Adsorption removal of arsenic from aqueous solutions and groundwater by isomeric FeOOH Open Access

The isomeric FeOOH has mainly three phases of α-, β- and γ-FeOOH. Among them, the α-FeOOH has been used in pigments, gas sensors and sources of magnetic materials (Sani et al. 2019). While β-FeOOH is usually used for functional building blocks for iron-deficient drugs, and anode materials for capacitors (Huang et al. 2021). The unstable γ-FeOOH phase can been used for mineral transformation, catalyst for Fenton reaction and the production of functional ceramics (Li et al. 2020a, 2020b). In environmental study fields, the isomeric FeOOH and the other iron (oxyhydr)oxides ubiquitously existing in both natural and engineered environments have great retention capacities of metal oxyanions such as arsenic oxyanions due to their high surface areas and reactivity (Shi et al. 2020). Arsenic removal has been a huge challenge since it threats to the freshwater resource quality and its oxyanions can generate enormous health and public concern. Arsenic (As) species has arsenate(V) and arsenite(III). Due to high mobility of As(III), it is 60 times more toxic and more difficult to remove than As(V) (Liu et al. 2021). Recently, the present situation of As removal has the key topics on nano-technological and biological process and current progress and future perspectives of possible mitigation options (Maity et al. 2021).

The sequestration of As(V) and As(III) by FeOOH is one of the most vital geochemical/chemical processes controlling their environmental fate, transport, and bioavailability (Shi et al. 2020). Biological processes for As removal from water and/or soil environment are effective and ecofriendly (Maity et al. 2021) and it can also be a feasible approach for the in-situ remediation of As-nitrate contaminated groundwater and surface waters with high concentrations of Fe(II) and Mn(II). In treatment of As-rich waters, the green and biologically-driven pathways to synthetize new nanostructured FeOOH filters (Casentini et al. 2019; Kim et al. 2022) and Fe-Mn oxides (Xiong et al. 2017; Cho et al. 2018; Yan et al. 2021; Hu et al. 2022) as effective, low cost and selective technology are always becoming more attractive. The other effective strategies by integrating nanotechnology, electrochemical processes of filtration (Liu et al. 2021), coagulation (Maldonado-Reyes et al. 2015; Bora & Dutta 2021), oxidation (Pervez et al. 2021a, 2021b; Peng et al. 2022), and membrane separation (Qiu et al. 2020; Wang et al. 2020) have been attended. During electrocatalytic As(III) oxidation into less toxic As(V) in aqueous bicarbonate solutions, three α-, β-, and γ-FeOOH polymorphs have specific complexation with bicarbonate under alkaline conditions, whereas the complexation disappears at pH 6.7 (Wang & Giammar 2015; Guo et al. 2019; Kim et al. 2021).

Isomeric α-, β-, and γ-FeOOH have always been considered as the selected adsorbents. Iron minerals of goethite (α-FeOOH) can load Sb of up to 3.14 wt% and As of 1.29 wt% (Jurkovic et al. 2019). A novel α-FeOOH modified wheat straw biochar (α-FeOOH@BC) material for removal of As(III) from aqueous solutions has its maximum adsorption capacity of 78.3 mg/g (Zhu et al. 2020). The prepared goethite impregnated graphene oxide (GO)-carbon nanotubes (CNTs) aerogel (α-FeOOH@GCA) shows excellent adsorption capacity of 56.43 mg/g for As(V) at the widely favorable application pH. The phosphate and silicate anions can compete with the arsenic species for active adsorption sites due to their similar anionic structure. Based on FTIR spectra, arsenic species are proven to form inner sphere complex on the surface of α-FeOOH@GCA through different ligand exchanging mechanism greatly dependent on the molecular structures of arsenic species (Fu et al. 2017). The maximum As(III) adsorption capacity on goethite quantum dots impregnated graphene oxide hybrids (α-FeOOH QDs@GO) is 147.4 mg/g, which is 2.52 and 4.60 times larger than those of β-FeOOH@GO and β-FeOOH, respectively. The arsenic adsorption mechanism on α-FeOOH QDs@GO reveals that hydroxyl and acetate ligand exchange are the main pathways for arsenic adsorption (Pervez et al. 2021a, 2021b).

