June of Programme of of

Isotopic Fractionation of Chlorinated Ethenes During Reductive DebaJogenation by Zero Valent Iron

by

HannahDayan

Athesis submitted to the School of Graduate Studies

inpartial fulfilment of the requirements for thedegreeof

Master of Science

EnvironmentalSCienceProgramme Memorial University of Newfoundland

June1998

Acquisitionsand BibIiograpllic5ervices

395~s.­

=ONK1AllN'

8ibIiothilquenationale ducanada Acquisitionsel selVices bibliographiques 395... WelinglOn Ot!lhraON K1AON<l

~

The author bas granted a non- exclusive licence allowing the National Library of Canada to reproduce, loan, distnbute or sell copies of this thesis in microform., paper or electronic formats.

The author retains ownership of the copyrightinthis thesis. Neither the thesis nor substantial extractsfromit maybeprinted or otherwise reproduced without the author's permission.

L'auteur a accorde one licence non exclusive permettant

a

la Bililiotheque nationale du Canada de reproduire, preter, distnbuer ou vendre des copies de cette these SollS la forme de microfiche/film.. de reproduction sur papier ou sur format electronique.

L'auteur conserve la propriete du droit d'auteur

qui

protege cette these.

Ni 1a these ni des extraits substantiels de celle-ci ne doivent

etre

imprimes ou autrement reproduits

sans

son autorisation.

D-612-36111-X

CanadtL

Abstract

Many surfaceandgroundwateIS are contaminated by chlorinated ethenes as a result of theirwidespreaduseindifferentindustries..One promising technique fortheremoval of thesecompoundsisreductive debalogenation using iron metal as the reducing agent.The method involves introducing an iron metal barrier intotheflow of groundwater.Zero valent iron reactswith thechlorinated etbenes byeitherp.etimination or dehydrogenation.

depending onthepH and Eh ofthe system. Thepurpose ofthisinvestigation was to determinethe overall magnitude and direction of isolOpic fractionationwithrespect tothe stable isotopes of carbon

e:zc.

13(:)during reductive dechlorination of selected chlorinated ethenes.Byquantifyingthefractionation of reductive dehalogenation as

wen

as other reactions. the processesthatmaybedominantintheremoval ofthepollutants canbe detennined. Replicateexperiments

were

done using low concentration solutions of trichloroethylene(fCE),perchloroethylene(PCE),1.2-cis..<fichloroethyeJene (c-DCE) and 1,2-tranS-dichloroethylene (t·DCE) andacid washedironpowderusingcompoundspecific isotopicanalysis.Thechangesinthecarbon isotope composition(aile)andthe consumption oftheetheoes were followed until they were no loDger detectable by gas chromatography isotope ratio mass spectromeuy (for 169. 345, 366 hours respectively).

Thepresence of dissolved oxygen allowed foracidicconditionsandthe precipitation of ferric hydroxide.Theconsequence oflhiswasthe ''nal:uraf'buffering ofthesystemto pH between5 and 6.ThetraIlSition from Ikliminationtodehydrogenationas the major

reaction pathway alsoliesin thispH range.1be pH also had an effect on the reaction rates. During the course of the reaction, enrichmentin

He

was observedintheisotopic composition of each of the compounds. Kinetic orRay~ighprocesses can be used to explain

this

effect. Fractionation factors were calculated usingtheRayleigh model and found to be 1.0146, 1.0036, 1.0087, and 1.026 for c-DCE, t-DCE,TeEandPCE respectively. Future studies should include changing parameters such as temperature, as well as using buffer solutions;

these

studieswill

be

useful inorder to obtain a more comp)ete picture of processesinnatural waters.

iii

Table of Contents

Abstract ...•.

Listof Tables ListofFigures .

Listof Abbreviations UsedillText Admowledgments .

..n

vi . .... vii ix

Chapter1 . Introduction . . . .. 1

1.1Trichloroethylene . . 6

1.2 Perchloroethylene . . ...•. 9

1.3Dicbloroethylene .. • . • • • • • • • • . • • • • . •It

1.4RemovalProcesses. . 12

1.4.1 In Situ Bioremedialion . . .. 14

1.4.2 In SituChemical Remediation . . . .. 17 1.4.3Reductive Debalogenalion using Zero Valent Iron .20

1.5 IsotopeChemistty . . .24

1.5.1lsotopeExchange . . . 26

1.5.2KineticEffects .. 27

Chapter 2 •Experlmeotal Procedure . . 31

Chapter 3 - Results . . 41

3.1 pH ..41

3.2~.. .Q

3.3 Products. . . 43

3.4Rate Constants. . . • . . . . 44 3.5 Isotopic Composition. . . • . . • . • • • • • . • . • . . • . . • . . . • . . • . . . 53 Chapter 4 - Discussion ...••.•••.

4.1 PrecisionandError 4.2 Isotope Effects 4.3 Reaction Rates

. 58

... 58 .... 59 .. 61

Chapter

s .

SummaryaDdCouduslons . . .... 82 ... 88

ListofTables

Table1A: Physical andcremicalPropertiesofTCE and PeE .. . 4 Table18: PhysicalandCbemicalPropeniesofc-DCE andt-DCE . 5 Table2: VapourPressurefor TCE, PCE and DeE

caIcu1atedfrom Henry'sConstant. . .. _. . . .. . . . • . . •• . • • . . 33

Table3: Temperature Program for TCE, PCE and DCE 38

Table4A: FIrStOrderRateConstants forReductiveDehalogenatioo of

Chlorinated Etbenes . _ 50

Table4B:NormalizedHalflivesof Reductive Dehalogenatioo of Chlorinated

Etbeoes . . 50

TableS:Second OrderRateConstants for Reductive Debalogenatioo of

ChlorinatedEtl1enes . . . 52

Table 6:Isotopic Fractionatioo Constants for ChlorinatedEtbenes . . 65 Table7A: Standard Reduction Potentials for Dehydrogenation

ofChlorinatedEtbenes . . . 79

Table 78: StandardReductiooPotentialsfor~.Elimination

ofChlorinatedEtbenes .. . 79

Ust

of Figuns

Figure l:Molecu1ar Structures of the Chlorinated Ethenes Used

in

this Study _.. _.. _. 3 Figure 2: Cycling of TeE

in

the Environment

Figure 3: Application of Chemical Agents for Abiotic Remediation . Figure

4:

Reaction Pathways for the Reductive Dehalogenation of

Chlorinated Ethenes .

... 8

. 19

. .... 22 Figure 5: Schematic of the Difference

in

Dissociative Energies for Isotopic Species .. 29

Figure 6: SPME Assembly 34

Figure 7:

Schematic of

GC-C-lRMS Figure

8:

Consumption of c·DCE Over

Time

Figure

9: Consumption oft-OCEOver Time

Figure

10:

Consumption ofTCE Over Time:

Figure

11:

Consumption of

PeE

Over

Time .

Figure

12:

Isotopic Fractionation of c·DCE over Time ...•••..

Figure 13: IsotOpic Fractionation oft-OCE overTime

.... 37

.... 46

.47 ... 48

..49 . .... 54

.... 55 Figure

14:

Isotopic Fractionation ofTCE over Time

Figure

15:

Isotopic Fractionation of

PCE

over Time .

... 56

. 57

Figure

16:

Plot of Isotopic Ratio Versus Fraction Remaining for c-DCE 66 Figure 17: Plot of Isotopic Ratio Versus Fraction Remaining for t-DCE .

Figure 18: Plot of Isotopic Ratio Versus Fraction Remaining for TeE

. ... 67 ... 68

Figure 19; Plot of Isotopic Ratio Versus Fraction Remaining for PCE Figure 20: Phase Diagram for Iron Hydroxide Species and PCE

Reductive Dehalogenation Reactions .

Figure 21: Phase Diagram for Iron Hydroxide Species and TCE Reductive Dehalogenation Reactions .

... 69

... 73

..74 Figure 22; Phase Diagram for Iron Hydroxide Species and TCE

Dehydrogenation . . . .. . .. 75

Figure 23; Phase Diagram for Iron Hydroxide Species and c-DeE Reductive Dehalogenation Reactions .

