rcm7220

Improvement of analytical method for chlorine dual-inlet isotoperatio mass spectrometry of organochlorines

Tetyana Gilevska1, Natalija Ivdra1,2, Magali Bonifacie3 and Hans-Hermann Richnow1,2*1Department of Isotope Biogeochemistry, Helmholtz Centre for Environmental Research UFZ, Permoserstr. 15, D-04318Leipzig, Germany2Isodetect GmbH Company for Isotope Monitoring, Deutscher Platz 5b, D-04103 Leipzig, Germany3Group of the Geochemistry of Stable Isotopes, The Institute of Earth Physics of Paris, Sorbonne Paris Cit, Universit ParisDiderot, UMR 7154 CNRS, F-75005 Paris, France

RATIONALE: The development of compound-specic chlorine isotope analysis (Cl-CSIA) is hindered by the lack ofinternational organochlorine reference materials with isotopic compositions expressed in the 37Cl notation. Thus, areliable off-line analytical method is needed, allowing direct comparison of the 37Cl values of molecularly differentorganic compounds with that of ocean-water chloride, to refer measurement results to a Standard Mean Ocean Chloride(SMOC) scale.METHODS: The analytical method includes sealed-tube combustion of organochlorines, and precipitation andsubsequent conversion of the formed inorganic chlorides into methyl chloride (CH3Cl) for the determination of 37Clvalues by Dual-Inlet Isotope Ratio Mass Spectrometry (DI-IRMS). A sample preparation step most sensitive to thesample size dissolution of the inorganic copper chlorides formed by combustion of -HCH was identied.RESULTS: Recovery of 94 5% of chlorine could be reached by applying determined optimal conditions for thedissolution, implying good external precision of 37Cl values (0.18 0.03, 1, n = 3). Validation of the optimizedmethod by the analysis of the produced and initial CH3Cl sample with known 37Cl values vs SMOC resulted in adifference of 0.11 0.04 (1), conrming the external precision and accuracy of the entire method.CONCLUSIONS: The efciency of the sample preparation method for CH3Cl-DI-IRMS analysis is independent both ofthe chemical structure of the chlorinated compound and of the amount of chlorine in the sample. This method has thepotential to be applied to a broad range of chlorinated organic compounds, e.g. reference material for the calibrationof methods for Cl-CSIA against SMOC. Copyright 2015 John Wiley & Sons, Ltd.

Organochlorine pollutants are of great interest to environ-mental scientists due to their persistence, bioaccumulationand frequent toxicity.[1] Isotopic compositions analysis hasfacilitated better understanding of the environmental fate oforganic chlorinated compounds (OCs) by tracing their sourcesand transformation processes.[2,3] Although the determinationof chlorine stable isotope composition (37Cl values) has thepotential to help identify the sources and degradation ofOCs in the environment, its application is mainly limited bythe currently available analytical techniques. In contrast tothe number of efcient high-throughput methods developedand routinely applied for compound-specic isotope analysis(CSIA) of carbon, hydrogen, oxygen and nitrogen, methodsfor the CSIA of chlorine in organochlorine compounds stillneed to be improved for routine applications.[4,5] In theexisting analytical set-ups mixtures of organochlorine

compounds are generally separated by gas chromatography(GC) and then transferred by a carrier gas either to a conver-sion unit for production of the HCl by high-temperaturecombustion (GC-HTC) which is then used for the analysis ofchlorine isotopes,[6,7] or directly to the mass analyzer unit formeasurement of the chlorine isotope ratios.[8,9] Currently,Isotope Ratio Mass Spectrometry (IRMS),[8,10] QuadrupoleMass Spectrometry (qMS)[8,1113] and Multiple CollectorInductively Coupled Plasma-source Mass Spectrometry(MC-ICPMS)[9] are used as mass analyzers for thedetermination of 37Cl values. The majority of the existingCl-CSIA methods demand comparison with molecularlyidentical reference compounds to refer isotopic ratios of targetanalytes to the Standard Mean Ocean Chloride (SMOC) scale,and for the quality control following the principle of identicaltreatment.[14] Thus far organic international standards forCSIA of chlorine are not available. Therefore, the chlorineisotope composition of reference materials must be previouslydetermined by off-line methods which allow direct referen-cing with ocean-water chloride samples.Thermal-ionization mass spectrometry (TIMS) and dual-

inlet isotope ratio mass spectrometry (DI-IRMS) are the massspectrometric methods in current use for the determination of37Cl values directly related to the SMOC scale. As reported

*Correspondence to: H.-H. Richnow, Department of IsotopeBiogeochemistry, Helmholtz Centre for EnvironmentalResearch UFZ, Permoserstr. 15, D-04318 Leipzig, Germany.E-mail: [email protected] authors contributed equally to this work

Copyright 2015 John Wiley & Sons, Ltd.Rapid Commun. Mass Spectrom. 2015, 29, 18

Research Article

Received: 10 February 2015 Revised: 27 April 2015 Accepted: 28 April 2015 Published online in Wiley Online Library

Rapid Commun. Mass Spectrom. 2015, 29, 18(wileyonlinelibrary.com) DOI: 10.1002/rcm.7220

1Journal Code Article ID Dispatch: 08.05.15 CE:

R C M 7 2 2 0 No. of Pages: 8 ME:

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

66676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

by Rosenbaum et al. for seawater samples,[15] DI-IRMS is bestsuited for samples containing >10 mol Cl, allows 37Clvalues to be measured with a precision of 0.1 (2). TIMSis more sensitive and applicable for samples containing~0.10.3 mol of Cl with achievable uncertainties of 0.2.However, these uncertainties do not directly refer to the37Cl values of OCs since both methods require off-lineconversion of organochlorine compounds into inorganicchloride, which then can be processed in parallel withstandard samples of ocean-water chloride for the analysis.Prior to DI-IRMS analyses, chlorine from OCs is rst

transformed into CuCl via combustion with CuO (SchemeS1 1,path a)[16,17] or into NaCl by trapping Cl into a sodiumcarbonate solution after an oxygen combustion bomb(Scheme 1, path b).[18] Alternatively, organic chlorine can betransformed into NaCl via reductive dehalogenation.[19,20]