The magnetic nanocomposite of β-FeOOH (74%) and Fe₃O₄ (26%) have the estimated maximum adsorption capacities for As(III) at pH 5 are about 9 mg/g, this eliminating the filtration step by applying a magnetic field (Cunha et al. 2019). A produced β-FeOOH@GO-COOH nanocomposite via carboxylic graphene oxide (GO-COOH) decorated with β-FeOOH provides high adsorption capacities of 77.5 mg/g for As(III) and 45.7 mg/g for As(V) within a wide range of pH 3–10, respectively. Adsorption efficiencies of 100% and 97% are achieved for five successive operation cycles for the removal of 100 mμg/L As(V) and As(III) in five fresh portions of aqueous solution (1.0 mL for each) with 3 mg nanocomposite (Chen et al. 2015). The prepared composite (FeOOH/CuO@WBC) has the maximum adsorption capacity of 76.1 mg/g at pH3.5 when As initial level is 150 mg/L (Liu et al. 2020). The polyacrylonitrile/ferric hydroxide (fiber) has the highest arsenic(III) adsorption capacity (11.31 mg/g) under the baseline conditions. The maximum adsorption capacity increases to 12.41 mg/g at about pH 9.0. After three adsorption cycles, the As(III) removal rate of each composite membrane reaches 90% (Luo et al. 2021).

In this work, the synthetic α, β, γ-FeOOH polymorphs (goethite, akaganéite, lepidocrocite) were used to remove As(III) from aqueous solutions and treat simulated As-polluted groundwater. The goal was to systematically investigate and compare adsorption behaviors and the maximum removal efficiency of As(III) on isomeric FeOOH by adsorption kinetic and isotherm studies. All adsorption experiments were conducted in batch conditions to obtain the effect of variable parameters such as the initial arsenic concentration and adsorbent dosage, and different pH and electrolyte solution. The results can expectantly provide theoretical proofs for investigation and extended application of adsorption materials in treatment of arsenic-polluted waters.

In all experiments, the used reagents were analytical grade and commercially available. As(III) and As(V) adsorbate solutions were obtained from various diluents of Na₃AsO₃ and Na₂AsO₄ as mother solutions, respectively, and the solvent was deionized water. The identified isomeric α, β, γ-FeOOH minerals (including goethite of Gth1and Gth2, akaganéite of Aka1 and Aka2, and lepidocrocite of Lep in this work, see Figure S1 in Supplement were prepared chemically and used as adsorbents for As(III), and the concisely chemical synthesis methods were shown in Table 1 (Schwertmann & Cornell 2000; Xu et al. 2013).

These adsorbent samples of 0.1 g mixed in electrolyte solutions resulted in the aimed adsorbent solutions. The pH of the mixed reaction solution was adjusted to 7.0 by adding a certain amount of acid/base (HNO₃/NaOH) solutions corresponding to the electrolyte (0.1 M NaNO₃ solution). All adsorption experiments were conducted in triplicate.

For the kinetic experiments on As(III) adsorption by goethite, akaganéite and lepidocrocite, 0.02 g adsorbent samples (Aka1, Aka2, Gth1, Gth2, and Lep, their content was 1.0 g/L) were added into a background electrolyte of 0.01M NaNO₃ in 50-mL polyethylene centrifuge tubes. The resulting adsorbent mixtures were adjusted to the constant pH of 7.0 and then mixed with Na₃AsO₃ mother solutions and the total volume of the mixed reaction solutions was 20 mL. In the above reaction mixtures, the initial As(III) concentration was selected as 20 mg/L, which mainly resulted from the case that the adsorption was fast to reach equilibrium at the lower levels of initial arsenic, while slow at the higher levels (Amrani et al. 2020). The centrifuge tubes were shaken at 180 rpm and 28 °C. Some reaction solutions were extracted out at different time intervals (i.e. 2, 5, 10, 20, 40, 60, 120, 240, 480, 720, 1,080 and 1,440 min) and filtrated through a 0.45 μm filter membrane for examination of the residual As(III).