Figure 24: Phase Diagram for Iron Hydroxide Species and t·OCE Reductive Oehalogenation Reactions .

... 76

. ... 77

List of Abbreviations UsedinText

atomic massunits

canadian Environmental ProtectionAct compoundspecificisotopic analysis Celsius

change over l.2-cis-dichIoroethyiene dense non aqueous phaseliquid dichloroethyleoe flame ionization detector gas chromatograpb gas chromatograpby-combustion-

isotope mass spectromeuy gas chromatograpby-mass spectrometry lethalconcentrationrequiredtoobserve

mortalityin50%ofa test population Molar Concentrationinmo1esll methane-monooxygenase partsper million PeeDee Belemnite perchIoroethylene

per

mil (partsper thousand) personalcommunication polymethyIsiloxaoe pounds per square inch solidphasemicroextraetioo 1.2-uans-dichloroethylene trichloroethylene vmylcbloride World Health Organization

amu

CEPA CSlA

·c

C/O c-DCE DNAPL DCE FID GC GC-C-IRMS GC-MS

LC~

M MMO ppm POB PCE '/~

pers.comm.

PDMS psi SPME I-DCE TCE VC WHO

Admowledgments

I wouldJ:iketotakethisopportunity tothankDr. Iun Abrajano for giving me the opponunity toworkin hislab. Ialsowouldliketothank Dr.Bob HelIeurandDr.Henry Longericb for beingmOStpatientwithme duringthefirstayat aMaster'sproject.

I am also grateful to Linda Wmsor and JerryPulcbanatthe Biogeochemistry Facility at MUN. Without them,therewouldprobablybenoresultsto speak of

ThankstoDr.Neil. Sturchio andthegroupatArgonne Laboratory whodid theBET an&ys;s-

ThankstoDr.Lynn Roberts fortheODeemailthat helpedthings progress duringthis project.

I would alsoliketothankAldenWoodward, who gave me a shoulderto cry onwhen things were looking prettygrim.

Thanks

to

Billiethe

cat

whogivesme unconditional love, and purrs.

Lastandcertainly not least, manythanksto Mom and Dadforalways being thereforme.

llris one's for}'Qu.

Chapter- I -utrodactioD

TheiDc:reasedincidence of reported health problemsandecological damageassociated withcontaminatc:dsurfaceandgroundwatershas lead toa coocc:rted effort,inscientific andengineering communities.. to develop effective watCl" quality characterisation and water treatment methods. Chlorinated ethenes are the focus ofthistypeofre:search,since theirpresence has become ubiquitousinthe environment withtheirextensiveusein industry. These compounds are highly toxic to most organisms, and do not readily degrade naturallyinthe environment.

Reductive deha1ogenation usingzerovalent ironisanewandpromising techniqueforthe treatmentof water that is contaminatedwithchJoroethenes. Current research focuses on the ratesandmechanisms involved inthisreactionandbowthereductive dechlorination processcanbeengineered into contaminated sites.

Thepurposeofthis studywasto determine themagnitudeanddirection of the changein the carbon isotope signature of chlorinated etbe:nes during reductive dehalogenation. To date, very few studies of thistypehavebeenundertaken. Pollutants from different sources have unique isotopic compositions and these "fingerprints" canbeused to trace pollutants to their sources (van Warmerdamet aI.,1995). lbis approachispreatised on the assumption that no alteration of the isotopic composition occurs between the release and sampling of the contaminantJustas compounds can have soucce isotopic

"fingerprinls", eachchemicalreaction involving lhe compound of interest couldaffect.me isotopic composition in a predictable way. By understanding theextentof fractionation in reactions such as reductive deba1ogenation. metypeand extent ofreaction of chlorinated ethc:ues canbeunderstoodinfield situatioDS.

Bench scale experim.enls were conducted to modellbc: cubon isotopic variationslhat.

would occur during reductive dehalogenation. Four chlorinated etbenes were selected:

ttichloroetb.ylene (fCE), perch1oroethylene (PCE), 1,2-cis-dicbloroethylene (c-DCE) and l,.2-trans-dichloroemylene (t·DCE) (Figure I). The environmental chemistry oflhese compounds isdiscussedin the following sections and asummaryofmeir physical and chemical constants is giveninTables lAandlB.

Trichloroethylene

Perchloroethylene

1,2.frans-Oichlotoethylene l,2-ci,yDichloroelhylene

Figure 1: Molecular Structures ofthe Chlorinated Elhenes Used in thisStudy.

TableIA: Physical and Chemical Properties ofTCEandpeE

Trichloroethylene(sqH)

Property VoJ~

Physical State colourless liquid

MolecuJar Weight IJl.Samu

MeltingPointO _87°C

BoilingPoint* 87°C

Solubility inWa~ 1100mgIL Density" 1.46glml(20 "c) Henry's Constant* 1.03xlO·:attn~m)/mole

Perchloroethylene(SCI,)

Property Value

PhysicalState MolecuJar Weight·

MeltingPoint*

Boiling Point·

Solubility in Warer- Density"·

Henry's Constant·

• source: Howard, 1990

•• source:Lide,l99S

colourless liquid 16S.8amu

·t9 ·C 121°C 1503 mgIL

\.62g1ml(20°C)

Table18: Physicaland Chemical Properties of c-DCE and t-DCE 1,l-ds~Dichloroethylene(SClflJ

Value Physical State

Molecular Weight Melting Point"' Boiling Point"' Solubility in Water*

Density"'·

Henry's Constant"'

colourless liquid 96.9amu

~80.S

ac

60.3

ac

3.Sgfl.

124 ghnl (20 "q 0.00337 atm-m'/mole

1,l-17ans-Dichloroethylene(CP~,)

Property Value

Physical State Molecular Weight

Melting Point·

Boiling Point"' Solubility in Wattt+

Density*·

Henry's Constant*

• source: Howard, 1990

•• source: Lide, 1995

colourless liquid 96.9unu

~soac

48 "C 6.3gIL 1.26glml(20"q 0.00672 atm-m'/mole

1.1 Tridilorodhyleae

TrichJoroetbylene (TCE)isa colourlessdensenon aqueous phase liquid (DNAPL).Itis moderately solubleinwater,andbas a higherdensitythanWllter, forming "puddles" of solventinnatun1 waters. TCE hasbeenplaced on the Priority Substances List of the Canadian Environmental Protection Act (CEPA) (Mooreet al.,1991) because of its toxicityandstabilityinthe environment.

Trichloroethylene isusedin both industrial and household products (Govenunent of Canada, 1993a, Mooreet al..1991).Industrialuses for TCE include degreasing of fabricated metal parts and cleaning of electronic components.Itis also used in the production of adhesive polymers, textile manufacturing and as an extractive solvent in foods. Paint coating removers and dry-cleaning fluids contain TCE as an active ingredient.

Several investigatiODS have been undertaken in order to determine the toxicological effects ofTCE. Effects include malignant tumours, changes in composition of lipids and proteins, and alteration of behaviour in mice and rats. Complete mortality of zooplankton requiredthreedaysintest ponds (at 110 mgfL). Continuous exposure ofTCE to fishalso showed adverse effects on growth and swvival rate at exposures to concentrations of 0.21

mWL

for 120 days (Government of Canada, 1993a).

Effects on humans have been identifiedinpostmortem examination ofadolescenlS who were addicted to inhaling the solventinglue or cleansing fluid (AviadoeIal., 1976).

liverandbraindamage were observedandinsome casesdeathwascausedby cardiac arrest. Kidney failure also occwred as a result of accidental poisoning.

Therelease ofTCE to the enviromne:ntis exclusively from mthropogenic sources as therean:no known natural sources oflbis compound. The cycling ofTCEinthe environment is shown in Figure2.Infiltration ofTCE into soil and natural waters is usually aresult ofeither spills, leakages of storage tanks, or leaching from landfills.It basbeenestimated that approximately60%ofthetotal world production ofTCE is releasedto the environment (Mooreel a1.,1991).

