A series of Fe–Mn binary oxides with different Fe:Mn ratios (1:1, 3:1, 6:1, 9:1) were synthesized to investigate the optimal Fe:Mn ratio for the removal of As(III) and As(V). Batch experiments were performed to determine the rate of adsorption and equilibrium isotherms. Adsorption kinetics were well described by the pseudo-second-order kinetic model for both As(III) and As(V). The adsorption equilibrium data fitted well to Langmuir and Freundlich isotherms. The maximum As(V) sorption capacity was observed at an Fe:Mn ratio of 6:1 (65.0 mg/g), whereas maximum As(III) uptake was at Fe:Mn ratio 3:1 (46.9 mg/g). Arsenic levels in real water samples were reduced from 37 μg/l to below the EU Water Framework Directive limit (10 μg/L) after treatment with Fe–Mn adsorbents.

INTRODUCTION

Arsenic is a toxic and carcinogenic chemical element widely distributed in the atmosphere, soils and natural waters, and arsenic contamination is considered to be one of the most serious environmental problems today (Chandra et al. 2010). The presence of arsenic in groundwater used for water supply is a current national and global problem. Typically, widespread arsenic contamination of groundwater is caused by natural processes and a range of anthropogenic activities (Smedley & Kinniburgh 2002). The toxicity, mobility, and bioavailability of arsenic are highly dependent on its oxidation state and chemical speciation (Deschamps et al. 2003). In natural water, arsenic is primarily present in inorganic forms and exists in two predominant species, arsenate (As(V)) and arsenite (As(III)). As(V) is the major arsenic species in well-oxygenated water, whereas As(III) is the dominant arsenic in groundwater (Smedley & Kinniburgh 2002). As(III) is much more toxic, more soluble, and more mobile than As(V). Although the conversion of As(III) to As(V) in oxygenated water is thermodynamically favoured, the rate of the transformation may take days, weeks, or months depending on specific conditions (Zhang et al. 2007).

Among the various As removal technologies such as coagulation, iron-exchange, ultrafiltration, reverse osmosis and biological treatment, adsorption processes are considered some of the most promising, with the advantages of having high removal efficiency, simple operation, low cost and high recycling rate without harmful by-products, meaning systems can be very cost-effective (Ma & Zhang 2008; Zhang et al. 2012; Cui et al. 2014). Among potential adsorbents, iron (hydr)oxides and iron-containing substances have been widely focused on for arsenic species removal due to their strong affinity and high selectivity for the sorption of inorganic arsenic species (Pierce & Moore 1982; Driehaus et al. 1998; Dixit & Hering 2003). However, their removal of As(III) is less effective than that of As(V). Oxidation pretreatments are thus usually employed to convert As(III) into As(V) prior to adsorption (Bissen & Frimmel 2003). Manganese oxides have been extensively used as the oxidizing agent, but their adsorption capacity is low (Lenoble et al. 2004) and this limits their application. Conceivably, a Fe–Mn binary composite that combines the oxidation property of manganese dioxide and the high adsorption of iron oxides would be able to oxidize As(III) and have high adsorption capacity for As(V) simultaneously.

Therefore, the main objectives of this research were to synthesize a series of Fe–Mn binary oxides with different Fe/Mn molar ratios by an oxidation and coprecipitation method, characterize the adsorbents and finally evaluate the adsorption mechanisms of As(V) and As(III) on the Fe–Mn binary oxides in synthetic and real water matrices.

MATERIALS AND METHODS

Materials

All chemicals were of analytical grade and used without further purification. Reaction vessels (glass) were cleaned with 1% HNO3 and rinsed several times with deionized water before use. As(III) and As(V) stock solutions were prepared with deionized water using As2O3 and As2O5 (Alfa Aesar GmbH, 99.9% purity), respectively. Arsenic working solutions were freshly prepared by diluting arsenic solutions with deionized water.

