding municipal and industrial dis-chargers. Although both drinking water and wastewater areusually treated to make them safe, many of the treatmentprocesses are not fully effective. This has resulted in thewidespread contamination of river waters snd, hence, drinkingwaters. It is obviously of critical importance to know theidentities and abundances of these contaminants in order toassess health effects and to devise rational control proce-dures,

There is now a growing knowledge of the organic com-pounds in the drinking water of many cities �-4! and, to amore limited extent, in the wastewaters of several types ofindustries �-7!. There have been, however, few studies of therivers themselves. This is unfortunate since the river is theever-present, connecting link between wastewater anddrinking water. Furthermore, the river can act as a carrier,sink, or reactor causing transformation of compounds to moreor less hazardous species.

This paper reports on a detailed study of the organic com-pounds in the Delaware River. We selected the Delaware forseveral reasons: Over 120 major chemical manufacturingplants see Figure 1! are located along its banks and manydischarge wastewater, either directly or indirectly, into theriver 8!. The Delaware is a major source of drinking water formany of the cities and counties in the area; for example, itprovides 50% of Philadelphia's water 9!. The incidence of

Experimental

In August 1976, 11 grab samples �.5 L each! suitable forsolvent extraction were collected from the center channel ofthe Delaware River between Marcus Hook, Pa. river mile 78!,and Trenton, N.J. river mile 132!, In October, two additionalsamples �,0 L! were taken from the shore for volatile analysis,

Environmental Science 4 Technology

235

Linda S. Sheldon and Ronald A. HlieeDepartment of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Mass. 02139

IHRH78HStouuar 132stapap IRr.r

Pal ~ ruon Papor IRR, IPurou »8 8 IRC3 Hoollar

Achro t liaac 117 II ~ 3 HatCulc ~Pupoht 115 117 7 tcnnoca

198 AlrcaI CO Aolllll ~ Hoot Hellad

ICO A.P. Otoun101.4 Alltcd999 A CA95.8 SAP, »archon

Phrla d orpit u

sco» popar as.4 ~ Oa-a93 Snoii,ein,pauiaharo,SPCii878-87.2 Racon, Holiti thl,acaon, Horculoo

oas 88 3 Air productc, puponias 8-84 9 woc»nohouao, Union coroldo, Alrao ooc80 2-75.8 SP all, PHC, Honuolki

Sun cll, A iliad 78.2

Wrltn nalolr S2.7 Diamond Shamrock82 2 Stauff tr

Figure 1. Delaware River between river mlles 60 and 140 showing lo-cations ol chemical companies side not slgnlllcant!Semollng Sltea: O = CCS8Ctad Attsopt 1976, rrk = COIISCted OCtaber 1976, 0= collected hisrch 1971

Figure I gives water sampling locations. This particular riversegment was of interest because it is the most heavily in-dustrialized area along the river and is a direct source ofdrinking water for several of the cities in the region includingPhiladelphia. In addition, because of high tidal flows, thissegment is well mixed, but it is far enough upstream to avoidestuarine salinity effects.

Grab samples were collected in 1-gal amber glass bottleswith Teflon-lined caps at depths of 0.5 � L0 m, All centerchannel samples were collected aboard the Aquode/phia, theboat used by the City of Phfladelphia for its own riversampling program.

Samples for volatile analysis were packed in ice to slow bi-ological degradation prior to analysis, while samples for sol-vent extraction were immediately preserved by acidifying topH 2 with hydrochloric acid and by adding approximately 250mL of dichloromethane. Addition of the organic solvent alsostarted the extraction process. Sample workup was begun assoon as possible after returning to the laboratory, usuallywithin 24 h. In all cases, samples were kept refrigerated untilanalyzed. Analytical techniques for concentration, separation,and identification using solvent extraction, gas chromatog-raphy, and mass spectrometry have been discussed in detailelsewhere �!.

The GC and GC/MS analyses of the initial samples indi-cated low levels of all orgsnics in the sub-ppb range! and onlygradual qualitative changes in the sainple composition as afunction of river location. Therefore, only five grab sampleswere collected in early March 1877; these were larger in vol-ume �1 L! and were more widely spaced than the first group.Their locations are given in Figure L

Initial results indicated that the samples were extremelycomplex, containing not only mixtures of industrial and nat-ural organic compounds but also high background levels of gaschromatographicaUy unresolvable materials. Organic back-ground interferences were removed from the extracts usinga silica gel chromatographic cleanup procedure. Dichloro-methane extracts were evaporated to dryness, and transferredto a column � X 0.6 cm i,d.! packed with deactivated silica gel�96 water!. The sample was then fractionated by successivelyeluting with l0 mL each of hexane, benzene, and methanol,To remove fatty acid interferences, fractionated extracts weredissolved in dichloromethane and extracted with aqueousNaOH pH 9-11!. GC/MS analyses were run on all samplesprior to each of these cleanup procedures to verify that sample

integrity was being maintained, that contamination was notintroduced, that major compounds were not lost, and thatsample components were not degraded. In addition, blanksrun for all of the concentration and cleanup steps showed nosignificant contamination.

IdentiTication of compounds in the river water extracts wasbasbd on coincidence of gas chromatographic retention tunesand on equivalence of electron impact and chemical ionizationmass spectra with those of authentic compounds. Thosecompounds not commercially available were synthesized inour laboratory.

We should emphasize that this study was primarily quali-tative; our principal goal was to identify compounds in theDelaware I iver rather than to exactly measure their abun-dance, For this reason, the concentration data are only semi-quantitative. These data were based on GC peak heightsmeasured from chromatograms of the CHzClz extracts beforeany cleanup procedures were applied. The GC response factorswere determined for nine repres'entative compounids andranged from 0.085 to 0,20 ng/mm, Since solvent extractionefficiencies were not determined, the concentrations reportedbelow are minimum levels, Other errors in quantitation resultfrom losses of volatile compounds such as toluene and chlo-robenzene! and from poor GC resolution due to high back-ground interferences in some nonfractionated samples, Takinga of these factors into account, we estimate that the errorsin quantitation range from +5099 for the more abundantcompounds ppb! to an order of magnitude for some of thevery trace level compounds �.01 ppb!, On the other hand,the Pe ati ue concentrations of a given compound which weremeasured at different river locations or times are more accu-rate because all samples have the same experimental bias.

One of the major experimental problems associated withGC/MS analysis of environmental samples is the possibilityof artifact formation during sampling, concentration, oranalysis procedures, Because we are dealing with a complexand undefined sample matrix, it is often impossible to predictreactions of individual compounds within this matrix or to runcontrolled experiments to test for their occurrence. However,we have taken care to both minimize these effects in terms ofsampling, handling, and cleanup techniques and to considerthem during data interpretation. For example, since liquidchromatographic cleanup separated sample components intothree groups, the number of interactive effects which couldoccur was reduced. Furthermore, in an LC fraction onlycompounds of a given polarity should appear. If within a givenfraction, a specific compound is identified which is outside ofthe proper polarity range, artifact formation was consid-ered.

Resu ts and DiscussionThe organic compounds identified in the Delaware River

water samples are listed in Table I, The data include theconcentration range and the location of the maximum con-centration for each sampling season, Structures for a selectedgroup of compounds are given in Figure 2.

Figure 5 shows high-resolution gas chromatograms for asingle water extract after sample workup. An inspection ofthese results demonstrates the high GC resolution and dy-namic range resulting from our procedures. Without this highdynamic range, only a few of the most abundant compoundswould have been identified.

The compounds listed in Table I are derived from threeprincipal sources; natural products, municipal wastes, andindustrial contaminants. Examples of each source are includedin the following discussion.

Compounds I, 2, and 2 are naturally occurring compoundsresulting from the normal biological processes taking placein the river. The first compound, 6,10,14-trimethy!-2-penta-

236Volume 'l2, Number 10, October 1976

isoprenoids1. 5, I 0, 14-trlmethyl-2~tadecanone2. o-terplneol'3. chhtrophys v

steroids4. cholesterol5. cholestene6. cholestanol

fatty acids and eaters7. stearic acid8. palmisc acid9, methyl stesrate

10. methyl palmltate11. methyl myrtstale

aromatic hydrocarbons12. benzene1S. IOluene14. Cz benzenes15. Cs benzenes16, C4 benzenes17. Cs benzenes18. styrene19. rr-methylstyrene20. Cs unsaturated benzene21, naphthalene22. methylnsphthalenes23. Cz naphthalenes24. Cs naphthalenes25. Cs naphthalenes26. pyre no27. fluoranthene28. anthracene29. phenanswene30. methylphe~31. chrysene

phenols32. phenol33. cresols34. Cr phenots35. Cs-phenols36, Cg~37. Cs-phenols36. pf 1,1,3,3-tetramethylbutyl!phenol'Sa. nonyiphenols40. phenylphenol41. cumylphenol'42. methylisoeugenol

chlorinated compounds43. chiorobenzene44. dichlorobenzene45. trichlorobenzenes46. chlorotoluene47. benzyl chloride48. dichloromethane49. chloroethylene50. chloroform51. Irichloroethyhne52. Ielrachkuoethylene53. dichlorophenois54. Irichlorophenols55. trlfluoromethyi!chloroanillna

NDs0.5-4 �!v4-6 �!

0 8-2 f1!98 ND98 3-16 �1!

108

106

5-10 �!ND

4-9 �!

939393

3-8 �1!trace

1-2 �1!

78

78

98as

NQNONQNONO

939393

NDND174

NDNDNDND

0.7-0.9 �!0.4-1 �!1-5 �!2-5 �!

0.2-0.5 �!'tracetracetracetracetracetrace

NDNDNDNDNDNDNDNDNDNAoNANANANANANANANANANA

D D D D D D DD D DD D D D

98as

as76, 98

787878

2-4 �!2222

0.4-2 �!1-2 �!1-2 �!

0.3NDND

98 ND98 NDas98 ND98 ND

105 trace98 0.2-2 �1!

105 0.04 � 1 �1!98 ND

NQ

9698

96as9698D D D D D D

D D D D

7.00.4

0.5-1�!3

NDNDNDNDND

0.32

trace-2 �!

98as78

Environmental Science 8 Technology

237

Table I. Compounds Found In the Oeiawere River

Table I. Continued

races.

ND0.2-2 8!0.1-1 8!

traceHA

tracetrace

r sees,

2-3 �!NQNQNQNQND

trace

78SS

56. trit luoromethyl!chloranltrobenzene51. bis chlarophsnyl! ketone58. bis chlarophenyl!methanol'59. chloraphenyiphsnyimsthanol'60. 1, t-bis chlarophenyl!-2,2<ichloroethylene'81. chloromethylacetophenone62, CqaHq iCIsOS

ethylene glycal derivatives63, bls�-chloroethyl! ether64, 1,2-bis�-chlorosthoxy!ethane'65. 1-�-chIcraethoxy!-2~henoxyethsneBe. 1-chlora-2-[2'-1', 1',3',3'-tstramethylbutylphenoxyjethoxy]ethane67, 1-chlora-2�-[2- p-I',1',3',3'-tetramethytbutytphenoxyjethoxy]ethoxyj-

ethane'68. 2- p-1',1',3',3'-tstramsthylbulylphenoxy!ethanol'69. 2-[2- p-1',1',3',3'-tstramsthylbutylphenoxy!ethoxy]ethanol70. 2i2-[2Ep-1',1',3',3'-tetramsthylbutylphenoxy!sthoxy]ethaxyjsthsnol'71. bis�-{2- nbutoxy!ethoxy]lethoxy! methane

eaters plastlcizsrs!72. tri ferf-butyl! phosphate73. trl�-butoxysthyl! phosphate74. triphenyl phOSphate75. dibutyl phlhalate76. dioctyl phthalstss77. butylbsnzyl phthalates78. dimethyl terphthalate79. di�-ethylhexyl! adlpate80. dgisobutyl! azelate81, dl�mthylhexyl! sebacats82. tetrasthyleneglycoi dl�wthylhexanoate! '83. tetraethyleneglycol dl�-msthylheptanoate!'84. tristhyleneglycal dl�-ethylhexanoste!

others85, 2mthylhexanol88. 2,2,4trlmethyl-t,3-pentanedlol-f-isobutyrate87. 2,2,4trimethyl-t,3~tanedkd-3-lsobutyrate88. phenyl-2~ropanoi89. isophorone90. nitroxylens91. a-phsnytsnisokr92. blnsphthyl sulfonss93. caffein94. methylcyclohexane95. methyl isabutyl ketone96. sthylthiopyridine97, phthallic acid98. 1,1,1-trlphsnylsthane99. Iluorsnone

939393

7878

tr'ace15 �!

