Introduction

River sediments and suspended particulate matter (SPM) are contaminated by heavy metals containing wastes and wastewaters discharging to the rivers (Saeedi et al. 2004; Calmano et al. 1993). Although sediments are assumed to be the main sink of heavy metals due to their adsorption capacity, they could also be considered as one of the main sources of metals pollution in aquatic systems affecting the physical–chemical (e.g., pH) properties of water bodies (Saeedi et al. 2004, 2011; Calmano et al. 1993). Therefore, metal mobility, sorption, and the possibility of release are important issues in metal pollution study of sediments and particulate matter.

Given the importance of ecological and environmental significance of river systems, much research has been focused on heavy metals pollution and mobility in soil, SPM, and sediment in watershed systems in recent decades (Li et al. 2009a, b; Shi et al. 1998; Rath et al. 2009; Muller 1969; Jain 2004; Saeedi et al. 2003; Hooda and Alloway 1998; Baptista Neto et al. 2000; Hatje et al. 2003). Retention of metals onto solid phases (such as sediments and SPM) is an important issue in the study of heavy metals pollution. Most metals retention study and sorption data have been obtained using batch equilibrium techniques and adsorption isotherm models (Veeresh et al. 2003; Morera et al. 2001). Adsorption isotherms provide useful information about metals adsorption. However, the adsorption mechanism and the interaction of metals with chemical compounds in solid media cannot be described by isotherm models (Green-Pedersen et al. 1997; Morera et al. 2001; Veeresh et al. 2003). To overcome this limitation and gain better information on the mechanism of sorption, research has been performed to evaluate the risk of metal release from metal-loaded solids; it is useful to combine investigations on metals sorption with sequential extraction procedures (Morera et al. 2001; Veeresh et al. 2003).

Sequential extraction procedures are prevalent in studying metal partitioning with solid media (Keller and V'edy 1994). These procedures are based on the principle that heavy metals are in different chemical fractions and can be extracted from geochemical phases using appropriate extracting reagents (Morera et al. 2001; Veeresh et al. 2003). The sequential extraction method developed by Tessier et al. (1979) has been applied by many researchers in studying metal partitioning (Li et al. 2009a, b; Morera et al. 2001; Veeresh et al. 2003; Jain 2004; Rath et al. 2009).

During the last decade, the potential risk of metal release from sediment into the aquatic environment has been a topic of focus. A code for assessment of the risk of sediment pollution in aquatic environment was first presented by Perin et al. (1985). The risk assessment code (RAC) resulted from the sequential extraction method and different strengths of metal bonds (Jain 2004). The potential release risk for metals in this method is evaluated based on metal fractionation and the percent of metal present in different chemical bonds within the sedimentary phase. Metals which contribute to weak bonds are released more easily than those which contribute to strong ones. Therefore, it is supposed that the higher metal fraction in exchangeable and carbonate bonds is related to higher remobilization and release risk. Jain (2004), who studied metal fractionation in sediments of the River Yamuna in India, used RAC based on fractionation results and found that cadmium and lead were in the high risk category. Liu et al. (2008) also used RAC to evaluate contamination of Moshui Lake sediments in China, finding very high risk of Zn and medium risk of Cu and Ni in sediments. They concluded that rapid urbanization and industrialization were the main reasons for the pollution. Other recent research (Nemati et al. 2011; Singh et al. 2005) have also been conducted based on metal fractionation studies and application of RAC.

Although chemical partitioning of metals in sediments, release risk assessment, and sorption of metals onto sediment in different parts of the world have been studied previously, studies covering metal contamination assessment, SPM and sediment capacity for metals sorption, and evaluating the release risk of sorbed metals in different capacities through sequential chemical extraction and RAC have not yet been conducted widely.