The obtained inorganic chlorides are then converted intoAgCl (Scheme 1, paths c and d) and subsequently into CH3Clfor DI-IRMS (Scheme 1, path e).[16] Alternatively, CuCl,obtained in the combustion reaction, can be directlyconverted into CH3Cl for

37Cl measurements by DI-IRMS(Scheme 1, path f).[2123]

Different physical and chemical processes during theconversion of OCs could generate isotopic fractionation,which needs to be thoroughly quantied. It has been reportedthat losses in the conversion of CuCl into CH3Cl and duringthe purication of CH3Cl could bias the obtained

37Cl value,leading to a correlation between the recovery of chlorine in thesample preparation procedure and the accuracy of the 37Clvalues.[17,21] Many attempts have been made to optimizedifferent sample preparation steps in order to reach thehighest possible recoveries of chlorine and the reproducibilityof 37Cl values for the entire procedure.[17,22,23] However, noclear relationship between losses at each separate conversionstep and changes in the isotopic compositions has beenreported up to now. In addition, all previously reportedmethods were optimized for specic amounts of combustedchlorinated compound, so the applicability and efciency ofthese procedures for samples containing different amountsof chlorine stayed unrevealed.

In this study we present a sample preparation method priorto CH3Cl-DI-IRMS, which holds the potential to be used forthe determination of 37Cl values of organochlorines asreference material for the calibration of the Cl CSIA methodsagainst SMOC. We selected off-line conversion of OCs intoCH3Cl via CuCl and AgCl, which allows one to obtain isotopicprecisions of 0.15 and high overall recoveries of chlorine(>97%).[16,24] The rst step of the procedure combustion ofOCs to CuCl is applicable to a broad range of organiccontaminants from small molecules, such as chlorinatedmethanes and ethenes,[21,23] to chemically persistent complexmolecules, such as chlorinated pesticides.[17,22]

We have chosen -hexachlorocyclohexane (-HCH, Lindane)as a model compound for the optimization of the samplepreparation method. -HCH and other isomers of HCH wereused worldwide as agricultural insecticides until they werebanned or restricted to specic applications by the StockholmConvention on persistent organic pollutants.[25,26]

In the course of this study, we investigated in detail eachtransformation from -HCH into CH3Cl to determine theinuence on the accuracy of obtained 37Cl values of chlorinerecoveries at different sample preparation steps. The leastefcient steps, leading to changes in the isotopic compositions,were identied and optimized. Our sample preparationmethod is applicable for different sample sizes, as we proposethe optimal water/chlorine ratio for the dissolution step withthe possibility of recalculating the necessary volume of waterfor each particular amount of chlorine. We applied theoptimized procedure to a CH3Cl sample with a known

37Clvalue relative to SMOC. Finally, we compared the obtained37Cl value of the recovered CH3Cl and that of non-processedCH3Cl to conrm the accuracy and precision of the optimizedmethod.

EXPERIMENTAL

Solvents and chemicals

Acetone (99.5%), potassium nitrate (KNO3; 99%), citric acidmonohydrate (C6H8O7*H2O; 99.5%), silver nitrate (AgNO3)and copper(I) chloride (CuCl; 99.9%) were purchased fromCarl Roth (Karlsruhe, Germany). Ultra-high purity (uhp)-water (resistivity of 18 M) was prepared with a MerckMilli-Q A+ system from Millipore (Billerica, MA, USA).-HCH (99.1%) was purchased from HiMEDIA (Mumbai,India); CH3Cl (99.90%, #N30) was obtained from Air Liquide(Paris, France); copper oxide (CuO; >98%) and potassiumphosphate dibasic dihydrate (Na2HPO4*2H2O; 98%) wereobtained from Sigma Aldrich, (Steinheim, Germany). Nitricacid (HNO3; 69%) was purchased from Merck (Darmstadt,Germany). Quartz glass (Schott) tubes (20 mm o.d.) wereprepared at the glass workshop of the Helmholtz Centre forEnvironmental Research UFZ (Leipzig, Germany).

Initial procedure for conversion of -HCH into CH3Cl

-HCH was converted into CH3Cl prior to Cl DI-IRMS. Weup-scaled the method to 250 mol of Cl to ensure reliableDI-IRMS analysis over all optimization steps and tominimize the inuence of a blank on the chlorine sample.The sample preparation procedure consisted of the following

Scheme 1. Chlorine from OCs is rst transformed into CuClvia combustion with CuO (path a)[16,17] or into NaCl bytrapping Cl into a sodium carbonate solution after an oxygencombustion bomb (path b).[18] Alternatively, organic chlorinecan be transformed into NaCl via reductive dehaloge-nation.[19,20] The obtained inorganic chlorides are thenconverted into AgCl (paths c and d) and subsequently intoCH3Cl for DI-IRMS (path e).