3trace-2 �!0.2-4 �!

98 NO98 ND98 NQ98 0.01-0.2 �!98 0.03-0.1 �!

101101'l01

NQ

NQ78 2 3�1! 115

NQNQNQ

1-3 �!

78 0 06-0.4 �0!78 0.4-2 �1!78 0, 1-0.4 �1!

0.1-0,4 �1!0.08-2 �1!0.3-0.3 �1!

HD0.02-0.3 �1!

98 ND98 ND78 1-4 �1!78 0.1-0.3 �1!78 ND

888893

0.4-2 �!0.3-3 �!

0.1-0.3 �!0.2-0.6 �!

3-5 �!0.4-1 �!

0.0e0,08-0.3 �!

NQtrace

1-14 �!0.1-0.3 �!o.e-f �!

106108

981878787898

3-5 �!1-8 �!1-4 �!2-3 �!trace0.3

traceNQ

traceNOHDNDNONDNQ

NDNDNDNDNDHDNDNQNDNOND

traceNQNQNQ

9878

D D103

water extracts, but it was a minor component in the winterwater samples. This is not surprising since chlorophyll comesfrom algae and phytaplanktan which are at higher levels inthe river during the summer months,

Municipal waste effluents are characterized by high can-centzatians of sterols, fatty acids, and fatty acid esters �7!.

dscanane, probably results fram the oxidative degradationof phytal; other workers have found this ketone in varioussediments �4! and plants 1 5!. Chlorophyll was observed ino',." GC/MS analyses as phytadienes which are produced inthe injection port by pyrolysis of the phytal ester part ofchlorophyll �!. Chlorophyll was abundant in the August

Volume 12, Number 10, October 1918238

' D: compound was detected in the vapor stripping analysis of the October samples; quantitation was not passible. s ND: compound was notdetected. Number indicates the number of sampkrs oul of 11 where ths compound was found. When no number is shown the compound wasdetected in oniy one sample. v Number Indicates the number of samples oul of 5 where the compound was tound. Chkrrophyll was identifiedfrom phytadisnes 18! which are its volatile pyrolytic degradadon products, 'NQ: compound was detected but for various reasons lt was notquantitated. r NA: analysis ol the LC fraction containing these compounds was not carried out. "Exact structtxe not known. For structure, seeFigure 2.

CNO CNONICN CN Ot�~iC � CNC � C-CN

CN! C0 CN ~

CN ~NO-C ~ CVO

0 0 CNC CNONO~0-CN,- -C ~,

Ctll C ~

OO 1 CI ..0~ I 0 C'I . '0OO 0 OlOO N ON ~ 0tO ~ ~ ~ 0

CHI.O~i � Q

Cl 0tC IC»ICNICHPOC ll ~ OC IC CO I ~ C ~ l

'll C NC ~ ~ CN ~ CNIIIC INCH ~ 0ll POC+'

0~0~Ii

0 0N CO CHIC ~ OI�C 0

0 ~ N CH CNP 0 ~ CNC 0CNP

l ~ CH 0 CCNCC ICNOCC

*0 0 N Ctt C INC II CHICO ~CNPC

CW I0 I ~ONCO~i

Cl CNI COO~ 00

CNP

CICI CHICNNOC ~ C CCICI CCC ~ OC ~ NCttlOC IICNCCICt INC CNCOC 0 COP 0 ~ 0 C NC CH ~

239

Figure 2. Structures of selected organic compounds found in the Del-aware River see Table I!

These compounds no. 4 � 11! were found at high levels in mostof the samples from the Delaware River. Far example, cho-lesterol was usually one of the most abundant compounds inthe water. The concentration profile for cholesterol in theAugust water samples showed a maximum at river mile 93; thisis consistent with locations of municipal sewage plants in thePhiladelphia-Camden area. Fatty acids were not quantitateddue to their poor chromatographic resolution, but they werepresent at very high levels in all samples,

The anthropogenic chemicals were by far the most nu-merous group of compounds and are the compounds ofgreatest concern to this study. In reviewing concentratian andsource data for these chemicals, it becomes apparent that theyare of three types: those found in industrialized urban areaswith no specific production source; those commonly used inmanufacturing processes with multiple sources; and thosespecific to a single industrial site and traceable to thatsource.

Included in the first group of general industrial contami-nants are all of the aromatic hydrocarbons no. 12-31!, mostof the phenolic compounds no. 32-42!, most of the chlori-nated species no. 43 � 54!, and some industrial solvents no.85, 88, 89, 95!. Almost all of these compounds have been iso-lated and identified in urban watersheds �8-20!; they appearto arise from automobile emissions, water chlarination, andgeneral urban activities.

Source identificatian for several of the phenols, notablyp-�,1,3,3-tetramethylbutyl!phenol, and the nonylphenalisomers is more difficult. Concentration data show highestlevels around Philadelphia, implicating general urban activityas the primary source; however, there are several high pro-duction chemical companies in the area, one of which producesthese phenals commercially �1!. Under these circumstancesno definite source can be identified.

Perhaps the most ubiquitous of all environmental con-taminants are the plasticizers no. 72-84!. These compoundscan be found in the wastewater from a large number of in-dustrial sources �8!. The most common plasticizers phtha-lates and adipates! show no concentration maxima along the

Environmental Science a Technology

lc 0 ot cl w tc cc e ll '0 Il co 00IO CC 00 % u OO Ill IOI W Ol Nl IPC 00 CO leI I 'Ct

~ ~ ~ I 0 0 PI u 8 W !l e ll IO ll Cl PP 4 e t!IO Nt IO W 'll I I OI OO II W W III 00 HC Wt ~ I CtFigure S. Gas chromatograms of benzene A! and base extractedmethanol �! fractions of Delaware River water collected March 1 977 river mge 99!.Run on SE-52 open tubular �5 m X 0,22 mm! staaa ~lttay column. Numberedpeaks Idantitled ln Table I

river and may be assumed to enter from multiple locations.Some of the less coinmon plasticizers tend to maximize inparticular river segments, suggesting single point sources. Forexample, tri ferf-butyl! phosphate, and tri�-butoxyethyl!phosphate, maximize near river mile 78 in the winter. Theseconcentration data are consistent with commercial productionsites along the river �1 !,

The plasticizer, tetraethyleneglycol di�-ethylhexanoate!,no. 82 on Figure 3, is interesting because it was both the mostabundant compound in the river and the most challenging toidentify, The electron impact mass spectrum for this com-pound Figure 4! shows an intense ion at m/e 171 with lessabundant iona atm/e 127,99,87, and 57. High-resolution massspectrometry established the elemental compositions of m/e171 and 127 see Figure 4!. The small fragment ion at m/e 45and the large neutral loss of 44 mass units �71 to 127! arecharacteristic of ethylene glycol compounds, The GC retentiontime suggested a rather high molecular weight despite theabsence of sny high mass fragments in the EI mode. MethaneCI gave no additional information on molecular weight; butisobutane CI showed an M + 1 ion at 447. An elementalcomposition for the molecular ion af CzsHssoz was hypothe-sized based on the rather saturated coinposition of m/e 171.Gas chromatography using a nitrogen-phosphorus detectordid not contradict this hypatheais. A search of the EPA TSCAlist �2! and Chemical Abstracts for industrial compoundscorresponding to this molecular composition indicated thattetraethyleneglycol di�-ethylhexanoate! a plasticizer pat-ented and produced by one of the companies along the river!was a possibility. In fact, river water concentrations for this

CIHB 0, 0 CIHBICI HB 15 C OCB5 OCBHI OCI HB OCI HB 0 C CH CB Hd 015dt's 55[121] m

0. 10

0058

2 0 BI24

316

dO 80 I DO 120 140 IdO 01 ~ 200

Figure s. Electron impact mass spectrum of tetraethyleneglycol dl�-ethylhexanoate!, compound 82E kxnsntsi composibonE of m! 0 f 27 and 1 7 1 established by hftffvresoiuaon massspectromesy; a values indicate error lin miffi455sss units! between measuredsnd calculated exact masses

50 50 100 110 120 130IIIVER IIIEE

Figure 8. River water concentrations of compound 67 as a functionof river mile for samples collected ln August 197ft summer! end March1977 winter!

CD I~rCHI OCHI Cdt DCHtCHI D 0 � CH2CHB

63 1 01 151 Z85

CHI� 0 � CHBCHI

51

CI C ~ I

60 50 I DO r20 140 I 50 150 tQQ 2 ZO I IQ 260 280 400Figure 5. Electron impact mass spectrum of 1-chloro2�-[2~pf',1',3',3'-tetramethyfbutyfphenoxy!ethoxy]ethoxy!ethene, compound 97,Relative intensities expended by factor of 4; Intensities of off4mttfe peaks are l35 sssfl!, 285 foots!. 287 �7%!

Volume 12, Number 10, October 1978

2ft0

compound were highest in that sample taken adjacent to thesuspected discharge site �7 !. Identification and approximateconcentrations of the compound were verified using the au-thentic commercial product. The methylheptanoate isomerand the triethylene glycol homolog compounds 83 and 84,respectively! were identified in a similar manner.

The other polyethylene glycol compounds no. 63-71! arealso influstrial chemicals which are specific to a single sourceand which are traceable to that source. Identification of1,2-bis�-chloroethoxy!ethane in the river water near Phila-delphia initiated a search for a possible source. According tothe 1974 U.S. Tariff Commission Report �1!, one of thecompanies in the area is the sole commercial producer of thiscompound and holds a patent for its production �3!. Simi-larily, 1-�-chloroethoxy!-2-phenoxyethane, bis�-chloro-ethyl! ether, and compounds 68, 69, and 70 are produced orpatented by the same company,

Identification of two other chloroethers no. 66 and 67! wasfacilitated by their spectral and structural similarity to theabove compounds. Figure 5 shows the FI mass spectrum ofcompound 67; the elemental composition of rn/e 285 obtainedfrom HRMS! is included. An electron impact fragmentationpattern of 63, 65, 107, 109, 151, and 153 is characteristic of amonochlorinated ion with 44 mass unit adducts. Previousidentifications of chlorinated ethylene glycols suggested thatthis should be a similar coinpound with m/e 63 due toCICHECHE, m/e 107 to CICHECHEOCHECHE, and m/e 151 toCICHECHEOCHECHEOCHECHz. Iona at 77, 91, and 135 arecharacteristic of Cz-phenolic compounds. A combination ofthese fragments accounts for the base peak at 285 see Figure5!. A mass chromatogram indicated a very weak inolecular ionat m/e 356 suggesting that a C3H « fragment should be addedto the 285 ion to give compound 67. The hypothesized struc-ture was synthesized by chlorinating the hydroxy compound no. 70! with PCls �3!. The GC retention time and massspectrum for the unknown compound were identical to the

synthetic compound. Compound 66 was similarly identi-fied.