Adsorbent preparation

Fe–Mn binary oxides with different Fe/Mn molar ratios (1:1, 3:1, 6:1 and 9:1) were prepared at room temperature (25 °C) in the laboratory, according to a method modified from that proposed by Zhang et al. (2007). In deionized water, 200 ml solutions of 11.85 g/L KMnO4 were prepared; 200 mL solutions of 1:1 FeCl3:FeSO4 were also prepared, with combined iron concentrations of 20, 60, 120 and 180 g Fe/l. Under vigorous magnetic stirring, the FeCl3:FeSO4 mixtures were added into the KMnO4 solutions, and 5 M NaOH solution was simultaneously added to keep the solution pH in the range of 7–8. After addition, the formed suspension was continuously stirred for 1 h, aged at room temperature for 24 h, and then washed repeatedly with deionized water. The suspension was filtered and dried at 110 °C for 4 h. The dry material was crushed and stored in a desiccator for use. The obtained material appeared in the form of fine powder.

Adsorbent characterization

The specific surface area was measured by nitrogen adsorption using the Brunauer–Emmett–Teller (BET) method with a Quantachrome Autosorb™ iQsurface area analyzer. The particle sizes were observed using a scanning electron microscope (SEM) (JEOL JSM 6460 LV). The point of zero charge (pzc) was determined according to the inert electrolyte titration method described by Kinniburgh et al. (1975): the Fe–Mn binary oxides were suspended in 0.01 M NaNO3 for 24 h, after which the rate of pH change with time was very slow; 20 mL of suspension was then adjusted to various pH values with 0.1 M NaOH or 0.1 M HNO3 solution. After agitation for 60 min for equilibrium, the initial pH was measured; then, 1.5 g of NaNO3 was added to each suspension to bring the final electrolyte concentration to about 0.45 M. After an additional 3 h, the final pH was measured. The results, plotted as ΔpH (final pH–initial pH) against final pH, yielded the pzc as the pH at which ΔpH = 0.

Batch adsorption tests

Sorption experiments for the kinetic study were conducted as follows: 10 mg of Fe–Mn binary oxide was suspended in a 20 mL solution containing 10 mg/L of As(III) and As(V). Solution pH was adjusted to 7 ± 0.2 with 0.1 M HNO3 or 0.1 M NaOH, and the vials were placed on an orbital shaker at 180 rpm. At predetermined times, samples were taken and filtered with a 0.45 μm membrane filter and the concentration of arsenic in the filtrate determined.

Adsorption isotherms of As(III)/As(V) on 1:1, 3:1, 6:1 and 9:1 Fe–Mn binary oxides were obtained using batch experiments at pH 7.0 ± 0.2. Initial arsenic concentrations varied from 0.2 to 50 mg/L. In each test, 10 mg of the adsorbent sample was loaded in the 40 mL glass vessel, and 20 mL of solution containing differing amounts of arsenic was added. In order to keep the pH level around 7.0 ± 0.2, 0.1 M of NaOH or HNO3 was added, accordingly. The vessels were shaken on an orbital shaker at 180 rpm for 24 h at 22 ± 1 °C. After the reaction period, all samples were filtered by a 0.45 μm membrane filter and analyzed for arsenic.

Analytical methods

Arsenic concentrations were determined by inductively coupled plasma mass-spectrometry (Agilent 7700). Prior to analysis, the aqueous samples were acidified with concentrated HNO3 in an amount of 1% and stored in acid-washed glass vessels. All samples used in our analysis were analyzed within 24 h of collection.

RESULTS AND DISCUSSION

Adsorbent characterization

The BET surface area measurements of the Fe–Mn binary oxides are given in Table 1. The Fe–Mn binary oxides all have high surface areas (109.6–301.0 m2/g), especially when compared to natural Fe–Mn minerals, which have relatively small specific surface areas, of the order 17–40 m2g−1 (Deschamps et al. 2005). Other authors (Zhang et al. 2007) have reported that the surface areas of both natural and synthetic iron oxides were around 6.4–320 m2g−1, higher for amorphous FeOOH, and lower for goethite and hematite.

Table 1

Physicochemical parameters of Fe–Mn binary oxides

Adsorbents BET surface area (m²/g) Micropore volume (cm3g−1Mesopore volume (cm3g−1
Fe–Mn 1:1 301.0 0.021 0.403 
Fe–Mn 3:1 250.4 0.040 0.338 
Fe–Mn 6:1 203.1 0.034 0.232 
Fe–Mn 9:1 109.6 0.009 0.144 
Adsorbents BET surface area (m²/g) Micropore volume (cm3g−1Mesopore volume (cm3g−1
Fe–Mn 1:1 301.0 0.021 0.403 
Fe–Mn 3:1 250.4 0.040 0.338 
Fe–Mn 6:1 203.1 0.034 0.232 
Fe–Mn 9:1 109.6 0.009 0.144 

The SEM showed the Fe–Mn binary oxides synthesized are nanomaterials (20.0 to 20.8 nm for FeMn 1:1; 18.4 to 21.9 nm for FeMn 3:1; 30.7 to 35.9 nm for FeMn 6:1; 16.6 to 19.2 nm for FeMn 9:1). The lack of magnetism of the adsorbents is in agreement with their presumed structures given above.

