Concentrations of Cd, Cu and Mn measured in situ by DGT in river- and lake-waters was first compared with the total dissolved concentra- tions measured in filtered samples in four Australian waters [25]. Over a 72 h deployment period, the mass of Cd and Cu in two rivers increased linearly with time and matched the total dissolved concentrations. This indicated that biofouling was not a concern and that concentrations of strong complexing ligands were negligible in these waters. However, in two other rivers, the fraction of Cu and Cd measured by DGT was only 30% and 50% of the total dissolved concentration. DGT was used to obtain an in situ depth-profile of Mn in a stratified estuary, which showed a pronounced concentration maximum associated with redox- associated, reductive mobilization. Zhang and Davison [26] were able to obtain more quantitative speciation information. They made the simplifying assumption that natural waters can be regarded as having two classes of compounds that can be measured by DGT (a) labile inorganic species, including the free ion, of total concentration C i with a mean diffusion coefficient D i and (b) labile organic species of total concentration C o with a mean diffusion coefficient D o . This assumption is reasonable in humic-rich fresh-waters where the metal speciation is dominated by complexes with fulvic acid. The mass of metal ( M DGT ) accumulated by a DGT device is then the sum of the contributions from both labile inorganic (M i ) and organic complexes (M o ). M DGT ¼ M i þ M o (11.17) According to Fick’s first law of diffusion, the inorganic and organic contributions to M DGT are given by Eqs. (11.18) and (11.19), respectively. M i ¼ D i C i At Dg (11.18) M o ¼ D o C o At Dg (11.19) Combining Eqs. (11.17)–(11.19) gives M DGT ¼ ðD i C i þ D o C o ÞAt Dg (11.20) It can be rearranged to M DGT Dg D i At ¼ C i þ D o D i C o (11.21) In situ monitoring and dynamic speciation measurements 267 Chapter 12 Use of ceramic dosimeters in water monitoring Hansjo ¨ rg Weiß, Kristin Schirmer, Stephanie Bopp and Peter Grathwohl 12.1 INTRODUCTION Passive sampling with ceramic dosimeters allows for time-integrated monitoring of dissolved chemicals in ground and surface water. The purely diffusion controlled device is based on a porous ceramic mem- brane. This membrane is in the shape of a tube. The ceramic tube functions as a diffusion barrier and at the same time serves as a con- tainer to hold a solid sorbent. The latter can be selected according to compounds of interest and time scale needed for monitoring. The sor- bents are required to have a high affinity and capacity for the uptake of the chemicals of concern combined with an easy extraction at high analyte extraction recovery rates. As long as such sorbents can be found, ceramic dosimeters fit any analytical need. Diffusive transport of chemicals across the ceramic membrane at steady state can be described by Fick’s first law. Thus, the accumulated mass of a chemical at the end of an exposure period can be used to calculate the time-weighted average (TWA) concentration at which this chemical was present over the entire sampling time. Based on this, the ceramic dosimeter allows for quantification of chemical concentrations over extended periods, without the need for calibration or frequent snapshot sampling. The idea of the ceramic dosimeter was first conceived by Grathwohl [1] and by now, a number of laboratory experiments as well as explo- rations in the field have proven the suitability of ceramic dosimeters for time-integrated, long-term monitoring. Applications to date include the sampling of polycyclic aromatic hydrocarbons (PAHs) using Amberlite IRA-743 sorbent (available from Sigma-Aldrich) as solid receiving Comprehensive Analytical Chemistry 48 R. Greenwood, G. Mills and B. Vrana (Editors) Volume 48 ISSN: 0166-526X DOI: 10.1016/S0166-526X(06)48012-0 r 2007 Elsevier B.V. All rights reserved. 