This study examines the metal contamination in river SPM and sediments. It determines the metal retention capability and mobility in SPM and sediments. Fractionation of adsorbed metals has been carried out to identify metal fractions in different chemical bonds within the sedimentary phase. Potential metal release into the environment under different river sorption capability scenarios is also evaluated. RAC based on the results of fractionation study is applied to evaluate and compare the risk of metal release in natural and metal-loaded sediments and SPM. Langmuir and Freundlich adsorption equilibrium isotherms are also used to model sorption of metals onto different geochemical phases. New information on heavy metal contamination of riverine sediment and SPM, chemical partitioning of metal-loaded sediments and SPM, chemical sorption isotherm models, and risk assessment of metals release from natural sediment and SPM as well as those of metal loaded is provided.

Study area

The Caspian Sea is the largest lake on earth. It is located where south-eastern Europe meets the Asian continent, between latitudes 47.07′N and 36.33′N and longitudes 45.43′E and 54.20′E. It covers an area of 386,400 km2, being 1,030 km long, with a width ranging from 196 to 435 km. Its average depth varies between about 25, 780, and 1,035 m in the northern, central, and southern parts of the lake, respectively. The salinity ranges from 4‰ in the northern parts to 13‰ in the southern parts, higher than in any other freshwater lake, but still much less saline than sea and ocean water (Saeedi et al. 2003). Given its salinity level and long-term isolation from world seas, the Caspian Sean has developed a unique vulnerable marine ecology. Among its important commercial fish species, sturgeon is the most valuable, from which famous Caspian caviar, a valuable economic resource, is produced. It is also the habitat of one of only two freshwater seal species in the world. Extensive coastal wetlands offer a popular stop-off during migrations for many bird species, bringing eco-tourists. Hence, the Caspian Sea represents enormous economic and environmental potential for its region.

The Caspian environment, on the other hand, is under different high pressure. Large oil and gas reserves are beginning to be fully developed. Urbanization, coastal industries, and tourism are major sources of discharging wide varieties of pollutants into the Sea.

Major rivers flowing into the Caspian through its south coast contributing flow, transporting sediment transport, and affecting the ecology are the SefidRud, Chaloos, Babolrud, Haraz, Talar, Gorganrud, and Tadjan Rivers (Saeedi et al. 2003). Tadjan River is an important Iranian river emptying into the Caspian Sea, with a catchment area of about 4,000 km2, a length of about 170 km, and mean flow of about 13.7 m3 s−1 (Fig. 1). This river is among the southern Caspian Catchment Rivers frequented as habitat and hatchery by fish, including sturgeon. The river is exposed to different types of pollutants such as heavy metals by industrial, urban, and agricultural waste disposal (Saeedi et al. 2003). There are extensive agricultural activities in the Tadjan watershed as well as industries which discharge wastewater into the river. The most important and largest industrial plant is the Mazandaran wood, pulp, and paper plant. Tadjan River also drains storm water and domestic waste water from Sari, the capital city of Mazandaran Province. Thus, the river is exposed to a wide range of contaminants including heavy metals from agriculture and industry (Saeedi et al. 2004, 2008).

Fig. 1
figure 1

Sediment and SPM sampling site within study area

Full size image

Many studies (Karbassi et al. 2008a, b; Saeedi et al. 2003, 2004; Saeedi and Karbassi 2008; Karbassi et al. 2007, 2008a, b; Charkhabi et al. 2005) have been conducted on pollution in the southern part of the Caspian and the corresponding rivers flowing into the Sea. However, the mechanism of metals sorption on sediments and SPM, metals fractionation, and risk of their release into the water body within the Caspian and receiving rivers have not yet been studied.

Materials and methods

Sediment and SPM sampling

The major aim of this study was to examine the sorption capacity and fractionation of natural as well as metal-loaded sediment and SPM. Determination of average concentration of metals in the particulate phase of the river is not included in the present study. Sediment and SPM samples were collected from one location downstream of the most important river tributaries in terms of flow, sediment load, and subcatchment erosion. Sampling was at a location at which most tributaries merge and prior to locations where most domestic and industrial wastewaters from Sari discharge into the river. Therefore, we could study the potential capacity of sediment and SPM to adsorb dissolved metal contaminants downstream. At the same time, we could study the potential release risk of those sorbed metals and compare fractionation of metals in sediment and SPM before and after metal contamination.