[16] Alternatively, CuCl, obtainedin the combustion reaction, can be directly converted intoCH3Cl for

37Cl measurements by DI-IRMS (path f).[2123]

T. Gilevska et al.

wileyonlinelibrary.com/journal/rcm Copyright 2015 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2015, 29, 18

2

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

66676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

steps: step 1: high-temperature combustion of -HCH to CuClin an evacuated sealed quartz ampule with excess of CuO(Eqn. (1)); step 2: disproportionation of Cu(I)Cl to formCu(II)Cl2 and dissolution of the formed soluble CuCl2 inwater (hereafter called dissolution) (Eqn. (2)); step 3:precipitation of chlorides as AgCl (Eqn. (3)); and step 4:halogen exchange reaction of AgCl with an excess of CH3I toform CH3Cl (Eqn. (4)). The obtained gaseous CH3Cl was thenchromatographically separated from excess CH3I and puriedbefore DI-IRMS at the Institut de Physique du globe de Paris(IPGP Laboratoire de Gochimie des Isotopes Stables, Paris,France) following the procedure routinely used to performDI 37Cl measurements (see section 2 of the SupportingInformation).[2732] In step 1 the ratio of CuO to the mass ofCl in the sample was 200 mg CuO per 1 mg of Cl (that is~7.1 mg for 1 mol of Cl); in accordance with Holmstrandet al.,[17] the combustion temperature was 620 C[16,17] andthe volume of water for the dissolution process in step 2 was25 mL.[16] The amounts of chemicals for the precipitation ofAgCl in step 3 were calculated proportionally to the amountof Cl reported by Jendrzejewski et al.[16] (detailed amountsand further experimental details of the procedure can befound in section 1 of the Supporting Information).

step 1 : HCH 24 CuO9 Cu2O 6 CuCl 6 CO23 H2O (1)

step 2 : 2Cu I ClCu Cu II Cl2 aq (2)

step 3 : CuCl2 aq 2 AgNO3 aq Cu NO3 2 aq 2 AgCl (3)

step 4 : AgCl CH3I excess CH3ClAgI (4)

Optimized procedure

The entire sample preparation procedure after optimizationconsisted of the following steps: step 1: organic sample,containing 250 2 mol of Cl was transferred to the quartzampule (pre-heated at 700 C for 1 h), containing 1.8 g ofCuO (pre-heated at 800 C for 1 h). The ampule was thenevacuated to ~1200 mbar, and sealed, while the lower partwas immersed in liquid nitrogen. The ampule wasafterwards heated in the furnace (with a temperatureincrease of 5 C/min towards 620 C and kept at 620 C for1 h), before being allowed to cool to room temperature (r.t.),washed with 10% HNO3 and uhp-water. Step 2: the ampulewas broken and all solid residues together with quartz culletwere transferred to a screw-capped bottle, containing 80 mLuhp-water, followed by vortex-mixing for 1 min and 1 h ofsonication at 50 C. The solutionwas decanted from the solidresidues, ltered through a nylon lter (0.22 m) and theltration residues were rinsed with an additional 20 mL ofwater to obtain a total volume of 100 mL for the rst extract(H2O/Cl ratio 0.4 mL/mol). Subsequently, 80 mL of freshuhp-water were added to the solid residues, the dissolutionprocedure was repeated, nished with washing of theltration residues with an additional 20 mL of water, and thetwo extracts were combined to obtain 200 mL of the nalsolution. Step 3: 20 g of KNO3 and a pH 2.2 buffer (4.5 g ofcitric acid*H2O+140mg of Na2HPO4*2H2O) in dry formwereadded to the obtained CuCl2 solution and heated at 80 C until

complete dissolution of KNO3 and buffer reagents.Afterwards, 3.75 mL of 1 M AgNO3 solution were added.The solution was left in darkness to cool down for 1 h andthe suspension was ltered on a glass microber lter anddried in the darkness at r.t. The obtained AgCl on the lterwas split into 24 subsamples to give approximately 50 molof chlorine on each part of the lter. Step 4: each lter partwas then placed in a borosilicate glass tube, excess of CH3I(100 L) was added, the tube sealed and left at 80 C for 72 hto form CH3Cl. CH3Cl was then separated from excess CH3Iby preparative gas chromatography using two Porapak-Qlled columns and analyzed by DI-IRMS as described byEggenkamp[32] and Bonifacie et al.[31]

Dual-inlet isotope ratio mass spectrometry (DI-IRMS)

The 37Cl measurements of CH3Cl gas were performed usinga triple collector gas-source dual-inlet mass spectrometer(Delta plus XP; Thermo Fisher Scientic, Bremen, Germany)at IPGP. The 37Cl values were obtained by determining thesignal intensity of m/z 52 (CH3

37Cl+) and m/z 50 (CH335Cl+)

using two different collectors with resistances of 1109and 3108, respectively. One measurement consisted of aseries of 10 individual comparisons of the ratio 52/50 in thesample CH3Cl to that of the CH3Cl gas used as a laboratorystandard. The reference gas is compared with CH3Clprepared from seawater chloride at least twice a day, andtypically following each 5 to 6 samples. This procedurechecks for instrumental drift during the day, and allows directreferencing of the 37Cl values of unknown samples to theSMOC scale.[2830] The chlorine isotope composition of eachproduced CH3Cl sample was determined by DI-IRMS twice.

Gas chromatography/mass spectrometry (GC/MS)

The concentration of the organic compounds was determinedby GC/MS to test the conversion of -HCH into copperchloride. Details of the GC/MS method are available insection 3 of the Supporting Information.

Ion chromatography (IC)

IC analyses of chlorine concentration were performed at theDepartment of Analytical Chemistry at UFZ on a DionexICS-2000 ion chromatograph (Thermo Fisher Scientic).Detailed information can be found in section 4 of theSupporting Information.

Chloride blanks

Chloride blanks were analyzed to investigate the total addedamount of chlorine at the dissolution step and through theentire conversion of OCs into AgCl. Two experiments wereperformed, using 1.8 g subsamples of commercially availableCuO. The rst subsample was suspended in 200 mL of uhp-water (as for two subsequent extractions of 250 mol of Cl).The optimized dissolution procedure was then applied andthe concentration of the chlorine extracted from the mixturewith CuO was measured by IC. The second subsample ofCuO was heated in a sealed evacuated ampule andsubsequently extracted with 200 mL of uhp-water asdescribed above for the optimal procedure. A 1 mL aliquotof the ltered solution was taken for the determination of

Chlorine stable isotope analyses of organic compound

Rapid Commun. Mass Spectrom. 2015, 29, 18 Copyright 2015 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

3

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

66676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

chlorine concentration. The rest of the solution was subjectedto the AgCl precipitation procedure. The mass of the driedAgCl precipitates and the chloride ion concentration in theltrate were determined. From these experiments the amountof chlorine which may be added over all steps of the entireprocedure was determined.