Swf:ac:espillsofTCEwillusually evaporate quickly into the atmospherebecause of its high vapour pressme. Photochemical reactionsinlhe atmosphere between TCE and hydroxyl radicalsoccurwilb balflives of approximately 3 daysand2weeksin the summerand wintermonths.,respectively (Government ofCanada, 1993a). TCE thatdoes not volatilize from surface water maybetransportedto groundwater. Natural removal of TeE from groundwater isthroughbiodegradation with balflives of a few months toa few years (Government ofCanada,1993a).

InCanadian surface waters, the rangeinconcentration ofTCE was found tobefrom

i ^• i

'0

c

~ ~

Figure 2: Cycling ofICEinthe Environment (After Mooreet at.,1991). The above schematic showsthatapproximately 10% oftbe TCE released to the environment is foundin surfaceandgroundwaters, however.thiscan differ fromspill tospilL

below the limit of detection to as high as 0.09 mgIL. Groundwater near landfillsitesbad concentrations from 4.4-7.2 mgIL. The guidelines set by the Government of Canada wen:

0.02 mgIL for watersusedfor recreation and 0.05 mg/L for raw water fordrinkiDg supplies (Mooreet al.,1991). World Health Organization (WHO) guidelines for drinking water are set at 0.07 mgIL (WHO, 1993).

1.2Perch1oroetbyleac

The major use of percbJoroethylene (PCE) in Canada is in lIte dry cleaning industry.

Several other industries. such as textiles andinkmanufacture also use PCE. It is also foundinproducts such as water repellents, adhesives. paint removers andwoodcleaners (Government of Canada, 1993b).

Adverse effects of PCE on kidney, reproductive and central nervous systems wen:

observedintoxicological testing on rodents (Government of Canada, 1993b). LCso (lethal concentrationrequiredto observe mortalityin50% of a test population) values for aquatic and marine biota were also investigated (Government of Canada, 1993b). The 24 hour exposure LC₅₀value for rainbow trout was approximately 5 mg/L. The 96 hour LCso for fathead minnow larvae was between 13-24 mg/L.

Kidney and liver failure have been observedinaccidental exposure ofPCE on humans.

Indoor air pollution from PCEiscommon., and as a result, increased incidence of cancer tothe liver and bladder have been observed. The risks ofthese effects areincreasedfor individuals employed in the dry cleaning industry (Government ofCanada. 1993b).

Anthropogenic activity is the only known source of PCE in the environment. Distillation vents, industrial liquid efIluenlS and leaching from landfill sites are common sources to soil and groundwater.

PCE is more volatile than TCE. and will evaporate more quickly from surface water. In the atmosphere. PCE will react with hydroxyl radicals with half lives between 27-58 days in the summer months (Government of Canada., 1993b).

PCE has a higher density than water andwilltend to sink to the bottom of bodies ofwater inwhich it is present. Laxger spills will form "puddles" of solvent at the bottom ofwater bodies. These may break into smaller droplets which can be resuspendedinthe water column and eventually volatilize (Government of Canada, 1993b).

Concentration ofPCEinsurface water is generally quite low (between 0.01-0.02 mg IL) inCanadian waters. High concentrations havebeenfoundingroundwater near dIy cleaning and land fill sites (0.06-200 mgIL). The WHO guideline for drinking water is 0.04 mgIL (WHO, 1993). The curmtt guidelines for PCEindrinking waterinCanada for

PCEis0.03mg/L(Government of Canada, 1996)

1.3 Dicbloroetbylcnc

l)-eis-dichloroetbyiene (c-DCE) and 1).trans-dichJoroethylenc (t-DCE) areusedas solvents in organic synthesis. They arealso used in perfumes, andinthe polymer industry in thermoplastics andrubbermanufacturing(Verschueren, 1983).

Sources ofDCE intotheenvironment includeairemissions or industrial liquid effluent (Howardet al. 1990).Isomers ofDCE mayalsobe present as degradation products of PCEand TCE.

DCE is highly volatile, and most sutface water contamination canbeexpected tobelost throughvolatilization. Reactionsinthe atmosphere with hydroxyl radicalscanoccur withhalflives of approximately eight days (Howard,1990).

Toxicology studiesinrats have shownthatat low exposure levels(0.33mmollkglday), some liver abnormalities were observed (McCauleyetaI.,1995). Congenitalheart defects were observedinrat fetuses whose mothers were exposed to DCE during pregnancy (DawsonetaI.,1993).Theseexperimentswere based on findings thatinsome areasinthe United States, where drinking water was contaminated byTeEand DCE,

cardiac defects were: common in children.. Leghorn chicks also develop heart defects fonowingexposureto5}llD.olIL (approximately0.052J!gIL)ofOCE (GoldbergeraJ., 1992). Thereislittle data on the effects on aquatic biota, although rainbow trout responded (crTatic movements and coughs) after one hour of exposure of 10 IlgIL ofc- OCE (Kaiseret ai.,1995).

Concentrations ofOCE found in waters around the United States vary between 0.001- 0.01 mgIL in surface waters. Concentrations in groundwater of 0.1 mgll. have been found. There are no guidelines for DCE available in Canada. WHO has set guidelines at 0.05 mgll. as a combined concentration of both cis and trans isomers (WHO, 1993).

1.4 Removal Processes

Current research into the environmental removal ofthese compounds, as well as other chlorinated solvents can be divided into biological and abiotic degradation processes.

Biological degradation, or bioremediation involves the use of microbial or enzyme activity to degrade these compounds. Abiotic processes make use of chemicals such as Fe{O), KMnO.., H10:z.. as well as ultraviolet light (Committee on Ground Water Cleanup Alternatives, 1994). These methods may be appliedin situ,where the reaction takes placewithintheaquifer;orex situ,where water mustbepwnped to the surface for trealmont

Differencesinsite geology, hydrology and level ofcontamination demand certain requirements ofcontaminant removal methods. Information required about the geology oCthe contaminated site includes presence of fracturesinthe rock as well as permeability of the layers of the aquifer. Low permeabilityandthe presence offractures can confine contaminants to an area making treatment more difficult (Committee on Ground Water Cleanup Alternatives. 1994).

The solid media (e.gsandsor days) present in the aquifer mustalso beanalysedto determine the retardation factor (Committee on Ground Water Cleanup Alternatives.

1994). This value quantifies the interactioo betWeen the solid media and the contaminant.

When there is high organic content, organic solvents such as PCE, TCE and DCE will tend toadsorbto these particles. The mobility of the solvent through the aquifer also depends on the hydrophobicity which is quantified by

K....

(Octanol-Water equilibrium constant). The chlorinated solvents used in this study are moderately soluble in water and are therefore quite mobile.

A contaminant plume forms because ofheterogeneites (such as pore size) within the aquifer. along with dispersion (lateral flow) and diffusion (gradient flow) processes.

These processes are affected by the hydrology of the area (rate offlow ofwater through the aquifer). The longer the amount oftime the contaminanthas beenpresent, the larger the area the plume may occupy.

The following sections are a brief description of current research in biological and abiotic reductive dehalogenation of chlorinated ethenes. Other removal methods are available and a description of these maybefoundin the book published by the Committee on Ground Water Cleanup Alternatives (1994).

1.4.1 In Situ Bioremediation

In situ bioremediation processes have been the focus of many laboratory and pilot scale

experiments. The main advantage of this type of remediation, is that there is conversion of the contaminant to harmless products. Site geology and intrinsic microbial activity are important parameters to consider for bioremediation.Asdescribed in the previous section, the mobility of contaminants through an aquifer may limit the efficiency of removal. The presence of appropriate populations of microbes must also be assessed.

Nutrients and sources of energy may need tobeadded to the subsurface in order to enhance microbial activity.

Several studies have shown that TCE, DeE and vinyl chloride(VC) can

be

degraded by methane-utilizing bacteria (methanotrophs) in a process known as cometabolism (Schaffneret aJ., 1996; McCarty et ai., 1991).Incometabolism, bacteria expend energy to oxidize a substrate (in this case, chlorinated ethenes) with no added benefit to the microbes.

Barrio-Lage et aI. (1986) and Moore et aJ. (1989) found that methanotrophs degrade DCE isomers. The former used anoxic conditions, and determined that processes otherthan reductive dehaJogenation were involved in the breakdown ofOCE.Itwas also determined that DeE was converted 10 VC only, a known carcinogen. When oxidizing conditions were used (Moore et aI., 1989), DeE isomers were mineralized 10 carbon dioxide.Itwas also observed that by the addition ofmethane, oxygen and other nutrients degradation rates were improved.