Surface sediment and SPM samples of Tadjan River were collected on 22 April 2008 (Fig. 1). Five surface sediment samples were collected with a mini Birge–Ekman type grab sampler, extracting portions from the center of the dredge with a polyethylene spoon to avoid metal contamination. Equal weights of samples from each were mixed to obtain a composite sample and transferred to the laboratory in plastic bags under 4°C. SPM samples were prepared by on-site filtering of river water samples at the same sampling site, through a 0.45-μm filter, and then collecting the filter residues. Sediment and SPM were wet filtered through a 63-μm-pore-diameter sieve. Sieved samples were dried at room temperature to a constant weight in flat plastic dishes. The dried sediments were stored at 4°C in polypropylene containers until analysis and other experiments.

Characterization of SPM and sediments

Sediment and SPM particles smaller than 63 μm were used for all analyses and sorption experiments. Organic matter was estimated by recording the loss on ignition (LOI) of sieved samples for 4 h at 550°C in a muffle furnace (Glasby and Szefer 1998). The pH of water and mixtures was determined using a calibrated Cyber Scan PC 510 pH meter. Sieved, dried, and powdered sediment and SPM samples were also subjected to X-ray diffraction analysis to characterize the mineralogical composition and clay mineralogy of samples using an X-ray diffractometer, PHILIPS Expert-Pro XRD, with CuKα radiation and Ni filter. The acceleration voltage and current were 40 kV and 30 mA, respectively. Powder XRD scans were performed with the fine powder samples and 4° to 60° 2θ.

Sorption experiments

Sorption experiments were conducted by mixing of different concentrations (0, 2.5, 5, 7.5, 10, and 12 mg L−1) of metals (Cu, Ni, Mn, and Zn) with constant doses of particulate matter (3.3 g L−1 for sediment and 2 g L−1 for SPM) at room temperature (18–20°C). Different concentrations of metal solutions were prepared by dilution of 1,000 mg L−1 stock standard solution. Sorption tests were conducted in natural river water solutions at pH 7.1 to 7.3, close to the original pH of the river, 7.2 to 7.4. After adding the adsorbent, the solutions were stirred mechanically for 3 h. They were then left aside for 30 min for the particles to settle. Supernatants were sampled and filtered through a 0.45-μm Whatman paper filter before analyzing the metal concentrations.

Sequential extraction procedure

The sequential extraction procedure for metal fractionation in five steps developed by Tessier et al. (1979) was conducted on sediment and SPM samples before and after all sorption tests. The five steps were performed in the following order: (F1) weakly bond/exchangeable metals fraction (10 mL of 1 M MgCl2; pH = 7.0); (F2) carbonate bond fraction (10 mL of 1 M sodium acetate; pH = 5.0, adjusted with acetic acid); (F3) Fe–Mn oxide bond fraction [20 mL of 0.04 M NH2 OH·HCl in 25 % (v/v) acetic acid]; (F4) organic bond fraction [3 mL of 0.02 M HNO3 and 5 mL of 30 % H2O2; pH = 2 with HNO3, and also 5 mL of 3.2 M ammonium acetate in 20 % (v/v) HNO3 to prevent re-adsorption of extracted metals onto oxidized sediment]; and (F5) residual fraction (digestion procedure with HCl/HNO3/HClO4).

Metal analyses

Sediment and SPM samples before sorption and after all sorption tests were subjected to determination of total metal contents. They were digested with HNO3/HCl/H2O2 according to the U.S. EPA 3050B method (U.S.EPA 1996) to determine the total recoverable metals content. Elemental concentrations in supernatants of all digested samples as well as all extraction solutions, after each step of the fractionation experiments, were determined based on atomic absorption spectrometry (Varian AA-30 model) according to U.S. EPA 7000 s series method.