RESULTS AND DISCUSSION

Chloride blanks

The chlorine concentration, measured by IC, in both preparedblank solutions was determined as 0.11 mg/L, whichcorresponded to 0.1% of the amount of Cl in the initialorganochlorine compound. Furthermore, there was nomeasurable amount of AgCl recovered on the lter after theprecipitation procedure. Thus, the chloride blanks both ofthe dissolution step (Eqn. (2)) and of the entire optimizedconversion procedure showed no signicant amount ofexternal chlorine.

Initial procedure: efciency and accuracy

The initial procedure for the conversion of -HCH into AgClshowed low recovery yields of chlorine (23 3% (n= 2)calculated from the mass of precipitated AgCl) (TableT1 1). Thisis in sharp contrast to the nearly complete conversion yields

reported by previous studies.[16,17] Such low recovery yieldsof chlorine suggest that the methods previously developedcannot be applied to the higher amount of -HCH withoutadjustments. The measured 37Cl value of CH3Cl, producedfrom -HCH by the initial procedure, was 1.15 0.34(n = 2) (Table 1). -HCH is generally produced by thechlorination of benzene with Cl2 gas, derived from brines,with 37Cl values typically between 0.5 and 0.[30,33]

The signicantly lower 37Cl value of -HCH that weobserved in experiments with only 23% yield, than thevalue of brine, could thus result from chlorine isotopicfractionation due to preferential loss of 37Cl isotopes. Inaddition, the external precision of 0.34 is much largerthan usually achieved for the sample preparation, i.e. fromseawater (0.08, 2), indicating that the samplepreparation procedure used was not optimal under theapplied conditions and for the amount of chlorine taken.Thus, a revision of all conversion steps with respect to the37Cl precision and accuracy was conducted, as presentedbelow.

Revision and optimization of sample preparation steps

The efciency of each individual step (Eqns. (1) to (4)) wastested by determining the recovery yields (dened as therecovered amount of Cl, measured by weighing AgClprecipitates or determination of the chlorine concentration

Table 1. Optimization of sample preparation method with -HCH

Optimization parametersWater ratio,mL/mol Cl

Chlorinerecoveries, %

37Cl valuea

(n = 2),

Average 37Clvalue from multiple

ampulesb,

Initial procedure 0.10 22.2 1.39 1.150.34 (n = 2)21.6 0.91

Sonication at the dissolution procedureVortex mixing for 1 min 0.06 1Sonication for 2 h 0.06 32Sonication for 1 day 0.06 38Sonication for 2 days 0.06 34Repeated extraction 0.10 57.1 0.60 -0.630.10 (n = 2)

44.9 0.70Enhanced oxygen supply 0.06 31Elevated temperature 0.06 49Water volume 0.28 53

2.00 695.00 73 0.12 0.170.08 (n = 2)

76 0.22Conversion of -HCH under optimized conditions

82 0.291st extraction 0.40 93 0.25 0.260.02 (1,n= 3)

90 0.257 0.66

2nd extraction 0.40 6 0.89 0.920.28 (1,n= 3)6 1.22

Combinedc 94 5 0.180.03 (1,n= 3)aDifference between duplicate measurements of the same prepared CH3Cl sample was typically below 0.02.b external precision of sample preparation procedure from different subsamples of -HCH.cCombined extract represents the sum of recoveries from two subsequent extractions and calculated 37Cl values, based onthe mass balance.

T. Gilevska et al.


4

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

66676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

in the solution by HPLC/the expected amount of Cl). Thepotential impact of these losses on the obtained 37Cl valueswas then taken into consideration.Step 4 transformation of AgCl, obtained from seawater

standard solutions, into CH3Cl (Eqn. (4), following exactlythe protocol of Jendrzejewski et al.[16] and routinely used inIPGP[2830] (Supporting Information, section 2)) proved tobe complete (overall yields close to 100%, including thepurication of CH3Cl from CH3I by preparative GC) and thusshould not inuence the accuracy of the isotope values in theinvestigated sequence of steps.Therefore, for the further revision of the previously

suggested procedure and its adaptation for large samples of-HCH, we took into account the following parameters: (1)incomplete combustion of -HCH in step 1; (2) incompleteprecipitation of AgCl in step 3; (3) inefcient dissolutionprocess in step 2.

(1) Incomplete combustion of -HCH may lead to theformation of partly dechlorinated organic by-productsassociated with lower yields of obtained CuCl andsubstantial changes in the 37Cl values of CH3Cl. We testedthe conversion of -HCH into CuCl by combusting -HCHas described in the initial procedure. After reaction, theampule was slowly cooled down to +5 C, then crackedand the copper oxide particles were dispersed in 10 mL ofacetone to dissolve any unreacted organic materialpotentially remaining in the ampule. GC/MS analysis ofthe residues after combustion showed that 0.01% of theinitially loaded compound was recovered, conrmingcomplete conversion of -HCH into inorganic reactionproducts. No traces of other organic products of -HCHdegradation were detected. These results were in the goodagreement with reports on complete combustion forDDTs[17] and polychlorinated hydrocarbons[16,34] andconrm that the selected combustion conditions aresuitable for structurally different OCs.

(2) Efciency of the AgCl precipitation procedure (Eqn. (3))was tested by applying it to a solution of NaCl in 25 mLof water with chlorine ion concentration equal to theCuCl concentration after 100% conversion of 250 molof Cl from -HCH. A test of the precipitation efciencyresulted in 36.6 mg of AgCl being precipitated from theNaCl solution, corresponding to 102% (the efciencyhigher than 100% could result from the uncertainty ofweighting NaCl and/or AgCl). This completeprecipitation of chloride is reinforced by the fact that notraces of remaining dissolved chlorine were detected byIC analysis of the ltrate after precipitation.

