Active Sampling Methods for Water and

Active sampling techniques represent the most widely used approach for the collection and extraction or trapping of contaminant residues in water. "Active sampling" refers to those methods that require physical intervention or external energy input for sample collection and/or residue extraction or trapping. Often, samples are excised or removed (i.e., grab sampling) from the exposure medium before residues are extracted. Traditionally, grab samples of water were nearly always extracted with organic solvents (i.e., liquid-liquid extraction, LLE). Using the LLE approach, the data generated are limited to the total chemical concentration in all waterborne phases, which includes microorganisms and algae, particulate organic carbon (POC), dissolved organic carbon (DOC), inorganic particulates, and the dissolved phase. Thus, the fraction of the total waterborne residues represented by the dissolved phase (i.e., the most readily bioavailable phase) is not distinguished from generally less bioavailable residues in other waterborne phases. To reduce the use of organic solvents and to discriminate between aqueous residues associated with POC-DOC and the more bioavailable dissolved phase, solid-phase extraction (SPE) systems were developed. This method is based on the percolation or pumping of water or air samples through columns, tubes or cartridges of sorbents consisting of polymeric phases bound to silica cores or various types of polymeric beads or foam. The SPE approach also includes Empore extraction disks (Kraut-Vass and Thoma, 1991). These disks or membranes consist of a Teflon fibril network loaded with SPE sorbents, but the particle size of these sorbents is smaller than those described for standard SPE columns. Although methods using SPE cartridges or columns and Empore disks are categorized as active sampling, the sample extraction involves diffusional and partitioning-sorptive steps. Thus, at some fundamental level active sampling involves passive (defined later) processes.

In cases where water is turbid, samples are generally filtered through glass fiber filters (GFF) prior to percolation through sorbents. This step recovers waterborne particulates and microorganisms with average diameters >0.7 |im, which are analyzed separately. However, chemicals associated with colloid-sized particulates and DOC are not removed by GFFs.

More recently, solid-phase micro extraction (SPME) fibers have gained widespread acceptance as equilibrium samplers for the extraction of water samples (Arthur and Pawliszyn, 1990). The SPME fiber consists of a small-diameter fused silica fiber coated with one of several polymeric phases for the sorption of analytes. The polymeric-film thickness for commercially available SPMEs generally ranges from 7 to 100 |im with the total phase volume of a 1 cm segment being 0.028 to 0.612 |L. Traditionally, SPME methods are only applied to excised samples and stirring is used to expedite analyte extraction or times to equilibrium. Also, analyte concentrations in SPMEs generally represent the total residues in a water sample (Mayer, 2003). The approach has a number of advantages over active sampling methods, which include the elimination of pre-filtration and organic solvent extraction steps, typically required for sample preparation, and the direct injection of the total sample into a gas chromatograph (GC) for analysis.

Nearly all grab sampling methods suffer from potential problems with sample preservation such as losses due to volatilization, sorption to container walls, and chemical degradation. For water samples, some of these problems can be avoided by the addition of an appropriate "keeper" solvent immediately after sample collection. However, the use of SPEs and SPMEs is thereby precluded and the ability to discriminate residue distribution among water-borne phases is lost. Data from grab samples and other active sampling methods represent only a single point or small window in time, which does not account for temporal variations in contaminant concentrations at study sites. Thus, adequate assessment of organism exposure requires labor-intensive multiple sample collections. Generally, the volume of collected samples is limited to <5 L, due to the difficulty in the handling, transport, processing, and extraction of large amounts of water. Consequently, method quantitation limits (MQL) may not be adequate for the analysis of trace (< 1 |xg L-1 or mg m-3) or ultra-trace (<1 ng L-1 or |xg m-3) levels of hydrophobic organic chemicals (HOCs). These relatively low quantitation limits are especially needed for assessing the environmental significance of HOCs that bioconcentrate (uptake from water by respiration or skin absorption) or bioaccumulate (uptake via skin absorption, respiration, and diet), and for some highly toxic organic compounds, which may or may not bioaccumulate.

