Passive Sampler for Dissolved Organic Matter inFreshwater Environments

Buuan Lam and Andre J. Simpson*

Department of Chemistry, University of Toronto, Scarborough Campus, Toronto, Ontario, Canada M1C 1A4

A passive sampler for the isolation of dissolved organicmatter (DOM) from freshwater environments is described.The sampler consists of a molecular weight selectivemembrane (1000 kDa) and an anion exchange resin(diethylaminoethylcellulose (DEAE-cellulose)). NMR in-dicates the samplers isolate DOM that is nearly indistin-guishable from that isolated using the batch DEAE-cellulose procedure. In a comparative study DOM isolatedfrom Lake Ontario cost ∼$0.30/mg to isolate using thepassive samplers while DOM isolated using the traditionalbatch procedure cost ∼$8-10/mg. The samplers havebeen shown to be effective in a range of freshwaterenvironments including a large inland lake (Lake On-tario), fast flowing tributary, and wetland. Large amounts(gram quantities of DOM) can be easily isolated byincreasing the size or number of samplers deployed.Samplers are easy to construct, negate the need forpressure filtering, and also permit a range of temporal andspatial experiments that would be very difficult or impos-sible to perform using conventional approaches. Forexample, DOM can be monitored on a regular basis atnumerous different locations, or samplers could be setat different depths in large lakes. Furthermore, they couldpotentially be deployed into hard to reach environmentssuch as wells, groundwater aquifers, etc., and as they areeasy to use, they can be mailed to colleagues or includedwith expeditions going to difficult to reach places such asthe Arctic and Antarctic.

Dissolved organic matter (DOM) represents the largest poolof mobile carbon on the Earth and is a fundamental link betweenthe terrestrial and aquatic environments. If aquatic DOM expe-rienced a 1% oxidation in a year, the CO2 released would be greaterthan that generated annually from the combustion of fossil fuels.1

DOM is a heterogeneous, complex mixture, ubiquitous in theenvironment, including soil, sediment, rain, and oceans, andrepresents one of the largest reservoirs of carbon on earth.2,3 This,combined with the natural cycling of DOM from terrestrial toaquatic environments,2 makes DOM a significant contributor tothe global carbon cycle,4,5 an important mediator in the physical

and chemical interactions that govern the fate and transport ofmany contaminants, and an integral player in climate change.6

However, despite its importance, many of the basic structuralcomponents in DOM have yet to be elucidated, due in part to thedifficulty in its isolation.

Traditionally, the isolation of DOM has required large samplevolumes to overcome the low concentrations in natural waters.7,8

These techniques, however, require extensive on-site practicesthat are labor-intensive and time-consuming, often requiringchemicals, columns, filters, generators, etc., to be used in the field.Traditionally, these methods often employ nonionic, macroporousresins such as XAD to bind and extract DOM.8-10 However, a largemajority of these resins must be maintained at low pH toadequately bind components.9 Since DOM from natural waters ispresent at near-neutral pH, the use of these resins requiresacidification that could potentially alter the chemical compositionof the isolated DOM. Diethylaminoethyl-cellulose (DEAE-cel-lulose), however, has been shown to isolate up to 90% of DOMfrom freshwater within the pH range of natural waters.11 Thischaracteristic gives DEAE-cellulose the potential to be an effectiveresin for isolating and concentrating DOM in its natural form.Promising resins alone, however, are not sufficient to effectivelyenhance the isolation of DOM for analysis. Large sample volumesare still required to overcome the low natural abundance of DOM,and filtering is critical to remove insoluble and biological species.To prevent biological alteration during transport, it is beneficialto at least filter samples, or carry out the entire isolation, in thefield.

In this study, we introduce the development and applicationof a passive sampler to effectively isolate and concentrate DOMfrom freshwater environments in situ. The use of passive samplersin environmental chemistry is far from new. However, the vastmajority of samplers thus far have focused on the concentrationof contaminants.12-14 The sampler described here, is to the

* To whom correspondence should be addressed. Tel +1 416 287 7547. Fax+1 416 287 7279. E-mail address andre.simpson@utoronto.ca.(1) Hedges, J. I. In In Biogeochemistry of Marine Dissolved Organic Matter;

Hansell, D. A., Carlson, C. A., Eds.; Academic Press: New York, 2002; pp1-33.