The values of pzc of the Fe–Mn binary oxides with Fe/Mn molar ratios 1:1, 3:1, 6:1 and 9:1 were 6.16, 6.54, 6.78 and 6.84, respectively, suggesting lower pHpzc value with increasing Mn/Fe molar ratio. The pHpzc of most iron oxides is pH 7–9 and that of MnO2 is 2–3 (Kanungo & Mohapatra 1989; Masue et al. 2007). Clearly, the pHpzc value lowered as MnO2 content increased in the binary oxides, due to its lower pHpzc value, which is in agreement with Zhang et al. (2012).

Adsorption kinetics

The adsorption kinetics of As(III) and As(V) at pH 7.0 ± 0.2 and room temperature (20 ± 2 °C) are shown in Figures 1(a) and 1(b). As we can see from Figures 1(a) and 1(b) a contact time of 720 min (12 h) was enough to achieve adsorption equilibrium for As(III) and As(V). A mixing time of 24 h was therefore used in the other batch adsorption experiments for both As(III) and As(V).
Figure 1

Adsorption kinetics of (a) As(III) and (b) As(V) onto Fe–Mn binary oxides. Initial concentration As(III)/As(V) = 10 mg/L, adsorbent dosage = 0.5 g/L, pH = 7.0 ± 0.2, ion strength = 0.01 M NaNO3.

Figure 1

Adsorption kinetics of (a) As(III) and (b) As(V) onto Fe–Mn binary oxides. Initial concentration As(III)/As(V) = 10 mg/L, adsorbent dosage = 0.5 g/L, pH = 7.0 ± 0.2, ion strength = 0.01 M NaNO3.

In order to better describe the removal kinetics of As(III) and As(V), three different kinetic models were used to model the experimental data (pseudo-first-order, pseudo-second-order and Elowich models). Based on the estimated correlation coefficients (R2), it is observed that the pseudo-second-order model (R2 > 0.9990) fits better than the pseudo-first-order (R2 = 0.8821–0.9236) and Elowich models (R2 = 0.8952–0.9490) for As(III) sorption on Fe–Mn sorbents. The same trend is observed for As(V); the pseudo-second-order model (R2 = 0.0876–0.9994) fits better than the pseudo-first-order (R2 = 0.6666–0.9775) and Elowich models (R2 = 0.9150–0.9901).

Generally, the pseudo-second-order rate constant (k2) decreases with increasing initial load of As(III) and As(V) in solution per gram of Fe–Mn binary oxides, which means that the As(III)/As(V) sorption is a more favourable process at lower than at higher solute concentration (Table 2). Similar results were observed for adsorption of As(III) and As(V) on Fe–Zn binary oxide where equilibrium was achieved after 25 h (Ren et al. 2011).

Table 2

Kinetic parameters for adsorption of As(III) and As(V) onto Fe–Mn binary oxides

  As(III) pseudo-second-order parameters
 
As(V) pseudo-second-order parameters
 
Adsorbent R2 k2 g mg−1 min−1 qe mg/g R2 k2 g mg−1 min−1 qe mg/g 
Fe–Mn 1:1 0.999 0.000738 13.6 0.988 0.00062 9.79 
Fe–Mn 3:1 0.999 0.000597 15.7 0.991 0.00038 15.5 
Fe–Mn 6:1 0.999 0.000517 13.1 0.999 0.00153 15.1 
Fe–Mn 9:1 0.993 0.000798 9.60 0.994 0.00041 12.6 
  As(III) pseudo-second-order parameters
 