To overcome the limitations of grab sample preservation and method sensitivity, in situ large-volume SPE systems, equipped with submersible pumps, have been developed (e.g., the Infiltrix Column System by AXYS Environmental Systems, Sidney, BC, Canada) for sampling aquatic environments. An alternative method for ultra-trace analyses of HOCs in water is the Goulden large sample extractor developed at the Canada Centre for Inland Waters (Forbes and Afghan, 1987). Detection limits are much lower for these types of systems (Rantalainen et al., 1998), but significant concerns still exist about sample contamination, an-alyte losses to exposed surfaces, filter plugging in turbid waters, the use of toxic chlorinated solvent, and procedurally mediated changes in the distribution of contaminants among phases constituting environmental waters. For many studies or programs involving multiple sites, in situ large-volume samplers are often too labor intensive and costly to use at all sites. Thus, simultaneous replication of sampling for statistical purposes is seldom performed.

Methods used for the active sampling and extraction or concentration of organic vapors in air are generally related to those described for water. For example, samples of volatile organic compounds (VOCs) and semi-volatile organic compounds (SVOCs) are often collected and concentrated by in situ pumping of air through SPE tubes or cartridges. Even when grab sampling is the method of choice (e.g., VOC sample collection in Summa-polished canisters and Tedlar bags), SPEs are often used for sample preconcentration or trapping prior to instrumental analysis. In all cases, GFFs are used when discrimination between the vapor and particulate phases is needed to estimate the relative contributions of the two-exposure pathways (e.g., SVOCs with log Koas > 8.5). The SPE sorbents used to concentrate vapors of trace to ultra-trace levels of SVOCs in large volumes of air include polyurethane foam plugs, Tenax and XAD-2 resin (Ockenden et al., 1998). These sorbents are also used for the extraction of ultra-trace levels of dissolved-phase waterborne residues (Rantalainen et al., 1998).

Many of the shortcomings listed for active sampling of waterborne residues apply to the analysis of VOC and SVOC vapors in air. Also, sampling analytes in an equivalent mass of air and water requires about a 103 larger volume of air, thereby practically limiting the applicability of grab sampling for the analysis of trace and ultra-trace SVOCs in air. Furthermore, because sampling sites for assessing global-atmospheric transport of contaminants are often in rugged terrain in remote locations, the choice of sampling methods may be limited by the size, weight and portability of the sampling apparatus and the need for electrical power.

In summary, active sampling and extraction or trapping methods provide reasonably reliable information on the total waterborne and airborne concentrations of HOCs, but only during one point in time or a relatively brief interval of time, which is in marked contrast to the exposure duration of most organisms. Most of these methods permit some discrimination between analyte concentrations in material trapped by the filter, representing the total residues associated with POC, inorganic particulates, and microorganisms (e.g., algae and spores), and dissolved and vapor phase residue concentrations in filtrates. Unfortunately, the potential effects of sampling, transport and filtration on the environmental distribution of contaminant residues among the various phases in water and air are difficult to predict a priori. For example, solutes and vapors can adsorb on GFFs and be misidentified as part of the particulate phase, while residues associated with fine particulates can desorb and be incorrectly identified as part of the dissolved phase (Mackay, 1994). Thus, residue concentrations in filtrates are not necessarily representative of the dissolved or vapor phases in the undisturbed sample media. With the exception of programmable in situ active sampling systems (e.g., the Infiltrix water sampling system and high volume [HiVol] air samplers) sample size may not be adequate for the analysis of trace and ultra-trace levels of contaminants. Also, active sampling methods (excludes the Infiltrix system) are generally relevant only to a few points in time and do not provide time-weighted average (TWA) concentrations. Strictly speaking, measurement of TWA concentrations during a specified time period requires continuous, additive extraction (i.e., integrative sampling, where the extraction medium acts as an infinite sink) of an exposure medium. TWA concentration data are useful indicators of organism exposure to HOCs.

+1 0

Post a comment