(3) Dissolution of CuCl. After conrmation that both thecombustion and the precipitation processes are completeat selected conditions, both in agreement with previouslyreported results for entire procedures,[16,17] we identiedthat CuCl2 formation and dissolution is the least efcientstep of the conversion of OCs into AgCl. We testedseveral modications of this particular step: (i)sonication, (ii) repeated extraction, (iii) elevatedtemperature, (iv) oxygen supply, and (v) water volume.

(i) As previously suggested by Jendrzejewski et al., insufcientdissolution time may lead to incomplete extraction ofwater-soluble chlorine species.[16] Thus, introduction of a

sonication step allowed us to signicantly increase thetotal recoveries of chlorine in the dissolution process andto shorten the overall time needed for this step. Wedetermined the concentration of dissolved chlorine in thesolution after 1 min, 2 h, 1 day and nally after 2 days ofsonication, corresponding to 1, 32, 38 and 34% of theinitially loaded chlorine, respectively (Table 1). This testsuggests that increasing the sonication time does notsignicantly increase the chlorine recovery yields.

(ii) Subsequently, we tested if the yield of the dissolutionprocess can be increased by repeated extraction. Aftercombining two extracts and washings after 2 h ofsonication and subsequent ltration, 51% of the chlorinewas recovered in precipitated AgCl and then convertedinto CH3Cl (Table 1). A 13% increase of the efciency inthe combined extract relative to the result achieved in theprevious experiment by a single extraction after 2 h ofsonication showed that it might be necessary to repeatan extraction from the same solid residues aftercombustion. DI-IRMS analysis of combined extractsshowed an elevated 37Cl value of 0.63 0.10 (n=2)compared with the value of 1.15 0.34 (n=2) obtainedthrough the initial procedure and associated with morethan two times lower overall yields (23%). These resultssuggest that the isotopically heavier fraction of chlorinecan still remain undissolved in the residues after anincomplete dissolution process, leading to biased lower37Cl values of extracted chlorine. This observationconrms that it is crucially important to achieve completetransformation of the dissolution process for the accuratedetermination of the 37Cl values.

(iii) During the dissolution step (Eqn. (2)) the water-insolubleby-product copper oxychloride (Cu3Cl2(OH)4) can beformed by the oxidation of CuCl by atmospheric oxygenin the presence of water (Eqn. (5)).

4 CuClO2 2 H2OCu3Cl2 OH 4 CuCl2 aq (5)

On the other hand, copper formed in the disproportionationreaction at the dissolution step (Eqn. (2)) may be thenoxidized with oxygen, shifting the reaction (Eqn. (2))equilibrium towards the formation of the desired product(water-soluble CuCl2). Thus, by introducing a gentle oxygenow for 2 h into the CuOCuCl water suspension aftercombustion, we tested whether an enhanced oxygen supplyaffects the prevalence of the disproportionation process (Eqn.(2)) over the competing oxidation reaction (Eqn. (5)) to obtainmaximum possible recoveries of chlorine. An enhancedoxygen supply caused only insignicant improvement ofthe efciency of dissolution, allowing us to recover 31% ofchlorine in the 15 mL of the extraction solution (Table 1).Thus, the negative effects of the possible side-reaction (Eqn.(5)) were negligible or compensated for by the improvedkinetics of the dissolution (Eqn. (2)) due to the better mixingof the solution. Thus, an enhanced oxygen supply was notconsidered as an optimization measure.

(iv) We tested the impact of elevated temperature on theefciency of the dissolution process by heating thesuspension of CuOCuCl, formed in step 1, at 90 C for2 h in a water bath. Formation of HCl in a further reactionof the combustion products CuCl and water (Eqn. (6)),



5

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

66676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

followed by its volatilization, was reported by Holt andSturchio,[21] as probably being the major source of thelosses of chlorine, associated with more variable 37Clvalues.

2 CuClH2OCu2O 2 HCl (6)

To prevent any losses of chlorine in the formofHCl, the heatingwas performed in screw-capped bottles. Elevated temperatureat the dissolution step allowed us to reach 49% conversion ofchlorine (compared with 34% conversion, obtained with thesame water/Cl ratio by 2 days of sonication at the roomtemperature), showing the temperature-enhanced kinetics ofthe dissolution process (Eqn. (2)). Thus, we used a temperatureof 50 C in the ultrasonic bath during the sonication forfurther experiments.

(v) The most signicant effect on the efciency of thedissolution process was caused by changing the volumeof water during extraction of the soluble chlorine species(Eqn. (2)) after combustion. By increasing the water/Clratio from 0.1 to 5 mL/mol 76% Cl recovery was reached.The recovered Cl had a 37Cl value of 0.17 0.08 (n= 2),that is 0.98 higher than the value of 1.15 0.34obtained for the lowest recoveries of chlorine in the initialprocedure, and thus exhibiting the same trend towardsincreasing 37Cl value with increasing chlorine recoveryyields. In an additional set of experiments with dissolutionof commercially available crystalline CuCl in differentvolumes of uhp-water (ratio mL/mol Cl from 0.36 to8.00, section 5, Supporting Information) we determinedthe optimal water/Cl ratio as 0.4 mL/mol, resulting in96% chlorine recovery (Supplementary Table S-1,Supporting Information). Thus, a water/Cl ratio of0.4 mL/mol was used for the further optimization steps.

