Introduction

Diffusive gradients in thin-films (DGT) was first used in the mid-1990s as an in situ technique for dynamic trace metal speciation measurements [1,2]. It has since been developed as a general monitoring tool for a wide range of analytes in addition to the transition and heavy metals originally measured, including the major cations, Ca and Mg [3], stable isotopes of Cs and Sr [4], radionuclides of Cs [5] and Tc [6], phosphate [7] and sulphide [8]. In a comprehensive study, Garmo et al. [9] demonstrated the capabilities of DGT to measure 55 elements with a Chelex® 100-based resin-gel.

As its name implies, DGT relies on the quantitative diffusive transport of solutes across a well-defined gradient in concentration, typically established within a layer of hydrogel and outer filter membrane. The filter membrane is exposed directly to the deployment solution and acts as a protective layer for the diffusive gel. Once diffusing through these outer layers, solutes are irreversibly removed or chelated at the back side of the diffusive gel by a selective binding agent, typically Chelex 100, which is immobilized in a second layer of hydrogel. The hydrogels used in DGT are typically made of polyacrylamide, which can be fabricated with a range of properties, including almost unimpeded diffusion due to the gel having a water content as high as 95% [10].

The pre-filter, diffusive gel and binding-gel layers are assembled into an all plastic sampling device comprised of a base and cap (Fig. 11.1). The cap is push-fit onto the base to provide a water-tight seal and has an opening or 'viewing window' that exposes a known area of the filter to the deployment solution. The theoretical basis for the use of DGT

Comprehensive Analytical Chemistry 48

R. Greenwood, G. Mills and B. Vrana (Editors)

Volume 48 ISSN: 0166-526X DOI: 10.1016/S0166-526X(06)48011-9

© 2007 Elsevier B.V. All rights reserved. 251

DGT Sampling Device

DGT Sampling Device

Chelex 100

Fig. 11.1. The DGT sampling device is composed a base and cap, which contains the pre-filter, diffusive gel and resin-gel layers. The DBL extends out from the device face into the bulk water with a concentration (Csoln).

Chelex 100

Fig. 11.1. The DGT sampling device is composed a base and cap, which contains the pre-filter, diffusive gel and resin-gel layers. The DBL extends out from the device face into the bulk water with a concentration (Csoln).

in aqueous solutions has been systematically developed over the past decade [1,7,11-14].

Transition and heavy metals have usually been measured using a binding layer of Chelex 100, known as the resin-gel layer. To lay the groundwork for their measurement by DGT, initial experiments were carried out to establish performance characteristics, including the capacity of the resin layer [2,15], the diffusion coefficients of metals and their complexes [16,17], pH dependence [2,15,18], the elution factor of metal ions from the resin-gel [2,9,10,19-21] and the effects of solution composition, flow and deployment time [18,21]. Effects of waters with very low ionic strength (< 1 mmol L-1) on DGT measurements reported by several workers [19,20,22] have been shown to be largely due to incomplete washing of the gels [23].

DGT has been deployed in situ in a wide range of natural waters, including fresh-waters, such as soft Canadian shield lakes [19], a eutro-phic hard-water lake [24] and several rivers [16,25-28], coastal sea-water

[29] and open ocean sea-water [1,2]. The early development of DGT, including applications in waters, soils and sediments, has been reviewed

[30]. Here, we focus on its further development and applications in aqueous systems, which include its use (1) for speciation measurements [12,16,24-29,31-35] and kinetic tools [11], (2) for bioavailability studies [36-40] and (3) for routine environmental monitoring [41-44].