As(V) pseudo-second-order parameters
 
Adsorbent R2 k2 g mg−1 min−1 qe mg/g R2 k2 g mg−1 min−1 qe mg/g 
Fe–Mn 1:1 0.999 0.000738 13.6 0.988 0.00062 9.79 
Fe–Mn 3:1 0.999 0.000597 15.7 0.991 0.00038 15.5 
Fe–Mn 6:1 0.999 0.000517 13.1 0.999 0.00153 15.1 
Fe–Mn 9:1 0.993 0.000798 9.60 0.994 0.00041 12.6 

The overall adsorption process may be controlled by either one or more steps, including outer diffusion, intra-particle diffusion and adsorption of the adsorbates onto active sites. The last step was considered to be rapid and thus cannot be treated as the rate-limiting step in the adsorption process (Yu et al. 2012). Consequently, the adsorption rate might be controlled by outer diffusion, inner diffusion or both. Since the above general kinetic models could not identify the rate-limiting step of As(III) and As(V) on Fe–Mn binary oxides, the intra-particle diffusion model based on the theory proposed by Weber and Morris and particle diffusion was used to analyze the rate-limiting step of adsorption. The values of R2 for the intra-particle pore diffusion model are closer to unity indicating that intra-particle pore diffusion of adsorbate contributes more towards the rate-determining step (Table 3).

Table 3

Various diffusion parameters for As(III) and As(V) adsorption by Fe–Mn binary oxides

  As(III)
 
As(V)
 
  Intra-particle diffusion model
 
Particle diffusion model
 
Intra-particle diffusion model
 
Particle diffusion model
 
Adsorbent R2 ki mg g−1 min−0.5 R2 kp min−1 R2 ki mg g−1 min−0.5 R2 kp min−1 
Fe:Mn 1:1 0.936 2.716 0.881 6.38E-04 0.882 0.2286 0.785 3.61E-04 
Fe:Mn 3:1 0.945 2.944 0.897 8.77E-04 0.884 0.3900 0.793 8.05E-04 
Fe:Mn 6:1 0.963 2.516 0.919 6.02E-04 0.713 0.2984 0.832 9.36E-04 
Fe:Mn 9:1 0.908 1.786 0.828 4.96E-02 0.918 0.3116 0.818 5.44E-04 
  As(III)
 
As(V)
 
  Intra-particle diffusion model
 
Particle diffusion model
 
Intra-particle diffusion model
 
Particle diffusion model
 
Adsorbent R2 ki mg g−1 min−0.5 R2 kp min−1 R2 ki mg g−1 min−0.5 R2 kp min−1 
Fe:Mn 1:1 0.936 2.716 0.881 6.38E-04 0.882 0.2286 0.785 3.61E-04 
Fe:Mn 3:1 0.945 2.944 0.897 8.77E-04 0.884 0.3900 0.793 8.05E-04 
Fe:Mn 6:1 0.963 2.516 0.919 6.02E-04 0.713 0.2984 0.832 9.36E-04 
Fe:Mn 9:1 0.908 1.786 0.828 4.96E-02 0.918 0.3116 0.818 5.44E-04 

However, in the case of the intra-particle diffusion model the lines do not pass through the origin (Figure 2), which reveals that the adsorption of As(III) and As(V) on Fe–Mn binary oxides is a complex process involving surface adsorption, inter-particle diffusion and intra-particle diffusion all contributing towards the rate of sorption (Dhoble et al. 2011).
Figure 2

Intra-particle mass transfer plots for As(III) and As(V) adsorption on Fe–Mn binary oxides.

Figure 2

Intra-particle mass transfer plots for As(III) and As(V) adsorption on Fe–Mn binary oxides.

Adsorption isotherms

Arsenic sorption capacities of Fe–Mn binary oxides with different Fe/Mn molar ratios were evaluated using adsorption isotherms at pH 7.0 ± 0.2 at constant ionic strength (0.01 M NaNO3). The adsorption isotherms were modelled using both Freundlich and Langmuir models (Figure 3).
Figure 3

Freundlich sorption isotherms of (a) As(III) and (b) As (V) on Fe–Mn binary oxides in 0.01 M NaNO3 at pH 7.0 ± 0.2.

Figure 3

Freundlich sorption isotherms of (a) As(III) and (b) As (V) on Fe–Mn binary oxides in 0.01 M NaNO3 at pH 7.0 ± 0.2.

