Portal de Periódicos da CAPES

Development of semipermeable membrane devices ( SPMD s) and polar organic chemical integrative samplers ( POCIS ) for environmental monitoring

2013; Wiley; Volume: 32; Issue: 10 Linguagem: Inglês

10.1002/etc.2339

1552-8618

David A. Alvarez,

Environmental Toxicology and Ecotoxicology

Since the publication of Rachel Carson's book Silent Spring 1 and passage of the Clean Water Act 2 and other laws to protect the environment in 1970s, there has been an increased awareness of the need to monitor chemical contaminants in the aquatic environment. The initial focus was on persistent organic pollutants (POPs), because they had the highest potential for bioaccumulation and biomagnification through the food chain. Analysis of POPs in the water column can be difficult, because the dissolved water concentrations of POPs are frequently extremely low (parts per billion or lower) as the result of their high octanol–water partition coefficients (KOWs). Limitations of analytical instrumentation with sufficient sensitivity to measure such low, but potentially biologically relevant, water concentrations, often prevented the use of common active sampling methods such as grab or discrete water samples. As a result, tissues and sediments became the primary matrices for monitoring POPs in the environment. Tissue samples are useful because they provide a direct measure of exposure, but they can underestimate the total exposure to the multitude of chemicals potentially in the aquatic environment. Issues of metabolism, depuration, low bioaccumulation potential of a chemical, avoidance of contaminated areas, and death of the organism can all result in a chemical not being identified. Contaminants in the water column may also occur episodically, making them difficult to measure without using an exhaustive, repetitive sampling program. These limitations necessitated the development of alternative sampling techniques to provide information on the true profile of dissolved chemical constituents in the water. Huckins et al. 3 defined a “passive” sampler as a human-made device with which the concentration of chemicals is mediated by the diffusion of chemicals from one matrix to another. The first passive sampling devices were the personal diffusional monitors developed in the early 1970s for determining occupational exposure to volatile organic chemicals (VOCs) in air 3. These monitors used a sorbent or reactive material separated from the surrounding environment by a semipermeable membrane 4. The potential exposure was measured as diffusion of the VOCs into the monitor, which allowed for the estimation of time-weighted average concentrations. The first report of using a passive sampling technique for organic contaminants in water was by Byrne and Aylott 5, who used a nonporous membrane containing a nonpolar solvent. The idea of using a passive sampler as a mimic for bioaccumulation was first reported by Södergren 6, who used a regenerated cellulose bag containing hexane to sample POPs from the water. In 1988, Huckins evaluated several combinations of nonporous membranes and solvents in an attempt to optimize the Södergren device for environmental monitoring 3. This research led to replacing the solvent, which in nearly all cases permeates the membrane and is lost to the environment, with a purified neutral lipid (triolein) contained within a low-density polyethylene (LDPE) layflat tube. This configuration maintains the desired high enrichment factor and more closely mimics accumulation into an organism 3. This lipid-filled device, referred to as a semipermeable membrane device (SPMD), immediately proved to be an extremely useful tool for measuring trace (sub–parts-per-trillion) concentrations of POPs in water. However, its acceptance as a quantitative tool was hampered by bias in the estimation of the time-weighted average concentrations from site-specific environmental conditions. The original models describing uptake into the SPMD, which were used to estimate time-weighted average concentrations, relied on sampling rates for each chemical measured, which were determined under specific conditions in the laboratory 7. When these laboratory-derived sampling rates were applied to field deployed SPMDs, the actual in situ sampling rates were often different, because diffusion into the SPMD is affected by water turbulence (which thins the aqueous boundary layer at the membrane surface), temperature, and the buildup of a biofilm on the membrane surface. Research by Booij et al. 8 (ranked 71 on the list of Top 100 most cited articles in Environmental Toxicology and Chemistry; Supplemental Data, Table S1) evaluated the uptake and release of hydrophobic chemicals with log KOWs from 4 to 8 in SPMDs under different flow regimes. Their findings showed that exposure standards added to the SPMD during construction could predict the effect of environmental conditions on the uptake kinetics of hydrophobic organic contaminants. They also demonstrated that the rate-limiting step in diffusion of chemicals in and out of the SPMD was the water boundary layer at the membrane surface and not the membrane itself, as previously hypothesized. Diffusion of a chemical through the SPMD is an isotropic process in which the uptake rate (ku) is the same as the elimination rate (ke; i.e. ku = ke). This finding laid the foundation for the development of the performance reference compound approach, which is critical in improving the accuracy of time-weighted average concentration estimates from passive samplers 9. Since the development of the SPMD, numerous other sampling devices for nonpolar organic contaminants