Quality control

All experiments were conducted in duplicate. Standard sediment sample (MESS-3), procedural blanks, and duplicates were run with the samples for quality assurance of the laboratory analyses. Concentrations of the elements were in satisfactory agreement with reported data (Table 1). Blank samples of spiked river water containing no adsorbents were also analyzed during all experiments to check both recovery and interferences, such as from colloidal particles, found to be negligible.

Table 1 Published and measured metal contents for MESS-3
Full size table

Risk assessment code

RAC was used in this study to evaluate the likely risk of metal release from sediment before and after sorption. This index is independent of background concentration and is applied to assess the metal bioavailability. Table 2 presents the criteria to assess the risk of sediment pollution according to RAC (Perin et al. 1985). The percentages of metal partitioning in exchangeable and carbonate fraction provide the RAC criteria. Less than 1 % participation of total metals in exchangeable and carbonate fraction is considered safe for the environment, whereas very high risk has to be considered when metal release in these fractions exceeds 50 % of the total metal content (Jain 2004; Rath et al. 2009).

Table 2 Criteria for risk assessment code (RAC) (Perin et al. 1985)
Full size table

Result and discussion

Characteristics of SPM and sediments

Weight fractions of fine (<63 μm) sediment and SPM particles in the samples were 72.1 % and 91.2 %, respectively. The loss on ignition (LOI) of sieved samples was 3.9 % for sediment and 3.7 % for SPM. XRD spectra of sediment and SPM samples are shown in Fig. 2. The presence of large quantities of quartz (42 %) and calcite (33 %) as major phases of sediment can clearly be distinguished from the X-ray spectra. While the major phases of SPM are also quartz and calcite, the percentage of calcite in SPM (27 %) is lower than that of the sediment. Dolomite, albite, chlorite, muscovite, and montmorillonite are the main clay minerals in the sediment as well as SPM. The total fraction of clay minerals in the sediment is 25 %, while it is 30 % for SPM. Bhattacharyya and Gupta (2008) reported that soil and sediment containing montmorillonite clay are usually good adsorbents because of the existence of several types of active sites on the surface- and ion-exchange sites. Montmorillonite is the most abundant clay in studied samples with 8 % and 11 % for sediment and SPM, respectively. Table 3 summarizes the weight percentages of mineral phases in the river sediment and SPM. Overall, the mineralogies of the sediment and SPM samples are similar, so great differences in metals adsorption capacity and behavior of sediment and SPM in this river are not expected.

Fig. 2
figure 2

X-ray spectrum of sediment and SPM samples of Tadjan River

Full size image
Table 3 Summary of XRD results for sediment and SPM of Tadjan River (%)
Full size table

Sorption capacity of sediment and SPM for metals

Sorption isotherms of different metals onto the Tadjan River sediment and SPM are plotted in Fig. 3.

Fig. 3
figure 3

Metals sorption capacity in Tadjan river sediments and SPM (mg kg−1)

Full size image

The sorption capacity of both sediment and SPM for metals rises with increasing concentration of metals in the river water, and, as expected, the sorption capacity approaches a constant value. Maximum sorption capacities of river sediment for the metals studied were Cu (2,191 mg/kg) > Mn (1,954 mg/kg) > Ni (1,395 mg/kg) > Zn (311.3 mg/kg). The maximum sorption capacities of river SPM were in the following order: Cu (2,082 mg/kg) > Ni (1,571.8 mg/kg) > Mn (1,125 mg/kg) > Zn (305.2 mg/kg). It is apparent that the sorption capacities of sediment and SPM for metals (Cu, Ni, and Zn) are almost equal, as they have similar mineralogy. However, sediments showed somewhat higher capacities. Concentrations of Cu, Mn, and Zn in natural sediment of the river were lower than for SPM (Table 4), whereas organic content (i.e., LOI) and calcite content were higher for sediments than for SPM. Thus, it seems that the higher organic matter, higher calcite content, and lower initial concentration of metals cause sediments to absorb more metals than SPM in this river. As seen from Fig. 3, around 43 % more than Mn is adsorbed by sediment than by SPM. This may be related to the larger size of particulate materials of sediment (27.9 % coarse material) compared with SPM (8.8 % coarse material) and to the tendency of Mn to accumulate and adsorb in larger particles. Greater adsorption of Mn by sediment than by SPM has been reported previously (Szefer et al. 1998).

