Introduction

Anthropogenic activities lead to accumulation of heavy metals in soils and thus extensive contamination of surface soils in many countries, which are poisonous for all living organisms (Lamb et al. 2009). The nature of sorption and desorption processes affects metal solubility and bioavailability in soil (Krishnamurti et al. 1999; Jiang et al. 2012). Three mechanisms: (1) inner-sphere-complexation, (2) outer-sphere-complexation, and (3) diffuse ion swarm may be responsible for ion sorption (Sposito and Schindler 1986; Bradl 2004; Tack 2010; Marchi et al. 2006). The most important factor which affects sorption and desorption of heavy metals in soil is pH (Elliot et al. 1986).

Cadmium (Cd) is a nonessential element (Wagner 1993; Yang et al. 2009), but nickel (Ni) is an essential element for plants that are naturally present in soils. Cadmium and Ni contamination of soils is an important subject as a consequence of anthropogenic activities. Human activities such as atmospheric deposition, phosphate fertilizers, and sewage sludge disposal are the main sources of soil Cd and Ni contamination (Smolders and Mertens 2013). It has been reported that iron (Fe) is one of the important metals in soils and aquatic environments (Sunda and Huntsman 1997; Maldonado and Price 2001; González et al. 2014).

Several previous studies have taken into account one or more factors affecting Cd sorption and desorption which include pH, the content of soil organic matter (OM), clay, oxides of Fe and Mn, calcium carbonate (CaCO3), and temperature (Tipping et al. 2003; Wang et al. 2004; Li et al. 2011; Jalali and Moradi 2013). Bolton and Evans (1996) studied the retention of Cd by soils with a wide range of characteristics. These authors expressed that with increasing pH and Cd concentration, the sorption of Cd was increased in all soils.

Schulthess and Huang (1990) found that the sorption of metals by clays was dependent significantly on solution pH and clay mineral characteristics such as silicon and aluminum oxide surfaces. Soil pH is the main factor affecting on Ni solubility, mobility, and sorption, while clay content, Fe and Mn oxides, and soil OM being in the secondary of significance (Anderson and Christensen 1988; Ge et al. 2000; Suavé et al. 2000; Tye et al. 2004; Iyaka 2011).

Iron mainly is found as ferric-hydroxide under aerobic state in neutral and alkaline soils, so it is insoluble and consequently low phytoavailable in these soils (Vert et al. 2002). In well-aerated calcareous or saline soils, high pH and oxidation condition produce Fe deficiency (Schulin et al. 2010). Jalali and Hemati (2013) indicated that pH, OM, and CaCO3 were the main factor affecting Fe availability in paddy soils.

Due to the fact that the goal of this study was to verify the influence of pH on metal sorption, thus, taking into account the hydrolysis constant (pK) is important because the proportion of hydrolyzed species will change with changing soil pH. In other words, proportion of free ions decreases with increasing pH and eventually hydrolyzed species emerge and gradually increase. Metals which have lower pK will be sorbed strongly.

In addition to heavy metal sorption and desorption, the speciation and saturation index (SI) are severely affected by pH. Knowledge of metal speciation and the complex interactions between metals and sorbent surfaces is required for realization of the fate and transport of metals in soil (Yang et al. 2006). For instance due to the adsorption and precipitation of metals by carbonate in calcareous soils, carbonate has an important role in immobilization of metals (Moral et al. 2005).

So far, many studies have been investigated the influence of soil solution pH on sorption, desorption, and speciation of metals, but little research has been done about (1) Fe sorption and desorption in the wide pH range (previous studies have evaluated them at acidic pHs) and (2) the effect of CaCO3 and OM on sorption and desorption of Cd, Ni, and Fe as a function of pH by calcareous soils. Thus, the purpose of this study was to evaluate the influence of pH on sorption, desorption, speciation, and SI of Cd, Ni, and Fe by four calcareous soils with different properties.

Material and methods

Soil samples

The surface horizons (0–30 cm) of four calcareous soils varying in their properties from Hamedan Province, Iran, were sampled, air-dried, ground, and passed through a 2-mm sieve prior to being used in this study. Soil pH and electrical conductivity (EC) were determined in deionized water with a soil solution ratio of 1:5 (Rowell 1994). Organic matter content, cation exchange capacity (CEC), and calcium carbonate equivalent (CCE) were determined by the dichromate oxidation, neutral 1.0 M NH4OAc saturation, and acid neutralization procedures, respectively (Rowell 1994). Percentages of sand, silt, and clay were determined by the hydrometer method (Bouyoucos 1962). Total Cd, Ni, and Fe concentrations were measured using 4 M HNO3 for 12 h (Sposito et al. 1982). These physicochemical properties of the soils were presented in Table 1.

