Structural elucidation of hexavalent Cr adsorbed on surfaces and bulks of Fe₃O₄ and α-FeOOH

Abstract

Magnetite (Fe₃O₄) and goethite (α-FeOOH) were synthesized via a hydrothermal approach and utilized as adsorbents for Cr⁶⁺ removal in an aqueous medium. The typical crystal structures of the synthesized Fe₃O₄ and α-FeOOH were confirmed by XRD and TEM. Fe₃O₄ in a spherical shape with a surface area of 32 m² g⁻¹ was established. While α-FeOOH had a rod-like form with a larger surface area of 84 m² g⁻¹. Cr⁶⁺ removal in an aqueous solution was studied in various conditions to evaluate thermodynamic and kinetic parameters. The adsorption isotherms on both adsorbents fit the Langmuir model indicating monolayer adsorption. Fe₃O₄ showed a better adsorption ability than α-FeOOH even though it had a lower surface area. XAS and XPS analysis strongly evidenced the production of stable Cr³⁺ species of Fe_(1−x)Cr_xOOH and Fe_(3−x)Cr_xO₄ by Cr⁶⁺ reduction and migration processes into the bulk structure. Thus, the existence of stable Cr-species in Fe₃O₄ structure strongly affected Cr-adsorption ability rather than the surface area of the adsorbent. However, the precipitated Cr₂O₃ and HCrO₄⁻ molecules electrostatically adsorbed on the outer surface of α-FeOOH without bulk transformation. The presence of physisorbed FeO–HCrO₄ species on α-FeOOH led to low reducibility and adsorption capability of Cr⁶⁺.

---

Introduction

Hexavalent Cr⁶⁺ ions are toxic contaminants because of their mutagenicity, carcinogenicity, and teratogenicity.^1–5 Toxic Cr⁶⁺ species generally exist in negatively charged forms, such as the ions of hydrogen chromate (HCrO₄⁻), chromate (CrO₄²⁻), and dichromate (Cr₂O₇²⁻), depending on solution pH and concentration.^6,7 However, no conclusive spectroscopic proof of structural changes in adsorbents and adsorbates is available.^8,9 X-Ray absorption spectroscopy (XAS) is a powerful tool for the element-specific characterization of local structure and chemistry. The electronic and local structures or geometry of chromium species in water will be explored. Understanding the stabilized Cr-species on the adsorbents is beneficial to the development of new materials from the waste treatment process. Wei et al.¹⁰ illustrated the determination of Cr-contamination in kitchen wastewater by XAS. The Cr-species distribution in the solution was estimated. Likewise, Meena and Arai¹¹ could explore the Cr³⁺/Cr⁶⁺ ratio and Cr-structures on Fe₃O₄ through the Cr-adsorption process from groundwater by using X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) analysis.

Different techniques, such as chemical precipitation, adsorption, photocatalysis, and reduction using cationic materials, are available for the removal of Cr⁶⁺ from wastewater.^7,12–16 Utilization of natural-cationic materials, such as activated carbon, chitosan, chelate resins, and natural zeolites for Cr⁶⁺ adsorption has been widely investigated due to their low cost, feasibility, handling simplicity, and flexibility.^17–20 However, the removal of the adsorbed Cr⁶⁺ species by those adsorbents still needs a further reduction process to complete the Cr disposal.

Chemical reduction using environment-friendly materials, such as iron oxides, is an interesting alternative method for Cr⁶⁺ removal from wastewater.^21,22 Iron (oxy) hydroxide of goethite (α-FeOOH) containing Fe³⁺ is a potential Cr⁶⁺ adsorbent due to its high surface area and thermal stability.^7,23 Moreover, octahedral sites of Fe³⁺ in orthorhombic α-FeOOH structure play a significant role in Cr⁶⁺ adsorption. In addition, the presence of a small amount of Fe²⁺ species on the α-FeOOH surface provides a great reduction potential for Cr⁶⁺ ions.²⁴ However, a drawback of polymorphs of the synthetic α-FeOOH formed in acidic conditions leads to an additional protonation, resulting in structure destabilization.^7,25,26 Hence, the preparation of a stable α-FeOOH structure with uniform crystalline needs to be investigated. Additionally, the metal oxide of Fe₃O₄ (magnetite) or ferrous-ferric iron oxide nature is interesting for Cr⁶⁺ removal because it has a suitable redox potential for Cr⁶⁺ reduction and excellent magnetic properties.^27,28 It is widely known that Fe²⁺ in magnetite can reduce Cr⁶⁺ to Cr³⁺ creating Cr³⁺-hydroxide or mixed substances via an adsorption and reduction process, namely reductive precipitation.^29,30 This adsorption behavior on the Fe₃O₄ surface was explored and Cr³⁺ and Fe³⁺-(oxy)hydroxides were passivated on the adsorbent surface.³⁰ The adsorption ability and structural stability of Fe₃O₄ are strongly affected by solution pH and the homogeneity of iron oxide particles. The effect of the passivation layer on bulk structure deserves our study. In addition, the relationship between adsorption ability and phase stability in bulk-iron oxides needs to be investigated. Understanding the stable Cr-forms on the ion-oxides is promising for catalyst applications.³¹