Methanotrophs are also able to oxidize TCE using methane-monooxygenase (MlVIO; an enzyme) (Anderson, 1993). MMO is inactivated following reaction with TCE, which is a drawback of thistypeof remediation. Methane, oxygen as well as nutrients must also be addedinorder to enhance microbial activity (Comminee on Ground Water Cleanup Alternatives. 1994).

Biological reductive dehalogenation is typically the process by which more halogenated species are removed. For this process to be effective, it is important that the chlorinated hydrocarbons be the only electron donors present (Scha.ffuer et aI.1996).

peE was observed to undergo reductive dehalogenation to ethylene, a harmless product (also a plant hormone) by Freedman and Gossen (1989). Methanotrophs were believed to be the active microbes in the removal ofPCE.Ina study by Major et al. (1991), it was

foundthatmethanotrophs may Dot be the only microbes responsible for PCE removal.

The low population of these microbes andhighconcentration of acetate demonstrated the possibilitythatacetogens (acetate-utilizing bacteria) may play an active role in PCE degradation.

Because the potential for reductive dehalogenation decreaseswithdecreasing number of chlorine atoms. some investigators have focussed on the use of two stage reactors for the complete degradation ofthese chlorinated solvents(FathepureandVogel. 1991;

SchaffiJ.eretat.1996). Thefirststage uses anaerobic conditions to degrade PCE and TCE; the second stage uses aerobic conditions to enhance degradation of the less saturated chlorinated ethenes.

The half lives for these processes depend on conditions used. FathepureandVogel (1991) observed in their two stage reactor over 90"10 removal ofTCEinless than two days (initial concentration ofPCE was-100J.LgIL).Mooreel at.(1989) conducted experimentsina simulated aquifer and found almost complete removal oft-DCE in approximately100 hours (initial concentration cfDCE was 1 mgIL). Fliermanset 0/.(1988)conducted experiments using a mixed culture (aerobic) over a period of 1-2 weeks. and observed 99% conversion ofTCEtocarbon dioxide.

There are several drawbackswiththe use ofbioremediation processes. Water quality of

the aquifer maybeaffected by the addition of nutrients or theincreasedmicrobialactivity (Committee on Ground Water Cleanup Alternatives, 1994). These factors mayintum dissolve metals from the surrounding bedrock. Bacterial metabolites mayalso affect the taste ofdrinkingwater. Finally, most contaminated sites usually have morethanonetype of contaminant Palumboet al. (1991) observed that the presence ofOCE and methylene chloride in a contaminated aquifer affected the degradation rates ofTCE.

1.4.2[,,-SituChemicalRemediation

Abiotic removal of chlorinated ethenes by hydrolysis or dehydrohalogenationhaslong halflives (O.85-2.lxIO^IOyears for OCE, O.49·l.3xI0⁶years for TCE, and 3.g-9.8xlO' years for PCE) (Barbee, 1994). These compounds are therefore quite stable innatural waters. Added chemicals, however, can react with the contaminantsand convert them to harmless products. The advantage of chemical remediation is that unlike bioremediation processes, the oxidizing and reducing agents are non specific and cantreatmixturesof contaminants (Committee on Ground Water Cleanup Alternatives, 1994). Chemical additives may also react with substances that cannot be biodegraded. The main drawback to using this method is that water quality (such as pH, dissolved metals, redox conditions) can be alteredwiththe large amounts of reactant injectedintothe subsurface.

Chemical additives maybeaddedto aquifers either by injection. or by installing reactive

barriers (Committee on Ground Water Cleanup Alternatives, 1994). (Figure 3). The fonner involves injection of the reactants to the subsurface via pipes or wells. Reactive barrier technology is still in developmental stages. Funnel walls are built to direct the hydraulic flow in the aquifer to a penneable reactive barrier (sheet or slurry) made of a reactant of choice. This type of remediation may be more easy to design and maintain than other methods.

a)

~·IIItI···~SurfaceSoiI

UnsaturatedZllne b)

Funnel:

Irnpermeable Wall

Figure3: Application of Chemical Agents for Abiotic Remediation (after Committee on Ground Water Cleanup Alternatives, 1994). a) Injection of reactantstocontaminated water.b)Reactive Barriers. Two impermeable walls act as funnels, todirectthe flow of water towards the permeable reactive wall.Onthe other side of the wall. the contaminants arc cODvmed to harmless products.

1.4.3 Reductive DehalogenationusiDeZero Valent lrou

Oxidizing (such as H101or OJ or reducing agents (such as Fe) are commonlyusedfor the abiotic removal of chlorinated solvents (Committee on Ground Water Cleanup Alternatives,1994).The discussion belowwillbe limited to the use ofzero valent iron for the reductive debalogenation ofchlorinated solvents.

Iron can be applied in a number ofways to the contaminated site: by the use of reaction vessels with pwnp and treat technology; reactive barriers; or colloid injection. The advantage of using iron metal is that it is a low cost material. Industrial waste ironfilings can be used to treat contaminated groundwater. Reductive debalogenation with iron works well with saturated chlorinated hydrocarbons such as TCE or PCE, as well as other contaminants such as dyes, pesticides and heavy metals (Wilson,1995;Tratnyek,1996).

Pumping is not required for this technology, allowing for relatively low operation and maintenance costs (Wilson,1995).This method can alsobeusedincombination with bioremediation.

The main disadvantage is that water quality must be carefully monitored or the reaction may not go to completion. Tratnyek (1996) found that the presence of oxides and carbonate ions on the surface of me iron may stop the reaction. As a result, toxic chlorinated products, such as vinyl chloride may accumulate.

The ability of iron to degrade organochlorine compounds was originally discoveredinthe seventies. These results were reconfumed more recently by Reynolds etaJ.(1990) wben galvanized and stainless steel grotmdwatcrmonitoringequipment reacted with cbJorinated solvents. The basis oftbc:se reactionsiscotTOsion chemistry, andthissimple concept sparked a great deal of interest and research. Many stUdies have investigated the kinetics oflhese reactionsinorder tominimizethe reactiontime,as well as to determine the extent to which these reactions go to completion.

There are two pathways for which reductive dehalogenation ofchlorinated ethenes by iron can occur (CampbeU, 1997). Reaction I shows the steps for hydrogenolysis. The last halfreaction can be replaced by Reaction 2,

p..el.im.ination.

Figure 4 is a schematic of the reaction pathways. Products from the reductive dehalogenarion of chJorinaled etbenes include:ethene,ethane and acetylene, as well asC)and C. hydrocarbons.Ina stUdy by DengetaJ.(1997), it

was

foundthatasmallpercentage ofcarbides presentinthe iron may contribute to baclcground coocentration of aliphatic compounds.

2FeO -2Fe'·+4e- 38,0 - 3R"+30g-

2N"+2e- -8, (1)

X-Cl+

n" ..

2e- -X-g+

cr

C\...JCI C ( \ C I

CI-C-C-Cl __-_~-

H CI

H>-\'

~

(a) correspond$toreductlYe~imination:

R(X}+R(X) +2e"::> R=R +2X"

(b) correspondstohydrogenolysis RX+2e"+~0'RH+X·

(el corresponds to reduction of a triplebond to a doUble bond RiER +2e'"'2W0'R{H)=R(H)

(dlcorresponclstoreductionof triplebond 10single R",R ... 46'+4H"0'R(H, )-R(H,)

Figure4: Reaction Pathways for the Reductive Dehalogenation ofCblorinated Ethenes (after Campbell eraI.,1997). Xinthe reactions represents the halogen,inthis case, Cl.

RC1=RCI+2e - - R!!R +2Cl- (2)

There arethreepossibilities as to which species is involved in the dehalogenation: Fe(O), Fe²",or H2.Gillham and O'Hannesin (1994) confirmed that the reaction was indeed taking place on the surface of the zero valent iron particles. Several studies have also shown that the rate of the reaction depends on the surface area of the iron available to the chlorinated species (Gillham and O'Hannesin, 1994; Orth and Gillham, 1996). Johnson el a/.(1996) detennined using previously published data, that the total surface area, as well as the reactive surface area were important parameters.