Table 1 Physical and chemical properties of studied soils
Full size table

Sorption of Cd, Ni, and Fe as a function of pH

One gram of air-dried soil was weighed in duplicate into 50-ml polyethylene centrifuge tubes with 20 ml of 0.6 mM of metal solution (Cd or Ni or Fe) (Merdy et al. 2009) prepared in electrolyte background concentration of 0.01 M NaCl. Suspension pH values were adjusted to 2–9 (Merdy et al. 2009) with increments of 0.5 and 0.7 M HCl or 0.05 M NaOH. The suspensions were shaken for 24 h, and their pH was measured again. A preliminary study of the metal sorption indicated that the suspensions have to be stirred for 24 h after addition of an acid or base to ensure a steady state. Samples were centrifuged at 4000 rpm for 15 min, and supernatant was filtered. Cadmium, Ni, and Fe concentrations in the supernatant were measured by atomic absorption spectrophotometry. Amounts of sorbed Cd, Ni, and Fe were calculated by the difference between the applied and final concentrations in the equilibrium solution.

Desorption of Cd, Ni, and Fe as a function of pH

In order to perform the desorption experiment, the soil separated from the supernatant solution was washed by distilled water to remove the residual metal in the remaining solution (Harter 1983). Desorption of Cd, Ni, and Fe from the soils was determined using 20 ml of 0.01 M NaCl on remained soil samples from the sorption experiments (Najafi and Jalali 2015). The soil in the tubes was then mixed using a vortex mixer and was shaken for 24 h. The equilibrated suspensions were centrifuged and filtered as previously described in the sorption experiment. Total metal concentrations in solutions were determined by atomic absorption spectrophotometry. Metals desorbed as a percentage of that sorbed was calculated for different pHs.

Speciation and saturation index

In order to estimate the speciation and SI of Cd, Ni, and Fe in the solution phase, the computer program Visual MINTEQ version 2.30 (Allison et al. 1991) was used. In addition of Cd2+, Ni2+, and Fe3+, concentrations of Ca2+, Mg2+, K+, Na+, HCO3 −, and Cl− in the equilibrium solutions were measured according to the standard methods (Rowell 1994) and used as input to the program for calculating SI and speciation (Jalali and Moradi 2013; Jalali and Hemati Matin 2015).

Results and discussion

Sorption experiments

In this study, the sorption of metals was dependent significantly on equilibrium pH (Fig. 1). Increasing solution pH leads to increase the negative surface charge on soil as well as decrease the competition of metal cations with H+ for sorption sites (Mahdavi et al. 2013), resulting in a more appropriate sorption of metal cations by soils (Garcia-Miragaya and Page 1978; Afkhami et al. 2010). Sorption of Cd and Ni showed the predictable trend of increased metal sorption with increasing equilibrium pH (the minimum and maximum sorption of them occurred at pH 2 and 8, respectively in all soils). Many authors suggested that with increasing pH, the retention of the metal cations on soil surfaces increases due to the different mechanisms (Sposito 1989; McBride 1994; Sparks 2003; Appel and Ma 2002; Usman 2008).

Fig. 1
figure 1

Sorption of metals as a function of equilibrium pH in different soils

Full size image

The narrow pH range (a transition range) in which rapid increases of sorption occur is called the sorption edge (Loganathan et al. 2012). Comparing the sorption of metals indicated that Cd and Ni initially had a gentle slope in soils 1, 3, and 4, and their sorption edge usually occurred around pHs 5–8 or 6–8, whereas for soil 2, it occurred in the lower pH range (Fig. 1). This would suggest that sorption edge decreases as the OM increases. The result agrees with the observations by Bolton and Evans (1996) that proved that most of the sorption at low pH values occurs on organic surfaces. Also, results of sorption edge and sorption isotherm study of Lee et al. (1996) indicated that different soils had very different sorption capacities for Cd. According to the average metal sorption at all equilibrium pHs, the order of Cd and Ni affinity for soils was soil 2 > soil 3 > soil 4 > soil 1. This sequence demonstrates that the presence of OM generally causes an increase in with these metals sorption. Some studies reported that the presence of high OM in soils mostly results in decrease of Cd and Ni extractability and increase of Zn extractability (McGrath et al. 1988; Arnesen and Singh 1998; Karaca 2004).