(2) Hedges, J. I. Mar. Chem. 1992, 39, 67-93.(3) Ogawa, H.; Tanoue, E. J. Oceanogr. 2003, 59, 129-147.

(4) Trumbore, S.; Druffel, E. Role of Nonliving Organic Matter in the Earth’sCarbon Cycle; John Wiley & Sons: Chichester, 1995.

(5) Baldock, J. A.; Masiello, C. A.; Gelinas, Y.; Hedges, J. I. Mar. Chem. 2004,92, 39-64.

(6) Keeling, C. D.; Whorf, T. P.; Wahlen, M.; Vanderplicht, J. Nature 1995,375, 666-670.

(7) Leenheer, J. A. Environ. Sci. Technol. 1981, 15, 578-587.(8) Thurman, E. M.; Malcolm, R. L. Environ. Sci. Technol. 1981, 15, 463-466.(9) Aiken, G. R.; Thurman, E. M.; Malcolm, R. L.; Walton, H. F. Anal. Chem.

1979, 51, 1799-1803.(10) Maccarthy, P.; Peterson, M. J.; Malcolm, R. L.; Thurman, E. M. Anal. Chem.

1979, 51, 2041-2043.(11) Miles, C. J.; Tuschall, J. R.; Brezonik, P. L. Anal. Chem. 1983, 55, 410-

411.

Anal. Chem. 2006, 78, 8194-8199

8194 Analytical Chemistry, Vol. 78, No. 24, December 15, 2006 10.1021/ac0608523 CCC: $33.50 © 2006 American Chemical SocietyPublished on Web 10/25/2006

authors’ knowledge, the first passive sampler designed for theisolation of DOM from the environment and the first to use a size-selective membrane (to negate the need for filtering and distin-guish between biological species and dissolved species) incombination with a selective resin (in this case DEAE-cellulose)to concentrate the negatively charged DOM species. This passiveisolation provides a means to collectively isolate, filter, andconcentrate DOM in the field, eliminating problems associatedwith traditional on-site extractions and large sample volumes.Additionally, these samplers have applications in spatial andtemporal studies to monitor changes in DOM at various depths(for example, numerous samplers can be deployed at specificdepths using a single mooring) and locations (potentially hundredsof samplers could be deployed on the same day by just oneresearch team (the same is true with recovery)). The samplersthus provide an economical and easy means by which researcherscan attain a vast array of DOM samples in a short time, critical tounderstanding its local/global variability, transformations, andreactivity.

MATERIALS AND METHODSSampler: General Design and Preparation. The passive

sampler is simple in both concept and design as highlighted inFigure 1. Three components comprise the passive sampler: (1)DEAE-cellulose (Sigma Aldrich, Order No. D6418)), a selectiveresin that adsorbs negatively charged species at neutral pH, (2)a poly(vinylidene fluoride) (PVDF) porous membrane with a

molecular weight cutoff (MWCO) of 1000 kDa (∼0.1 µm, (Spec-trapor, Order No. 138525)), and (3) a high-density polyethylene(HDPE) casing with predrilled holes (constructed in-house). Note,the authors have found simple “Nalgene” style HDPE disposablescrew cap bottles can be drilled and used as a protective casingif specially designed casings are not available.

Prior to use, DEAE-cellulose was precleaned using a cycle ofacid, base, and distilled water washings. Specifically, the cleaningregiment consisted of 0.1 M hydrochloric acid, 0.1 M sodiumhydroxide, and distilled water washings in between acid/basecleanings. A minimum of 10 full acid-water-base cycles wereperformed, followed by a minimum of 100 rinses with excessdistilled water, and finally freeze-drying. Note, the resin settlesfairly quickly (in ∼30 min). Cleaning can be easily carried out bymixing in a tall container (such as a large measuring cylinder)and then carefully decanting the supernatant after settling. Thecomplete process (including cleaning and freeze-drying) will takeclose to 1 month. The authors advise cleaning a large batch ofresin, which can be stored in its freeze-dried form in an airtightsterile container for later use.