First order kinetics with respect to the hydrocarbons are involvedinreductive dehalogenation. Johnsonet aJ. (1996) also found that the reaction was first order with respect to the surface area of iran. Sorption processes also playa significant role in the disappearance ofTCE and PCE in aqueous solutions (Burriset a/., 1995). Orth and Gillham (1996) hypothesized that TCE actually remains adsorbed to the iron particle until it is completely dehalogenated. The basis for this assumption was the fact that there was low concentration of chlorinated hydrocarbon products.

The reactions maybeenhanced by doping iron with other metals such as nickel

23

(Appleton, 1996) orpalladiwn (Muftikian, 1995). Pilot scale studies were done using nickel plated iron, with promising results. Thinner reactive barriers were required using the doped iron (10 inch thickness for iron alone as opposed to only 2 inches for the nickel plated iron), and the reaction proceeded ten times faster than for iron alone. Bench scale experiments were conducted by Muftikian et al. (1995) using palladiwn coated iron.

Halflives were less that one hour for TCE.

1.S Isotope Chemistry

Along with the treatment of contaminated waters, it is also important to trace the source of the pollutant, to prevent further contamination, or to charge those who fail to uphold environmental laws and guidelines. Organic pollutants will have specific stable isotopic signatures, depending on their source or mode of fonnation. Compound specific isotopic analysis (CSIA) can be commonly used to allocate sources using the stable isotopes of either carbon, hydrogen, oxygen, nitrogen, chlorine or sulfur. It has been found, for example, that chlorinated solvents from different sources have distinct carbon and chlorine isotopic signatures (van Wannerdam et al., 1995). The variations in isotopic compositions ofPCE and TCE from different manufacturers were presumed to have been caused by the processes used to manufacture the chlorinated solvents (van Wannerdam et al., 1995). These variations apparently arise due to variations in synthesis substrates or during the manufacture of the solvents.Inreductive dehalogenation, the C·Cl bond is

brokenandbeDce isotopic fractionation may be observed for isotopes ofcarbon and chlorine. Forthisstudy,the two stable and most abundantisotopesofcarbon, LlCand

°c

were of interest. Fractiooarion will be most evident betweenisotopestha1are most abundant.

Reactions, such

as

photolysis. oxidation or reduction,that

occur

in thenatural environmentcanchange the isotopic composition of poUutants. presenting complications tosimple tracing or apportioning ofsources. Giventhatthe isotopic consequence of biological and abiotic reactions

can

besmdied,CSIAcan also be used to study the mechanisms involved in biological or biotic transformations of compounds of intc:re:st.

Transformation ofpoUulants in the environment can then be predicted, using the information on the direction and magnitude ofilie change in isotopic composition from experimental data.

There are two main processes that can cause isotopic fractionation: isotope exchange reactions and lcinetic processes. The former involvestheequih"briWDisotopicdistribution between different compounds, phases or molecules. The latter depends on reaction rates of isotopic molecules.

1.5.1 Isotope Exchange

A general fonn of isotopic exchange reaction can be represented by the following equilibrium:

(3)

where A and B are different species, and the subscripts represent 1 and 2 represent the lighter and heavier isotopes, respectively. A common example is the exchange of¹⁸0in the equilibrium between carbon dioxide and water:

The ratio of heavy to light isotope in a compound is represented as:

(4)

(5)

For the stable isotopes of carbon, the ratio would beBC/12

e.

The fractionation factor ex is defined relates the isotopic ratio of two compounds in equilibrium:

(6)

whereR~and R_Bare the isotopic ratios of two compounds definedinequation 3.

For reaction 4. a would be:

(7)

The value

a

is defined as the isotopic composition of a compound:

a.. ,. (~ -

1)x10'

R~ (8)

whereR,.is the isotopic ratio of a standard, and 6 is expressedinper mil~/OO>.The stable carbon isotope composition is represented as 6^1lC.Thereferencestandardthat is usually used for carbon isotopes is PeeDce Belemnite (PDB).

1.5.2KineticEffects

Kinetic effects are observed in fast, incomplete or unidirectional reactions such as dissociation reactions. These effects arc quite important, since they provide information about the mechanisms of chemical reactions (Hoefs. 1987).

Reaction 9 is an example of a rearrangement reaction studies by Bartholomewet al (1954). The rate determining step is the dissociation of the bond between C·Clinterr-

butyl-chloride and itwasfound that »CI reacted faster thanneI.

The dissociation energies of molecules containing the heavy isotope are higher than those containing the lighter isotope (O'Neil, 1986). For example, bonds between llC_CI are more easily broken then llC_CI (Figwoe 5) This can be more easily understood byusing the hannonic oscillator model for the vibrational frequency v ofa molecule:

(10)

where, k is a force constant, and IJ. is the reduced mass for a diatomic molecule made up of masses m_land m₂andisgiven by:

m,m₂

.. ^{- -}

m ....'"2

(11)

Ifm_lis the lighter isotope of the element of interest (for example,l2q,it can easily be seen that ifmlis replaced by the heavier isotope,IJ.will increase and the vibrational frequency will decrease (k does not change). Since v is lower for the heavier isotope, the bond dissociation energy is greater than for the lighter isotope. Figme 5 is a schematic

Interatomic Distance

Fipre 5:Schematicofthe DifferenceinDissociativeEnergies forIsotopic Species (after Hoefs. 1987). The IlC·Cl bond is much more stable and has a higher dissociative energy than theI~C-C1bond.

representation of this. The implication is that a reactantwillbecome isotopically heavier during the course of a reaction. as the molecules containing the lighter isotope are consumed first.

Kinetic isotopic effects have been found to depend on the activation energy for the reaction(Buistand Bender, 1958). For example, in Buist and Bender (1958) the ratio between reaction rate constants,k1t'k1.was detennined for different reactions suchas decomposition of substitutedureasand reaction of methyl iodide and substituted tertiary amincs (wherek_l1represents the reaction rate constant where the bond withUCis broken, and similarly,kl•represents the reaction rate constant where the bond withI·Cis broken).

These ratios were plotted with the activation energies of the different reactions. The higher the activation energy, the larger the isotopic effect observed (i.e k12/kHincreased, indicating that the reaction with the lighter isotope was faster). This also showed that there is a strong dependence of the isotopic effect on temperature (since activation energies are temperature dependent).

Chapter 2 - EIperimentai Procedure

TCE (99.5%; Spectrophotometric Grade), PeE (99.9%; HPLC Grade), and DCE (98%

mixture of c-DCE and t-DCE) were

an

purchased from Aldrich Chemical Company. Prior [0useinlhereactions, Nanopure wale[' (17.5Me)was boiled to remove anymicrobial activiJ.y. Electrolytic iron was purchased from Asher ScientificInc.1be iron was cleaned withrnelhanol forthirtyminutes (Roberts..pers.comm.),thenwith 1M HQsolution for1 bour,rinsedwithNanopure water (17.3Me)and dried under nitrogen.Thecleaned iron was stored under nitrogen until use.Thesurface area ofthe iron was 4.17 m²/g, determinedbyBETanaI)'is.

Forthe reactions, 254 ml capacily bonJes wereusedwithMininen valves.. All glassware used in these experiments was washedwithmethanoland heated in a 100 "C oven until

Four replicates

were

prepared for eacb of TCE. PCE and DCE solutionsinwaleI'.

Volumes of2~l,15~1and 15~lofTCE.PCEandDCE, respectively were measured using a 10~1(±l'ii)syringe (Hamilton). The volume of water used was 245ml,andthe displacement of iron was 3 ml, allowing for 6mlof beadspace. Using thedensitiesin Tables lA and lB,theconcentrations calculated for TCE, PCE and DCE (combined isomers) were 12 ppm. 99 ppm and 76 ppm. respectively. Vapour pressures calculated

from Henry's Constants are given in Table 2.

InthreeoCthe vessels, 20goCiran was placed.~fourth vessel was a blank solution containing onlythe analyte of intereSt and water.Amagnetic stir bar (displacement of2-3 ml)was added toeachblanktocompensatefor thedisplacement causedby iron.Eachof the vessels

were

placedon a Burrelwristaction shaker.Thetempemure at whichthe reactions took place was 20 °C.