We should point out that compounds 66 and 67 are not ar-tifacts formed by the chlorination of compounds 68 to 70during the course of sample analysis. Since these two groupsof compounds were separated during LC fractionation priorto GC analysis, there was no opportumty for their intercon-version.

Concentration profiles for compound 67 are given in Figure6. Although the relative effects of dispersion due to tidal flowand of downstream movement due to the net river fiow are notprecisely known, it is clear from these data that compound 67comes from a point source located near river mile 100. In fact,the company which produces the related alcohols no. 68-70!,the chlorinated ethylene glycol no. 64!, and the Cs-phenol no.38! discharges its effluent at river mile 104,

The presence of these compounds in the Delaware Rivermay have some health implications. If the discharge site atriver mile 104 is correct, then these compounds would enterthe river only six miles downstream from the inlet for Phila-delphia'5 drinking water. Tidal action is sufficient to carrythese chemicals upstream to the inlet and, in fact, the volatileethers, bis�-chloroethyl! ether, and 1,2-bis�-chloroethoxy!-ethane, have been found in the drinking water supply �4!.Health effects, notably the carcinogenic activity, of thesecompounds are not knovm. It should be stressed that thehigher molecular weight compounds no. 65-70! have not yetbeen detected in the drinking water, nor have their healtheffects been evaluated.

The chlorinated compounds no. 57-60!, bis chlorophenyl!ketone, bis chlorophenyl!methanol, chlorophenylphenyl-methanol, and 1,1-bis chlorophenyl!-2,2-dichlo-roethylene see Figure 2 for structures!, represent anotherimportant group traceable to a single industrial source. Al-

though none of these compounds is manufactured commer-cially, the insecticide I,l-bis p-chlorophenyl!-2,2,2-trichlo-roethanol is produced by the same company which manu-factures most of the ethylene glycol compounds:

We suggest that compounds 57 to 60 are either manufacturingby-products from the production of this insecticide or are itsenvironmental degradation products. Model experiments withthe commercial insecticide demonstrated that compounds57-60 were not formed during our analytical procedures.

Finally, a few miscellaneous compounds which were iden-tified in the Delaware River and which have not been previ-ously reported as water contaminants will be discussed: Trifluoromethyl! chloroaniline and trifluoromethyl! chlo-ronitrobenzene no. 55 and 56! were identified in the water;they had maximum concentrations at, river mile 78. Bothcompounds represent common substructures in various pes-ticide and dye molecules, and several of the companies locatedalong the river have patents using these compounds �5 � 27!.It is possible that these trifluoromethyl!chloro compoundsare actually present in the river water as such, but it is alsopossible that they are formed in the GC injection port by py-rolytic degradation of larger pesticide or dye molecules, Allthree binaphthyl sulfone isomers no. 92! were identified inthe river water near Philadelphia, Product literature for oneof the companies in the area indicates production of con-densed sulfonated polymers derived from naphthalene sul-fonic acid and maleic anhydride. It seems likely that the bi-naphthyl sulfones could be formed as by-products duringpreparation of this commercial product.

River water samples were collected both in August and earlyMarch; this allowed us to compare results fiom two samplingseasons. Generally, the two data sets were qualitatively similarsuggesting that pollution sources remained stable over the testperiod; however, two major changes were observed. First,winter samples contamed high levels of volatile organics�0 � 20 ppb! which were not detected in the summer water.Most likely high water temperature �5 � 27 'C! and turbulentriver flow volatilized organics from the river during the sum-mer months.

The second change was the three- to fourfold increase in thelevel of almost all orgarncs in the winter samples, This ob-servation was corroborated by weekly data on nonspecificorganic levels COD, TOC, etc,! collected by another labora-tory during the same sampling period 9!. Winter sampleswere collected during a period of high stormwater runoff andwere very turbid in nature. It is possible that high levels ofparticulate matter were responsible for increased organicconcentrations in the water column. These particles couldhave been sedimentary organic compounds, or they could haveprovided favorable adsorption sites within the water columnfor dissolved organics. As an alternative, municipal and in-dustrial waste treatment systems may have been adverselyaffected by the coM winter temperatures, resulting in signif-icantly higher organic loads entering the river system.

ConclusionsThe organic compounds in any river system will be a com-

plex mixture of natural products, municipal wastes, and in-dustrial contaminants with the predominance of any typedependent on river hydrology, discharge sources, and generalriver conditions. In a large river, flow volumes are usually or-

dere of magnitude greater than incoming discharge streams;thus, the concentration of most industrial chemicals will bein the sub-ppb range, The analysis of these low levels of ex-tremely complex mixtures is not an easy task. The analysis oforganics in an industrial wastewater, for example, is usuallymuch simpler.

We have noticed the predominance of ethylene glycol de-rivatives in the Delaware River. Compounds 63-71 and 82-84are all based on ethylene glycol and are among the mostabundant anthropogenic compounds in the river, Since fewmass spectra of these compounds are in reference collections,their proper identification is frequently difficult and over-looked. In most cases the presence of an ion at rn/e 45 togetherwith abundant neutral losses of 44 amu should indicate to themass spectral interpreter that an ethylene glycol derivativemay be a good structural hypothesis.

Acknow iedgmentsThe cooperation of the Delaware River Basin Commission

and the help of the Water Department of the City of Phila-delphia are gratefully acknowledged.

Literature Cited�! "National Water Quality Inventory: 1974 Report to Congress",

EPA 440/9-74-0DI, Val 1, Office of Water Planning snd Standards,Washington, D.C., 1974.

�! Dowty, B. J., Carliele, D, R., Laeeter, J. L., Environ, Sci. Tcchnoi.,9, 762 �975!.

�! Coleman, W. E., Lingg, R. D., Me ton, R, G., Kapfler, F. C,, in"Identification and Analysis of Organic Pollutants in Water", L.H. Keith, Ed., pp 3D5-27, Anu Arbor Science, Ann Arbor, Mich.,1976.

�! Keith, L, H., Garrison, A, W., Allen, F. R., Carter, M. H., Floyd,T. L�Pape, J. D., Thureton, Jr�A. D., ibid., pp 329-73.

�! Jungclaue, G. A., Games, L. M., Hites, R. A�Anat, Chem., 48, 1894�976!.

�! Games, L. 54., Hitce, R. A., i bid,, 49, 1433 �977!.�! Jungclaue, L. A., Lopez-Avila, V., Hitee, R. A., Environ, Sci.

Techno! 12 88 �978!, 8! "Industrial Diechacge Inventory", Delaware River Basin Com-

mission, Trenton, N.J., 1975. 9! City of Philadelphia Water Department, private communication,

1977.�0! Hoover, R� IVIaeon, T. J., McKay, F. W., Fraumeni, Jr., J. F.,

Science, 189, 1005 �975!.�1! Hoover, R., Fraumeui, Jr�J. F., Environ. Ree., 9, 196 �975!.�2! Harleman, D.R.F�Lee, C. H., Tech. Bull. No, 16, Committee of

Tidal Hydrology, Corps af Engineers, U,S, Army, 1969.�3! Harleman, DRF., Ippen, A. J�Proc. ASCE, 95,9 �969!,�4! Ikan, R., Baedecker, M, J�Kaplan, I. R., Nature, 244, 154

�973!.�5! Kami, T�J: Agric. Food Chem., 23, 795 �975!.�6! Hitee, R. A., J. Org. Chem�39, 2634 �974!.�7! Garrison, A. W., Pope, J, D�Allen, F. R., in "Identification and

Analysis of Organic Pollutants in Water", L, H. Kcith, Ed., pp517-56, Ann Arbor Science, Aun Arbor, Mich., 1976.

�8! Shacke ford, W. M., Keith, L, H., "Frequency of Organic Com-pounds in Water", EPA-600/4-76-062, Nat. Tech, InformationService, Springfield, Va. 1976.

�9! Grab, K�J. Chromatvgr., 84, 255 �973!.�0! Grab, K., Grab, G., ibid., 90, 303 �974!.�1! U.S. Tariff Commission, "Synthetic Organic Chemicals: United

States Production and Sales, 1974", T.C. Pub, No. 614, U.S.Printing Office, Washington, D.C., 1973.

�2! "Toxic Substance Control Act, PL 94-469, Candidate List ofChemical Substances", U,S. Environmental Protection Agency,Office of Toxic Substances, Washington, D.C., 1977.

�3! A bright, R, L., McKeever, C. H., U.S, Patent 3294847 �966!.�4! Suffet, I. H., Radziul, J. V,, J, Am, Water Works Assoc., 68, 520

�976!.�5! Theiesen, R. J., German Patent 2261918 �973!.�6! Gerhard, W., German Patent 2203460 �973!.�7! Stoffel, P, J., U.S. Patent 3746762 �973!.

Received for revieic December 22, 1977. Accepted May 15, 1978.Project supported by the Research Applied ta Ivatianal Weedsprogram vf the National Science Foundation grant number Eiv V-75-13069!.

Envkonmental Science 8 Technology 24!

Sources and Movement of OrganicChemicals in the Delaware River

L. S. Sheldon and R. A. Hites

Reprinted with permission from Enofronmental Scienceand Technology, 1S�!:d74-79 May 1979!. Copyright

1979, American Chemical Society.

Sources and Movement of Organic Chemicals in the Qelaware River

Linda S. Sheldon and Ronald A. HlteaDepartment of Chemicai Engineering, Massachusetts Institute of Technology, Cambridge, Mass. 02139

Experimental

g45Environrnentsi Science tt Technology

8 The transport of industrial organic chemicals from theirsource, into the Delaware River, through various treatmentfacilities, and into Philadelphia's finished drinking water wasstudied using water samples collected in August 1977. Solventextraction, liquid chromatographic cleanup, and gas chro-matographic mass spectrometry were used for compoundseparation and identification. Results confirmed dischargesources for many previously identified coinpounds, Further-more, it was shown that many of these compounds circulatedinto Philadelphia's drinking water, and that the various waterand waste treatment facilities had s minimal effect on theorganic levels. For all chemicals, dilution processes were re-sponsible for the largest reduction in organic concentrations.Results were substantiated by a 10-week sampling programdesigned to monitor seven selected waste chemicals.

Nearly 100 organic compounds of biological, municipal, andindustrial origin have been identified in the Delaware River I !, Among the industrial contaminants, several compoundsseemed to be coming from a specific plant in the Philadelphiaarea, Furthermore, relatively high levels of anthropogenicchemicals were observed in the river near the Philadelphiaarea I !, indicating that they may be entering the city' sdrinking water. We have, therefore, traced the movement ofvarious industrial chemicals from their origin, through theriver, and into Philadelphia's drinking water. We have alsoconducted a 10-week, continuous sampling program to mon-itor seven selected compounds in the aquatic system. Thispaper is a report on these studies.

The Sampling Area. Only a small segment of the DelawareRiver, lying just north of Philadelphia, was studied. A sche-matic diagram of the complete sampling area is shown inFigure 1. General flow and hydraulic characteristics of theriver have been discussed previously I !. The box in the upperleft-hand corner of Figure 1 represents a plant in the Phila-delphia area which we will refer to as plant A, This plant doesnot discharge its wastewater directly into the river, but ratherinto the city sewer along with several other industrial users.These combined industrial wastes are treated at the City ofPhiladelphia's Northeast Sewage Treatment plant usingclassical secondary treatment methods �!. The treated ef-fluent is then discharged into the Delaware River at river mile104.

