T.T. Yen Le

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/62911

#### **Abstract**

Unique properties of metals that are largely different from the characteristics of organic substances should be considered in risk assessment. The bioavailability and toxicity of metals depend on their chemical speciation, that is, physical-chemical forms, in the environment, which is largely influenced by the environmental chemistry. Since metals in the environment are not always available while organisms have developed different processes to actively regulate the body burden, assessment of metal bioaccumulation might provide a better understanding of potential risks. Metal bioaccumulation is a prerequisite for metal toxicity, but is not the only determinant of metal toxicity. In addition to metal accumulation, metal toxicity is influenced by the subcellular partitioning of metals, which is controlled by the capacity of organisms to sequester and to detoxify metals. Different modelling approaches have been developed to address some of these factors. Both empirical and mechanistic equilibrium models have been developed and applied for characterising metal speciation in the environment. Metal bioaccumulation has been predicted by biodynamic models. The ability of organisms to detoxify metals has been taken into account in assessment based on the induction of metallothionein (MT) or subcellular partitioning. In addition, the interactions between organisms and metal ions have been taken into consideration in assessment of metal toxicity based on the accumulation of metal ions at biological surfaces.

**Keywords:** Metal, modelling, risk assessment, speciation, bioaccumulation, subcellu‐ lar partitioning, toxicity

### **1. Introduction**

Environmental pollution with metals is a serious problem in many areas in the world, and assessment of metal bioaccumulation and toxicity is of high concern. In such assessment, unique properties of metals that are largely different from the characteristics of organic substances

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

should be considered. The bioavailability and toxicity of metals are controlled by their chemi‐ cal speciation, which is heavily influenced by environmental chemistry [1–5]. Metals are an intrinsic component of the environment, naturally occurring with varying background concentrations [6, 7]. Several metals are essential elements, that is, the elements that are present in living organisms and able to interact with the living system, and an insufficient amount of these elements leads to preventable or reversible effects on certain biological function, which do not occur at a physiological amount of the elements [8]. There is a limit to the regulation by the homeostasis mechanisms. Biological functions of organisms are affected by either the deficiency (below the limit) or toxicity (above the limit) of the essential metals. Because of these unique properties, organisms have developed various regulatory mechanisms for uptake, metabolism, distribution, storage, and elimination in response to metal exposure and to maintain metal homeostasis. The responses of organisms depend on a number of factors, from environ‐ mental conditions, bioaccumulation, to internal detoxification [9–11]. Metal bioaccumulation, a prerequisite for metal toxicity, is determined by the chemical speciation of metals in the environment, which depends on environmental conditions. In addition to metal accumula‐ tion, metal toxicity is influenced by metal subcellular partitioning. Different modelling approaches have been developed to address some of these factors as described in more detail in the following sections.

#### **2. Metal availability in the environment**

The bioavailability and toxicity of metals depend to a large extent on the chemical speciation of metals [12–15]. A widely recognised definition of the chemical speciation is that the chemical speciation of an element is the identification and quantification of the different, defined species, forms, or phases in which the element is present [16, 17]. Each form of the element is defined by its isotopic composition, electronic or oxidation state, or molecular structure [18, 19]. The mobility of metals in soil depends on the partition of metals between the solid and solution phases [20], which is determined by properties of the metals as well as composition of both solid and solution phases [21]. The most important soil properties include the total metal content, pH, cation exchange capacity (CEC), contents of soil organic matter (SOM), clay, oxides, and exchangeable Ca and Mg [22–26]. Besides the chemical extraction for direct determination, the solid-solution partitioning of metals in soils can be characterised by equilibrium models, which allow for predicting the speciation of a metal based on soil properties. Generally, two main approaches have been applied for modelling the solidsolution partitioning of metals in soils. They are based on: (1) empirical relations or transfer functions between the concentration of metals in soil solution or in a specific form and soil and solution properties [22, 24, 26, 27] or (2) complex process-based (mechanistic) multi-surface complexation models incorporating a detailed simulation of soil surface complexation and solution chemistry [23, 26, 28–31].

#### **2.1. Transfer functions**

The partition of metals between the solid and solution phases in soil has been expressed by adsorption isotherms or a distribution coefficient (*K*d; L/kg):

$$k\_d = \frac{\mathbf{c}}{\mathbf{c}\_{ss}} \tag{1}$$

where *K*d (L/kg) is the distribution coefficient; *C* (mg/kg or mol/kg) is the total or reactive concentration of the metal in soil; *C*ss (mg/L or mol/L) is the solution concentration.

In aerobic conditions, metals occluded in mineral particles and organic matter that are not readily available are included in the total pool, whilst the reactive fraction (extracted with 0.43M HNO3) represents the metals available for exchange with soil solution with a particular time span [32]. Therefore, the reactive concentration has been used to substitute the total concentration in the above equation [32, 33]. The partitioning of metals in soil is usually expressed by sorption isotherms as described by a Freundlich equation:

$$\mathbf{C} = \mathbf{k} \cdot \mathbf{C}\_{ss}^{u} \tag{2}$$

where *k* and *n* are the Freundlich parameters. The exponent *n* reflects the variation in the binding strength with varying total pools. Equation 2 can be rewritten and extended to take into account the influence of soil properties [22, 32, 34–36]:

$$
\log C = \log k + n \cdot \log C\_{ss} + \sum a\_i \cdot \log \left( X\_i \right) \tag{3}
$$

where *ai* is the coefficient that reflects the influence of soil parameter *Xi* on the partitioning of metals between the solid and solution phases. The distribution coefficient has also been related to soil properties:

$$
\log K\_d = a\_0 + a \cdot \log C\_\infty + \sum a\_i \cdot \log \left( X\_i \right) \tag{4}
$$

Empirical relationships have been established between the distribution coefficient or the dissolved concentration, on the one hand, and the total/reactive concentration and soil properties, on the other. Different equations have been derived, depending on the empirical data that they have been based on, whilst a full description of soil properties is usually not available. Among soil properties, pH is the most commonly integrated parameter in transfer functions, followed by soil organic carbon (SOC), while other soil properties are rarely included in. This reflects the importance of pH and organic matter in determining metal speciation as well as the partitioning of metals between the solid and solution phases.