Table 4 Quality of sediment and SPM of Tadjan River at sampling location
Full size table

Quality of SPM and sediments

Total metals

Concentrations of metals of interest in natural sediment and SPM of Tadjan River before conducting sorption experiments and loading by excess metals are presented in Table 4. The background concentration of studied elements in the Tadjan River watershed presented by Saaedi (2003) is also included in Table 4. It can be seen that the concentration of all metals studied are lower in natural sediment and SPM of the river than in the earth’s crust. Most of the Tadjan watershed is formed by carbonate soils, and calcite contents in sediment and SPM of river are high (Table 3). So, metal contents in particulate phase of the river are lower than in the earth’s crust. On the other hand, compared with background levels of metals in the watershed, the concentration of studied metals are rather high. This is because the sampling point is located downstream of the main tributaries that provide most of the agricultural drainage in the watershed. Hence, it is expected that the sediments and SPM should have higher concentrations than background values.

Geochemical distribution of metals—SPM

Metal fractionation of sediment and SPM samples were conducted according to the procedure outlined by Tessier et al. (1979). Results depicted in Fig. 4 show that more than 70 % of heavy metals (87.3 % of Cu, 70.1 % of Ni, and 71.7 % of Zn) are present in the residual fraction of SPM. This indicates that the mobility of heavy metals in the natural SPM of the river is limited. As a result, none of the Cu, Ni, and Zn could be released easily into the aqueous phase given the physicochemical variations of the river water. However, it should be noted that the participation of Mn in the exchangeable and carbonate phases is considerable (i.e., 50.1 %). This could be expected as Mn is known to be a relatively mobile element in the water–sediment interface in the environment (Karbassi et al. 2008a, b). It can have different chemical valences and participate in various oxidation–reduction reactions, making it reactive and mobile in contact with chemical compounds and extractors.

Fig. 4
figure 4

Metals fractionation in Tadjan River natural sediment and SPM

Full size image

Geochemical distribution of metals—sediments

More than 70 % of Zn, Ni, and Cu (81.3 %, 75.8 %, and 76.9 %, respectively) are also bonded in residual fraction of sediment so that they could not be extracted easily (Fig. 4). The participation pattern of metals in the five fractions in sediment is nearly as same as in SPM. However, residual-fraction-bonded Mn in sediment (35.5 %) is lower than for SPM (49.1 %) so that Mn in river sediments is mobilized more easily than for SPM. Most mobile Mn in both sediment and SPM are associated with carbonate bonds (fraction F2 in Fig. 4). It is 30.4 % for sediment and 23.9 % for SPM. Large amounts of calcite in sediment and SPM, higher in sediment than in SPM, could explain the higher percentage of carbonate-bonded Mn in sediment than in SPM.