Cleaned DEAE-cellulose (250 mg) was slurry packed withdistilled water into 7-cm-length (24-mm-width) PVDF porousmembranes, which were presoaked in 0.1% sodium azide for aminimum of 48 h. Preliminary experiments (not described here)demonstrated that 250 mg/7-cm length of tubing was the mostefficient use of both resin and PVDF membrane; this ratio allowedresin mobility within the sampler such that its entire surface areawas available for DOM sorption. Packed membranes were placedinto the constructed HDPE casings to form the passive samplers.

Laboratory Studies. Initial laboratory studies were performedto test the applicability of the samplers for the isolation/concentration of DOM. Two separate studies were performed. The

(12) Vrana, B.; Mills, G. A.; Allan, I. J.; Dominiak, E.; Svensson, K.; Knutsson, J.;Morrison, G.; Greenwood, R. TrAC-Trends Anal. Chem. 2005, 24, 845-868.

(13) Stuer-Lauridsen, F. Environ. Pollut. 2005, 136, 503-524.(14) Namiesnik, J.; Zabiegala, B.; Kot-Wasik, A.; Partyka, M.; Wasik, A. Anal.

Bioanal. Chem. 2005, 381, 279-301.

Figure 1. (a) Schematic showing the components of the passive sampler. (1) The 1000-kDa MWCO poly(vinylidene fluoride) membrane. (2)Porous HDPE casing to house sampler unit (designed in-house) to prevent large organisms (fish, etc.) and debris from compromising themembrane. (3) DEAE-cellulose resin. (b) Expanded region showing the resin/membrane/water interface. (4) Dissolved negatively charged DOMenters the membrane and is sorbed onto the resin and concentrated manyfold, (5 + 6) Dissolved neutral or positively charged species (forexample, metals in the case of positively charged species) can enter the membrane but are not retained (note the vast majority of DOM isnegatively charged). (7 + 8) Large species including particulate organic matter and biological species cannot enter the membrane. The use ofthe membrane removes the need for filtering.

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006 8195

first used the International Humic Substances Society (IHSS)Standard Suwannee River natural organic matter, isolated fromthe Suwannee river by reverse osmosis,15 and the second dissolvedorganic matter isolated by the XAD resin procedure from Meth-wold Fen, Norfolk, United Kingdom (see Simpson et al.16 forisolation details and characterization). Samplers (250 mg of resin,in 7 × 24 mm) were suspended (using fishing line) in 10 ppmDOM solutions (no additional salts were added), which werecontained in stoppered 1-L bottles and continuously mixed usinga magnetic stirrer. The pH of the Suwannee River and NorfolkDOM samples were 7.1 and 6.9, respectively. Each experimentwas carried out in triplicate. Samplers were left to equilibrate overa period of 2 weeks, and the concentration of DOM remaining insolution was determined by total organic carbon (TOC) analysis(see below).

Field Studies. A field study was carried out at the LyndeShores Conservation Area (representative of a wetland) close tothe mouth of Lynde Creek, Ajax, Ontario. Passive samplers wereplaced in triplicate in a steel cage and suspended 50 cm belowthe surface of the water. Samplers were removed from the cageon days 1, 3, 7, 14, and 28, extracted, as described below, andDOM yields determined gravimetrically. In addition, a conven-tional DEAE-cellulose batch extraction was performed. Briefly 40L of water was collected, filtered through 0.22-µm Teflon filters(Millipore), batch extracted with DEAE-cellulose,11 and recoveredas described below. In addition to Lynde Shores ConservationArea samplers were also placed in highland creek (a fast-flowingtributary), and Lake Ontario (to represent a large freshwater lake).