Water flow in this segment of the river is dominated by tidalmovement rather than by dovmstream river flow; tidal vol-umes are an order of magnitude greater than downstream riverflows. During periods of normal flow, effluents discharged intothe river travel approximately 7 miles upstream during hightide �!. Under these conditions, water flow in the upstreamdirection is sufficient to transport industrial chemicals fromthe sewer outfall upstream to the intake pipes of Philadel-phia's Torresdale drinking water facility at river mile 110 �!,Intake valves for this plant are open only during high tide,making industrial waste contamination of the city's drinkingwater not only possible but probable �!. Water entering thedrinking water plant is treated using standard techniques �,5!: prechlorination; settling; coagulation ferric chloride, alum,and lime!; disinfection; flocculation; and filtration rapid sandfilters!, After a final chlorination step, drinking water is dis-tributed throughout the city. Water from this treatment fa-cility provides the city of Philadelphia with approximately

5096 of its finished drinking water �!. All present drinliingwater standards are being met at this water treatment plant�!.

Samples were collected in late August 1977 from sites a toh, as shown in Figure 1, Our purpose wss to follow a 24-h slugof industrial wastes through the cycle from plant A to thefinished drinking water, The sampling scheme was designedto account for retention times between the various samplinglocations, as well as for tidal movement in the river �, 4!.Details of this sampling regime are out1ined in Table I.

The composite sample from plant A was taken from a 5-galcontinuous sampler after the 24-h sampling period. All othersamples were composites of individual grab samples collectedat a particular location. River water samples were collectedapproximately 100 yards from the western shore at the des-ignated river mile and at a depth of about 0,5 m,

Another set of samples was collected weekly over the 10-week period extending from January 15 to March 28, 1978,from points c, g, and h see Figure 1! and from the centerchannel of the Delaware at river mile 98. Samples from sitesc, g, and h were composites of 200-mL grab samples collectedevery 8 h beginning Tuesday 8 a.m., Tuesday 8 p.m,, andWednesday 8 am�respectively. River samples were 1-gal grabsamples taken every Wednesday morning.

All samples were collected in glass bottles fitted with Tef-lon-lined screw caps, Methylene chloride and hydrochloricacid were added to the water samples at the collection site inorder to minimize biological degradation and to start the ex-traction process. Since waste effluents from plant A do notsupport microbial activity �!, sample preservation in the 24-hcontinuous sampler was not needed,

All samples were stored in the dark, Small samples werekept on ice during transport to the laboratory. Larger sampleswere refrigerated as soon as possible after collection,

Analytical techniques and instrumentation for the con-centration, separation, and identification of sample compo-nents have been discussed in detail elsewhere I !, In general,analytical techniques used in this study included solvent ex-traction, liquid chromatographic fractionation, high-resolu-tion gas chromatography, computerized gas chromatographicmass spectrometry GC � MS! in both the electron impact EI!and chemical ionization CI! modes, mass spectrometric se-lected ion monitoring SIM!, and high-resolution mass spec-trometry HRMS!.

For the initial phase of this study August, 1977!, concen-tration values were semiquantitative and were based onstandard curves for selected compounds. Estimated errors inquantitation are approximately +2096 in plant A's waste ef-fluent, +5096 in the Northeast influent and effluent and theriver water, and an order of magnitude in the finished drinkingwater.

During the second phase of this study January-March,1978!, experimental procedures were developed to moreprecisely quantitate seven previously identified compounds.Concentration values were measured using selected ionmonitoring SIM! performed on the unfractionated, combinedneutral and acidic extracts for each sample. Sample concen-trations were calculated by comparing the computer-inte-grated peak areas of selected masses with those obtained fromstandard solutions containing the seven compounds.

Voluntes, Types,

vol, L0.50,51

23232344

8/23 12 p.m. to 8/24 12 p.m8/24 2 a.m. to 8/25 2 a.m.8/24 8 a,m. to 8/25 8 s.m.8/25 10 a.m.8/25 10:30 a,m.8/25 11 s.m.8/25 8 s.m. to 8/26 8 a.m.8/25 8 p.m, to 8/26 8 p.m.

1977.

continuousgrabgrabgrabgrabgrabgrabgrab

�! plant A effluent b! Nartheazt influent c! Northeast effluent d! river mlle 106 e! river mlle 108 f! river mile t18 g! Torresdakl Influent h! Torresdale effluent' All samples taken In August

7 R ~ OH, n ~ IS. R OH nvz9. R OH, n ~ 310. R ~ OH, n 4

R OH, n ~ 512. R ~ C I, n ~ 213. R ~ C I, n 3

nvw nnl ~110

nvnv mnaI O4

151 !

CH3C � NH � C � C ~CH

0 CH3

IS,

CI

CH3

CH

CH3

CI19. C � N

0

0II

oc � C=OHI 2CH3

0CH2= C � C OCH2 CHsin

CH3

Volume 13, Number 5, Msy 1978246

Table I. Satnp lng Scheme Glv}ng Os alla of Tltnlng,

~ - NEr FLOW OELAW4RE RlvER NET FLOWFigure 1. The sampling area, showing collection sites. R!ver mlleagesare rrwasured upstream from the mouth; net flow proceeds from rightto left

Solvent extraction efficiencies were measured for theseseven campounds. Preextracted water samples were spikedwith a known aliquot of a standard solution. Spiked sampleswere extracted and quantitated using the above procedures.Tests were run in triplicate using water samples from all foursampling locations. Recoveries were better than 7593 in allcases. Reported concentration values were corrected for sol-vent extraction efficiencies and have errors af less than d:20th,excluding sampling errors.

Resu!ts and DiscussionAll of the compounds identified in the industrial waste-

water, the municipal sewage effluent, the river water, andPhiladelphia's finished drinking water are listed in Table II.Some structures are given in Figure 2. Estimated concentra-tions have been included for most of the abundant com-pounds, The compounds in Table II are listed according tolocation of first appearance. Within each of these groups,chemicals have been subdivided by compound type. This ar-rangement allows for both a quick identification of specificpollution sources and for a facile appraisal of the movementof these chemicals in the aquatic system.

For an overview of the occurrence and environmental sig-nificance of many of the compounds listed in Table II, thereader is referred ta our previous paper on the Delaware River�!. During the following discussion, only those compoundswhich were nat previously identified in the Delaware Riveror which gave some insight into the movement of chemicalsthrough the various treatinent processes and in the DelawareRiver will be considered.

Identification of Contamination Sources. The firstob-jective of this study wss ta verify that plant A was the specificsource for a set of previously identified compounds. Thesecompounds included 1,2-bis chloroethoxy!ethane �!, thephenyl glycols � � 11!, the chlorinated phenyl glycols �2 and13!, DDE �7!, dichlorobenzophenone �6!, and the binaph-

and I.ocaflons See Figure 1!oovnssns llo orhzonrol. h ~ onlploo

14 74 7

111

4 74 7

CH3 CH3I I

R CH CH 0 I2 2 n i / 2 3C � CH � C � CH

CH3 CH3

CI2 CH3

20. C � N CH

0 CH

64. n 265, n ~ 366 n ~ 4

Ffgure 2, Structures of selected organic compounds found ln the Del-aware River see Table II!

thyl sulfones �7!.Our data see Table II! verify that these chemicals are, in

fact, being discharged from plant A along with various otherphenalic compounds �-5!, chlorinated compounds �8-20!,and esterified species �5 and 26!. All of the above compoundsare either commercial products manufactured at plant A arare process byproducts.

The commercial herbicide �! 2,5-dichloro-N- l,l-di-methyl-2-propynyl!benzamide �8! was discharged in plantA'2 waste effluent in relatively high concentration �00 ppb!.We should point out that this compound was not detectedduring our earlier work �!, but plant A operates in a batchmode �! and does not consistently discharge the same mix

i. plant AA. phenols

1. phenol2. cresol8. cctylphenolsv~ nonylphenolso5. ~1-2,Mi-ter-butylphenol

B. ethylene glyCOI derhratlveaB. 1.2+is�Mloroe&oxy!ethane7. 2- p-I',I',3',3'-Ietramethylbutyiphenoxy!ethsnolm8. 2-[2'-I', I',3'.3'-letramethylbvtylphsnoxy!ethoxy]ethanolm9. 2/2-[2- p-I',I',3',3'-tetramedtylbulylphenoxy!ethcxy]-sthoxy!ethanol

10. 2-[2�-[2- p-I', I',3',3'-Istramethylbutylpfwnoxy!ethoxy]-elhoxy!ethoxy]ethanol

11. 2-�-[2-�-[2- p-I', I',3',3'-tetramethylbulyiphenoxy!ethoxy]-elhoxy!ethoxy]ethoxy!ethanol m

12. 1-chloro-2-[2- p-1',1',3',3'-tetranwthylbutylphenoxy!-e8wxy]ethane m

I-chi ore-2-�-[2 p-1',1',3',3'-tstrttmethyibutylphenoxy!-ethoxy]ethoxy!ethane

C. chlorinated compounds1 ~, tetrachlorostyrenes'15. hexachfcroedtylbenzetwo18. dfchhrobenzophenorwss17. 1,1+is chlorophenyl!-2,2-dichloroethyfetw DDE!18. 2,5-dfchloro-hHI,Isflmethy!-2~yl!benzamlde~18, chloro-hHI, I<llsopropyl!benzsmides ~20. dlchlcro-hf�,1-dllsopropyl!benzamkfeo nr21. dlchlorobenzeneso22. chlorotolueneo28. Irlchlombenzensso24. Ietrachlorobenzenes'

D. plssdctzers25. bls�-ethylhexyl! adlpste28. dkwtyl sebacate27. Iris wrt4utyf! phosphate

E. hydrocarbons28. Cq benzenes'29. Co berlzsnsso80. naphthalene$1. methylnsphthalenes82. Cr naphthalenes'$8. Co naphthalenes'$4. C4 benzenes'$5. CI+zoI$6. CIo!too

F. others8'7. binaphthyl svlfones'88. isophcrons

II. Northeast influentA. phenofs

88. pfwnylphenol~ 0. cvmylphenol41. tsrt-butylmsthoxy phenol

C, chlorinated compovnds42. dichiorophenols'48. Irich crcphsnofso44, bls chkxophenyl!methanol o

D. phsticlzers45. triphenyl phosphate48. dSxttyf phthahte47. butylbenzyl phthalste48. bis�~lhexyf! phthalate

o0.3 0.3 to2020

20040un

700050

5000800200

60uA400Wlun

2 040.02 0.02

0.01

I0.02 50.02 50.002 t

100200100400

un 0.80.4

I t

0.30.2 vn

50uA

10

un

NAr NA200 VA Vn Wl

NA NA NA NA NA NA

0.4 0.3 0.2

Vn

2000 200 80 0.6

0.20.5 0.31500 120 50 0.6

0.5 0,08400vn

1000180050050un

100un

200200

BOvn

11020040

t

100t

20I

20

603020

t t t 10t

1 020,4 0.30.4 0.2t t0.2 0.08

v" vv vt tt t

0.1 0.1

0.02 0.01

0.04 0.02

0.2 0.04 0.02 0.002200020050

10

un 0.5 0.4 0.3 0.4 0.8

1000

500t

2002000

1004020

t t2001030

v v2 08t t0.2 0.02I II I2 I

v1040.4

II40UA10

0.6un uA 3 06uA

'IOOun10

un un un0.3 0.01

t t t

UAtt

0.01t t I

04 040.1 0.15 0.7 0.2

t Wlt un I

0.1 0.002

unt3

2 0.3 0.225 0.6 0.4

100 O.B 0.3100 1 I

0.2 0.03 I0.1 0.1 0.30.3 0.1 0.40.5 0,8 0.5

165040

200

247Environmental Science 6 Technology

Table II. Cofnpounda and Their ConcentraSona ppb! abaerved al the Var oua Safagl lnS Sltea See Fiynre 1!Ixont Na IR AM IMI Tor Tor RMA In oot 'Iee 1IN In oak 11 ~

Table II. Continuedvar stw rwh eel 11S

W W nh tulh ha 1es 1es

t t t tt t t tt t t tt t t tt t t It I t I

t tt � tI tt � t

O,S � 0.91 � 0.9

2 05h

h h0.4 0.21 10.6 0.3

m0.020.002t

t 30,12

0.03

0.2 0.20.2 0.30.5 0.3

2 70.04 0.02I It t UA

produced commercially using the reaction schema outlinedin Figure 3 8!. DDE is the unrsacted starting material; te-trachlorostyrene and hezachlorosthylbenzene are probablycleavage byproducts formed during the initial chlorinationstep or from the reaction intermediate l,l-bia p-chlorophe-nyl! tatrachloroethane. Dichlorobenzophenone could formduring alkaline hydrolysis of either the tosylate ester inter-mediate or the pesticide itself. Two other structuraHy relatedcompounds, bis chlorophenyl!methanol �4! and chloro-

of waste chemica!s. Concentrations of compound 18 around0.003 ppb were found in drinking water samples during our10-week Iluantitation study see below!.