Comparison of different studies

Table 5 presents a brief comparison of mean total metal contents and fractionation in sediments and SPM of some rivers around the world. While the total content of Cu in the Tadjan River sediment is almost the lowest among these rivers, association of copper with exchangeable, carbonates, and reducible bonds in sediments is considerable compared with the Odiel River in Spain, Bruntte River in Canada, and the canals of Delft in the Netherlands. At the same time, sediments of Yamuna, Haraz, Gediz, and Odra Rivers show higher Cu contents. On the other hand, the total concentration of Ni in Tadjan sediment is almost equal to that in another northern Iranian river (Haraz), lower than in the Gediz River, and higher than in the Odiel and Odra rivers. Partitioning of Ni showed that sediments of all of these rivers contain greater fractions of total Ni in their sediment than the Tadjan. Total contents of Zn as well as zinc partitioning in first three geochemical phases of sequential extraction in sediments of Tadjan River are also significantly lower than for most of the other rivers listed in Table 5. Total contents of Mn in different rivers sediments differ widely, but Tadjan sediments have highest content of Mn among the listed rivers, while in all river sediments Mn content are similar in the first three geochemical phases (e.g., around 50–60 %). Note that in comparison with sediments, none of the studies included in Table 5 contained information on the SPM of those rivers.

Table 5 Comparison of the natural sediment quality of Tadjan River with a few other rivers around the world
Full size table

Mobility of spiked samples

The metals fractionation study reveals the state of their partitioning in weak to residual bonds. Although the particulate phases of rivers are considered to be a sink for heavy metals due to their sorption capacity, they could be a source of metal pollution by releasing metals from weak bonds when environmental conditions change. Metals partitioning after sorption onto the particulate phase gives insight into possible metals release. In this study, after sorption experiments, sequential extraction was conducted to reveal metals fractionation during sorption in sediment and SPM of the river. Fractionation results are presented in Fig. 5.

Fig. 5
figure 5

Heavy metals fractionation in sediments and SPM following sorption experiments

Full size image

Geochemical distribution of metals in spiked samples—SPM

As shown in Fig. 5, during the first steps of loading metals, when SPM has capacity for additional sorption, metals mostly absorb into relatively loose bonds, mainly exchangeable carbonates and Fe/Mn oxides. As the total sorption increases and approaches a constant value, although there is still some sorption into those bonds, the percentage sorption onto those bonds remains approximately constant. This indicates that in the very first steps of dissolved metal addition into the river, most metals adsorb onto exchangeable and easily reducible phases. While association of Cu and Mn to organic/sulfur bonds is low and decreases with increasing total sorption, this geochemical phase shows more association for Ni and Zn. Geochemical fractions for Cu rank in order F1 > F5 > F2 > F3 > F4 for the first sorption step, while the order is F5 > F1 > F3 > F2 > F4 at saturation. Although the normalized pattern of distribution indicates more association with the residual phase, it should be remembered that the total amounts of Cu in other fractions will be much higher at the maximum capacity of SPM for Cu sorption. In the case of Mn, it seems that there is no significant change in the order of Mn association with the geochemical phases. It shows a uniform pattern of distribution from the first step of sorption to saturation, with rank order F1 > F2 > F3 > F4 > F5 indicating that most of Mn is adsorbed in non-residual phases, particularly in exchangeable and easily reducible bonds. Ni and Zn show normalized geochemical phase distribution patterns similar to Cu, the main difference being that more than 90 % of sorbed Ni and Zn accumulate in the non-residual phases. Their association with exchangeable and easily reducible phases is more than 70 %, leading to higher risk of release from contaminated SPM.

Geochemical distribution of metals in spiked samples—sediments

In the case of sediment, the general patterns of normalized distribution of metals in geochemical phases are somewhat similar to those of SPM. The major difference is that metals associated with residual phase show higher values than for the same metal in SPM during sorption. However, when maximum sorptions of metals take place, more than 85 % of Mn, Ni, and Zn are associated with the non-residual phase, with about 80 % in exchangeable and easily reducible phases. This indicates that in the case of river contamination, although sediments show a significant capacity (greater than SPM) to absorb metals, Ni, Zn, and Mn are at higher risk of release into the water body due to physical–chemical changes in water quality. In the case of Cu, the ultimate proportion in the residual phase is about 40 %, which is higher than for the other metals studied. Among the metals investigated, only Zn shows more association with easily reducible phases than with exchangeable ones.