In situ monitoring and dynamic speciation measurements 11.2 METHODOLOGY 11.2.1 Gel preparation

Details of the procedures for preparing the diffusive gel and the resin-gel used for trace metal measurements are well documented [2,30] and can be obtained directly from DGT Research Ltd. (Lancaster, UK). Procedures for preparing specific binding gels for other solutes can be found in the individual scientific publications. The diffusive gel most commonly used is prepared from a polyacrylamide gel cross-linked with an agarose derivative and is referred to as APA2 [17]. It is prepared by mixing 3.75 mL of acrylamide solution (40%) and 4.75 mL of deionized water with 1.50 mL of DGT cross-linker (DGT Research Ltd.) (0.3%, measured by weighing 1.50 g). This 10 mL of gel solution is well mixed using a pipette and to it a further 70 mL ammonium persulphate solution (10%, 0.1 g in 1g of H2O) and 25 mL of N,N,N'N'-tetramethylethylenediamine (TEMED) (99%) are added and well mixed using a pipette. The solution is then cast between two acid-cleaned glass plates, separated by a 0.5 mm plastic spacer (for 0.08 cm diffusive gels), and immediately placed into a 45°C oven for 1h. Once the gels are completely set, they are removed from the glass plates and placed into a deionized water bath (two gel sheets per litre of deionized water), which is changed repeatedly until all the excess polymerization products are removed, i.e. the pH of the wash solution is equal to that of the deionized water it is stored in. Finally, the diffusive gels are conditioned in a separate solution of 10-30 mmol L 1 NaNO3 before use.

The Chelex® 100 resin-gel used for measuring trace metals is prepared by mixing 3-4 g of Chelex® 100 (200-400 mesh), available from Bio-Rad Laboratories (USA), with 10 mL of gel solution consisting of 3.75 mL of acrylamide (40%) (BDH Electran®), 1.5 mL of DGT cross-linker (DGT Research Ltd.) and 4.75 mL deionized water. To this solution, 50 mL of ammonium persulphate (BDH Electran®) and 15 mL of N,N,N'N'-TEMED (BDH Electran®) are added. The resin-gel solution is immediately cast between two glass plates, which are separated by a 0.025 cm acid-cleaned plastic spacer and held together with plastic clips. Once cast, the glass plate assembly containing the gel with incorporated resin, known as resin-gel, is placed into an oven set at 45°C for ~1h. Once the gel sheets are set, they are placed into ultra-pure water and allowed to hydrate before use by repeatedly changing the wash solution until all the excess polymerization products are completely washed from the gel. As with the diffusive gels, the resin-gels are also placed into

10-30 mmol L 1 NaNO3 before use. For low-level trace metal work, all handling and processing of gels, up until the time of deployment, should be carried out in a laminar flow Class-100 clean bench using 'ultra-clean' trace metal techniques.

11.2.2 Diffusive gel variants

DGT has generally used a diffusive hydrogel prepared from an acryl-amide monomer cross-linked with a patented agarose derivative (DGT Research Ltd.). Originally, the diffusion coefficients of metal ions in this gel (DGel), known as APA1, were found to be very similar to their diffusion coefficients in water (Dw). However, changes in the manufacturing process of the cross-linker resulted in a gel with a slightly smaller pore size, which is now referred to as APA2. These two gels are notionally the same, i.e. 15% acrylamide with 0.3% agarose cross-linker, but the diffusion coefficients of metal ions in APA2 are approximately 0.85 x Dw. A third type of gel in common use, APA3, uses a reduced amount of cross-linker (0.12%) and has a pore size closer to the original APA1 gel. Additional gel types include a restricted gel (RG) with a much smaller pore size that appreciably retards the diffusion of complexes with fulvic and humic acids [10,17]. It uses bis-acrylamide (BDH Electran®) as the cross-linker (0.8%) and 7 mL of ammonium persulphate and 2 mL TEMED per mL of gel solution.

DGT holders have also been used with a chromatographic paper as a diffusion layer [45], but required individual calibration at ionic strengths ^5mmol L 1 due to increased effective diffusion coefficients, likely due to the paper having a negative charge. The authors found them easy to prepare and handle. DGT has also been successfully used with a membrane filter as the diffusion layer [1,21].