Sample Extraction. Extraction of bound DOM from thepassive samplers was performed by cutting and removing the resinfrom the PVDF membranes. The resin was then placed in 50-mLTeflon centrifuge tubes and extracted using ∼40 mL of 0.1 Msodium hydroxide. The tubes were centrifuged at 10000g for 10min to pellet the resin, and the supernatant was decanted. Thepellet was then resuspended, and the previous steps were repeatedfour more times, or until the extracting solvent was clear, to ensurecomplete extraction of DOM from the resin. The extracted DOMwas then ion-exchanged using Amberjet 1200H Plus resin (Ald-rich) and freeze-dried. The freeze-dried sample was resuspendedin deuterated dimethyl sulfoxide (DMSO-d6) or deuterium oxide(D2O (with 2 µL of 40% by weight NaOD added, to ensurecomplete solubility)) for NMR analysis.

Sample Analysis. NMR is a well-established tool for theanalysis of dissolved organic matter.17-19 Samples were analyzedusing nuclear magnetic resonance (NMR) spectroscopy on aBruker Avance 500 equipped with a 1H BB-13C 5-mm tripleresonance broadband inverse probe. Unless otherwise stated, 1-Dsolution state 1H NMR experiments were performed with 512scans, a recycle delay of 3 s, and 32 K time domain points. Solventsuppression was achieved by presaturation utilizing relaxation

gradients and echos (PURGE).20 Spectra were apodized bymultiplication with an exponential decay corresponding to 1-Hzline broadening in the transformed spectrum and a zero fillingfactor of 2. Blank 1H NMR were run on completely assembledpassive samplers soaked in distilled water, and no backgroundsignals from the samplers were detected.

TOC analysis was carried out on a Shimadzu TOC-V Seriesanalyzer. Samples were acidified to pH 2 followed by sparging toremove dissolved inorganic carbon. The sample aliquots were thensubjected to high-temperature catalytic oxidation to form CO2 andsubsequent CO2 concentrations measured with a nondispersiveinfrared detector.21

RESULTSInitial laboratory experiments demonstrated that 72 ( 5

(Suwannee River) and 89 ( 4% (Norfolk) DOM was removed fromthe 10 ppm laboratory solutions after 2 weeks of equilibration.Figure 2A shows the NMR spectrum of the Suwannee river DOMisolated on the passive sampler compared to the original sample(Figure 2B). The NMR spectra are generally similar, indicatingthat the passive samplers give a reasonably representativeoverview as to DOM components present in the environment. Thelargest difference in the material concentrated on the passivesampler is the reduction of small molecular weight sugars(indicated by sharp signals in the 3-4 ppm region). Such smallcomponents are highly soluble and may be less likely to perma-

(15) IHSS, http://www.ihss.gatech.edu/, accessed Aug 2006.(16) Simpson, A. J.; Tseng, L. H.; Simpson, M. J.; Spraul, M.; Braumann, U.;

Kingery, W. L.; Kelleher, B. P.; Hayes, M. H. B. Analyst 2004, 129, 1216-1222.

(17) Simpson, A. J. Magn. Reson. Chem. 2002, 40, S72-S82.(18) Kaiser, E.; Simpson, A. J.; Dria, K. J.; Sulzberger, B.; Hatcher, P. G. Environ.

Sci. Technol. 2003, 37, 2929-2935.(19) Kim, S.; Simpson, A. J.; Kujawinski, E. B.; Freitas, M. A.; Hatcher, P. G.

Org. Geochem. 2003, 34, 1325-1335.

(20) Simpson, A. J.; Brown, S. A. J. Magn. Reson. 2005, 175, 340-346.(21) Wetzel, R. G.; Likens, G. E. Limnological Analyses, 3rd ed.; Springer: New

York, 2000.

Figure 2. 1H NMR spectra showing DOM dissolved in D2O/NaOD.(A) Fraction from the Suwannee River reverse osmosis standardcollected on a passive sampler after 2-weeks equilibration. (B)Suwannee River reverse osmosis standard dissolved directly in D2O/NaOD. PURGE20 water suppression was applied to the residual watersignal at ∼4.7 ppm.