Overall for sediments and SPM, the amount of absorbed Ni, Zn, and Mn increased as total metals absorption increased. However, the general pattern of metals partitioning in different bonds did not depend significantly on the change in total metals sorption. In other words, although the increase in initial metals concentration led to an increase in the metals sorption, the normalized metal distribution among the different chemical bonds did not change significantly. Moreover, participation of Fe–Mn oxide and organic bond in metals sorption experiments revealed that not only the physical adsorption (adsorbed by weak bond and carbonate) but also chemical absorption occurred during the sorption of metals onto the particulate phases of the river. This was reported previously by Morera et al. (2001). It should also be emphasized that the large amounts of metals adsorbed to weak bonds (exchangeable and carbonate fractions) in sediment and SPM resulting from river water pollution make the aquatic system very vulnerable and sensitive to water quality changes, which may cause redissolution of adsorbed metals. Hence, although the particulate phase of rivers (sediment and SPM) may reduce the potential risk of dissolved heavy metals spills, the risk of releasing metals into water remains very high compared with the case of natural particulate matters in the river before the spill. It is clearly important to study the risk of release of adsorbed metals into the aqueous phase.

Risk assessment code

The order of the bioavailability of the heavy metals in sediments may be approximated by determining the metal fractionation. The distribution of heavy metals in weak and strong chemical bonds can reveal the risk of metals release. Gibbs (1977) believed that metal bonds could be considered as more bioavailable to adsorptive, exchangeable, and carbonate phases. Changing environmental conditions may cause release and remobilization of metals. Figure 6 depicts the RAC values of the metals studied in natural sediment and SPM samples of Tadjan River as well as for metal-spiked sediment and SPM at different loading rates. The risk is assessed based on the percentage of metal in exchangeable and carbonate bonds according to Table 2.

Fig. 6
figure 6

a RAC values for studied metals in natural sediment and SPM prior to sorption experiments. b RAC values for studied metals following sorption experiments

Full size image

Based on Fig. 6a, the risk of Mn in natural sediment is higher than in SPM. For both sediments and SPM, based on the RAC criteria, Mn falls in the high-risk category, while Ni in SPM shows medium risk, and the other metals in sediment and SPM are in the low-risk category. Although the total concentration of studied metals in natural sediment and SPM of the river were lower than in the mean crust and mean sediment (Table 4), they still have some risk of release. This clearly shows that detailed geochemical fractionation studies and risk assessment are needed to determine sediment contamination by metals.

Due to metals participation in weak bonds (exchangeable and carbonate fraction), all metals have higher release risk than before the sorption tests (Fig. 6b). RAC values for Cu and Zn in both sediment and SPM fall in the high-risk category at all metal loading scenarios. Mn and Ni show higher risks of loading excess dissolved metals into the river, both of these metals falling in the very high risk of release category in the case of an increase in dissolved metal contents in the river water. Although both Mn and Ni show very high risk level in sediment and SPM of the river, Mn is at a higher level of risk of release than Ni. However, note that heavy metals like Cu, Ni, and Zn are more toxic and probably have more significant ecological and health risks in river sediment and SPM than Mn. RAC only provides an index for assessing risk of release from solid phase and overall environmental risk of sediment and SPM contamination of the river.

RAC values for all metals in spiked samples show that although metals can absorb onto all geochemical bonds within sediment and SPM, they are more likely to accumulate in weak bonds than in the other geochemical phases that lead to higher risks of release in contaminated sediments and SPM of a river.