Based on the estimated correlation coefficients (R2), it was found that both models were suitable for describing the adsorption behaviour of As(III) and As(V) on Fe–Mn binary oxides. However, slightly better correlation coefficients for adsorption of As(III) and As(V) on Fe–Mn binary oxides were obtained with the Freundlich isotherm model (Table 4). This is probably due to the heterogeneity of the adsorbent surface with the presence of manganese dioxide (Zhang et al. 2012).

Adsorption of As(V) increased with increasing Fe:Mn molar ratio and reached a maximum of approximately 50 mg/g (maximal removal capacity corresponding to the isotherm plateau in Figure 3) when the Fe:Mn ratio was 6:1 and then reduced with further increases in Fe:Mn molar ratio. The As(V) adsorption capacity decreased to 20 mg/g at a Fe:Mn molar ratio of 1:1. The lower As(V) adsorption with Fe:Mn 1:1 could be attributable to the presence of a high concentration of manganese dioxide which has very low arsenic adsorption ability (Figure 3).

In the case of As(III), adsorption capacity reached a maximum when the Fe:Mn ratio was 3:1 and then reduced with increasing Fe:Mn molar ratio. The lowest As(III) adsorption was observed with Fe:Mn molar ratio 9:1 and may be due to the low concentration of manganese dioxide which can oxidize As(III) (Figure 3(b)). The higher As(III) sorption capacity of the Fe–Mn binary oxides with the 3:1 and 1:1 molar ratios could be explained as follows. Fe–Mn binary oxides combine the oxidation properties of manganese dioxide and the high As(V) adsorption features of iron oxides. Consequently MnO2 effectively oxidizes As(III) to As(V), and the resulting As(V) is adsorbed by the original iron oxide adsorption sites. Furthermore, during As(III) oxidation the MnO2 is reduced and Mn2+ released into solution, resulting in the production of fresh adsorption sites at the solid surface, favouring adsorption of the As(V) formed (Deschamps et al. 2005; Zhang et al. 2007). The As(III) uptake was therefore enhanced. On the other hand, due to the low As(V) adsorption ability of manganese dioxide alone, the high Mn concentrations present in the binary oxide with Fe:Mn molar ratio 1:1 were not beneficial to As(V) uptake (Zhang et al. 2012). The sorption of As(III) on Fe–Mn binary oxides with Fe–Mn ratios of 6:1 and 9:1 was similar to that of Fe–Mn 1:1 and Fe–Mn 3:1, except with lower capacities, indicating that part of the As(III) was converted to As(V). This suggests that the Fe–Mn binary oxides with higher MnO2 contents can oxidize As(III) more effectively (Zhang et al. 2012).

The obtained qmax values for the Fe–Mn adsorbents suggest they are effective for both As(V) and As(III) removal. The essential features of the Langmuir isotherm may be expressed in terms of equilibrium parameter RL, which is a dimensionless constant referred to as the separation factor or equilibrium parameter. The RL value indicates the adsorption nature to be unfavourable if RL > 1, linear if RL = 1, favourable if 0 < RL < 1 and irreversible if RL = 0 (Dada et al. 2012). From the data calculated, RL is greater than 0 but less than 1 for the sorption of As(III) and As(V), indicating that the Langmuir isotherm is favourable. In the present case, the values were in the range 0.11–0.99 for As(III) and 0.19–0.99 for As(V), showing the favourability of the process and the good fit of the Langmuir isotherm (Table 4).

Table 4

Freundlich and Langmuir isotherm parameters for As(III) and As(V) adsorption on Fe–Mn binary oxides at pH 7.0 ± 0.2

  As(III)
 
As(V)
 
Fe:Mn ratio 1:1 3:1 6:1 9:1 1:1 3:1 6:1 9:1 
Freundlich 
KF (mg/g)/(mg/L)n 4.41 8.71 3.94 3.91 3.45 6.68 7.81 5.23 
 1/n 0.462 0.462 0.555 0.534 0.490 0.541 0.549 0.519 
R2 0.9802 0.9677 0.9931 0.9926 0.9870 0.9945 0.9919 0.9904 
Langmuir 
qmax(mg/g) 37.77 46.93 41.59 37.07 25.94 57.58 64.95 43.65 
KL 0.070 0.154 0.055 0.061 0.080 0.076 0.086 0.074 
RL 0.22–0.99 0.11–0.97 0.27–0.98 0.18–0.98 0.20–0.98 0.21–0.99 0.19–0.98 0.21–0.99 
R2 0.9639 0.9656 0.9813 0.9814 0.9684 0.9933 0.9897 0.9903 
  As(III)
 