In the present study, iron oxides of Fe₃O₄ and α-FeOOH were synthesized by the hydrothermal method and used as Cr⁶⁺ adsorbents. The crystalline structure, morphology, and surface area of the bare and adsorbed materials were characterized. The Cr-adsorption ability and behavior were explored. The adsorbent surfaces and bulk structures were elucidated by XPS and XAS.

Experimental

Materials and methods

Materials. The chemicals including iron(III) nitrate nonahydrate (Fe(NO₃)₃·9H₂O, 98%, Merck), iron(III) chloride (FeCl₃, 98%, ACROS), potassium dichromate (K₂Cr₂O₇, 99.9%, LOBA), sodium acetate (NaOAc, 99.5%, LOBA), potassium hydroxide pallets (KOH, 85%, LOBA), hydrochloric acid (HCl, 37%, QRëC™), ethylene glycol (EG, AR grade, QRëC™) and 1, 5-diphenylcabazone (50% diphenylcabazid, ACS grade, Merck) were used in this work.

Methods. The crystal structures of α-FeOOH and Fe₃O₄ were revealed by powder X-ray diffraction (XRD; PANalytical, EMPYREAN). The XRD patterns were indexed by Match! Software using the Crystallography Open Database (COD), which was available at Beamline 1.1 W (Multiple X-ray Techniques), Synchrotron Light Research Institute, Thailand, SLRI. The unit cell parameters are refined by using FullProf Software. Morphologies were detected by transmission electron microscopy (TEM; Tecnai G²). The oxidation states of the samples before and after adsorption were determined by X-ray absorption spectroscopy (XAS) under the 3.2 and 5.2 beamlines (SLRI) in the intermediate photon energy ranges of Cr K-edge (5.7–6.7 keV), Fe K-edge (7.0–8.0 keV), Fe L₃-edge (0.7–0.73 keV) and O K-edge (5.2–5.5 keV), respectively. The chemical states of Fe, Cr and O were analyzed by X-ray photoelectron spectroscopy (XPS; PHI5000 Versa Probe II, ULVAC-PHI, Japan) at the SUT-NANOTEC-SLRI Joint Research Facility, 5.2 beamline.

Synthetic procedures

Synthesis of α-FeOOH. Goethite (α-FeOOH) was synthesized by the following hydrothermal route.³² Iron solution was prepared by dissolving 4.85 g of Fe(NO₃)₃·9H₂O into 10 mL of deionized (DI) water. The solution was then slowly dropped into 10 mL of 4.8 M KOH solution. The mixture was transferred to a Teflon-lined autoclave and crystallized at 100 °C for 6 h. The crystallized sample was then washed with DI water and ethanol several times, and the resultant brown powder of goethite was dried overnight at 50 °C.

Synthesis of Fe₃O₄. Magnetite (Fe₃O₄) was synthesized by the following route.³³ 1.94 g of FeCl₃ solution was dissolved into 40 mL of EG in a two-necked round-bottom flask under N₂ atmosphere. The solution mixture was stirred for 30 min and then 10 mL of 1.35 M NaOAc in EG was then added. After 3 h, the mixture was transferred to a Teflon-lined autoclave and heated to 190 °C for 24 h. The obtained powder was separated by a magnetic bar. The resultant black powder of Fe₃O₄ was washed with DI water and ethanol several times and dried overnight at 50 °C.

Cr⁶⁺ adsorption analysis. A stock solution of 100 ppm Cr⁶⁺ was prepared from K₂Cr₂O₇. Working solutions with desired concentrations were prepared by diluting the stock solution with DI water. Adsorption experiments were carried out by adding each adsorbent to 50 mL of Cr⁶⁺ solution and stirring vigorously. After adsorption, the adsorbents were separated from the solution by centrifugation for α-FeOOH and using a magnet for Fe₃O₄. Cr⁶⁺ concentrations were determined by the colorimetric method. Each sampling solution was added to 1,5-diphenylcarbazide ligand with 10% (v/v) of H₂SO₄ (pH = 2) to form chromium complex. The absorbance of this complex was collected by using UV-Vis spectrophotometry at λ_max = 540 nm.^34,35 The Cr adsorption capacity of the absorbents (mg g⁻¹) was calculated as
where C₀ (mg L⁻¹) and C_e (mg L⁻¹) are the initial and equilibrium metal concentrations, respectively, V (L) is the solution volume, and W (g) is the adsorbent weight.