Theheadspace of the solutions

was

sampled using solid phase microextraCtion (SPME) using a polydimethylsiloxane (PDMS) fiber.TheSPME assemblyismade up of a coated fiber and a modified syringe (Figure 6). This simple extraction method was developed at the University of Waterloo (e.g. Cbai eraL,I993).Thefiberisexposed tothe beadspace untilthe analyteS equilibrate between the fiber and the sample matrix. For example, in the experiments for this StUdy, 20 minutes wasrequiredfor TCE. PCE and DCE inthe beadspace toequilibratewitha PDMSfiber(Hunt. 1996). Nofurtherpreparationis requiredandthe analyteS are tranSferred directly into the injector of a gas chromatograph (GC) where they thermally desorb from thefiberatthe injec1.or temperature of (250 0C).

Atthe injector, the desorbed compound.iscryofocused (350c) ontothecolumn for one minute prior to rampingthe column temperature.

Table 2: Vapour~for TCE. PCEandOCECakulaIedfromHenry's Constants

ProM

^{TCE PCE t·OCE· c·OCE·}

. .iume (PQ 2-0 15.0 6.7.5 8.25

mass<&> ^0.0029 0.024 0.0084 0.010

moles 0.000022 0.0001.5 0.000086 0.00011

concentration of solution 0.091 0.60 0.3.5 0.43 (moVm^l)

Pressure(atm.) 0.00093 0.0089 0.0024 0.0014

• Itwas found from GC.c·IRMS. usingthepeakareasinthe chromatograms. that the DCE solutionismadeup of approximately 55% c-DCEand 4.5% t·DCE.

- plunger cap

plunger --....,....

I

~----syringe

barrel

- needle

Flgure 6: SPME Assembly(afterChaittaL.1993).Theplunger capispushed andthe fiberisexposedtothesample.

ExperimentSin this study confirmed results from previousstUdies(Hunt. 1996;Diasand Freeman, 1997)that minimalcarbon isotope (<0.5%0> fractionation occurred during SPMEfibersampling. To determine iftbere was isotopic fractionation duringthe SPME sampling, each ofthesolvents were analyzed on a VGOptimamass spectrometer by dual inlet conventional analysis. to determinetheisotopic composition oftbe bulk SOlutioIl 11lis entailedreactingthe solventwithprecombustedCuOin evacuatedsealedtubes,at 500 °Ctoproduce COlandCuClTheCOlwas cryogenicallypurified, using a vacuum lineandthenanalyzed onthemassspectrometer for61~.The collectedgasisanalyzed directly on a mass spectrometer. Tbe mass spectrometerhasa reference and a sample bellows. orinlets.During

a

run,the bellows

are

alternated (opened and closed)

so

that eachsideisanalyudmultipletimes (for a more accurate isotopic composition measurement). WhentheCO:!entersthemassspectrometer. it is ionized anda magnetic field is used to separatethedifferent

masses

44. 45, 46 amu, representing1~01'"COl and1~leOI'O,respectively.Themasses are separated because differencesinmomentum wbichcauses adeflectionin themagnetic field.The lighter ions(iemass 44 amu) are deflectedthemost. Three ionbeamsare creaced and eachbeamis "collected" by Faraday cups.Whenthe ions strikethe cups. a current is produced whicbisconverted to isotopic compositions on by a computer. Tbe ratio of masses 45:44isused tocalculated61~.

TheresuItsobtained from measurements ofthebeadspace ofthestandard solutions using SPME

were

comparable to those from conventional analysis withinthe precision of gas

35

cbromalOgrapby-combustion-isotope ratio mass spectromeuy (GC-C-IRMS) measurements(0.5⁰/00>.

Tests were also doneinorder to determine whether there

was

a differenceinthe stable carbon isotope composition between the vapour and the liquidphasesinthe reaction vessels. It was found that there wasminimaldifference (<0.5⁰/00>between the beadspace andthesolvent solutioninboththe blank (withoutFe~and reaction vessels.Sampling the beadspace was preferredsincethe SPME tiber canbedegraded by the iron powder suspendedinthe solutions.

Compound.specificisotopicanalysis(CSIA) was used to foDowthestable carbon isotope composition(6^1JC with respect toPOB)ofthesolventduringthereaction.using gas chromatography-combustion-isotope ratio mass spectromeuy (GC.C-IRMS) (Figure 7).

TbeGC was a Hewlen Packard HP5890Series IV.andtheIR-MS

was

a

va

Optima.

Thecolumn used intheGCwas

a

Restek:RTX502.2100

m

column (Chromatographic Specialties Inc.).!becarrier gas was He at acolumn headpressureof 12psi.The temperature program andallother parameters are summarizedinTable 3. Identification of some of the products ofthereductive dehalogenation ofthechlorinated eIhenes was doneusinga Varian Saturn 3gaschromatograph-mass spectrometer (GC-MS),wherea fused silicacapillarycolumn wasused(30 m; DB 624; Chromatographic Specialties Inc).

Flgure 7: Schematic of GC-C-IRMS

Table 3: TemperatureProgramfor TCE. PCE and DCE

TeE

injcx::m" 250 "C initia1tempe1'alUl'e 3S"C

hold

''''''''''

2O"Clm.inutoe

---

^{250 "C}

hold 10 minutes

rola! 21.7SmiDutt:s

PCE

~<r 250 "C initialtemptntute 35 "C

hold 1 minute

25 "ClminuU:

---

^{250 "C}

hold 13 mimnes

rola! 22.6 minutes

DCE

Injoaa 2OO'C

initia1~ature 3S"C

hold 1 minute

2O"Clminnte

---

^{250 "C}

hold 7 minutes

rola! 18.75minutes

The sample injected intolheGC

can

beanalyzedeither by IRMS or byfIameioniution detector (FID). ForMSdetection. the compoundsarefirstseparatedon the column and then cornbusted at 800 ·Cinafurnace and converted to COzbyreactionwithCuD.TIle gasisthen cooled to-100 °C. to remove water. The change over (C/O) valve switches between referenceCO₂gas andthe gas from the combusted sample (see Ftgure 7). During a

run.

sample (eluting from the column) or reference gas can beanalyzedat pre- programmed intervals.. The carbondioxidegasis analyzedin themassspectrOmeterasin dualinlet analysis (see above).

Before theexperiments wereinitialed.severaltests were dODe to ensure thattheobserved changes inpeak.area(inthe chromatogram) and isotopic composition were in fact due to the reaction and not due to any other processes. Iron was tested both cleaDed and uncleaned and solutions containing only water and ironwerealso prepared.

Thebeadspaces of the reaction vessels were sampled duringthe course of the reaction. A typical

run.

consisted of samplingthebeadspace using SPME for twenty minutes.and then allowing the analyteS to desorb intheinjector forten minutes. (It was found thatthiswas sufficient time for desorption). Data collection was started. with injection ofthe sample.

During the analysis, referenceCO2

was

used(1l'3t:= -17.5 %0) to ensure that the instroment

was

performing properly.The software that accompanies the VG Optima calculatedthe isotopic composition ofthemajorpeaksinthe chromatogram. relative to

39

the

POB standard. The reaction was allowed to proceed until the reactantpeak

was no

longer detectable by the MS (-1 week for

TeE,

-2 weeks for PCE and DeE).