have been developed. Some of the more widely used devices include LDPE strips without lipid, silicone strips, polymers on glass, and solid-phase microextraction (SPME) devices 3. Each of these devices has proven useful as an environmental monitor with a specific niche, such as using the SPME to measure porewater concentrations in sediments. Largely operating using the same kinetic parameters as the SPMD, these devices often reach equilibrium much faster than SPMDs as a result of lower sorptive volumes (resulting in lower capacities to retain sampled chemicals), which can prevent their use in measuring the time-weighted average concentrations of chemicals over extended periods 3. Using a sampler that reaches equilibrium quickly does have the advantages of being able to be used over shorter exposure periods and simplified models to estimate ambient water concentrations. Although the popularity of the SPMD and passive sampling techniques grew throughout the 1990s, it was evident that these devises were not sampling a large proportion of the organic contaminants present in the environment. Chemicals with a log KOW < 3 or ionic in nature have extremely low diffusion rates into the LDPE membrane of the SPMD. In the late 1990s, research was initiated for a new passive sampler to complement the SPMD and to allow passive sampling of a wide range of organic contaminants with moderate to high water solubility. This research resulted in the development of the polar organic chemical integrative sampler, or POCIS, consisting of a triphasic admixture of solid-phase adsorbents contained between sheets of a microporous polyethersulfone membrane 10. Initial characterization studies focused on the chemicals atrazine, diazinon, and 17α-ethynylestradiol as model hydrophilic compounds not readily sampled by other devices. In 2002, a national reconnaissance of pharmaceuticals and personal care products in surface waters of the United States was published 11. This was the first major effort to identify the breadth of such products in the environment, ultimately leading to a shift in the paradigm of environmental monitoring from the traditional POPs research to classes of chemicals with low bioaccumulation potentials and often thought of as benign in the environment. This newfound interest in pharmaceuticals and personal care products required adaptations to classic sampling and extraction techniques and highlighted the potential strengths of the POCIS. Research with the POCIS was expanded to include a variety of model pharmaceuticals and additional polar herbicides. To improve the extraction efficiencies of chemicals with multiple functional groups and ionic moieties, the triphasic admixture was replaced with a more universal sorbent, Oasis® HLB (Waters Corp). This adaptation, along with information on the mass transfer into the device and the kinetic (sampling rate) data for the model compounds needed for the estimation of time-weighted average concentrations, greatly expanded the utility of the POCIS for environmental sampling 12. The seminal publication by Alvarez et al. 12 (ranked 77 on the list of Top 100 most cited articles in Environmental Toxicology and Chemistry; Supplemental Data, Table S1) led to an explosion of research with the POCIS, with more than 133 citations in the subsequent 9 yr 13. During this period, more than 300 individual compounds have been shown to accumulate in the POCIS; however, the ability to predict effects of environmental conditions on the sampling rates has proven difficult as the application of the performance reference compound approach has largely failed 14. The use of a solid-phase extraction sorbent as the sequestration medium results in anisotropic exchange of chemicals (ku ≠ ke), meaning the classical performance reference compound approach does not work, with the exception of a few very polar pesticides 14, 15. This limitation may increase the error in time-weighted average concentration estimates, rendering the POCIS to a semi-quantitative status as a monitoring tool. Currently, both the SPMD and POCIS are widely used across the world. These samplers' ability to provide information on contaminants at levels much lower than achievable by other means is highly sought after in many studies. A comparison of active (grab) sampling to passive (SPMD and POCIS) techniques for a series of POPs and emerging contaminants in waters of the Potomac River basin, USA, found that 39 chemicals were detected in at least one of the grab samples, compared with 100 chemicals in the passive samplers 16. Included in the 100 chemicals were 6 biogenic hormones and sterols that were present in the POCIS but not in any of the grab samples collected. The use of the SPMD and POCIS is largely as a research tool; however, agencies at local, state, and federal levels have begun to use SPMDs and POCIS for regulatory monitoring in limited cases. The SPMD is a better-characterized sampling device than the POCIS (e.g. performance reference compounds for site-specific uptake rates), leading to a greater confidence in the results and acceptance in regulatory and legal cases. A considerable amount of ongoing research should increase our understanding of the sampling kinetics of the POCIS, which will improve the accuracy in the time-weighted average concentration estimates and may result in an expanded role in regulatory matters. Any use of trade, firm, or product names is for descriptive purposes only and does to imply endorsement by the U.S. Government. Table S1. (49 KB PDF). All Supplemental Data may be found in the online version of this article. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.