Conversion of -HCH under optimized conditions

We applied the total set of optimized parameters (repeatedextraction by sonication for 2 h at 50 C with two times100 mL of water, corresponding to a water/Cl ratio of0.4 mL/mol) to prove the efciency of the full procedureboth in terms of complete chlorine recoveries and in thereproducibility of the determined 37Cl values (Table 1). Totest if the remaining residues of chlorine species after a non-complete dissolution process have the same chlorine isotopiccomposition as initially dissolved CuCl2, fractions from tworepeated extractions were separately analyzed for chlorinecontents and 37Cl values. The optimized extractionprocedure was performed in triplicate from three combustedsubsamples of -HCH. Chlorine concentration and DI-IRMSmeasurements showed that 88% of the chlorine, with a 37Clvalue of 0.26 0.02 (n = 3), can be recovered in the rstextraction of CuCl2, but the second recovered fraction ofsoluble chlorine salt (6%) was signicantly enriched in 37Cl,exhibiting a 37Cl value of 0.92 0.28 (n = 3). We havecalculated the combined 37Cl value of the two extracts withthe 94 5% of recovered chlorine (sum of recoveries from 1st

and 2nd fractions), based on the mass balance. The obtained

37Cl value of 0.18 0.03 (n = 3) more accurately representsthe isotopic composition of the organochlorine compoundstudied here. We hence conclude that incomplete chlorineextraction leads to analytically biased too low 37Cl values ofthe studied organochlorine compounds. Therefore, werecommend applying repeated extractions to obtain completechlorine recovery and thus achieve accurate and precisedeterminations of the Cl isotopic compositions of organo-chlorine compounds.

Conversion of CH3Cl under optimized conditions

The determined optimal conditions of the dissolution processwere incorporated into the full procedure for the conversionof commercially available CH3Cl gas with known isotopiccomposition (Table T22). The optimized conversion procedurethat we dened here allowed CH3Cl to be transformed intowater-soluble CuCl2 and back to CH3Cl with a slight off-setof 0.11 0.04 (1, n = 3) (Table 2), which shows animprovement in comparison with a decrease of 0.23 0.05 (1) for the entire sample preparation procedurewith 89% of recoveries reported by Holt et al.[21]

Isotopic precision of the method

The external precision of all the 37Cl values of seawaterindependently prepared and analyzed over the course of thisstudywas0.07 (standard deviation of 15measurements);the shift within 1 day was in the range from 0.02 to 0.07(n = 2). The difference between duplicate DI-IRMS analysesof two introduced subsamples of the same produced CH3Clwas 0.02. The external precision of the method,determined from different ampules with identical startingmaterial (-HCH), was 0.03% (1, n = 3) (Table 1). Similarly,the precision of the 37Cl values, obtained through the

Table 2. Validation of the overall optimized procedurewith CH3Cl

37Clvalue ofcommercialCH3Cl,

After the entire sample preparationprocedure

37Clvaluea

(n = 2),

Average37Cl

value 1b

from 3ampules, 37Clc,

1.181.060.02 1.13 1.170.04 0.11 0.04

1.20aDifference between duplicate measurements of the sameprepared CH3Cl sample, produced by subsequenttransformation of commercial CH3Cl to soluble CuCln,precipitation of AgCl and transformation back to theCH3Cl for the DI-IRMS analysis, was typically below 0.01.bExternal precision of sample preparation procedure fromdifferent subsamples of CH3Cl.cDifference between 37Cl value of unprocessed CH3Cl andthat of CH3Cl after the entire sample preparationprocedure.

T. Gilevska et al.


6

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

66676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

sample preparation procedure from CH3Cl, was 0.04(1, n = 3) (Table 2). This external precision for OCs is ofthe order of the uncertainty related to the preparation ofsea water standards (0.08, 2), when only inorganicchloride is processed to CH3Cl.Such a good external precision for the studied compounds

(-HCH and CH3Cl) proves the applicability of the optimizedmethod for the chlorine isotope analysis of organochlorines.

CONCLUSIONS

We present a method for the determination of the chlorinestable isotope composition of -HCH by DI-IRMS in 37Clnotation vs SMOC, using seawater for direct referencing. Inthe course of this study the links between losses of chlorine atevery separate conversion step and accuracy of the measured37Cl values were quantied. Dissolution of water-solublechlorine salts, formed by combustion of -HCH, was identiedto be the critical step,most sensitive to the amount of chlorine inthe sample and associated with signicant losses of chlorine.Incomplete chlorine extraction at this dissolution step led toinaccurate lower 37Cl values of the studied organochlorinecompound. Optimization of this step resulted in completechlorine recoveries through the entire sample preparationprocedure (945%) and more accurately determined theisotopic composition of -HCH (0.18 0.03, 1), whichwas enriched in 37Cl by 0.97 in comparison with the non-optimized initial procedure. Thus, we recommend thedeveloped optimized chlorine extraction method for thedetermination of the 37Cl compositions of organochlorinecompounds. The optimized sample preparation procedureshould be applicable for different amounts of chlorine in thesample. Validation of the procedure with a CH3Cl sample withknown isotopic composition proved the method to be accurateand precise and showed a total deviation of 37Cl results of lessthan 0.11 0.04 (1, n = 3).As the most critical step is not related to the oxidation of

the chlorinated compound and, thus, is independent of itschemical structure, the presented optimized procedure holdsthe potential for 37Cl determination of a broad range ofchlorinated organic compounds.

AcknowledgementsWe gratefully acknowledge H. G. M. Eggenkamp, ThomasGiunta, Gerard Bardoux and Michaela Wunderlich foranalytical support and helpful discussions on the improve-ment of applied methods, as well as Kristina Hitzfeld, JulianRenpenning and Angela Woods for critical comments. Weacknowledge nancial support from the European Unionunder FP7-People-ITN-2010 (Grant Agreement No. 264329).

REFERENCES

[1] A. Cincinelli, F. Pieri, Y. Zhang, M. Seed, K. C. Jones.Compound specic isotope analysis (CSIA) for chlorineand bromine: A review of techniques and applications toelucidate environmental sources and processes. Environ.Pollut. 2012, 169, 112.

[2] Y. Abe, R. Aravena, J. Zop, B. Parker, D. Hunkeler.Evaluating the fate of chlorinated ethenes in streambedsediments by combining stable isotope, geochemical andmicrobial methods. J. Contam. Hydrol. 2009, 107, 10.

[3] M. Blessing, T. C. Schmidt, R. Dinkel, S. B. Haderlein.Delineation of multiple chlorinated ethene sources in anindustrialized area A forensic eld study usingcompound-specic isotope analysis. Environ. Sci. Technol.2009, 43, 2701.