Chapter3·IlesuIts

3.1 pH

Duringthe

course of

the

reaction.

changesinpeakarea and isotopic composition were recorded..Physicalobservations noted were gas production{HJandtheformation of yeUow-white precipitate (presumably, Fe(OH)]).Thereaction pH was nOt monitored dwingtheindividualexperiments to avoidthepossibilityof reaction product loss or introduction of other experimentalartifacts. However,(1)the beginning andfinalpH weremeasuredinthePeEexperiments.and(2) pHwas monitoredin''blank'' (metalFe and water only) experimentsto identify the range of pH values likely encountered inall experimentS.Thelatter was performed ontheassumptionthatthereaction of water and mew Feis thedominant influence 00 solution pH. TheinitialpH ofthedeionized water was5.5.and the pHofthe blanksolutions remained constant at 4.5. ThefinalpH ofthe PeE solution was 5.2. Analternatephase.Fe(OH), was notedinmost ofthe experiments (see Reaction12).Vaillancourtetai.(1997)alsoobserved the precipitation of Fe(OHh duringthe reaction of metal Feinwater under similarlyacidicconditions. No precautions were taken to ensurethai:tbereactions would takeplaceunder an oxygenfree atmosphere. GiventhepHachievedinthese experiments(pHof4-6)and the presence of Fe(OH)"itis likelythatoxygen

was

present eitherindissolved forminthewater. or as oxides of iron. (NOTE: Under completely reducing conditions. Fe(OH)2 would

precipitate;GiIIbamand O'Hannesin, 1994).1becontinued formation ofFe:z...and electrons fromtheiron, alongwiJ.h theprecipitation oftbeferric hydroxide(anddissolved oxygen)in thesystem, provided a"natural" buffer so lhatthepHdidnotdrift.

(12)

3.2Iron

Inorder to confirm that the observed changesand thereaction products weretheresuhof the reductive dehalogenation reaction as opposed[0thefonnation of organic compounds from residual carbides presentin theiron(Dengerai.,1997), testswere done on iron- water blanks (no solvent). No organic: reaction productsweredeteCtable. Tests werealso done using bothcleanedanduncleanedironandTeE. Organic reaction products could notbedetected andthere

was

no shiftinthe isotopiccomposition of TCEusing the uncleaned iron.Athinorganic coating appears to havebeenpresent ontheFemetal powder as suggested bythefact that organic solvent washing readily produced reactiveFe surfaces(Robens,pets.comm..). Imerestingly,the lackof reaction betweentheuncleaned Fe andthebaloge.oatedcompounds,inspite of obvious presence ofFe2+inthesesystems, supportS suggestionsmadepreviouslybyothers, lhatthesurface ofFemetal powderis directlyinvolvedintheelectron transfer step (ie.heterogeneousreaction) (Burriser aL, 1995).Theorganic coatingapparently badan inhibiloryeffect onthesurfacereactions involvingtheFemetal powder.

3.3Products

Identification oftheorganic reaction products

was

ratherdifficult.1bemassspectraof theorganic products fromtheGC-MSanalysiscould notbematchedwithlibraryentries withhighdegree of confidence (Saturn GC-MS NIST92Library).It appearsthatC~.C4

andCs hydrocarbons were among^theproducts formed.Itwas confirmed however.that thethreeisomers ofDCEwereobserved fromthereductive dehalogenation ofTCE and PCE (c-DCEand1,1-dichloroethyiene were presentinapproximatelyequalamountS.and

£raCeamountS oft-DCE werealsopresent).Thisresult contrastswithprevious zero valentFedehalogenatioostudiesthatshowedboth1.1 DeE tobegenerally lackingandt~

DCEatvery low abundance (e.g. Campbell et ai.• 1997). Ethene and acetylene

were

typicalproducts observedinother studies(Gillhamand O'Hannesin. 1994. Robens etal.

1996).Theywere ootdetectedin our experiments, althoughtheirpresence cannotbe completelyruledoul. The conditions fortheGC-MSanalysis

were

not suitable for resolving acetyleneandethene.(Tbe.segases mustbeanalyzedat subambient temperatures).

A massbalance isusefulinaccounting for an processesthatare involvedintheremoval of chlorinated ethenes.Itwas not possible to perform isotopic massbalancecalculations for thestable carbon isotopes fromtheGC-C-IRMS data, becausetheproductpeakspresent for

an

thereactions

were

toosmalltobeisotopicallyanalyzed.Forexample. severalof

43

the isomers ofOCE

were

resolvedalsinglepointsin theTCE andPeEreactions.

however anaccuweisotope composition couldDOtbe determined because thepeak areas

were

toosmall.Futureexperime:nt.sshould includeiDcreasingthe coocenuation of the sol'YeDt., possibly focusing 00the productSasthereactionprogressed.

3.4RateCoDsIancs

Thereactionsfor TCE. PCE and DeE.

were

allowed to proceed for 169 homs. 345 boms. and336hours.respectively,untilthe compoundscouldnolODgerbeanalyzedby OC-e-IRMS. For DCE. thetransisomer was no longer detectable after 70 hours.

The plotS of disappearance of each compound overtime

were

fittedtofirstand second order

reaction

rate

cunoes as

showninFigures&-11_Curvefitsforfirstorder kinetics wereca.1cuJaIedbySigmaPlot4.()()Cl usingthefoIlawiDg equation:

A.=

Aoe-

^to

whereAispeak area attimet,

Ao

istheinitialpeak

area.

andIt istherateCOostanL (13)

For PCE. TCE. t·OCE. c·DCE, correlation coefficients (Rl values. where a perfect fitis R¹==1)of0.80.0.99, 0.97and0.99wereobtained forthe.fitto the first order rate equation.ForTeE.,theinitialfour pointswereignored for the fitsincetbcy have a great

deal ofvanation.,andare likely[0representslowapproach to equilibrium between TCE, solid surfaces and beadspace (c.l Robens etai,1996; Campbell erat.,1997). Similarly, fortheregression forPCE,theinitialfour points wereignored.TherateconstantS for the four compounds are showninTables 4A and 4B. Sincethreereplicates were run for each reaction, the standard deviation plotted as error bars are also showninFigures 8-11.

AlsoshowninFigures8-11 are the plots oftbe blank solution analysis. The blanks remainedrelativelyconstant over time, and variations are caused by fluctuationsinthe performance of the GC-C-IRMS. Only one blank (control) was run for each etbene so no error bars could be calculated.

1 2 , - - - ,

• •

• • •

300 200

Time (hours) 100

o+-

~----_----~---.J

o

~gure8: Consumption of c-DCE over Time.Thecirclesrepresentthedatafromthe reactionvessels,andthesquares representthec-DCE blank. (solventandwater only).

• • • •

_ 6

•

J!l

$ J,

i

⁵

'"

~4 '~

~

'-t::

tf.3

Oua..o;,-~

•

1

0 20 40 60

eo

Ture(hours)

Figure ,:Consumption of t-DCE over Time.The circlesrepresent the data fromthe reactionvessels.andthe squaresthet-DCE blank (solventandwater only).

47

12.,---,

10

• •

•

• •

60 80 120 140 160 180

Time (hours)

••

20

o-l---~-~-_-~-_-~_-_-__l

o

F1guft10: 1beConsumption orTCEoverTIlDe.ThecirclesIl:presentdatafromthe reaction vessels, andthesquaresrepresenttheTCE blank(solvent andwateronly).

• •

16

,.

".§'

12

•

"§

.e

¹⁰

~

'"

<{

i!!

'" '"

^Q)

0..

•

• ••

•

•

• • • •

• •

o+--~-~-~-~-~-~-~-~--I

o

50 100 150 200 250 300 350 400 450

Time (hours)

Figure

11:

The Consumption ofPCE over Time. The circles represent the data from the reaction vessels and the squares represent the PCE blank (solvent and water only).

49

Table 4A: First OrderRate CoDStallts for Reductive Debalogenation ofChlorinated Etbenes

Compound Rate halfMe halfMe halflife balflife

C o = (boUB) (hours)- (hours)b (hours)"

(br^l)

PCE

0.00252 275.1 3.6

-300

TCE 0.0086 80.6 8.6 -26

I-DCE 0.015 46.21 252.0 -2'

,-DCE 0.0053 130.78 82.0 -80

as reported byGiDhamandO'Hannesin (1994).

as reponed by Robertst'taL,(1996).

as reported by CampbeDt'taL(1997).

Table 48: NormalizedHalf lives of Reductive Dehalogenation of Chlorinated Ethenes*

Compound balflife (lrm¹ml·^l)

PCE 7665.9

TCE 2246.3

I-DCE 1287.8

c·DCE 3644.8

0.28 0.67 19.7 6.41

Halflives were normalized to 1 m¹/mlswface area as reponed by Gillham and O'Hannesin (1994).

'The datawerealsofitted tolhesecond order rate equation (fable 5);

A=_'_

Ia+

...!....