As(V)
 
Fe:Mn ratio 1:1 3:1 6:1 9:1 1:1 3:1 6:1 9:1 
Freundlich 
KF (mg/g)/(mg/L)n 4.41 8.71 3.94 3.91 3.45 6.68 7.81 5.23 
 1/n 0.462 0.462 0.555 0.534 0.490 0.541 0.549 0.519 
R2 0.9802 0.9677 0.9931 0.9926 0.9870 0.9945 0.9919 0.9904 
Langmuir 
qmax(mg/g) 37.77 46.93 41.59 37.07 25.94 57.58 64.95 43.65 
KL 0.070 0.154 0.055 0.061 0.080 0.076 0.086 0.074 
RL 0.22–0.99 0.11–0.97 0.27–0.98 0.18–0.98 0.20–0.98 0.21–0.99 0.19–0.98 0.21–0.99 
R2 0.9639 0.9656 0.9813 0.9814 0.9684 0.9933 0.9897 0.9903 

Effect of pH on As(III) and As(V) sorption

The influence of solution pH (from 3 to 11) on the removal of As(III) and As(V) by Fe–Mn binary oxides was investigated and the results are presented in Figures 4(a) and 4(b). The removal percentage of As(III) was maintained above 90% at pH values ranging from 3 to 8 for all four adsorbents. However, further increasing the pH from 9 to 11 resulted in removal percentages declining to 41–61% for As(III), depending on the adsorbents applied. As(V) sorption on Fe–Mn binary oxides followed similar trends (Figure 4(b)). As(V) sorption on Fe–Mn binary oxides decreases with increasing solution pH and the maximum percentage removal was observed under acidic conditions. Many other researchers have observed similar pH effects for the sorption of As(V) onto iron oxides or iron-containing oxides (Zhang et al. 2010; Ren et al. 2011; Zhang et al. 2013; Kong et al. 2014). Under the investigated pH range (3–11), H2AsO4and HAsO42– are the dominant As(V) species in solution. Lower pH is favourable for the protonation of the sorbent surface and increasing the number of positively charged sites, enlarging the attractive force existing between the sorbent surface and the arsenate anions, therefore improving the adsorption in the lower pH region. With an increase in solution pH, negatively charged sites gradually dominate, enhancing repulsion effects, and the amount of adsorption consequently drops (Zhang et al. 2012). On the other hand, As(III) adsorption on iron adsorbents is very different from that of As(V), since H3AsO3 is the dominant dissolved As(III) species below pH 9.2. However, the similar As(III) and As(V) adsorptions observed during this experiment suggest indirectly that the initial As(III) was oxidized to As(V) and then adsorbed by the Fe–Mn adsorbents. Generally, As(III) and As(V) removals by Fe–Mn sorbents were greater than 90% until the solution pH was increased above 9, indicating that the material should be effective for the majority of water supplies, which normally have a pH range of 6.5–8.5 (Gu et al. 2005).
Figure 4

Effect of initial solution pH on (a) As(III) and (b) As(V) removal by Fe–Mn binary oxides. Initial As(III)/As(V) concentration = 200 μg/L, adsorbent dose 0.5 g/L.

Figure 4

Effect of initial solution pH on (a) As(III) and (b) As(V) removal by Fe–Mn binary oxides. Initial As(III)/As(V) concentration = 200 μg/L, adsorbent dose 0.5 g/L.

The concentrations of Fe and Mn released into solution after reaction with As(III) and As(V) at different solution pH were also investigated (Tables 5 and 6). It can be seen that the obvious release of Mn was observed under acidic conditions (pH 3–5). The release of Mn was greater for As(III) sorption than for As(V) sorption. This can be explained by the As(III) removal mechanism of Fe–Mn binary oxides described by Zhang et al. (2007), which includes an oxidation step coupled with a sorption process and which is shown in Equation (1): 
formula
1
The above reaction states that MnO2 oxidizes As(III) to As(V), whilst releasing Mn2+. Consequently, the higher release of Mn2+ for As(III) sorption compared to As(V) can be attributed to the reductive dissolution of MnO2 caused by oxidation of As(III). At pH 3 and pH 4 the release of Mn observed was up to 1.87 mg/L, whereas at greater pH values no Mn release was found, possibly through the following mechanisms: (1) part of the As(III) was directly adsorbed onto the adsorbent without oxidation; (2) the redox reaction given in Equation (1) did not progress far enough to the right to generate excess Mn2+; and (3) the dissolved Mn2+ was re-adsorbed onto the adsorbent, which has been well elucidated elsewhere (Xu et al. 2011). Negligible release of Fe was observed in the solutions with initial pH values from 4 to 11, indicating that the Fe–Mn binary oxides are stable under the conditions investigated.
Table 5