[4] T. C. Schmidt, L. Zwank, M. Elsner, M. Berg, R. U.Meckenstock, S. B. Haderlein. Compound-specic stableisotope analysis of organic contaminants in naturalenvironments: a critical review of the state of the art,prospects, and future challenges. Anal. Bioanal. Chem. 2004,378, 283.

[5] M. Elsner, M. A. Jochmann, T. B. Hofstetter, D. Hunkeler, A.Bernstein, T. C. Schmidt, A. Schimmelmann. Currentchallenges in compound-specic stable isotope analysis ofenvironmental organic contaminants. Anal. Bioanal. Chem.2012, 403, 2471.

[6] K. L. Hitzfeld, M. Gehre, H. H. Richnow. A novel onlineapproach to the determination of isotopic ratios fororganically bound chlorine, bromine and sulphur. RapidCommun. Mass Spectrom. 2011, 25, 3114.

[7] J. Renpenning, K. L. Hitzfeld, T. Gilevska, I. Nijenhuis, M.Gehre, H. H. Richnow. Development and validation of anuniversal interface for compound-specic stable isotopeanalysis of chlorine (Cl-37/Cl-35) by GC-high-temperatureconversion (HTC)-MS/IRMS. Anal. Chem. 2015, 87, 2832.

[8] A. Bernstein, O. Shouakar-Stash, K. Ebert, C. Laskov, D.Hunkeler, S. Jeannottat, K. Sakaguchi-Soder, J. Laaks, M.A. Jochmann, S. Cretnik, J. Jager, S. B. Haderlein, T. C.Schmidt, R. Aravena, M. Elsner. Compound-specicchlorine isotope analysis: A comparison of gaschromatography/isotope ratio mass spectrometry and gaschromatography/quadrupole mass spectrometry methodsin an interlaboratory study. Anal. Chem. 2011, 83, 7624.

[9] M. R. M. D. Van Acker, A. Shahar, E. D. Young, M. L.Coleman. GC/multiple collector-ICPMS method forchlorine stable isotope analysis of chlorinated aliphatichydrocarbons. Anal. Chem. 2006, 78, 4663.

[10] O. Shouakar-Stash, R. J. Drimmie, M. Zhang, S. K. Frape.Compound-specic chlorine isotope ratios of TCE, PCEand DCE isomers by direct injection using CF-IRMS. Appl.Geochem. 2006, 21, 766.

[11] K. Sakaguchi-Soder, J. Jager, H. Grund, F. Matthaus, C.Schuth. Monitoring and evaluation of dechlorinationprocesses using compound-specic chlorine isotopeanalysis. Rapid Commun. Mass Spectrom. 2007, 21, 3077.

[12] C. Aeppli, H. Holmstrand, P. Andersson, O. Gustafsson.Direct compound-specic stable chlorine isotope analysisof organic compounds with quadrupole GC/MS usingstandard isotope bracketing. Anal. Chem. 2010, 82, 420.

[13] B. A. Jin, C. Laskov, M. Rolle, S. B. Haderlein. Chlorineisotope analysis of organic contaminants using GC-qMS:Method optimization and comparison of differentevaluation schemes. Environ. Sci. Technol. 2011, 45, 5279.

[14] R. A. Werner, W. A. Brand. Referencing strategies andtechniques in stable isotope ratio analysis. Rapid Commun.Mass Spectrom. 2001, 15, 501.

[15] J. M. Rosenbaum, R. A. Cliff, M. L. Coleman. Chlorine stableisotopes: A comparison of dual inlet and thermal ionizationmass spectrometric measurements. Anal. Chem. 2000, 72,2261.

[16] N. Jendrzejewski, H. G. M. Eggenkamp, M. L. Coleman.Sequential determination of chlorine and carbon isotopiccomposition in single microliter samples of chlorinatedsolvent. Anal. Chem. 1997, 69, 4259.



7

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

66676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

[17] H. Holmstrand, P. Andersson, O. Gustafsson. Chlorineisotope analysis of submicromole organochlorine samplesby sealed tube combustion and thermal ionization massspectrometry. Anal. Chem. 2004, 76, 2336.

[18] L. I. Wassenaar, G. Koehler. On-line technique for thedetermination of the delta Cl-37 of inorganic and total organicCl in environmental samples. Anal. Chem. 2004, 76, 6384.

[19] M. Numata, N. Nakamura, T. Gamo. Precise measurementof chlorine stable isotopic ratios by thermal ionization massspectrometry. Geochem. J. 2001, 35, 89.

[20] M. Numata, N. Nakamura, H. Koshikawa, Y. Terashima.Chlorine stable isotope measurements of chlorinatedaliphatic hydrocarbons by thermal ionization massspectrometry. Anal. Chim. Acta 2002, 455, 1.

[21] B. D. Holt, N. C. Sturchio, T. A. Abrajano, L. J. Heraty.Conversion of chlorinated volatile organic compounds tocarbon dioxide and methyl chloride for isotopic analysis ofcarbon and chlorine. Anal. Chem. 1997, 69, 2727.

[22] C. M. Reddy, N. J. Drenzek, T. I. Eglinton, L. J. Heraty, N. C.Sturchio, V. J. Shiner. Stable chlorine intramolecular kineticisotope effects from the abiotic dehydrochlorination ofDDT. Environ. Sci. Pollut. Res. 2002, 9, 183.

[23] O. Shouakar-Stash, S. K. Frape, R. J. Drimmie. Stable hydrogen,carbon and chlorine isotope measurements of selectedchlorinated organic solvents. J. Contam. Hydrol. 2003, 60, 211.

[24] C. Aeppli, D. Bastviken, P. Andersson, O. Gustafsson.Chlorine isotope effects and composition of naturallyproduced organochlorines from chloroperoxidases, avin-dependent halogenases, and in forest soil. Environ. Sci.Technol. 2013, 47, 6864.