A,

(14)

where Aisthepeak:area at timet,kistherate constant and

Ao

isthe initialpeakarea.It isbelieved. that other

processes

were affecting the rates of the reductive dehalogeoation of the chlorinated ethenes.Thesecan include: the presence of hydrogengasinthereaction vessel the acidic conditions or the presence of chlorinated reaction products (such as DCE isomers) competing for iron sudacearea.

5l

TableS:SecondOrder Rate CODSWUS forReductiveDehaJogenaUoDofChloriDa1ed E""",,,

Compound Rate COnstaDl halftife(!loUIS) R'

<,;gnaT""")

PCE 0.000324 474.2 0.78

TCE 0.00291 52.7 0.91

I-DCE 0.00329 46.6 0.89

c-DCE 0.00103 149.4 0.98

3.5IsotopicComposition

Largeshiftsinthestable carbon isotope compositionwere observedduringthereaction with eachofthesolvents.Plotsof isotopic compositionwith time areshowninFigures 12-15. The till(:ofPCE.TeE andc-DCEshiftedby24 °/00.12 °/00 and 24°/1»0 respectively. Eachcompoundbecameenrichedwith

1'<:.

t-DCE

was

anomalous.sinceit became lighter first,before becoming isotopically beavier(after2Shours).Therewas somedifficultyinthemeasurement ofthet-DeE isomer.The peakwas not well resolved andthereforethe integration andcalculation oftheisotopic compositionmaynothave beenvery accurate.Thesametrendwas observed for all replicateS ofthisexperiment. as wellas during a preliminary set of experiments. Theerrorbarsshown representthe Standarddeviation oftheisotopic compositioninthethreereplicates foreach time interval

The analysesoftheblanksolutionsare alsoplonedinFigures12-15.Theisotopic compositions oftheblank:solutionsremainedrelatively constant overtime(~0.5

°'00;

note:thisisalso the uncenaintyin any measurement onthe GC-C-IRMS).indicatingthat degradation ofthecbloroetbenes was caused bythepresence ofmetalliciron..Onlyone conuol

was

run.

so

DOstandarddeviation couldbecak:ulated.

53

-, -,.

~ •

.P •

"

^-20

_• • ^•

•

-25

, . • ^{.. . .. .}

-30

•

^{'00 200}

Time (hours)

300

Figure12:IsotopicFractionation of c-DCE over Time.Thecirclesrepresent data from thereaction vessels, andthesquares representthec·OCE blank (solvent and water only).

. .. ⁱ . ⁺

• t

rme{hours)

..

-,,-l---~----~---_---____l

o

F1gure13: Isotopic Fractionation or t·DCE over Tune.Thecirclesrepresentdatafrom thereactionvessels.andthesquares representthet·DCE blank (solvent andwateronly).

-16

^• + +

-18

-2<)

+

.l

^-22

+

Jr' -., .t

-26

t'·

::~ ^• _{•• ••} ^• _{• • • •}

-32

0 ^2<) 40 60 TIme (hours)60 180 120

'"

¹⁶⁰

Flgure 14: Isotopic Fractionation ofTeEover Time.The circlesrepresent data fromthe reactionvessels.andthesquares representtheTCEblank(solventandwater only).

·25 • • • •

•30

~t ... ,. ..- ....

t

t f

200 250 300 350 400 450 500

Time (hoursj

-35+--~~-~-~-~~-_-~-~-

o

50

Figure15: Isotopic Fractionation ofPCE over Time.The circlesrepresentdatafromthe reaction vessels. andthesquares representthePeE blank (solventandwater only).

Chapter 4 -Discussion

4.1PrecisionandError

Theexperimental control samples (blanks) areusedto determine me analytical precision (or reproducibility) of these measurements. lbe errorinthe isotopic composition(10 deviation) ofPCE.TCE.t-DCE and c-DCE controlswere ±O.43D/oo•±O.47 °/00' ±O.40G/oo and:10.25D/I»'respectively.Theexperimeotal errorinme reactioovesselswere PCE.

TCE. t-DCEandc·DCE were ±O.76 0'00> :10.50 G/oo• :10.55 °/00and±O.59°/_00,respectively Instrumental errorwithGC·C·IRMSis±D.5°/00,Since thepeak

areas were

not calibrated..itcould not be determined whether the10deviationisacceptable error. The relative error forPCE.TCE. t-OCE. and c·DCE contrOls were: 15.8%.9.24 %. 5.7%and 6.8'Ai,respectively.Therelative experiments error for thepeakareas ofPCE. TCE. t- DCE.and c·DCE reactiom; were: 11.9%.8.6'Ai.5.6%and 6.7 %. respectively.

Potential sources of errorinthese experiments include variations of temperatureinthe laboratory. isotopic fractionationduringsampling usingSPMEor betweenthevapour and liquidphasesinthereactionvesselsandinstrumentalerror. TestS have shown that isotopic fractionationduringsamplingandbetweenliquidandvapour phasesisminimal.

It isbelieved that the temperature variations may cause variationinthevapour pressure withinthe vesselsand affecttheconcentrations measurements of each compound.

4.2ReactionRates

Previous studies (GilIbamand O'Hannesin, 1994) found that underreduc:iDgcondition1.

the first order reaction rate decreased fromPCEtoDeE.lbe trendsinmeasuredrates of debalogenationandhalflives relative to the decreasingdegreeof halogenation fromPeE, TeE,toDeEinour experimentsareclearly opposite totheobservationsmadebyGiIlham andO'Hannesin (1994). ItisnOlewoIthy that recent experiments performed by Campbell t!!taL(1997) also showTCEdehalogenationratesthataresubstantially fasterthan the correspondingrates fOfPCE.results thatareconsistent with our observations..TIle reversetrend observedinour experiments and Campbell t!!faL(1997)relative to those described byGiIIhamandO'Hannesin (1994)isquiteinterestingastheybearontheissue of efficacy of abiotic remediation of halogenatedetheoes.

Comparison ofthenormalizedhalflives(normalized to1m²/ml surfacearea)inour e~riments.,with

those

published byGilIham and O'Hannesin (1994) and Johnsont!!taL (1996),clearly shows the same trend as the un-normalized values. (ThefasteStrateswere observed forPeE,whereasinour experiments,PCEhad the slowest rate of reaction). It isinteresting to note that despite the muchsmallersurface area oftheiron metalusedby GillhamandO'Hannesin (0.287m²/g), and thelarger surface areaofthe iron used in our experiments(4.17 m²/g),therates of the reactionsinourexperiment were several orders of magnitude slower. This observation maybethe result of a lower reactive surface area

available forthedehalogenation of chloroetbenesinnur StUdy,causedbytheoxidation of tbc:metallicironsurface..Johnson eraL(1996)proposedthatthereactivesurfacearea

was

more ofalimitingfactor thantotalSUIface area inreaction~of reductive dehalogenaJion.Thepresenceof oxygeninour reaction vessels may have oxidizedthe surfaceoftbc: ironpowder. Thereactionbetweenzero vaJenI ironwith thechlorinaIed etbenes

was

biDdered.andas aresult, thedehydrogenation reactiom

were

slowerthan expected.

OnekeydifferenceintheconditionsusedinGillham.andO'Hannesin (1994),Campbell er aL(1997) and our experimentsispH.TheIanertwo

were

either buffered near neutral or drifted toslightlymoreacidicvalues(<6).Incontrast.GiIIhamand O'Hannesin's (994)

~tsdrifted to more basicvalues (>9).Thedi1fereDce inpHtrendsappearstobe

a

resultofthe additionofDe0landtbelackofbeadspace inGillham andO'Hannes:in's (1994) exper1menl Itisunclear bowthisdifference inpHcan potentiallyaffectthe reactivity of balogenated ethenes.The effectof pHdeserves furtherinvestigation,andas willbediscussedbelow,itmay have a strongeffecton tbe reactionnte:sof reductive debalogenation.Ifa pHeffecton reaction rateS ofthedifferent chlorinatedethenescanbe demonstrated infuture experiments,itwillhave implications for tbe field application of zero-valent iron degradation forspecificcontaminant discharges.

Onefinalaspecttoconsiderisbeadspace. Althoughthebeadspaceinour systems was