Concentrations of Fe and Mn released into solution after reaction with As(III) at different solution pH

  pH
 
Adsorbent 10 11 
Fe release (mg/L) 
 Fe–Mn 1:1 0.06 N.D N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 3:1 0.09 0.06 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 6:1 0.07 0.06 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 9:1 0.09 0.07 N.D N.D N.D N.D N.D N.D N.D 
Mn release (mg/L) 
 Fe–Mn 1:1 1.87 0.29 0.048 N.D N.D N.D N.D N.D N.D 
 Fe–Mn 3:1 1.27 0.26 0.035 N.D N.D N.D N.D N.D N.D 
 Fe–Mn 6:1 0.87 0.12 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 9:1 1.87 0.10 N.D N.D N.D N.D N.D N.D N.D 
  pH
 
Adsorbent 10 11 
Fe release (mg/L) 
 Fe–Mn 1:1 0.06 N.D N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 3:1 0.09 0.06 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 6:1 0.07 0.06 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 9:1 0.09 0.07 N.D N.D N.D N.D N.D N.D N.D 
Mn release (mg/L) 
 Fe–Mn 1:1 1.87 0.29 0.048 N.D N.D N.D N.D N.D N.D 
 Fe–Mn 3:1 1.27 0.26 0.035 N.D N.D N.D N.D N.D N.D 
 Fe–Mn 6:1 0.87 0.12 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 9:1 1.87 0.10 N.D N.D N.D N.D N.D N.D N.D 

N.D represents ‘not detected’. The practical quantitation limits for Fe and Mn were 0.068 and 0.034 mg/L.

Table 6

Concentrations of Fe and Mn released into solution after reaction with As(V) at different solution pH

  pH
 
Adsorbent 10 11 
Fe release (mg/L) 
 Fe–Mn 1:1 0.08 0.07 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 3:1 0.13 0.08 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 6:1 0.18 0.06 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 9:1 0.19 N.D N.D N.D N.D N.D N.D N.D N.D 
Mn release (mg/L) 
 Fe–Mn 1:1 0.77 0.05 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 3:1 0.47 0.25 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 6:1 0.57 0.03 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 9:1 0.89 0.06 N.D N.D N.D N.D N.D N.D N.D 
  pH
 
Adsorbent 10 11 
Fe release (mg/L) 
 Fe–Mn 1:1 0.08 0.07 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 3:1 0.13 0.08 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 6:1 0.18 0.06 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 9:1 0.19 N.D N.D N.D N.D N.D N.D N.D N.D 
Mn release (mg/L) 
 Fe–Mn 1:1 0.77 0.05 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 3:1 0.47 0.25 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 6:1 0.57 0.03 N.D N.D N.D N.D N.D N.D N.D 
 Fe–Mn 9:1 0.89 0.06 N.D N.D N.D N.D N.D N.D N.D 

N.D represents ‘not detected’. The practical quantitation limits for Fe and Mn were 0.068 and 0.034 mg/L.

Arsenic removal from real groundwater sample

The ability of the Fe–Mn binary oxides to remove arsenic from naturally contaminated groundwater from Kikinda (37 μg As/L, pH 8.2, 5.1 mg DOC/L) was investigated to prove their practical application. The results obtained from the batch studies with the Kikinda groundwater are given in Figure 5. It is evident from Figure 5 that arsenic may be easily removed to below the Water Framework Directive drinking water standards (European Commission 2000) by Fe–Mn binary oxides. By increasing the adsorbent dose from 0.5 g/L through to 5.0 g/L, percentage removals of As increased from 21%–77% (Fe:Mn 1:1), 30%–86% (Fe:Mn 3:1), 38%–90% (Fe:Mn 6:1) and 24%–87% (Fe:Mn 9:1). However, beyond a dosage of 1.5 g/L, there was no significant change in the percentage removal of As, with the exception of the binary oxide with a 1:1 Fe:Mn ratio. In this case, dosing 2.5 g/L slightly increased the As removal percentage.
Figure 5

Arsenic removal from groundwater using Fe–Mn binary sorbents.