Adsorption isotherms were detected by the Langmuir (eqn (1)) Freundlich (eqn (2) and (3)), Dubinin–Radusckevisch–Kanager (eqn (4)–(6)) and Temkin (eqn (7)) models.³⁶

q_e = K_fC_e^1/n, (2)

ln q_e = ln q_(D–R)(−βε²), (4)

q_e = k₁ ln(k₂) + k₁ ln C_e, (7)

where q_e is the amount of adsorbed Cr⁶⁺ (mg g⁻¹), C_e is the equilibrium concentration (mg L⁻¹), b is the adsorption equilibrium constant (L mg⁻¹), Q_m (mg g⁻¹) is the maximum adsorption capacity for the Langmuir model, K_f is Freundlich constants (mg g⁻¹ (L mg)^−1/n) and n is heterogeneity factor, representing the adsorption capacity and the adsorption intensity, respectively. q_(D–R) is the theoretical adsorption capacity (mg g⁻¹), β is a constant related to adsorption energy for a mole of the adsorbate (mol² kJ⁻²), R is the ideal gas constant (0.008314 kJ K⁻¹ mol⁻¹), T is the absolute temperature (K), E (kJ mol⁻¹) is the mean free energy per molecule of adsorbate when it transfers from the bulk of the solution (infinity) to the adsorbent surface, k₁ and k₂ are Temkin constants, where k₁ is related to the heat of adsorption (J mol⁻¹) and k₂ is the equilibrium binding constant (L g⁻¹).

The adsorption thermodynamics of α-FeOOH and Fe₃O₄ was determined at the temperatures of 25 °C, 40 °C and 60 °C using 1 ppm Cr⁶⁺ solution and 0.02 g of each adsorbent. The effects of different initial Cr⁶⁺ concentrations (1–20 ppm), adsorbent doses (0.02–0.1 g), and contact times (10–90 min) in the Cr adsorption efficiency were investigated. The optimization of pH for Cr⁶⁺ removal was conducted in the range of 1–7, which covers the point of zero charges (PZC) of both absorbents.

The stability and adsorption behavior of Cr species on adsorbent surfaces were investigated by performing a reusability test. Each adsorbent was separated from the solution and dried overnight at 50 °C before starting a new cycle.

Results and discussion

Structures of α-FeOOH and Fe₃O₄

The crystal structures of as-synthesized and Cr-adsorbed Fe₃O₄ and α-FeOOH were determined by XRD as shown in Fig. 1a and b, respectively. The XRD pattern of α-FeOOH indicated its orthorhombic structure. The lattice constants of α-FeOOH structure were calculated as a = 4.611 Å, b = 9.962 Å, c = 3.024 Å (JCPDS no. 29-0713).^37,38 The diffraction pattern of Fe₃O₄ indicated its cubic structure (JCPDS no. 19-0629)^28,39 with a lattice constant of a = b = c = 8.364 Å. The phase indexing and the refined unit cell parameters were reported in ESI data S1† and their corresponding XRD refinement in S2–S5.† Specifically, the Fe₃O₄ and Cr@Fe₃O₄ samples were indexed as magnetite Fe₃O₄ phase (cubic, space group Fd3̄m), while the α-FeOOH and Cr@FeOOH samples were indexed as iron (III) oxide hydroxide goethite phase (orthorhombic, space group Pbnm).

Fig. 1 XRD patterns of as-synthesized and Cr-adsorbed (a) Fe₃O₄ and (b) α-FeOOH.

A closer inspection of the refinement results, the addition of chromium species into the materials led to a slight shrinkage of the unit cell parameters compared to the pristine materials. This observation suggested the incorporation of Cr³⁺ into the material matrix of Fe₃O₄ (or FeOOH) without the presence of an additional crystalline impurity phase. In these cases, the atomic site occupancy could not be precisely defined. In addition, XAS was used to propose the local structures of Cr and Fe species in the obtained materials.