The sorption edge for Fe (pH 2–3) was lower than that for Cd and Ni in four soils. Furthermore, Fe sorption trend was different and characterized by three phases (Fig. 1); the fast increasing rate at pH 2–3, followed by a slower or steady rate up to pH 7 (only a slight difference was shown in the sorption of Fe at this pH range) and decreasing phase at pH > 7. The main reason of downward trend at pH > 7 presumably was the presence of Fe as Fe(OH)4 − particularly at pH 9; hence, this negative charge prevented Fe sorption. A notable feature of Fe sorption trend was the greatest decreasing slope in soil 2 (with 4.3 % OM) and then soil 3 (with 2.1 % OM), confirming that the OM enhances the solubility of Fe at high pH. According to Fig. 1, the minimum sorption of this metal in soil 2 occurred at pH > 7. Soil solution pH has a significant effect on Fe solution concentration (Lindsay 1972), and also, there is a significant correlation between the percentage of organic Fe complexes and soil pH (Ammari and Mengel 2006). Mashhady and Rowell (1978) demonstrated that increasing soil pH from 7.2–9.1 by adding Na2CO3, led to increase of Fe solution concentration due to dispersed OM. Also, Gerke (1997) reported that increasing soil pH caused an increase in complexation of Fe by dissolved organic matter (DOM). The least increasing and decreasing slopes (the relatively steady trend) of Fe sorption trend were observed in soil 4 (with 21.5 % CaCO3 and 1.7 % OM). This could be related to Fe precipitation. As CaCO3 content increases, Fe is precipitated as Fe carbonates and hydroxides (Seatz and Peterson 1964; Rattan and Deb 1981). Selectivity sequence of Fe sorption in the four soils was soil 4 > soil 1 > soil 3 > soil 2, which agrees with the observations by Seatz and Peterson (1964).

We determined pH50 values for sake of comparing the sorption behavior of metals. The pH50 value is the pH at which 50 % of sorption occurred (Kinneburgh et al. 1976). Sparks (2003) and Violante et al. (2008) explained that strong sorption of the soil for metals may be considered significant when 50 % of the sorption of metals occurs at lower pH. Furthermore, they reported that the pH50 position of the sorption edge of metals is affected by hydrolysis or acid–base properties. Figure 1 shows that the pH50 values for Cd are 5.1 (soil 2), 5.6 (soil 3), 6.6 (soil 1), and 6.8 (soil 4), while for Ni are 3.8 (soil 2), 5.5 (soil 3), 6.4 (soil 4), and 6.8 (soil 1). The average of pH50 values for Cd and Ni were 6.0 and 5.6, respectively. So Ni had a lower pH50 value than Cd, which is accordance with Tiller et al. (1984) results. Since the sorption percent of Fe at all equilibrium pHs was higher than 80 %, we could not apply pH50 value for this metal (Fig. 1).

The most common sequence of metal sorption at pH 2–9 for four soils was in the order of Fe ≫ Ni > Cd. In general, the sorption of Fe at pH 2 was higher than 80 % in four soils, while in the case of Cd and Ni was less than 30 %. High tendency of Fe for sorption is because of the fact that this metal is trivalent. On the other hand, this sequence is not in accordance with sequences of hydrated radius (or ionic radius), atomic weight, electronegativity, and Misono softness value, but it is correlated with pK1. Based on the pK1 of Ni which is 9.9 as against the pK1 of Cd which is 10.1 (Baes and Mesmer 1976), the sorption of Ni was greater than Cd, as was showed by Antoniadis and Tsadilas (2007).

Speciation and saturation indices of Cd, Ni, and Fe

According to the results of MINTEQ speciation solubility program, Cd speciation was significantly affected by pH (Fig. 2). This program predicted that Cd species such as free Cd (30.6–67.3 %) (average of four soils), CdCl+ (20.7–36.5 %), CdHCO3 + (1.2–2.6 %), and CdCl2 (aq) (0.8–1.3 %) were observed at all equilibrium pHs. The species of CdCO3 (aq) (0.1–40.6 %) and Cd(CO3)2 2− (0.0–4.2 %) were present at pH ≥ 7, while CdOH+ (0.3–1.7 %) was present at pH ≥ 8 (Fig. 2). Naidu et al. (1994, 1997) found that Cd-hydrolyzed species were present at pH > 8, while the predominant species at the pH range of the sorption edge was free Cd. The sorption of metal hydroxyl species to surfaces is preferential than the higher valent-free metal ions (Loganathan et al. 2012). With regard to the percentages listed for Cd species, CdCl+ after free Cd had higher percentage. As our result, a number of authors reported that Cd in soils containing high Cl− is more soluble, causing formation of CdCl+ complex (McLaughlin et al. 1994; Li et al. 1995; Norvell et al. 2000).