#### **2.2. Mechanistic multi-surface complexation models**

Substantial progress has been obtained in simulating the sorption of metal ions to soil and dissolved organic matter (DOM) with the development of different models, for example, MINEQL [37], GEOCHEM [38], NICADonnan [39], and WHAM V and VI [40, 41]. Metal speciation in soil has been characterised by various models calculating metal movement and available fractions and simulating impacts of physicochemical properties on metal dynamics in soil. The movement of different metal species in an environment is characterised by fate and transport models, while the partitioning of metals into dissolved, colloidal, and particulate phases is predicted by speciation/complexation models. Assemblage models have been formed by incorporating various surface complexation models. These process-based models provide a detailed description of metal complexation on different sorption surfaces [39, 42]. Surface complexation models based on thermodynamic processes have been developed for describing reactions of metals on the surface of soil reactive phases, such as SOM [39, 41], oxides [42, 43], and clay minerals [44]. Multi-surface complexation models have been formed from individual surface complexation models, assuming that the adsorption properties of soil can be charac‐ terised by the sum of the individual adsorption properties of individual components. Metal concentrations in soil solution have been predicted by using mechanistic geochemical models accounting for interactions of metals with various reactive surfaces [23, 24, 26, 29, 30].

Originally developed to characterise metal speciation in oxic waters [40, 41], WHAM has been applied to simulate the solid-solution partitioning [23, 45]. This model consists of several submodels: Humic Ion-Binding Model V and models of inorganic solution chemistry, adsorptiondesorption reactions of fulvic acids, precipitation of oxides, and cation exchange reactions on clays. In the Humic Ion-Binding Model V, humic substances are represented by molecules containing proton-dissociating groups that can bind to metal ions [40]. The binding of metal ions to humic substances is described in terms of complexation at discrete sites while taking into account both electrostatic interactions and competition among ions. In the application to soil systems, a fixed charge mineral cation exchanger was included to describe the presence of clays [46].

Weng et al. [29, 47] have proposed another multi-surface or mechanistic model for character‐ izing metal speciation in which soil is considered a set of independent sorption surfaces, that is, organic matter, clay silicates, and iron (hydr)oxides. Metal binding to these surfaces is simulated using advanced adsorption, cation exchange models, and default binding parame‐ ters. Humic acid is considered an analogue for SOM. The site density of SOM is assumed to be equal to 31% of the density of humic acid [47], while metal binding to SOM is described by using the non-ideal competitive consistent Adsorption (NICA)-Donnan model [39]. Clays are represented by illite because of their importance in Dutch soils [48]. The average charge of illite is assumed to be pH dependent (0.25 mol/kg; [49]). An electrostatic Donnan model is used to describe metal binding to clay. For amorphous iron (hydr)oxides, their amount is predicted as the oxalate-extractable Fe, and their surface area is considered similar to that of hydrous ferric oxices (HFO; 600m2 /g) [42]. For crystalline iron (hydr)oxides, their surface area is assumed to be the same as that of goethite (100 m2 /g), and their amount is calculated as the difference between total (aqua regia) and oxalate-extractable Fe. Metal binding to the oxides is simulated using the two-site surface complexation diffuse double layer model [42]. The complexation of metals with carbonate and dissolved organic matter (DOM) is also taken into account in the model. DOM is assumed to consist of 30% humic acid and 30% fulvic acid [29].

The ORCHESTRA (Objective Representing CHEmical Speciation and TRAnsport) modelling framework [50, 51] is a combination of different sub-models. Thermodynamic data from the MINTEQA2 [52] were used to calculate solution speciation. The sorption of ions to SOM and DOM was simulated with the NICA-Donnan model [39, 53, 54] using generic binding param‐ eters and constants derived by Milne et al. [55, 56]. When data on DOM were not available, DOM was assumed to consist of 50% reactive humic substances. In addition, SOM and DOM were represented by humic acid assuming that humic acid consists of 50% carbon [56]. The sorption of ions on the surface of (hydr)oxides was described by the generalized two layer model (GTLM) of Dzombak and Morel [42]. Site densities of the amorphous iron and alumi‐ nium (hydr)oxides were calculated from the assumed specific surface area of 600 m2 /g for hydrous ferric oxide [42]. Crystalline iron (hydr)oxides were represented by the hydrous ferric oxide, assuming the same reactivity. Moreover, a specific surface area of 100 m2 /g was assumed for these (hydr)oxides [43]. The nonspecific sorption of ions to permanently charged clay surfaces was modelled with a Donnan model. The illitic clay mineral with an average charge density of 0.25 eq/kg and a fixed Donnan model of 1 L/kg were used to represent the clay mineral [49].