11.2.3 Alternative binding agents

Numerous variants on the diffusive and resin-gel types originally used in DGT have been used, including poly(acrylamideoglycolic acid-co-acrylamide) for selective binding of Cu2+ [46]. Under competitive binding conditions, this resin showed a stronger binding affinity for Cu. When tested as a binding agent for the DGT technique, it was shown to have a linear mass response with respect to time, with CDGT/Csoln ratios of approximately 1.

An alternative approach to using ion-specific resins is the use of co-polymer hydrogels composed of polyacrylamide-polyacrylic acid [47]. These gels selectively bind Cu2+ and Cd2+, over alkali and alkaline earth metals, within the hydrogel structure. DGT devices containing a liquid binding phase, poly(4-styrenesulphonate) (PSS) aqueous solution and a cellulose dialysis membrane (CDM) as a diffusive layer, have also been proposed [48,49], with the intent of excluding the availability of copper-organic matter complexes to the device.

A commercially available strong cation exchange membrane (Whatman P81) has also been used as the binding layer in DGT [50]. Like Chelex 100, it preferentially binds transition metals over competing divalent matrix cations such as Mg and Ca. DGT devices with this binding layer showed good agreement with theory and its performance was not degraded even after four consecutive uses. A systematic comparison of DGT sampling devices containing different binding agents showed that DGT concentrations measured by the different binding agents can be significantly different and suggests that DGT labile concentrations may depend on the binding strength of the binding agent [51]. However, as above a low threshold, the binding strength would not be expected to affect measurement of a complexed metal, further work is required to establish this observation [2].

DGT has been used to measure phosphate in natural waters using a binding layer of ferrihydrite embedded in gel [7]. This binding layer has also been used for the measurement of As [52]. Mason et al. [53] have used a mixed binding layer (MBL), consisting of both Chelex 100 and ferrihydrite, to measure both cations (Mn, Cu, Zn and Cd) and anions (molybdate and phosphate) in a single measurement. As all elements were measured by ICP-MS, including 31P, 1mol L 1 HCl was used for elution to minimize interferences from 14N16OH and 15N16O. Measurements of both cations and anions by devices with a MBL were similar to those made using standard DGT resin-gels containing either Chelex 100 or ferrihydrite. An alternative binding agent, suspended particulate reagent-iminodiacetate (SPR-IDA) available from CETAC Technologies Inc. (USA) with similar iminodiacetate functionality to Chelex 100, has been used [30]. While a systematic characterization of its performance has been carried out [15], due to the elevated cost of this resin type and its smaller bead size, it is most appropriate for laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) for determining trace metal distributions in sediments at high spatial resolutions in 2D [54].

K.W. Warnken, H. Zhang and W. Davison 11.3 DGT THEORY

11.3.1 DGT principles

At steady-state, the diffusive flux (F) of an ion is given by Fick's first law of diffusion where dC/dx is the change in concentration (g cm-3) occurring over the distance x (cm):

The diffusion coefficient measured at infinite dilution and a reference temperature of 25°C, D0, can be corrected to any in situ temperature, Dt, by applying the Stokes-Einstein equation, where T0 and Tt are in Kelvin (K):

The viscosity of water can be expressed by the following equation [55] where *0 is the viscosity of water at the reference temperature of 25°C and *t is at the in situ temperature t (°C):

In DGT with the commonly used APA2 gel, the diffusion coefficient of divalent metal cations in the diffusive gel layer, DGel, can be approximated by 0.85 x D [17]. At steady-state, the concentration gradient, dC/dx, is the difference between the concentration in the bulk solution and the concentration at the interface between the diffusive and resin-gel layers, C, which is 0 if the resin-gel layer is a rapid and effective sink. The distance x is the diffusional path length, which is the combined thickness of the diffusive gel layer, the protective membrane filter and the diffusive boundary layer in solution, DBL (see Fig. 11.1 and later). Here we simplify the system by assuming that the thickness of the DBL is negligible and the diffusion coefficient for the other two layers, with a combined thickness of Dg, is the same:

The flux (F) is equal to the mass (M, in g) of metal through an area (A, in cm2) per unit time (t, in s):

0 0

Post a comment