8196 Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

nently bind to the DEAE-cellulose than the high molecular weightmultidentate species that are thought to predominate in DOM. Itis important to note the relative yield of Suwannee river isolatedon the passive sampler (72 ( 5%) is less than that reported byMiles et al.11 (88%) by the DEAE batch procedure. Later in thispaper we demonstrate using NMR analyses that the passivesampler and batch DEAE procedures used to isolate DOM fromthe field are extremely similar. The difference in yields observedhere is likely to be, at least in part, a result of the considerableamount of salt that is concentrated alongside the organic com-pounds by reverse osmosis (the method used to concentrate theIHSS Suwannee River Standard). Negative ions (mainly chloride)will compete for binding sites on the DEAE-cellulose and slowdown or prevent the sorption of all DOM components. Saltconcentration is just one factor that will affect the rate of DOMuptake in the environment, DOM concentration, and waterdynamics, among others may also play key roles. Consideringthese factors, it is critically important to demonstrate that thepassive samplers can isolate DOM in reasonable quantities in thefield and that the DOM isolated is the same as that obtained byDEAE in conventional batch extraction, which is now becominga popular technique for isolating DOM from the environment.22-25

Figure 3 shows the yields of DOM isolated in the field at LyndeShores Conservation Area, Ajax, Ontario. The yields are expressedin terms of individual samplers each containing 250 mg of resin.The uptake follows the generalized uptake profile expected for apassive sampler device.26

After 1 month. 18 ( 2 mg of DOM had concentrated on eachof the three passive samplers, giving a total yield on all threesamplers of 54 mg. As a comparison, water was collected fromthe conservation area midway through the sampling period andthe DOM isolated by batch extraction. Figure 4A shows the NMRspectrum of the DOM collected on the passive sampler in

comparison with that extracted by the conventional batch proce-dure (Figure 4B). It is clear that the material is extremely similar,indicating that the passive sampler approach isolates the samecomponents as conventional DEAE-cellulose batch extraction.Slight variations in the NMR profile are expected considering thatthe water for batch extraction was collected on a single day inthe middle of the sampling study, whereas the DOM from thepassive samplers represents “the average” DOM that has beencollected over the period of the whole month.

Table 1 compares yields collected from various sites. FromLake Ontario, 36 mg of DOM/g of DEAE-cellulose was isolated,whereas nearly twice that, 60 mg, was obtained for the Lyndeshores conservation area. This is in line with TOC concentrationsfrom the two sites, which were 2.8 and 5.0 ppm, respectively. Inaddition to the relatively small scale studies described so far, alarge-scale study was carried out to ensure “bulk” amounts ofDOM could be isolated if required. In this case, 750 mg of resinwas packed into 21 cm × 24 mm tubes (such that the ratio of 250mg of resin/7 cm of membrane was preserved) to create muchlonger samplers. Sixty of these larger samplers (equivalent to 180of the smaller samplers) were deployed, and after 2 weeks, 2.79g of DOM was isolated. This equates to 62 mg of DOM/g of resinused and is consistent with the 60 mg of DOM/g of resin isolatedfrom the same site using the smaller samplers (although sampleda month later, see Table 1).

(22) Peuravuori, J.; Pihlaja, K.; Valimaki, N. Environ. Int. 1997, 23, 453-464.(23) Raastad, I. A.; Ogner, G. Commun. Soil Sci. Plant Anal. 1997, 28, 1311-

1321.(24) Peuravuori, J.; Ingman, P.; Pihlaja, K.; Koivikko, R. Talanta 2001, 55, 733-

742.(25) Peuravuori, J.; Monteiro, A.; Eglite, L.; Pihlaja, K. Talanta 2005, 65, 408-

422.(26) Mayer, P.; Tolls, J.; Hermens, L.; Mackay, D. Environ. Sci. Technol. 2003,

37, 184A-191A.

Figure 3. Field study showing the uptake of DOM on passivesamplers from Lynde Shores Conservation Area. Samplers weredeployed in triplicate each containing 250 mg of resin. Yields areexpressed per individual sampler, and bars indicate the standarddeviation between triplicate samplers.

Figure 4. 1H NMR spectra of DOM isolated from Lynde ShoresConservation Area using (a) passive samplers and (b) traditionalisolation, involving filtering and batch isolation on DEAE-cellulose.