An interesting case is presented by several of the mul-tichlorinated aromatic compounds �4 � 17!: tetrachlorosty-rene, hezachloroethylbsnzene, DDE, and dichlorobenzo-phenone. None of these compounds are produced commer-cially; however, plant A did manufacture the pesticide 1,1-bis p-chlorophenyl!-2,2,2-trichloroethanol. This pesticide was

2!I B Volume 13, Number 5. May 1979

E. hy*ocarbons4$, pyrene t t60. fluorardhsne t - t51. anthracene52, phenanthrene5$. methylphensnthreneSL chrysene

F. others55. cholesterol 400 200 2 ISL choiestsnol 800 300 $257. a-terplneol 80 $05S. 2-phenyl-2-propanol 70 7059. stearlc acid h - m"$0. palmltlc acid h81. benzll t Un Ull82. horny l acetate 100 50 0.18$, AA n+Utyl!benzenesulfonamlde Un UA t

lll. Northeast effluentB. ethylene glycol derivatives

64, diethyleneglycol dfmethecryfate w 10 02 ISS. trlethyleneglycol dl~latem 35 05 01 5SL tetraethyleneglycal dictate 700 10 3 0.5 0.02$7. chlorophenyfphenylmethanola 0.1 0.1 0.1 Un

F. others88. menthol 8 Un

IV. riverB. ethylene glycol derivatives

89. bis�- [2- nkutoxy!ethoxy! ethoxy~ un 1 270. trlethyleneglycol bls�-ethylhexanoste! 0.2 0.171. tetraethyleneglycol bls�~lhexanoate! 2 3 I

C. chlorinated compounds72. dimethyl 2,3.5,6-tetrachlor~blte UA Un lal

D. plastlclzers7$. 2,2,4-trlmethylpentane-1,3-dlol-1-isobutyrate t t74. 2,2,4-trimethylpentane-1.$dlo~sobutyrate t t7S. 2,2,4-trimethylpentane-1,3-dloldihobtdyrate I t

F. others78. chlorophyti' 4 � 977. fluorenone Un Un I78. ethylthlopyridinea I t t7$. I, 1-bls chlorophenyi!-2,2,2-lrlchloroethane DDT! t

V. drlnfdng waterC. halogenaied compounds

80. dlchlorolsopropenyltoluenee UA81. bromochlorophenole Un$2. dibromophenola lal8$. dlchlorobromophenole Un84, dibromochhxaphenote Un

' t Irxsestee that only trace leven were dele'. ' - bxsefsm col dehacM. ' Un sxseaw swt campaond was net resolved gss efvomahgraphkally snd. swrefom,was col qcanNated, r Tfw predominant species was p-I, I,s.a-tebametwlbutrlphsnaf, ~ other Isomers were present. ~ Several laomen present, ' NAIndIcates that analysis for Uvres ~ was nol carried out inot aoalrzedl. e learner Unknown. "v indicates votadie campound; these ~ waukl notbe retained in the water column der hg ew aommer manta. ' Qruebee vnxnaem, mal wt born CI, present h hexane fracgon, I h indicates very Ngh ~hns;these eompovnds give broad unreceived peaks whkh could not be quantthted. "m Indicates moderate ~lan. ' Chlorophyll wae observed as a eerier ofphrtadlenes see ref fl. See Flgore 2 for Ihe saoebaw of this compmaxl.

"U- M"Ar sos HileSO,

"C l-C3"0CI E

H 0

Figure 3. Reaction pathway for the commercial prOductlon of 1,1-bis rNchtorophenyl!-2,2,2-trlchloroethanoi see ref ty!

phenylphenylmethanol �7!, first appear in the Northeastinfluent and effluent water, respectively. We think that theseare probably degradation products of one of the above chlo-rinated species. We should point out that the pesticide itselfwss not detected in any of the wastewater or river watersamples.

Although some of the methyl substituted compounds�8 � 34! and chlorinated aromatics �1-24! and the solventisophorone �8! first appear in plant A's waste effluent, theyare common industrial chemicals which could also be enteringthe water system at various other points. This was confu medby comparing concentration data for these compounds withthe same data for the compounds specific to plant A. Theformer compounds show much smaller changes in concen-tration between sampling locations, suggesting multiple dis-charge sources.

Most of the compounds which appear for the first time inthe Northeast treatment plant's influent �9-63! are commonindustrial or municipal contaminants. They are not unusualand have been discussed in detail elsewhere �, 9-11!. X- n-Butyl!benzenesulfonamide �3! is interesting because ithas never been identified in environmental samples. Its majorcommercial use is as a plasticizer for polyamide materials�2-1 4!. It has also been patented as a star ting material in theproduction of sulfonyl carbamate herbicides �5!. The exactsource of this contaminate is not yet known.

Those compounds originally appearing in the treatmentplant's effluent water �4-68! were, of course, not present inthe influent water; this suggests that they were formed duringthe treatment process. The most striking example is thepoly ethylene glycol! derivative, tetraethyleneglycol di-methacrylate �6!. This particular chemical is commonly usedas a copolymer in many synthetic materials �6-18!, It seemspossible that a polymer entering the Northeast treatmentplant is being degraded to monomer units during treatment,or that residual monomer is being washed off polymers duringtreatment. This compound was the most abundant chemicaldischarged in the Northeast treatment plant effluent; thisleads to correspondingly high river water values. The di- andtriethyleneglycol homologues �4 and 65! were also identi-

10,000

cl CNA100

fe 210

e0.4

0.1.02

PA NE NE rm rm orr orrin out 106 IOS in out

Figure 4. Concentration levels of 2,64ichtoro-+1, 1 dimethyl-2-pro-pynyl!benzamlde �9! throughout the sampling system

t0,000

1000I I 0100 SO

u IO

Ol P A NE NE rm rm Torr TorrOut in Out IOS IOS in Out

Figure 6. Concentration levels of dlchorobenzophencne �6! ~the sampling system

fied.Compounds first appearing in the river water �9-79! may

be categorized into three groups according to source; first,those entering the river system from other industrial dis-charges such as various ethylene glycol derivatives �9-71! andvarious plasticizers; second, those compounds formed by thenatural biological activity in the river, for example, chlorophy!I�6!; lastly, compounds which enter the river via rainwaterrunoff, most notably the herbicide dimethyl 2,3,5,6-tetra-chloroterphthalate �2! �9!,

In the finished drinking water a series of halogenatedcompounds appears which were previously undetected. Itseems logical that these compounds, especially the halogen-ated phenols 81-84!, are formed during the chlorinationprocess �0!,

Movement of Compounds through the System. It iseasiest to assess concentration changes as various compoundstravel from industrial wastewater to finished drinking waterif the data are presented graphically. Figures 4 to 7 are a seriesof bar graphs showing concentration data for several com-pounds at each of the seven sampling locations. These par-ticular compounds were chosen because: a! they are uniquechemica!s entering from a single, well-defmed source, and b!they complete the sample loop and were found at all samplelocations. This second characteristic makes it possible to as-sess the effects of all treatment processes and of dilutionduring upstream river movement.

Figures 4 to 7 indicate several trends. Large changes inconcentration approximately four orders of magnitude! wereobserved between plant A's effluent and the finished drinkingwater, Obviously, this large decrease in organic concentrationis important when considering allowable discharge levels andtreatment processes, For all four compounds, a definite con-centration pattern developed over the sample system. Thegreatest concentration decreases occurred between plant A' seffluent and the Northeast Treatment plant's influent sitesa to b! and between the Northeast Treatment plant's effluentand the first upstream river sampling location sites c to d!,It is interesting that these large decreases in conoentration are

Environmental Science 6 Techtology 249

Table III. Median Cosiceatratlona ' and Relative Concentrattorla for the 1 0-Weett Study January to March, 1 878!and Grab Sample Concentrations {Auttuat 1077!

se-week eeeee eor'~ Il rover o

Terrriver o k»

0.10.40.60.20.20.10.08

~ a skier e ke ea200 3 0.4 0.0110 un un 0.02un 0.8 un 0.0280 0.8 0.3 0.250 0.6 0.3 0.230 0.4 un20 0.4 0.02 0.01

0.2 100 4 0.20,03 100 4 0. 40.06 100 5 0.50,04 100 2 0. 10.04 100 3 0.30.02 100 2 0.20.003 100 5 0.08

8 0.40,3 0.030.5 0,050.3 0.020,5 0.070.3 0,040.2 0.003

3 7 612131718

2008

'lo2020204

' Tres range of the individual meeouremsnls ls usually a faotar ol 3 above sod below Ne nwdlsa; for exemple, for 6 medkm of 20 ppb, ere range ls 7 ta Bg ppb.o See Table II. u River mlle 98, u River mlle 108.

la,000 10,000

lava

m 100u n 10u

� 100

! iaeu u u u uV

OlOi

Ol P4 NE NE rm rm Torr Torroui n oui l06 iae rn our

.Ol P 4 NE NE rm rm Terr Torrn Oui tas lae in OurFigure 7. COnCenirallOn leVelS Of 1-Ch arc-2/2-[2+1',1'.3',3r-tetramethylbuty!phenoxy!ethoxy]ethoxy!elhane �3! throughout iheSampling System

Figure 6. Concentration levels of 1-chloro-2-[2+ '.1',3',3'-tetsa-methy!buty!phenoxy!elhoxy]elhane �2! throughout the samplingSySlern

AcknowledgmentsThe cooperation of the Delaware River Basin Commission

and the help of the Water Department of the City of Phila-delphia are gratefully acknowledged,

Literature Cited

Received for review July 14, 1978. Accepted December 22, 1978, Thisproj eci ivas supported by the Chemical Thr eats 1 0 hfan program ofthe National Science Foundation Gnznt fifo. Efi/V-75-13069!.

volume 13, Number 5, May 1979250

caused solely by dilution. In the first case, plant A's effluentwas diluted with other industrial wastewaters; in the secondcase, the municipal waste effluent wss diluted with river water.In the two areas where treatment was performed, namelybetween the Northeast Treatment plant'8 influent and ef-fluent sites b to c! and between the Torresdale drinking waterplant's influent and effluent sites g to h!, only small con-centration decreases occurred. For these four compounds, atleast in this system, dilution is the most effective treatmentprocess.

On the other hand, the data in Table II show that there areseveral compounds where treatment processes, especially atthe Torresdale plant, are effective. These include the hydro-carbons, sterols, palmitic and stearic acids, some of the eth-ylene glycol compounds, the phenols, and chlorophyll. Un-fortunately, it appears that the compounds of greatest envi-ronmental significance may be the least affected by the wastetreatment processes.