The TEM images and lattice parameters of α-FeOOH and Fe₃O₄ are displayed in Fig. 2. The α-FeOOH exhibited short-range-order and rod-like structures (Fig. 2a). The selected area electron diffraction (SAED) pattern of rod-like particles confirmed the orthorhombic structure of α-FeOOH with the (130) and (061) characteristic planes. Its particle size was in the range of 620 (±219) × 55 (±21) nm as shown in S6.† The TEM image of Fe₃O₄ in Fig. 2b revealed spherical particles of 400 ± 124 nm with characteristic planes of (311) and (422).³³ The BET surface areas of α-FeOOH and Fe₃O₄ were 84 m² g⁻¹ and 32 m² g⁻¹, respectively. The larger particle size of Fe₃O₄ led to a lower surface area.

Fig. 2 TEM images of (a) α-FeOOH and (b) Fe₃O₄.

Cr-adsorption ability

Effect of pH. The PZC values of Fe₃O₄ and α-FeOOH were 5.95 ± 0.04 and 3.95 ± 0.06, respectively (S7†). The Cr-adsorption dependency on pH is shown in Fig. 3. When pH < PZC, surfaces of Fe₃O₄ and α-FeOOH are protonated to a positive charge. Thus, the adsorption behavior of Fe₃O₄ and α-FeOOH is facilitated by an electrostatic mechanism with negative Cr-compounds (HCrO₄⁻, CrO₄²⁻, and Cr₂O₇²⁻).⁴⁰ The adsorbent surface does not favor Cr⁶⁺ adsorption with a pH increase.⁴¹ The highest Cr-adsorption ability of both adsorbents was detected at pH = 3.

Fig. 3 Effect of pH on Cr⁶⁺ uptake by the Fe₃O₄ and α-FeOOH adsorption condition: Cr⁶⁺ initial concentration = 1 ppm, adsorbent dose = 0.4 g L⁻¹, time = 30 min.

Fe₃O₄ had a higher Cr-adsorption ability than α-FeOOH, indicating its stronger attraction to the negative compounds of Cr⁶⁺. Thus, pH = 3 was selected as the optimized value to reach the maximum capacity of the adsorbents. The maximum Cr⁶⁺ uptakes by α-FeOOH and Fe₃O₄ were determined as about 2.08 mg g⁻¹ and 2.87 mg g⁻¹, respectively. In addition, the effects of adsorbent dose, initial concentration and contact time of Cr-adsorption were discussed in S8–S10.†

Adsorption behavior of Fe₃O₄ and α-FeOOH

Adsorption kinetics. The kinetic parameters of Cr⁶⁺ adsorption on α-FeOOH and Fe₃O₄ are summarized in S13.† The given correlation coefficient (R²) of the adsorbents obeyed the pseudo-second order model. The model implies that the rate law for adsorption is described in terms of an initial concentration of Cr rather than in terms of adsorbent concentration. In addition, the chemisorption process is implicated in a rate-determining step of Cr on both ion-oxides.^45,46

Adsorption thermodynamics. Adsorption isotherms were fitted based on the Langmuir and Freundlich, Dubinin–Radusckevisch–Kanager and Temkin models (S11†), and the corresponding fitting parameters and correlation coefficients are listed in S12.† The Cr⁶⁺ adsorption behaviors of α-FeOOH and Fe₃O₄ obeyed the Langmuir model (higher regression coefficient (R²)) indicating that adsorption takes place at a specific homogeneous site, and monolayer coverage was formed on the adsorbent surface.^42–44 A low b value implied a strong binding ability.²⁸ Therefore, the lower b value of Fe₃O₄ indicated its stronger Cr⁶⁺ adsorption ability than that of α-FeOOH.

The thermodynamic parameters of α-FeOOH and Fe₃O₄ are listed in S14.† Gibb's free energy changes (ΔG°) for both adsorbents had negative values, confirming a spontaneous nature of adsorption.³⁴ An α-FeOOH had a negative standard enthalpy change (ΔH°) implying an exothermic process.⁴⁷ α-FeOOH had a negative entropy change (ΔS°) indicating the decreased randomness at the solid/solution interface.⁴⁸ In contrast, Fe₃O₄ showed an endothermic process from a positive ΔH° value indicating that an ion-exchange on the adsorbent surface might occur.⁴⁹ The positive ΔS° of Fe₃O₄ implied randomness at the solid–solution interface during the adsorption process due to adsorption and subsequent desorption.⁴⁸

The reusability of both adsorbents for five cycles is shown in Fig. 4 to support the thermodynamic hypothesis of α-FeOOH and Fe₃O₄. Different adsorption behaviors of the adsorbents were exhibited. The Cr⁶⁺ uptake of Fe₃O₄ dramatically decreased (90–95%) in the subsequent cycles. On the other hand, the Cr⁶⁺ uptake of α-FeOOH slowly decreased by 55–60% in the subsequent cycles. These results implied that the Fe₃O₄ surface was completely adsorbed by Cr-species and it was difficult to detach from the surface. In contrast, the α-FeOOH surface was weakly adsorbed by the adsorbent due to a weak interaction (physisorption) which probably detached from the surface.