[25] J. Vijgen, P. C. Abhilash, Y. F. Li, R. Lal, M. Forter, J. Torres, N.Singh, M. Yunus, C. G. Tian, A. Schaffer, R. Weber.Hexachlorocyclohexane (HCH) as new Stockholm ConventionPOPs a global perspective on the management of Lindaneand its waste isomers. Environ. Sci. Pollut. Res. 2011, 18, 152.

[26] H. van den Berg. Global status of DDT and its alternativesfor use in vector control to prevent disease. Environ. HealthPerspect. 2009, 117, 1656.

[27] H. G. M. Eggenkamp, J. J. Middelburg, R. Kreulen.Preferential diffusion of Cl-35 relative to Cl-37 in sedimentsof Kau Bay, Halmahera, Indonesia. Chem. Geol. 1994, 116,317.

[28] M. Bonifacie, C. Monnin, N. Jendrzejewski, P. Agrinier, M.Javoy. Chlorine stable isotopic composition of basementuids of the eastern ank of the Juan de Fuca Ridge (ODPLeg 168). Earth Planet. Sci. Lett. 2007, 260, 10.

[29] M. Bonifacie, J. L. Charlou, N. Jendrzejewski, P. Agrinier, J.P. Donval. Chlorine isotopic compositions of hightemperature hydrothermal vent uids over ridge axes.Chem. Geol. 2005, 221, 279.

[30] A. Godon, N. Jendrzejewski, H. G. M. Eggenkamp, D. A.Banks, M. Ader, M. L. Coleman, F. Pineau. A cross-calibration of chlorine isotopic measurements andsuitability of seawater as the international referencematerial. Chem. Geol. 2004, 207, 1.

[31] M. Bonifacie, N. Jendrzejewski, P. Agrinier, M. Coleman, F.Pineau, M. Javoy. Pyrohydrolysis-IRMS determination ofsilicate chlorine stable isotope compositions. Applicationto oceanic crust and meteorite samples. Chem. Geol. 2007,242, 187.

[32] H. G. M. Eggenkamp, Utrecht University 1994.[33] H. G. M. Eggenkamp, R. Kreulen, A. F. Koster van Groos.

Chlorine stable isotope fractionation in evaporites. Geochim.Cosmochim. Acta 1995, 59, 5169.

[34] N. Jendrzejewski, H. G. M. Eggenkamp, M. L. Coleman.Characterisation of chlorinated hydrocarbons fromchlorine and carbon isotopic compositions: scope ofapplication to environmental problems. Appl. Geochem.2001, 16, 1021.

SUPPORTING INFORMATION

Additional supporting information may be found in theonline version of this article at the publisher's website.

T. Gilevska et al.


8

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

66676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

USING e-ANNOTATION TOOLS FOR ELECTRONIC PROOF CORRECTION Required software to e-Annotate PDFs: Adobe Acrobat Professional or Adobe Reader (version 7.0 or above). (Note that this document uses screenshots from Adobe Reader X) The latest version of Acrobat Reader can be downloaded for free at: http://get.adobe.com/uk/reader/

Once you have Acrobat Reader open on your computer, click on the Comment tab at the right of the toolbar:

1. Replace (Ins) Tool for replacing text.

Strikes a line through text and opens up a text box where replacement text can be entered.

How to use it

Highlight a word or sentence. Click on the Replace (Ins) icon in the Annotations

section.

Type the replacement text into the blue box that appears.

This will open up a panel down the right side of the document. The majority of tools you will use for annotating your proof will be in the Annotations section, pictured opposite. Weve picked out some of these tools below:

2. Strikethrough (Del) Tool for deleting text.

Strikes a red line through text that is to be deleted.

How to use it

Highlight a word or sentence. Click on the Strikethrough (Del) icon in the

Annotations section.

3. Add note to text Tool for highlighting a section to be changed to bold or italic.

Highlights text in yellow and opens up a text box where comments can be entered.

How to use it

Highlight the relevant section of text. Click on the Add note to text icon in the

Annotations section.

Type instruction on what should be changed regarding the text into the yellow box that appears.

4. Add sticky note Tool for making notes at specific points in the text.

Marks a point in the proof where a comment needs to be highlighted.

How to use it

Click on the Add sticky note icon in the Annotations section.

Click at the point in the proof where the comment should be inserted.

Type the comment into the yellow box that appears.

USING e-ANNOTATION TOOLS FOR ELECTRONIC PROOF CORRECTION

For further information on how to annotate proofs, click on the Help menu to reveal a list of further options:

5. Attach File Tool for inserting large amounts of text or replacement figures.

Inserts an icon linking to the attached file in the appropriate pace in the text.

How to use it

Click on the Attach File icon in the Annotations section.

Click on the proof to where youd like the attached file to be linked.

Select the file to be attached from your computer or network.

Select the colour and type of icon that will appear in the proof. Click OK.

6. Add stamp Tool for approving a proof if no corrections are required.

Inserts a selected stamp onto an appropriate place in the proof.

How to use it

Click on the Add stamp icon in the Annotations section.

Select the stamp you want to use. (The Approved stamp is usually available directly in the menu that appears).

Click on the proof where youd like the stamp to appear. (Where a proof is to be approved as it is, this would normally be on the first page).

7. Drawing Markups Tools for drawing shapes, lines and freeform annotations on proofs and commenting on these marks. Allows shapes, lines and freeform annotations to be drawn on proofs and for comment to be made on these marks..

How to use it

Click on one of the shapes in the Drawing Markups section.

Click on the proof at the relevant point and draw the selected shape with the cursor.

To add a comment to the drawn shape, move the cursor over the shape until an arrowhead appears.

Double click on the shape and type any text in the red box that appears.

rcm7220

Documents

csia of chlorine

sample preparation method

isotope monitoring

organochlorine compounds

entire method

irms analysis

isotopic compositions

initial ch3cl sample