Table III presents median concentration values for sevencompounds which were found in plant A's effluent. These datawere collected over a 10-week period using selected ion mon-itoring GC/MS techniques. Table III also lists the concen-tration data for the summer grab samples see Table II! andrelative concentratian values for the 10-week study. A com-parison between the concentration values for the 10-weekstudy and the summer grab measurements shows goodagreement within the estimated error! for the two data sets.Data on the relative cancentration levels in the 10-weeksamples deinonstrate again that dilution is the most importanttreatment process for reducing industrial waste levels in thisaquatic system, No more than a 5098 reduction in concentra-tion is achieved at the Torresdale water treatment plant forany of the compounds. For the phenyl glycols � and 8!, achlorinated phenyl glycol �2!, and 2,5-dichloro-Af-�,1-dimethyl-2-propynyl!benzamide �8!, the treatment processappeared to have no effect at all.

�! Sheldan, L. S., Hites, R. A�Environ. Sci. Technol., 12, 1188�978!.

�! Suffet, 1. H., Brenner, L., Cayle, J. T., Cairo, P. R., Envi son, Sci.Techno l., 12, 1315 �976!.

�! Hsriemsn, D. R. F., Lee, C. H., Technical Bulletin NO. 18, Com-mittee of Tidal Hydrology, Carps of Engineers, U.S. Army, 1969.

�! Maaweriag, J. F., BisnkebshiP, W. Mu Miller, L�Vaigt, F., J. Am.Water Works Assoc., 69, 210 �977!.

�! Rsdziui, J. V., Water Sewage Works, 185,76 �977!.�! Csrio, P, R,, City of Philadelphia Water Department, personal

communication, 1977.�! Horram, B. W., U.S. Patent 3 534 098 �970!. 8! Wilson, H. F., U,S, Patent 2 812 362 �957!. 9! Shsckelfasd, W. M., Keith, L. H., "Frequency of Organic Com-

pounds in Water", EPA-600-4-76-062, National Technical Infor-mation Service, Springfield, Va�1976.

�0! Grab, K., J. Chromalvgrm 84, 255 �973!.�1! Grab, K., Grab, G., J. Chramatogr., 90, 303 �974!.�2! Herdwicke, J. E., French Patent 2 212 382 �974!,�3! Takeuchi, T., Suzuki, H., Japanese Patent 7 434 947 �972!.�4! Rogues, G., Vo Dinh hfan, French Patent 2 208 010 �974!.�5! Stephens, J. A., U.S. Patent 3 933 894 �976!.�6! Miller, L S., U.S. Patent 3 616 028 �971!.�7! Kosugi, F., Motoki, T., German Patent 2 263 193 �973!.�8! Bsssett, D. R., German Patent 2 331 141 �974!.�9! Windbalz, M., Ed., "The Mesck Index", 9th ed., Merck a Ca.,

Rsbway, N.J., 1976, p 371.�0! Morris, J. C., McKsy, G., "Formation of Helagenated Orgsnics

by Chlorination of Water Supplies", EPA-600-1-75-002, NationalTechnical Information Service, Springfield, Vs., 1976; Chem,Abstru 85, 182017w �976!.

Chloro-organic Compounds in the LowerFox River, Wisconsin

P. H. Peterman, J. J. Delfino, D. J. Dube, T. A. Gibsonand. F. J. Priznar

Reprinted with permission from Hydrocarbons and Halo-genated Hydrocarbons kn the Aquatic Enoironment B. K.

Afghan and D. Mackay, eds.!, pp. 145-80. New Yorh:Plenum Publishing Corp., 1980.

251

CHLORO-ORGANIC COMPOUNDS IN THE LOWER FOX RIVER, WISCONSIN

P.H. Peterman, J.J. Delfino, D.J. Dube, T.A. Gibson

and F.J. Priznar

Laboratory of Hygiene, University of Wisconsin-Madison,

465 Henry Mall, Madison, WI 53706

ABSTRACT

The Lower Fox River, Wisconsin is one of the most denselydeveloped industrial river basins in the world. During 1976-77about 250 samples were analyzed by GC and GC/MS including biota,sediments, river water and wastewaters from 15 pulp and/or papermills and 12 sewage treatment plants. A total of 105 compounds wereidentified in selected extracts by GC/MS with another 20 compoundscharacterized but not conclusively identified. Twenty of the 105compounds are on the EPA Priority Pollutant List. Other compoundsidentified in pulp and paper mill wastewaters, including chloro-guaiacols, chlorophenols, resin acids and chloro-resin acids havebeen reported toxic to fish by other investigators. Several com-pounds apparently not previously reported in wastewaters arechloro-syringaldehyde, chloroindole, trichlorodimethoxyphenol, andvarious 1-4 chlorinated isomers of bisphenol A. Concentrationsof the various compounds, when present in final effluents, rangedfrom 0.5 to ca. 100 p g/L. An exception was dehydroabietic acid,a toxic resin acid not found on the EPA Priority Pollutant List.It was frequently found in pulp and paper mill effluents in concen-trations ranging from 100 to 8500 pg/L. PCBs were found in allof the matrices sampled. Sixteen of the 35 fish exceeded the FDAlimit of 5 mg/kg while 31 of the 35 exceeded the Canadian limitof 2 mg/kg. Concentrations of PCBs and other chloro-organicswere related to point source discharges. There was a direct cor-relation of the concentrations of these compounds in wastewaterwith suspended solids values.

253

P. H. PETERMAN ET AL.

ZNTRODUCTZON

Concern over the sources, distribution and fate of organiccompounds in natural waters has increased considerably in recentyears. With the development of GC/MS/DS instrumentation, thousandsof compounds have been identified in industrial and municipal waste-waters, receiving waters and biota Donaldson, 1977!. Also, dueto the extensive use of chlorination, numerous chlorinated organiccompounds are being formed and these are now a matter of interestto researchers and regulatory officials Jolley, 1976! .

Because of this interest, the U.S. Environmental ProtectionAgency. Great Lakes Program Office, Region V! contracted with theState of Wisconsin Department of Natural Resources and Laboratoryof Hygiene! to assess the sources and distribution of organiccompounds, particularly polychlorinated biphenyls PCBs! and otherchloro-organics, in the 64 km Lower Fox River in northeasternWisconsin Figure 1!. This river drains into Green Bay-LakeMichigan and i.s one of the most densely developed industrial riverbasins in the world. Pulp and paper mills predominate; many ofthese use extensive amounts of chlorine. Five of the paper mills

0 lO

Figure l. Effluent discharges to the Lower Fox River.

CHLORO-ORGANIC COMPOUNDS

1.5 L

PH > 11 w KOH

FXTRACT 2 x w 100mL 15% CHZCI2 in Hexane1 x w 100 mL Hexane

SOLVENT PHASF AQUEOUS PHASEPH < 3 w 50L HCI

FLORISIL COLUMN CHROMATOGRAPHYEXTRACT 2 x w 100 mL CHCI3

CONC. Sr REDISSOLVE IN ACETONE

ELUTE w Hexane or ELUTE w BS Ether0% Ether in Hexane in Hexane 5 mL *lhiuot

ADD 0.2 mi. 5% KZCO3

METHYLATE w 1 mL CH31 for 30 min.

ANALYZE by GC/EC and GC/MS EXTRACT% Hexane

ANALYZE by GC/EC and GC/MS

Figure 2. Water and wastewater extraction sequence.

de-ink and recycle paper to produce pulp, leading to the dischargeof PCBs in their wastewaters.

EXPE REMENTAI

During 1976-77, ca. 250 samples were analyzed, including riverand lake bottom sediments, snowmelt, biota seston, clams and fish!,river water and wastewaters from 15 pulp and/or paper mills and 12municipal sewage treatment plants. Four of these municipal plantsalso treat pulp and/or paper mill wastewaters ~ Wastewaters com-prised the majority of the samples received and were analyzed asdescribed in Figure 2. Wastewater samples �.5 L! were extractedat pH > 11 with methylene chloride/hexane, fractionated on Florisiland screened by gas chromatography with electron capture detection GC/EC! using procedures for chlorinated base-neutral compounds USEPA, 1973!. The remaining aqeuous phase of each sample wasacidified to pH 3 and extracted with chloroform. The solvent wasevaporated and the residue dissolved in acetone. This fractionwas analyzed for chlorophenols, chloroguaiacols, and related chlori-nated compounds. Selected extracts were derivatized with methyliodide to facilitate analysis of acidic compounds and to confirmcompounds identified by GC prior to methylation. Fractions whoseGC/EC chromatograms exhibited significant unknown peaks were analyzedwith a Finnigan 3100D Gas Chromatograph/Mass Spectrometer GC/MS!and 6000 Data System.

255

P. H, PETERMAN ETAL.

Electron impact mass spectra ranging from m/z 35-500 vereacquired every 3-4 sec. at an emission current of 0.35 ma, electronenergy of 70 eV, amplification of 10 7 amp/V and electron miltipliersetting of 2.10 kV. Calibration with perfluorotertiarybutylaminewas carried out according to the instrument manual and to specifi-cations given by Carter �976! . PCBs and other base-neutral com-pounds were chromatographed on glass columns �.8 m x 2 mm i.d.!packed vith 3X SE-30 or 3X SP-2100 and temperature programmed from100 to 220 C. Nethylated and non-methylated acid fractions wereanalyzed on 1.8 or 3 m x 2 mm glass columns packed with Ultra-Bond20M and programmed either from 90 to 210'C or 110 to 250~C,respectively.

Compounds were identified by a! comparison of retention timeand mass spectrum of a suspected constituent with those of a stan-dard of that compound; b! comparison of the full or partial massspectrum 8 peaks! of a constituent with published spectra e.g.Ei ht Peak Index, 1974; or c! by interpretation of the massspectral fragmentation pattern. Nany of the compounds were avail-able commercially, while some were provided by other researchers.

One compound that was unavailable from any source was chloro-syringaldehyde. Therefore, an experimental chlorination ofsyringaldehyde was performed. A commercial standard of syringal-dehyde was added to a solution of 5.25X sodium hypochlorite com-merci.al bleach! in aqueous acetic acid. The reaction proceeded16 hours, after which time the reaction product was extracted withmethylene chloride, evapo-concentrated to dryness with a gentlestream of air, redissolved in acetone, and then in]ected into theGC/NS. A total ion chromatogram indicated both unreacted syringal-dehyde and newly formed chlorosyringaldhyde. A very small amountof dichlorosyringaldehyde vas also detected in the reaction product.

RESULTS AND DISCUSSIONS

Chlorosyringaldehyde was one of the 105 compounds identifiedby GC/MS Table 1!. Compounds in final effluents which were detec-ted several times by GC/NS were quantitated by GC/NS, GC using flameionization detection FID! or GC/EC and listed in Table 2. Variouseffluents and extraction efficiencies were experienced, thereforeonly concentration ranges are given. The concentrations of thesecompounds generally corroborate earlier investigations of pulp andpaper mill effluents Rogers, 1973; Keith, 1976!. Compoundsdetected and quantitated by GC/EC in fish, clams, river water,seston and sediments are also included in Table 2. A completeset of these data appears in a technical report WI DNR, in press! .

CHLORO-ORGANIC COMPOUNDS

Table 1 Compounds Identified but not Quantified inSamples from the Lower Fox River System

PHTHALATES

Dibutyl PhthalateDiethyl PhthalateDioctyl Phthalate

a aa

3-methoxy phenyl! orguaiacy1 acetone

RESIN ACIDSESTERS

c c c

b c c c c cb a c

RESIN ACIDS METHYL ESTERS

Methyl Dehydroabietate

Continued next page!