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006 8197

DISCUSSIONFundamental Differences: Passive Sampler versus Batch

Extraction. It is important to consider that even if samples wereisolated from the same location at the “same time” there wouldstill likely be small variances between DOM isolated using apassive sampler approach and those from more traditionalprocedures. First, with the more traditional isolation, sampling isoften carried out over just 1 or 2 days, providing a “snapshot” ofDOM in time. Whereas, the samplers are deployed over one ormore weeks, providing a more integrative DOM sample over adefined time period. Thus, the DOM isolated on a single day couldtheoretically be very different from that collected over a largerperiod, for example, over 2 weeks. One could argue that this isan advantage in that the DOM isolated by the passive samplerapproach is more representative of the material present on averageand less susceptible to specific daily fluxes. Alternatively, if hightemporal resolution is required (for example, DOM samples needto be isolated every few hours), the passive sampler approachwould not be suitable.

Second, and arguably more important, when using the passivesamplers a filtration step (most commonly done under pressure)is not required. It is quite feasible that pressure filtration (whichitself is an established method for lysing cells27 and nearly alwaysrequired for conventional isolations) may in fact rupture cells orsmall aquatic species (plankton, etc.), which it turn contributematerial that inadvertently becomes operationally defined as partof the DOM. While the contribution of such “mechanical forces”is very difficult to assess in the field and across a wide range ofenvironments, the passive sampler approach does provide asignificant advantage in that filtration is not required.

Cost and Labor. Traditional isolation methods may rangefrom various resin treatments, to ultrafiltration, or simple con-centration by lyphilization. Each has their advantages and disad-vantages;25 however, nearly all either involve the transportationof large amounts of water to a laboratory or require thetransportation of extensive apparatus, including generators, filters,resins, columns, and ultrafiltration apparatus to a field site. In turn,numerous personnel and specialized transport are required. In2003, the authors carried out a study isolating DOM from LakeOntario using the DEAE-cellulose method. Approximately 500 Lof water was filtered at three different sites (1500 L total).Disposable Millipore 0.22-µm filters were used with peristalticpumps (generator required), and the filters required replacement

after ∼25 L of water at a cost of ∼$US 200 each (∼$4000 persite). Large amounts of resin, 1.5 kg per site, were used (∼$1500per site). In addition, personnel (a team of 5), large-capacitycolumns, transport, and numerous smaller consumables (silicatubing, fitting, carboys, etc.) placed isolated costs for the threesites in excess of $20 000 and cost per site in the area of $8 000-$10 000. Approximately 1 g of DOM was isolated per site, resultingin an isolation cost of $8-10/mg using the conventional approach.

In a comparative study of Lake Ontario in 2005 using thepassive samplers, each (0.25 g) passive sampler required ∼7 cmof PVDF tubing (1 m ) $33) and 0.25 g of DEAE-cellulose resin(cleaned and freeze-dried cost, $1/g) and a polycarbonate casedesigned in-house ∼$0.1 each. Thus, each sampler costs ∼$2.70each. From Lake Ontario, over a 14-day period, 72 mg wascollected using eight samplers in total at an approximate cost of$0.3/mg or ∼$21 for the entire sample (∼30 times more cost-effective than the conventional approach). As a further example,in another study DOM for Lynde shores conservation area, 54mg was concentrated on only three samplers left for 1 monthduring midsummer. This equates to ∼$8.10 for the entire sampleor $0.15/mg.

Novel Studies. Financial grounds alone may be sufficient tojustify using passive samplers for the isolation of DOM fromfreshwater aquatic environments, but as well as providing a simpleand cost-effective alternative, the passive sampler also permits arange of temporal and spatial experiments that would be verydifficult or impossible to perform using conventional approaches.For example, DOM can be monitored on a regular basis atnumerous different locations, or samplers could be set at differentdepths in large lakes to monitor how DOM at depth changes withtime. They could be deployed into hard to reach environmentssuch as wells or groundwater aquifers, where removal of largeamounts of water, needed for conventional approaches, coulddetrimentally effect the environment or may be difficult to access.Furthermore, as the samplers are low cost and easy to use, theycan be mailed to colleagues or included with expeditions goingto difficult to reach places such as the Arctic and Antarctic.