279 TABLE 12.1 Minimum sampling times required to reach detection limits for selected chemicals in the ceramic dosimeter a PAHs BTEX CHCs Naphthalene Phenanthrene Benzene Toluene TCE c PCE d Minimum mass (mg) b 0.09 0.12 1.2 1.35 6 6.3 Assumed aqueous concentration Predicted required sampling times 0.1 (mgL À1 ) 330 d e 1.4 a f 9 a 11 a 61 a 73 a 1(mgL À1 ) 33 d 53 d 341 d 1.1 a 6.1 a 7.3 a 10 (mgL À1 ) 3 d 5 d 34 d 41 d 224 d 267 d 100 (mgL À1 ) 0.3 d 0.5 d 3.4 d 4.1 d 22.4 d 26.7 d a Calculations are based on Eqs. (12.3) and (12.4) with T ¼ 101C, m ¼ 2, e ¼ 0.305, Dx ¼ 0.15 cm, A ¼ 8.4 cm 2 . b Minimum analyte mass detectable in the ceramic dosimeter by the method described in Bopp et al. [2]. The extraction volume in this method is 30 mL, which could be significantly reduced to improve sensitivity if needed. Detection limit is set to three times the detection limit of standard instrumental analytical methods. c Trichloroethylene. d Tetrachloroethylene. e d ¼ days. f a ¼ years. H. Weiß et al. 288 Chapter 13 Passive diffusion samplers to monitor volatile organic compounds in ground- water Don A. Vroblesky 13.1 INTRODUCTION Diffusion samplers (DSs) have been used since at least the 1990s to sample volatile organic compounds (VOCs) in ground-water [1,2].DSs can be advantageous for sampling VOCs in ground-water primarily because they have the potential to reduce costs substantially compared with pumping approaches to well sampling. The depth-specific charac- teristic of the samples can also be advantageous in certain investigations. In general, the types of DSs used to examine VOCs in ground-water can be divided into sorption devices and equilibrium DSs. Sorptive devices are discussed briefly in the following paragraph; however, this chapter concentrates on equilibrium DSs. Sorption devices for measuring VOC concentrations in ground-water typically consist of a semi-permeable membrane enclosing a sorptive medium, such as hydrophobic carbonaceous resins or polymeric resins. In this type of sampler, dissolved VOCs partition into a vapor phase in order to cross the hydrophobic membrane (examples include low-density polyethylene (LDPE) and Gore-Tex s1 ) to the internal sorption material in an air space. A simple example consists of an air-filled glass vial containing a wire coated with activated carbon. The vial is enclosed in a plastic zip-lock bag and buried in the bottom sediment in a pond. This type of inexpensive DS has been used successfully to map the zone of VOC-contaminated ground-water beneath a tidal pond where the con- taminated ground-water was discharging to surface-water [1]. Because sorptive-type samplers continue to sorb analytes until the sorptive 1 The use of trade names does not imply endorsement by the U.S. Geological Survey. Comprehensive Analytical Chemistry 48 R. Greenwood, G. Mills and B. Vrana (Editors) Volume 48 ISSN: 0166-526X DOI: 10.1016/S0166-526X(06)48013-2 r 2007 Elsevier B.V. All rights reserved. 295 Fig. 13.2. Passive vapor diffusion (PVD) sampler showing (A) an empty vial, (B) a completed sampler consisting of the open vial enclosed in two poly- ethylene bags and attached to a survey flag, and (C) a crimp cap. Fig. 13.1. Typical water-filled polyethylene diffusion bag (PDB) samp lers used in wells, including (A) a diffusion bag with polyethylene protective mesh, (B) a diffusion bag without mesh, and (C) a bag and mesh attached to a bailer bottom. Passive diffusion samplers 297 . to M DGT are given by Eqs. (11.18) and (11. 19) , respectively. M i ¼ D i C i At Dg (11.18) M o ¼ D o C o At Dg (11. 19) Combining Eqs. (11.17)–(11. 19) gives M DGT ¼ ðD i C i þ D o C o ÞAt Dg (11.20) It. Toluene TCE c PCE d Minimum mass (mg) b 0. 09 0.12 1.2 1.35 6 6.3 Assumed aqueous concentration Predicted required sampling times 0.1 (mgL À1 ) 330 d e 1.4 a f 9 a 11 a 61 a 73 a 1(mgL À1 ) 33 d 53. al. 288 Chapter 13 Passive diffusion samplers to monitor volatile organic compounds in ground- water Don A. Vroblesky 13.1 INTRODUCTION Diffusion samplers (DSs) have been used since at least the 199 0s to sample