257

c c ca a a a c

Acetone, Tetrachloro-Acetovanillone

Aniline, Trichloro-Benzene, Dichloro-diethyl-Benzo ate, Dimethyl-Benzoate, Methyl-methoxy-Benzoic Acid

Benzoic Acid, Isopropyl-Benzophenanthrene, Methyl-

or Benzanthracene, Methyl-!BenzophenoneBenzyl AlcoholBipheny 1Biphenyl, Methyl-Bisphenol ABisphenol A, Chloro-Bisphenol A, Dichloro-

� isomers!Bisphenol A, Tetrachloro-Bisphenol A, Trichloro-Borneol, Iso-Caffeine

Camphor, Oxo-Carb az o le

Ch lo r dane

DDD

DDE

DDT

Dodecane

PATTY ACIDS AND THEIR METHYL

Heptadecanoic AcidLauric Acid

Myristic AcidOleic Acid

Palmitic Acid

Stearic Acid

Methyl PalmitateMethyl S t carat eGuaiacol

Guaiacol, Dichloro- � isomers!HeptadecaneHexachlorocyclopentadieneHexadecane

Indole, Chloro-p-Menth � 4-ene-3-oneNaphthalene, Isopropyl-Naphthalene, Methyl-Nonadecane

Octadecane

Pentadecane

Phenanthrene, Methyl-Phenol

Phenol, p-Tertiary Amyl-Phenol, Chloro-Phenol, p � ~ � chloroethyl!-Phenol, Decyl-Phenol, Ethyl-Phenol, Nonyl- � isomers!Phenol, Trichloro-dimethoxy-Phenol, Undecyl-Phenyl DecanePhenyl DodecanePhenyl UndecanePhosphate, Tributyl-

P ropan-2-one, � � y droxy-

6,8,11,13-Abietatetraen-18-oic Acid

8,15-Isopimardien-18-oic AcidOxo-dehydroabietic AcidPimaric Acid

Sandaracopimaric Acid

P. H. PETERMAN ET AL.

Table 1 Cont.!

RESIN ACIDS CHLORINATED

b Chlorodehydroabietic Acid � isomers!b Dichlorodehydroabietic Acid

RESIN ACID METHYL ESTERS CHLORIN TED

Methyl ChlorodehydroabietateMeth 1 Di,chlorodeh droabietate

c Salicyclic Acidc Syringaldehyde

Syringaldehyde, Chloro-Tetradecane

Toluene, Dichloro-Toluene, Trichloro-

c Vanillin

c Vanillic Acidc Veratrole, Dichloro-c Veratrole, Trichloro-

Xy lene, Di chio ro-Xy lene, Trichloro�

a Compounds on EPA Consent Decree Priority Pollutant Listb Compounds in paper mill wastewaters reported toxic to fishc Other compounds previously reported in paper mill wastewaters

To assess the significance of the compounds detected in thisstudy, certain classifications were assigned. Twenty of the 105compounds, including PCBs, appear in the EPA Consent Decree PriorityPollutant List USEPA, 1977! . Although the commercial use ofAroclar 1242 and other forms of PCBs in printing inks and carbonlesscopy paper apparently ended in 1972, PCBs are still being releasedinto the Lower Fox River Basin. Deinking-recycling processes offive paper mills are some of the main sources. PCBs were detectedin all of the various matrices sampled. All 35 fish fillet samples,consisting of both rough and sport fish, contained detectable levelsof PCBs which were correlated to their fat content. Sixteen ofthe 35 fish exceeded the U.S. Food 6 Drug Administration tolerancelimit of 5 mg/kg, while 31 of 35 exceeded the Canadian Pood 6 DrugDirectorate tolerance limit of 2 mg/kg.

PCBs are lipophilic and accumulate in fat tissue. Clamsseeded in the Lower Fox River for 9 � 28 days showed that PCBs canrapidly bioaccumulate WI DNR, in press!. The mean uptake ratesvaried from 10 to 24 pg/day. The higher PCB uptake rate in the

Table 2 Compounds Identified and Quantified inSamples from the Lower Fox River System

Environmental

Natrix

Concentration

Compound Un i ts

Anisole, Pentachloro-

Anisole, Tetrachloro-Benzothiazole

Benzothiazole, Hydroxy-c Benzothiazole, Nethylthio-b Dehydroabietic Acid

0.008

10

10

ug/L0. 04Wastewater

ug/L4015Was tewat ers

4g/Lmg/kg

ug/Lpg/L

40

0. 28

20

100

0.1

0.22

2

5

Was tewaters

Sediments

Was tewaters

Was t ewat ers

Phenol, Tetrachloro-a Phenols, Trichloro-

� isomers!a Polychlorinated Biphenyls

Aro c lo r 12 42, 12 48 and12 54!

a Polycyclic AromaticHydrocarbons Acenaphthene, Anthracene,Chrysene, Fluoranthene,Pyrene!

* Entire aqueous sample filtereda Compounds on EPA Consent Decree Priority Pollutant Listb Compounds in paper mill wastewaters reported toxic to fishc Other compounds previously reported in paper mill wastewaters

259

a Dieldrin

b Guaiaco1, Tet rachloro-b Guaiacol, Trichloro-

� isomers!a Hexachlorocy cl ohexane

Lindane!a Phenols, Dichloro-

� isomers!a Phenol, Pentachloro�

Was tewaters

Ri ver wat er

*Seston

Fish

Wastewaters

Wastewaters

Wastewaters

Wastewaters

Wastewaters

S ediment

Fish

Was t ewat ers

Was tewaters

Raw wastewaters

Final effluents

River water

*Seston

Sediments

Clams seeded

Fish

Wastewaters

0. 05

0. 002

0. 02

0 ~ 005

0. 04

10

10

10

100

0.2

0.1

0. 05

0. 002

0. 05

0.26

0.5

0.5

0. 38

0. 02

0. 05

0. 06

0. 08

30

30

40

8500

2.7

0.022

50

60

8200

56

0. 85

0. 029

61

0. 74

90

10

pg/Lpg/Lpg/L

mg/kgug/Lug/Lpg/Lpg/Lpg/L

mg/kgmg/kg

pg/Lug/L

ug/Lpg/Lpg/Lug/L

mg/kgmg/kgmg/kg

pg/L

P. H, PETERMAN ET AL.

clams occurred at locations having relatively high PCB concentra-tions in the river sediments. These locations were downstreamf rom discharges containing PCBs.

Other compounds identified in pulp and paper mill wastewaterswere those found to be toxic to fish by other investigators Rogersand Keith, 1974; Leach and Thakore, 1977! . These toxicants in-cluded chloroguaiacols, resin acids, chloro � resin acids and oleicacid. Our observations of chlorophenols corroborates work byLindstrom and Nordin �976! . The source of the chlorophenols inthe mill wastewaters investigated in this study has not yet beendetermined. Chlorophenols may have been used by paper mills forslime control or been present as wood preservatives. Phenoliccompounds could also have been chlorinated in the bleaching orwastewater treatment stages. Chlorocatechols may have been present,but the analytical method employed did not appear to give any sig-nificant recovery of these compounds.

Previous investigations of toxicants in paper mill wastewaterhave always involved a pulp mill that derives its pulp from wood,thus releasing wood extractives and lignin-derived compounds suchas resin acids, guaiacols and other phenolics, some of which becomechlorinated in the bleach plant. In our survey, the highest levelsof chloroguaiacols occurred at a sulfite pulp mill which producesmostly bleached pulp. These compounds eluted cleanly from anunmethylated acid extract on the Ultra-Bond 20M column which re-solved 3 apparent trichloroguaiacol isomers. Chloroguaiacolswere not detected in some Fox River paper mill wastewaters, presum-ably because these mills either do not bleach their wood-derivedpulp, or else they bleach purchased pulp and/or deinked recycledpaper. The last two should contain lesser amounts of the woodextractives and lignin-derived compounds.

The GC/MS analysis of a sample extract of a paper mill whichbleached either purchased bleached pulp and/or deinked recycledpaper is shown in Figure 3. The total ion chromatogram TIC! ofa methylated acid extract of the mill's final effluent shows 39identified compounds, most of which are methylated derivatives.Thus, chloroanisoles were originally present as chlorophenols,chloroveratroles as chloroguaiacols and the dimethyl ether deriva-tives as various bisphenol A isomers. Likewise, the fatty andresin acid methyl esters were originally present as the correspond-ing acids in the acid extract. The scale of the TIC has beenlimited to 30K of full scale to better show the compounds at lowerconcentrations. For reference, 75 ng of aldrin peak 15! wereinjected as an external standard. Peak 8, representing tetrachloro-guaiacol, was quantitated by GC/EC at 14 pg/L, while peak 25, rep-resenting the resin acid dehydroabieti.c acid DHA!, was quantitatedby GC/FID at 3200 pg/L. This concentration approached that of8500 pg/L which was seen in the aforementioned sulfite pulp mill's

CHLORO-ORGANIC COMPOUNOS

ISCl

CSI

EOeo

B',ISm ~SIas ca

'E 'EK K

I

4lg g'E Eesa- m

O C OCg O� amCSea IO a~ 4Cl

Cs s Cl

~ C4lOCS

IOCO O

lOIII

~ C 4lCC E

X CS 44

mNO eo

OE X~ 4ISSss

~ Cgos a~ g

C4 aCOCXa

am gm, 4IO CE oCE Ess CE

ISCa aca. g

assCS CD

Os~ ll

4IX .Clll

IamCC

44IaCO

KIISSR%~ t etamaC C4I 4I'a CEC C

COCIS

5 aIkl ~ IE oOC Ooa C

COCao $nK

ClammIa ISC C44 � mOH OIa «O.E'mc a mE «a«ac >

o c E sem IssD ISa- o'a IOT a wcac c ~.Cce Cc '5 a a+ I- 5 5 Z

O~ 4~ 4aa OEIOSI «m ISISE C om ca Ia

S 4-a

K R ~ 4Q « IN N

.8aas 0Crs Ea a

aCJ

eec4aeO Os

mc a~ 4Oas «CICO ' CCJ N CS

8OssCSCSI4Sl

a

COC1

ClCL.

X CD

m~ 4 ISC SsOm aaIS cs

~ 4C

O ~ 4

gC eaa u~ e mC «Cooaca a- Z

Ncaac chio ~ clem&«Ness ac«seal coN N N N N N N N Os oe ss CO ss 44 os CO 44N CO ac IO

~ 44lCE~ O

92

K ~ CCCC0COI ~ C

C!

CD

C!

30Allld Wlt

I

UJ

KIZ:~ CJC~ CC

ICDOCEKI'SCICI

CD

I�«C

IIll

«Ncoacsolor Soaso

E ~EE CS

E O- CJQ 4CSIONDC OaIaIas CS~ O� E

E

'cS aC IS43c m

EO

4I

Ia

mo ECES CEa O4

C a OlSlO

Ek

CO I

N 44g COR COmRm XO~ 4O'b SEIM maI'm a

'csO daa Q OCE 5

E3E3

C3lf3

K3CCCO

E ZU7

I4Iv

K3U3

final effluent. Since the mill whose effluent is represented inFigure 3 lacks a wood pulping process, the relatively large amountsof fatty and resin acids present, especially DHA, could have comefrom its use of resin sizing Merck Index, 1976! in the papermakingprocess. This water-intensive process could have diluted theavailable chlorine, thereby reducing the effectiveness of formationof chloro-resin acids peaks 31, 32 and 37, Figure 3!.

DHA appears to be the most stable of the resin acids Brownleeand Strachan, 1977; Fox, 1977!. The toxicity of resin acids tofish has been known since 1936. The 96 hour LC5 o concentrationsof DHA for young Sockeye Salmon are 2000 pg/L Rogers, 1973! and750 pg/L for Coho Salmon Leach and Thakore, 1977!. The latterinvestigators also reported even lower 96 hour LC5 ~ concentrationsfor mono- and dichlorinated DHA.