Applications are not just limited to the aquatic environment,and DOM collected from buried passive samplers in soil/sedimentcould be used to isolate the mobile fraction of DOM that is carriedthrough the terrestrial environment. Isolating organic componentsfrom soil that possess the greatest potential to be leached canhelp establish models to predict the fate and transport of chemicalcontaminants associated with this mobile fraction.28(27) Guerlava, P.; Izac, V.; Tholozan, J. L. Curr. Microbiol. 1998, 36, 131-135.

Table 1. Isolated DOM Yields Obtained with the Passive Sampler from Various Field Locationsa

sample locationnumber ofsamplers

samplingtime

totalyield

DOM yieldnormalized to

1 g of resin

Lake Ontario, Burlington, Ontario 8 2 weeks 72 mg 36 mgHighland Creek, Scarborough, Ontario 12 2 weeks 140 mg 47 mgLynd Shores Ajax, Ontario (July, 2006) 3 2 weeks 45 mg 60 mgLynde Shores, Ajax, Ontario (July-Aug, 2006)b 3 1 month 54 mg 72 mgLynde Shores, Ajax, Ontario (June 2006) 180c 2 weeks 2.79 g 62 mg

a The number of samplers is based on each sampler containing 250 mg of resin. Lake Ontario represents a large freshwater lake, HighlandCreek a fast-flowing tributary, and Lynde Shores a wetland system. b This field study was carried out in triplicate, and the yields of DOM variedby ( 2 mg for the different samplers. All other amounts are cumulative yields from all samplers combined. c These samplers were actually deployedas 60 × 750 mg samplers (equivalent to 180 × 250 mg samplers); the ratio of resin to membrane was kept the same (see text).

8198 Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

Future and Further Considerations. The advantages ofusing the passive samplers described here to isolate DOM froma freshwater environment are numerous. However, it is importantto discuss some potential drawbacks that may limit the universalapplication of this method for DOM isolation, at least in the nearfuture. First, and most importantly, the DEAE-cellulose resin isan ion-exchange-based resin. In a salt water environment, the highconcentration of Cl- ions will compete for binding sites on theresin. One sampler was deployed in the Irish sea, and while atrace amount of DOM could be recovered, the abundance of saltsmade subsequent analyses nearly impossible. The authors do notrecommend this present design to isolate DOM from salt water.Presently, work is underway to design a resin that would performbetter under saline conditions.

Another potential concern is biological growth on the mem-brane surface, this could block the pores in the sampler, and thebiological species could contribute exudates that are in turnconcentrated by the sampler. With the present design, nobiological growth was noted on any of the samplers with the PVDFmembranes presoaked in sodium azide. In fact, Figure 4 showsthat DOM isolated from the passive sampler is nearly identical tothat from batch extraction, indicating during a whole month of“summer sampling” in an algae-rich wetland, there is no evidenceof additional biological contributions to the DOM composition (ascompared to the DOM isolated by the traditional batch extraction

procedure). However, the “upper time limits” of the present designare not known and will vary between environments. Thus, futurestudies should be carried out with this in mind, especially if verylong exposure periods are planned or the membrane compositionor pretreatment is altered. With the present knowledge, theauthors recommend the samplers be deployed for a period of 2weeks, up to 1 month.

In summary, the passive samplers described here provide aneconomical and efficient device for the concentration of DOM fromfreshwater environments. Not only are they an attractive alterna-tive to traditional concentration approaches, but their easyimplementation permits a novel range of experiments that areimpossible using large-scale isolation.

ACKNOWLEDGMENTWe thank colleagues B.P. Kelleher and M. Alaee for help in

deploying samplers. C.M. Febria for help in TOC analysis, andProfessors Myrna Simpson, William Kingery, Dudley D. Williams,and Dr Emma L. Smith for their scientific and editorial input. Wethank the National Science and Engineering Research Council ofCanada (NSERC) for providing funding in the form of a discoverygrant (A.J.S.).

Received for review May 9, 2006. Accepted September 27,2006.

AC0608523(28) McCarthy, J. F.; Jimenez, B. D. Environ. Sci. Technol. 1985, 19, 1072-

1076.

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006 8199