Figure 5

Arsenic removal from groundwater using Fe–Mn binary sorbents.

CONCLUSION

Fe–Mn binary oxides were prepared relatively simply by simultaneous oxidation and co-precipitation. Adsorption kinetics for As(III) and As(V) followed the pseudo-second-order kinetic model. The adsorption of As(III) and As(V) onto Fe–Mn binary oxides is a complex process involving surface adsorption, inter-particle diffusion and intra-particle diffusion, all contributing towards the rate of sorption. Equilibrium data fitted well to the Langmuir and Freundlich models. As(V) adsorption increased with increasing Fe:Mn ratio and reached a maximum at a 6:1 Fe:Mn ratio. The maximum As(III) sorption capacity was observed when the Fe:Mn ratio was 3:1. In real groundwater samples, arsenic was readily removed from 37 μg/L to below 10 μg/L by Fe–Mn binary oxides. These results show that Fe–Mn binary oxides represent an attractive adsorbent for arsenic removal from water due to their high uptake of both As(V) and As(III) and low cost.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the support of the Ministry of Education, Science and Technological Development of the Republic of Serbia (Project No. III43005).

REFERENCES

REFERENCES
Chandra
V.
Park
J.
Chun
Y.
Lee
J. W.
Hwang
I. C.
Kim
K. S.
2010
Water-dispersible magnetite-reduced graphene oxide composites for arsenic removal
.
ACS Nano
4
,
3979
3986
.
Dada
A. O.
Olalekan
A. P.
Olatunya
A. M.
Dada
O.
2012
Langmuir, Freundlich, Temkin and Dubinin–Radushkevich isotherms studies of equilibrium sorption of Zn2+ unto phosphoric acid modified rice husk
.
IOSR-JAC
3
,
38
45
.
Deschamps
E.
Ciminelli
V. S. T.
Ramos
A. Y.
2003
Arsenic sorption onto soils enriched in Mn and Fe minerals
.
Clays Clay Miner.
51
,
197
204
.
Deschamps
E.
Ciminelli
V. S. T.
Höll
W. H.
2005
Removal of As(III) and As(V) from water using a natural Fe and Mn enriched sample
.
Water Res.
39
,
5212
5220
.
Driehaus
W.
Jekel
M.
Hildebrandt
U.
1998
Granular ferric hydroxide – a new adsorbent for the removal of arsenic from natural water
.
J. Water SRT-Aqua.
47
,
1
6
.
European Commission
2000
Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 Establishing a Framework for Community Action in the Field of Water Policy
.
Official Journal 22 December 2000 L 327/1
.
European Commission
,
Brussels
.
Kanungo
S. B.
Mohapatra
D. M.
1989
Interfacial properties of some hydrous manganese dioxides in 1-1 electrolyte solution
.
J. Colloid Interface Sci.
131
,
103
111
.
Kinniburgh
D. G.
Syers
J. K.
Jackson
M. L.
1975
Specific adsorption of trace amounts of calcium and strontium by hydrous oxides of iron and aluminum
.
Soil Sci. Soc. Am. Proc.
39
,
464
470
.
Lenoble
V.
Laclautre
C.
Serpaud
B.
Deluchat
V.
Bollinger
J. C.
2004
As(V) retention and As(III) simultaneous oxidation and removal on a MnO2-loaded polystyrene resin
.
Sci. Total Environ.
326
,
197
207
.
Ren
Z. M.
Zhang
G. S.
Chen
J. P.
2011
Adsorptive removal of arsenic from water by an iron–zirconium binary oxide adsorbent
.
J. Colloid Interface Sci.
358
,
230
237
.
Yu
F.
Chen
J.
Chen
L.
Huai
J.
Gong
W.
Yuan
Z.
Wang
J.
Ma
J.
2012
Magnetic carbon nanotubes synthesis by Fenton's reagent method and their potential application for removal of azo dye from aqueous solution
.
J. Colloid Interface Sci.
378
,
175
183
.