Other compounds previously reported in paper mill wastewaterswere also found in this study including acetovanillone, guaiacol,methyl thiobenzothiazole, syringaldehyde, vanillin and vanillicacid. Several compounds commonly used in industry that were iden-tified including benzothiazole, an antioxidant; bisphenol A, afungicide or an intermediate in the production of epoxy resins;and nonyl phenol, present in surfactants Merck Index, 1976! .

Several compounds which apparently have not been reportedbefore in the environment are chlorosyringaldehyde, 5 separatechlorobisphenol A isomers, and trichloro-dimethoxyphenol, whilechloroindole was found apparently for the first time in a sewagetreatment plant effluent. Chlorosyringaldehyde was identifiedtogether with syringaldehyde in a semi-chemical pulp and paper milluntreated wastewater, as seen in another TIC Figure 4!. Syringal-dehyde, a hardwood lignin degradation product would be anticipatedto come from a pulp mill using hardwood. The apparent chlorinationreaction within the plant compares with similar examples of thechloroguaiacols and chloro-resin acids in other plants. Chloro-syringaldehyde was identified in an acid fraction without derivati-zation when chromatographed in an Ultra-Bond 20M column. The massspectrum of chlorosyringaldehyde has isotopic molecular ions ofm/z 216 and 218 which are consistent for a compound with one chlo-rine atom Figure 5!. These two ions as well as the fragment ionsm/z 215, 201, 173, 145, 130, 127 and others have been shifted 34mass units higher than for syringaldehyde which is also consistentwith the addition of a chlorine atom to the benzene ring. In addi-tion, the laboratory chlorination of syringaldehyde described inthe Experimental Section yielded a mono-chlorinated compound whichmatched not only the identical retention time but also the massspectrum of the apparent chlorosyringaldehyde in the untreatedwastewater Figure 4!.

262

CHLORO-ORGANIC COMPOUNDS

TOTAL ION CHROMATOGRAM:ACID EXTRACT OF A PAPER MILL WASTEWATER INFLUENT

100

50 100 150 200 250 300 350 400ILIoP TIC: CHLORINATED SYRINGALDEHYDE PRODUCT

Column Conditions: Ultra-Bond 20M, 3m x2mm,Temperature Prottrammed ItO-215'C 4 4 C /min $7mNQALaf H7of

e 100

50 100 150 200 250 300 350 400 450SPECTRUM NUMBER

Total ion chromatograms: syringaldehyde andchio rosy ringaldehyde

Figure 4.

ACID EXTRACT OF PAPER MILL WASTEWATER INFLUENTSCAN dr350 � dt346 21.50 MIN.

127 2 l6100

GALDEHYDE

30 50 150PERCENT

OF

BASE CHLORINATED SYRINGALDEHYDE PRODUCTPEAK SCAN/347-& 343 21.36 MIN

l27100

IOO 250 M/2200

263

Figure 5. Mass spectra of chlorosyringaldehyde

A group of chlorinated bisphenol A compounds was identifiedin the same extract as that shown in Figure 3, which also containedchloroguaiacols and chloro-resin acids. Despite its widespreadindustrial use, bisphenol A has apparently only been recentlyidentified in the environment Matsumoto et al, 1977!. The massspectrum of its dimethyl ether derivation is compared to that ofpeak 2la Figure 3! and included here Figure 6! because it wasapparently not available in the literature. The mass spectrum ofits derivative shows the molecular ion at m/z 256 and the base peak -CH3!+ at m/z 241 shifted 28 mass units higher than that of bis-phenol A, which is consistent for methylation of both hydroxylgroups. The mass spectra of the 1 � 4 chlorinated isomers peaks28, 29, 33, 35 and 36, Figure 3! show similar upward shifts of thetwo main ions 34 mass units for each additional chlorine atom. Themass spectra of these compounds also show the respective isotopicclusters corresponding to the number of chlorine atoms present.Using Cl, the molecular ion and the base peak of monochloro-bisphenol A dimethyl ether were, respectively, 290 and 275; thoseof both dichlorobisphenol A dimethyl ether isomers were 324 and309; those of trichlorobisphenol A dimethyl ether were 358 and343; and those of tetrachlorobisphenol A dimethyl ether were 392and 377. A pure standard of tetrachlorobisphenol A was commer-cially available, as it is apparently used as a flame retardant.The mass spectrum of its methylated derivative is compared with

STANDARD. 'BISPHENOL A DIMETHYL ETHER241

100

15

PERCENT

OF 30 50BASE

PEAK PEAK

z PERCENTOF

15 TOTAL

150100241

30 50 100 150 200 250 M

Figure 6. Mass spectra of bisphenol A dimethyl ether.

264

PEAK 33, TETRACHLORO-BISPHENOL A DIMETHYL ETHERIll-Ig !+

I 00

PERCENTOF

BASEPEAK

PERCENTOFTOTAL

15030 50 IOO Mfz

STANDARD: TETRACHLORO-BISPHENOL A DIMETHYL ETHER

100

PERCENTOF

BASEPEAK

PERCENTOFTOTAL

30 50 IOO 150 200 2

Figure 7. Mass spectra of tetrachlorobisphenol A dimethyl ether.

Trichloro-dimethoxyphenol was tentatively identified in theacid extract Figure 3! . Its methylated derivative was peak 11which eluted just after a close congener, trichloroveratrole peak10!. Trichloro � dimethoxyphenol was another compound which wasdetected in the acid extract chromatographed directly without deri-vati.zation on an Ultra-Bond 20M column. It eluted just aftertetrachloroguaiacol. Its mass spectrum showed abundant molecularions at m/z 256-260 consistent for 3 chlorine atoms, a similarcluster at ions m/z 241-245 M-15!+ and m/z 198-202, M-58!+. Themass spectrum of the methylated derivative, peak ll Figure 3!showed abundant isotopic molecular ions now at m/z 270 � 274 withsubsequent fragments M-15!+, M-43!+ and M-58!+.

Chloroindole apparently has not been previously detected inwastewaters, but it has been isolated from a bacterium Pseudomonas

f lava ~la eanica! from the Pacific Ocean Higa and gcheoer, 1975! .In our study of untreated wastewater from a municipal sewagetreatment plant, this compound was detected in the 20/ ether inhexane Florisil eluate at a concentration of ca. 30 p g/L. Themass spectrum showed abundant isotopic molecular ions m/z 151 and153, base peak of m/z 89, and less abundant ions at m/z 124, 116

that of peak 33 Figure 3! in Figure 7. In a laboratory chlorina-tion of bisphenol A performed similarly to that of syringaldehyde,it was shown that 2,4,6-trichlorophenol was the main product formed,although various chlorinated bisphenol A isomers having 1-4 chlor-ines were also formed.

and 63. In comparing the compound from our sample with a commer-cial standard of 5-chloroindole, both mass spectra matched well buttheir retention times differed by several minutes.

In addition to the 105 compounds identified in this study,another 20 or so compounds were detected but not conclusivelyidentified to date. A group of related compounds was consistentlydetected. The most prominent were two apparent isomers with amolecular weight of 196. In wastewater samples of the paper millrepresented in Figure 3, these two isomers were followed by aboutnine chlorinated isomers with apparent molecular weights of 230,264 and 298. The mass spectra of all of these are included ina technical report WI DNR, in press! . Mass spectra of the twonon-chlorinated isomers are similar to diphenylacetaldehyde andtrans stilbene oxide. Although the compounds with molecularweight of 196 have been detected in various extraction fractions,they and the chlorinated isomers primarily have been found in thefirst Florisil eluate �X ether in hexane!. Their concentrationshave been sufficiently high to mask some of the PCB peaks detectedwith GC/EC. This mill's extensive deinking, recycling and bleach-ing processes could conceivably release the compounds with molecularweight 196 which ultimately become chlorinated.

This GC/MS study was aided by the use of a low-loaded, Ultra�Bond 20M column packing ca. 0 ' 3X Carbowax 20M! similar to thatfirst discovered by Aue �973! . Elutions were characteristicallysharp, with polar phenolic compounds eluting quite well. Baselineseparation of pentachloroanisole from tetrachloroveratrole wasachieved, contrary to the case for 3X SP-2100, 3X OV-17, or themixed phase packing 4X SE � 30/6X OV-210 designed for pesticideanalyses. Very low bleed on temperature programmed analysesaided background substraction resulting in optimum mass spectra.

The fate and long-term health and ecological implications ofmany of the 105 compounds identified requires further research.For PCBs and some other chloro-organics, sampling data and labora-tory experiments show a direct correlation of their concentrationsin wastewaters with suspended solids concentrations. Suspendedsolids reduction in wastewater treatment plant also reduces thechloro � organic concentration in the final effluent WI DNR, inpress! . For example, the untreated wastewater of a paper millwhich deinks and recycles paper contained 25 p g/L PCBs and 2,020mg/L suspended solids. Following primary clarification, concen-trations were reduced to 2.2 pg/L PCBs and 72 mg/L suspended solids.After secondary treatment the final effluent contained only 1.4 pg/LPCBs and 10 mg/L suspended solids. Now the final disposal of thetreatment plant sludge containing PCBs must be resolved.

266

ACKNOWLE DGEMENTS

This study was supported in part by a contract from the U.S.EPA Great Lakes National Program Office, Region V, EPA ContractNo. 68-01-4186! . Additional support was provided by the WisconsinDepartment of Natural Resources and the Wisconsin Laboratory ofHygiene.

REFERENCES

Aue, W.A., C. Hastings, and S. Kapila. 1973. J. Chromatog. 77,299-307.

Brownlee, B., and W.M.J. Strachan. 1977. J. Fish Res. Bd. Can.34, 838-843.

Carter, M.H. March, 1976. EPA Report No. 600.4 � 76 � 004.

Donaldson, W.T. 1977. Environ. Sci. & Technol, 11, 348.

Fox, M.E. 1977. J. Fish Res. Bd. Can. 34, 798-804 '

Higa, T., and P.J. Scheuer. 1975. Naturwissenschaften, 62, 395-396.

Jolley, R.L. Ed.! 1976. Proceedings of the 1975 Conference onthe Environmental Impact of Water Chlorination, ORNL CONF751096, Oak Ridge, Tennessee.

Keith, L.H. Ed.! 1976. Identification & Analysis of OrganicPollutants in Water. Ann Arbor Science Publishers, Inc.,

Ann Arbor, MI, 718 p.

Leach, J.M., and A.N. Thakore. 1977. Prog. Water Tech., 9,787-798.

Lindstrtim, K. and J. Nordin. 1976. J. Chromatog.,128, 13-26.

Mass Spectrometry Data Centre ~ 1974. Eight Peak Index of MassSpectra. 2nd ed., Mass Spectrometry Data Centre, Aldermaston,U.K., 2933 p.

Matsumoto, G., R. Ishiwatari, and T. Hanya. 1977. Water Res. 11,693-698.

Merck and Co., Inc. 1976. The Merck Index, 9th ed., Merck andCo., Inc., Rahway, N.J.

Neidleman, S.L. 1915. CRC Critical Reviews in Microbiology 3�!, 333-358.

267

Rogers, I.H. Sept. 1973. Pulp Paper Mag. of Can., 74 9!.

Rogers, I.H., and L.H. Keith. 1974 . "Organochlorine Compoundsin Kraft Bleaching Wastes. Identification of Two ChlorinatedGuaiacols", Environment Canada, Fisheries & Marine Service,Res. & Dev., Technical Report No. 465.

USEPA. 1973. Fed. Reg. Part II, Vol. 38, Nov. 28.

USEPA. 1977. "Sampling and Analysis Procedures for Screening ofIndustrial Effluents for Priority Pollutants", EMSL, Cincinnati,Ohio, 69 p.

WI DNR, "Investi.gation of Chlorinated and Nonchlorinated Compoundsi.n the Lower Fox River Watershed". EPA Contract No. 68-01-4186 in p ress! .

268