Further, for the experiments on the adsorption isotherm, the arsenic initial concentration was varied (2, 4, 8, 10, 15, 20, 30 and 40 mg/L) and other conditions were kept the same as the kinetic experiment. All mixed solutions were shaken for 24 h until adsorption equilibrium, and then filtered and measured for arsenic concentration in supernatant.

In this section, the dependent effect of four parameters on As(III) adsorption by goethite, akaganéite and lepidocrocite was studied using variable-controlling approach. For the effect of initial As(III) concentration (or adsorbent content), there was a series of designed data of 10, 20, and 30 mg/L (or 0.25, 0.50, 1.0 and 2.5 g/L) and then for the effect of pH, its values ranged from 3.0 to 12.0. Further to explore the effect of different electrolytes, the selected electrolytes were NaNO₃, NaCl, Na₂SO₄, Na₂CO₃ and NaH₂PO₄ and their ion strengths were designed as 0.001, 0.01 and 0.1 mol/L. In these experiments, the other reaction conditions were the same as those given in the isotherm adsorption.

Column experiments were carried out in a small glass columns (400 mm of height) with a layer of sand core (100 mesh opening size) in their bottom to prevent the sorbent discharged from the beds into the sampling tubes. There was a 100 mm depth for the fixed bed composed with 5 mm (depth/height) glass wool at its outlet end and then about 0.8 g of FeOOH adsorbent and the simulated groundwater with 200 μg/L arsenic species as the feeding solution in the test column. The simulated groundwater (about pH 7.0) contained components (mg/L) K⁺ (300), Na⁺ (371), Cl⁻ (273), NO₃²⁻ (200), SO₄²⁻ (300), CO₃²⁻ (200), SiO₃²⁻ (5), H₂PO₄⁻ (0.1) and HA (1) (Kim et al. 2022). A peristaltic pump (BT-200B, Shanghai Qingpu analytical instrument Co., China) was used to feed the aqueous solution into the packed FeOOH adsorbent at a constant flow rate of 4 mL/min. All column experiments were repeated twice.

The surface structure and bonding groups of adsorbent samples were determined by Nicolet 740 Fourier transform infrared spectrometer (FTIR), which is equipped with KBr spectroscopy and DTGS detector, at test background value of 400 mg KBr and a resolution of 4 cm⁻¹. Changes of elements and binding energies were detected by X-ray photoelectron spectroscopy (XPS, Escalab250Xi, Thermo, USA). The concentrations of As(III) in the mixed reaction solutions were determined by atomic fluorescence spectroscopy (AFS) and the inductively coupled plasma spectrometer (ICP).

By contrast adsorption capacity of five adsorbent samples for As(III), we could markedly observe that Aka2 was the strongest, Gth1, Gth2 and Lep was similar, while Aka1 was the weakest. It was obvious that the adsorption increased rapidly and mainly focused on the first 20 min, which might be due to the excellent surface activities of adsorbents (Wang et al. 2016). Followed by the slow adsorption stage, As(III) ions might enter into the inner structure of the isomeric FeOOH until saturation.

The pseudo-second-order rate equation can describe the kinetic adsorption of As(III) by isomeric FeOOH as shown in Figure 1(b) and Table 2. The experimental data agreed well with the fitted equation and all correlation coefficients reached 0.999. The saturated adsorption capacities of As(III) by Gth1, Gth2, Aka1, Aka2 and Lep were 6.79, 6.23, 5.02, 12.2 and 5.97 mg/g, respectively. Therefore, the adsorptions of As(III) by FeOOH were pseudo-second-order reactions since the measured values were close to the theoretical values.

It could be seen from Figure 2(b) and 2(c) and Table 3 that the experimental data were fitted to both Langmuir and Freundlich equations for As(III) adsorption by isomeric FeOOH. According to the fitting results, the maximum adsorption capacities of As(III) on Gth1, Gth2, Aka1, Aka2 and Lep calculated by the Langmuir equation were 12.3, 7.50, 6.29, 23.4 and 17.7 mg/g, respectively. Obviously, their adsorption capacities were arranged in declining order of Aka2 > Lep > Gth1 > Gth2 > Aka1.