Freundlich and Langmuir isotherm models for the experiments

The Langmuir isotherm (Langmuir 1916) and Freundlich isotherm (Freundlich 1906) models are the most common models to describe experimental sorption data. Applying these equilibrium models often provides some insight into both the sorption mechanisms and the surface properties and affinities of the sorbent (Ho et al. 2005). In the present study, both models were tested to describe the relationship between the heavy metal adsorbed and its concentration in solutions. The Langmuir isotherm model is based on the assumption of homogeneous distribution of binding sites on the adsorbent surface. This model also assumes that there is no interaction between adsorbed molecules. The linear form of the Langmuir isotherm model is expressed by:

$$ \frac{1}{{{q_e}}} = \frac{1}{a} + \frac{1}{{ab}}.\frac{{1}}{{{C_e}}} $$
(1)

where q e is the amount adsorbed at equilibrium (mg/g), C e is the final concentration of metal in the solution (mg/L), and a (mg/g) and b (L/mg) are empirical constants related to the maximum uptake and energy of adsorption. These constants can be evaluated from the intercept and the slope when 1/q e is plotted versus 1/C e .

The Freundlich model, the earliest known relationship describing the sorption, also considers monomolecular layer coverage of solute by the sorbent. A major assumption of this model is that heterogeneous sorption leads to different affinity for adsorption. The linear form of the Freundlich isotherm model by taking natural logarithms is expressed by:

$$ \ln \,{q_e} = \ln \,{k_F} + \frac{1}{n}\,\ln \,{C_e} $$
(2)

where k F and n are empirical constants related to adsorption capacity and adsorption intensity of the sorbent, respectively. The values of k F and 1/n can be determined from the intercept and slope, when lnq e is plotted versus lnC e . In this study, using data from sorption tests and metals fractionation after sorption, Freundlich and Langmuir adsorption isotherms were fitted to metals sorption data of exchangeable, carbonate, Fe–Mn oxides, and organic fractions (Fig. 7), the fractions that showed more participation and importance in the sorption of metals.

Fig. 7
figure 7

a Linearized Langmuir isotherm for adsorption of heavy metals by sediments in four fractions. Linearized Freundlich isotherm for adsorption of heavy metals by sediments in four fractions

Full size image

Figure 7a plots 1/q e vs. 1/C e for heavy metals adsorption on sediments. Fractions which are well fitted by straight lines indicate applicability of the Langmuir isotherm for those systems. Figure 7b plots lnq e against lnC e , showing that a straight line with slope 1/n and intercept lnk F is obtained. Figure 8a and b also shows that the linearized isotherm models of Langmuir and Freundlich fit sorption data for different geochemical phases in the river SPM. It seems that the Langmuir and Freundlich models, developed based on physical sorption assumptions, can be applied for sorption of metals onto different geochemical phases of sediment and SPM of rivers.

Fig. 8
figure 8

a Linearized Langmuir isotherm for adsorption of heavy metals by SPM in four fractions. b Linearized Freundlich isotherm for adsorption of heavy metals by SPM in four fractions

Full size image

Conclusions

In the present study, we have investigated the adsorbability of heavy metals (Cu, Ni, Mn, and Zn) on Tadjan River sediment and SPM and fractionation of metals before and after sorption. Results reveal that sediments contain more calcite and less clay mineral than SPM. As mineralogy of both sediments and SPM were similar, their sorption behavior and capacity also showed similarities.

Sequential extraction on sediment and SPM after sorption experiments revealed that although a considerable amounts of metals adsorb with weak bonds in particulate matter of the river (i.e., exchangeable and carbonate bonds), some metals absorb onto particulate matters in the form of chemical bonds with Fe–Mn oxides, organic matter as well as other stronger geochemical phases. Higher amounts of metals participating in weak bonds during the sorption process lead to the risk of metal release from sediment and SPM in case of pollution spills or to long-term sediment contamination by adsorption.

Data from chemical partitioning of adsorbed metals could be fitted well to both the Freundlich and Langmuir isotherm models. Although river sediments and SPM could be considered as main sink and self-purification mechanism of metals in riverine systems, the capability to remove dissolved toxic metal ions from aquatic systems should be relied on with caution.

This study shows that release risk of metals in sediment and SPM is significantly raised after sorption in comparison to the risks in natural sediment and SPM samples.