Fig. 2
figure 2

Cadmium species as a function of equilibrium pH in different soils

Full size image

At all equilibrium pHs, Ni predominantly was present as free Ni (40.0–93.8 %) (average of four soils), NiHCO3 + (3.9–16.7 %), and NiCl+ (0.1–0.2 %), while NiCO3 (aq) (0.1–49.6 %) and NiOH+ (0.1–3.6 %) were present at pH ≥ 6 and pH ≥ 7, respectively (Fig. 3). These are in line with that found by Merrikhpour and Jalali (2013) and Jalali and Moradi (2013) for Cd and Ni. They observed that in the pH above 7, free Cd and Ni were the major species in soil solution. As Cd, the percentage of free Ni declined markedly with increase of pH, while the percentage of carbonate and hydroxyl species increased (Fig. 3). Precipitation of metals may control their bound to the soil at the high pH values as the sorption is changed to precipitation; hence, metals are precipitated as carbonato-complexes and hydroxy-complexes (Christensen et al. 1996). Ge et al. (2000) observed that NiCO3 was the dominant species for Ni in the soil solutions at pH > 8.1. The apparent difference of Ni and Fe species with Cd species was the low affinity to form complex with Cl ion at all pHs. Christensen et al. (1996) reported that Cd–Cl complexes may exist at all pH values, while Ni–Cl complexes are imperceptible.

Fig. 3
figure 3

Nickel species as a function of equilibrium pH in different soils

Full size image

Figure 4 shows that free Fe (0.0–68.2 %) (average of four soils) is present just at pH 2–5 with a decreasing trend. The species of FeOH2+ (0.1–70.9 %) increased at pH 2–3 while decreased at pH 3–7, indicating that the Fe hydroxides were precipitated even at low pH values. At pH 2–7, Fe(OH)2 + (0.2–99.3 %) increased and then followed by a decreasing trend at pH 7–9. The percent of FeCl2+ (0.1–7.1 %) decreased at pH 2–4 and was not observed at higher pH range. The species of Fe(OH)3 (aq) (0.0–4.3 %) was present with an increasingly trend at pH ≥ 6. At pH ≥ 7, Fe(OH)4 − (0.1–88.2 %) had increasingly trend and was not observed at pH < 7 (Fig. 4). As reports by Lindsay (1979), our calculations showed that the three most abundant species in the pH range of 7–9 were Fe(OH)2 +, Fe(OH)3, and Fe(OH)4 −.

Fig. 4
figure 4

Iron species as a function of equilibrium pH in different soils

Full size image

As can be seen from Figs. 5, 6, and 7, SI of metals increased with increasing pH. In this respect, minimum and maximum SI were observed at pHs 2 and 9, respectively. The MINTEQ program demonstrated that Cd solutions were saturated with respect to otavite at pH ≥ 5 in all soils (Fig. 5). Otavite is a carbonate mineral with CdCO3 formula. So we can conclude that CO3 2− anion was responsible for Cd precipitation. It was indicated that in calcareous soils, Cd was precipitated as CdCO3 (Cavallaro and McBride 1978). Cadmium solutions were undersaturated with respect to Cd(OH)2 (s), CdCl2 (s), CdCl2:1H2O (s), CdCl2:2.5H2O (s), and CdOHCl (s) (Fig. 5).

Fig. 5
figure 5

Saturation indices of Cd minerals as a function of equilibrium pH in different soils

Full size image
Fig. 6
figure 6

Saturation indices of Ni minerals as a function of equilibrium pH in different soils

Full size image
Fig. 7
figure 7

Saturation indices of Fe minerals as a function of equilibrium pH in different soils

Full size image

For all soils, Ni solutions were saturated with respect to NiCO3 (s) and Ni(OH)2 (c) at pH ≥ 6 and pH 9, respectively, while were undersaturated with respect to Ni(OH)2 (am) at all equilibrium pHs (Fig. 6).

Unlike Ni and Cd, the MINTEQ program predicted no Fe mineral was undersaturated at pH > 4 in all soils (Fig. 7). Iron solutions were saturated with respect to hematite, goethite, and Fe(OH)2.7Cl0.3 (s) at all equilibrium pHs (the SI of them increased with increasing pH). Furthermore, solutions were saturated with respect to lepidocrocite (at pH > 2), ferrihydrite, ferrihydrite (aged) and maghemite (at pH > 3), and magnesioferrite (at pH > 4) (Fig. 7). This is not in line with that found by Moharami and Jalali (2013) who studied the sorption and desorption of Fe in contaminated calcareous soils in the presence of different cations and anions. They indicated that Fe solubility was controlled by siderite and vivianite at low and high added concentrations of Fe.