In Figure 3(b), it clearly exhibited that the adsorption capacity of As(III) on samples (Gth1/Gth2, Aka1/Aka2 and Lep) increased with the increase of arsenic concentration. When the concentration of As(III) in aqueous solutions was 30 mg/L, the adsorption capacities were about two times higher than those (10 mg-As(III)/L). This could be connected with the higher specific surface area of the adsorbents and their enough active sites (Samanta et al. 2018).

Obviously, the properties of surface acid sites (including type, concentration, and strength) play an important role in the activity of FeOOH adsorbents. However, the pH_pzc (the pH for the point of zero charge) is corresponding to the whole crystal in the FeOOH system, while the adsorption process is a microcosmic process mainly between arsenic and surface of materials (Wei et al. 2016). It is because the adsorption and photocatalytic activity of metal (hydro)oxides (such as FeOOH and TiO₂) largely depend on its surface atomic structure and the degree of exposed reactive crystal facets (Yan et al. 2016; Fu et al. 2017).

In summary, the IR results show that the arsenic species interacted with α-, β-, γ-FeOOH by forming the sphere complex. The band (at about 1,600 cm⁻¹) assigned to −OH deformation vibration of Fe − OH obviously weakens. This indicates the surface hydroxyl groups substituted by adsorbed arsenite. At the same, there are several new peaks at 620 − 900 cm⁻¹ after arsenic uptake, resulting from synergistic effects based on symmetric and asymmetric stretching vibrations of the As − O bond in the As − O − Fe linkage. Obviously, there is the surface complexation of arsenic species on FeOOH adsorbents, since arsenite can form innersphere complexes by chemical adsorptions on hydroxyl groups at surface sites of isomeric FeOOH (Fu et al. 2017). It is also documented that As(III)/As(V) species might be reasonably attributed to the formation of As − O − Fe chemical bonds and the replacement of M − OH radicals by arsenic species (Ge et al. 2017). If solution pH is down the point of zero charge, the outer-sphere complexes could occur due to electrostatic interactions of arsenite with Fe-OH₂⁺ groups (Kim et al. 2021).

The isomeric FeOOH samples (Gth1/Gth2 as goethite, Aka1/Aka2 as akaganéite and Lep as lepidocrocite) might be used as effective adsorbents for arsenic removal. The adsorption equilibrium achieved after 24 h and the isothermal saturated adsorption amount calculated by the Langmuir equation were 12.3 mg/g (Gth1), 7.50 mg/g (Gth2), 6.29 mg/g (Aka1), 23.4 mg/g (Aka2), and 17.7 mg/g (Lep). Both Freundlich and Langmuir models were applicable for the description of isothermal adsorption process, while Lagergren pseudo-second-order rate model was better fitted to kinetic experimental data. Additionally, adsorption efficiency was influenced by various factors such as solution pH, initial FeOOH content, arsenic concentration and electrolyte solutions. The increase of adsorbate concentration or adsorbent content was favorable to improve the adsorption capacity of As(III) within a certain range. And the inhibition of anions on the adsorption of As(III) by the isomeric FeOOH was H₂PO₄⁻ > SO₄²⁻ > Cl⁻, CO₃²⁻ and NO₃⁻. Herein, there were the appropriate parameters of pH 7.0, 1.0 g/L dosage of isomeric FeOOH and NaNO₃ as electrolyte solution in the adsorption process. In the filter system with simulated arsenic-containing groundwater at about pH 7.0, the strength of As removal abilities for FeOOH was Aka2, and then Aka1/Gth1, and finally Gth2/Lep. Comparing IR spectra, there were some differences in the original and As(III)-loaded isomeric FeOOH. These results could provide reference evidence in treatment of As(III)-polluted waters.

The authors acknowledge the National Natural Science Foundation of China (no. 41472034) and the Natural Science Foundation of Jiangsu Province (SBK20191444) supporting the present study.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.