Desorption experiments

Desorption of metals was done by using 0.01 M NaCl as a function of initial pH (equilibrium pH of sorption experiment). The amount of metals desorbed from soils was strongly affected by pH (Fig. 8). Generally, Cd and Ni desorption decreased with increasing initial pH, as was observed by Gray et al. (1998) and Mellis et al. (2004). McBride (1989) indicated that at low pH, some of the adsorbed metals can be displaced by H+ ions. Furthermore, Serrano et al. (2009) indicated that in natural soils, Cd sorption at low and high pHs occurs via ion exchange and specific sorption, respectively. Thus, as stated above, Cd desorption at low pHs was greater than higher pH values. In general, Cd and Ni desorption curves were characterized by three phases: (1) the greatest desorption at pH 2, (2) the low desorption at pH 3–7, and (3) the least desorption at pH > 7. In contrast to the Cd and Ni, desorption of Fe as a function of pH has an irregular trend (a valley-like curve) and in general, when the pH increased an upward trend was observed except for soil 3 (Fig. 8). Some studies have reported this type of curve for desorption of lead (Pb), copper (Cu), and Cd as affected by organic acids (Gao et al. 2003; Yang et al. 2006; Najafi and Jalali 2015).

Fig. 8
figure 8

Desorption of metals as a function of initial pH (equilibrium pH of sorption experiment) in different soils

Full size image

The average desorption values of Cd, Ni, and Fe by four soils at pH 2–9 were 1.2 to 5.9, 1.4 to 3.9, and 1.6 to 7.5 %, respectively. Maximum desorption of Cd and Ni occurred at pH 2 and in soil 2, whereas minimum desorption of them occurred at pH 8–9 and in soil 4 (Fig. 8). So, Cd and Ni sorption in soils containing high amounts of CaCO3 was less reversible in comparison to soils containing high OM, despite the higher sorption capacity of these soils. It has been reported that the sorption of metals by carbonate is irreversible (Morse 1986). The comparison of the Cd and Ni desorption shown in Fig. 8 shows that Cd desorption is lower than Ni at higher pH values in three soils (except in soil 1 due to the low CaCO3 content), indicating that calcite has a high tendency for Cd sorption (McBridge 1980; Shaheen et al. 2013). Zachara et al. (1991) explained that the hydration energy of Cd is less than Ni, so the sorbed Cd and Ni on calcite form surface precipitate and surface complex, respectively. In general, Ni sorption was more irreversible than Cd at low pH values. Uren (1992) found that Ni is sorbed preferentially on OM than inorganic surfaces. Moreover, Stevenson (1982) reported that at pH lower than 6.0, less Cd organic complex is formed. For Fe, maximum desorption was observed at pH 9 and in soil 2 (Fig. 8). As stated previously, high pH increases the solubility of OM (Han and Singer 2007) and thus, metal sorption on soil surfaces decreases when metal–DOM complexes increase. This was proved for Pb in a study by Suavé et al. (1998) and for mercury (Hg) in a study by Yin et al. (1996). Minimum desorption of Fe occurred at lower pH ranges and in soil 4 (Fig. 8), confirming that irreversibility of Fe sorption in soil containing high CaCO3 is greater than other soils, as was observed for Cd and Ni.

Conclusions

According to the results of present experiments, the following conclusion can be achieved: (1) sorption of metals was greatly dependent on pH with a similar sorption behavior of Cd and Ni and dissimilar sorption behavior of Fe. (2) Sorption and desorption selectivity of soils for metals were dependent on pH, CaCO3 and OM content, and nature of metals. (3) Sorption experiments demonstrated that Cd and Ni sorption on soil having high OM content was higher than that on other soils, while Fe sorption on soil having high CaCO3 was higher than soils with high OM. (4) In general, the selectivity sequence of metals sorption in the four soils was Fe ≫ Ni > Cd. (5) Desorption experiment illustrated that sorption of metals on CaCO3 was less reversible than OM because sorption on carbonate minerals is often dominated by strong chemical interactions. (6) Speciation and SI of metals were influenced by pH. Free metal percentage decreased as pH increased, while the percentage of carbonate and hydroxyl species increased. (7) Cadmium and Ni solutions were undersaturated with respect to most minerals, except otavite for Cd and NiCO3 (s) and Ni(OH)2 (c) for Ni at high pH range. In contrast to the Cd and Ni, the MINTEQ program demonstrated that Fe solutions were saturated with respect to all of Fe minerals at pH > 4.