Fig. 4 Reusability for Cr⁶⁺ removal of Fe₃O₄ and α-FeOOH adsorbents with Cr⁶⁺ initial concentration = 1 ppm, adsorbent dose = 0.4 g L⁻¹, pH = 3, time = 30 min.

Analysis of the adsorbed species on adsorbents

Surface analysis of α-FeOOH and Fe₃O₄ by XPS. The oxidation states of surface elements in the pristine and Cr-adsorbed adsorbents were analyzed by XPS (Fig. 5). The C 1s peak at 284.8 eV was used as a reference to correct electrostatic charging. The binding energies of Fe 2p_3/2 and Fe 2p_1/2 in Fig. 5a and b corresponded to Fe₃O₄ and α-FeOOH. The peaks of octahedral Fe²⁺, octahedral Fe³⁺, and tetrahedral Fe³⁺ at 710.7 eV, 711.7 eV, and 714.1 eV were determined in the pristine Fe₃O₄ sample corresponded to the typical Fe₃O₄ structure. The corresponding energies of Cr@Fe₃O₄ were found at 710.3 eV, 711.7 eV, and 713.9 eV, respectively, which were slightly lower than those of the pristine Fe₃O₄. In addition, the area ratio of Fe²⁺_Oh/Fe³⁺_Oh in Cr@Fe₃O₄ was higher than the pristine Fe₃O₄. The result indicated a decrease of Fe³⁺_Oh in Fe₃O₄ after Cr-adsorption. The Fe 2p energies at 710.7 eV and 712.3 eV corresponding to Fe₂O₃ and α-FeOOH components, respectively, were observed on both the pristine α-FeOOH and Cr@FeOOH.⁵⁰ It was implied that surface species of α-FeOOH were not changed after Cr-adsorption.

Fig. 5 Fe 2p XPS spectra before and after Cr⁶⁺ adsorption of (a) Fe₃O₄ and (b) α-FeOOH and O 1s spectra before and after Cr⁶⁺ adsorption of (c) Fe₃O₄ and (d) α-FeOOH.

The O 1s peaks of pristine Fe₃O₄ at 530.5 eV, 531.8 eV, and 533.1 eV appeared from lattice oxygen (O²⁻_lattice), oxygen in the adsorbed-hydroxyl group (OH_ads), and oxygen in water-adsorbed molecules (H₂O_ads), respectively (Fig. 5c and d). After Cr adsorption, the binding energies of O²⁻_lattice and OH_ads slightly decreased to 530.1 eV and 530.9 eV, respectively, indicating a presence of O²⁻_lattice in CrO(OH) over the Fe₃O₄ surface.^51–53 The results exhibited a partial ion-exchange Cr³⁺ in Fe³⁺ sites which well corresponded to the decrease of Fe³⁺_Oh constituent as mentioned in the Fe 2p energy. Moreover, no peak corresponding to water-adsorbed species on Cr@Fe₃O₄ was determined because Cr-adsorbed species fully covered the Fe₃O₄ surface. The O 1s peaks of α-FeOOH at 530.5, 530.5 and 531.5 eV corresponded to O²⁻_lattice, OH_ads and adsorbed H₂O species, respectively. The adsorbed H₂O was not determined on Cr@FeOOH. In addition, the binding energy of O²⁻_lattice in Cr@FeOOH was slightly higher than that in α-FeOOH. These results possibly indicate the presence of precipitate oxide form of Cr₂O₃ on the α-FeOOH surface.

The binding energies of Cr 2p_3/2 and Cr 2p_1/2 in Cr@Fe₃O₄ and Cr@α-FeOOH are shown in Fig. 6. The Cr 2p_3/2 to Cr 2p_1/2 splitting energies of both samples were about 10 eV which was typically found in a Cr³⁺/Cr⁶⁺ mixture.⁵⁴ The Cr 2p_3/2 exhibited the characteristics of Cr³⁺ in Cr₂O₃ (576.6 eV), CrO(OH)/Cr(OH)₃ (577.7 eV) and Cr⁶⁺ in CrO₃ (579.9 eV).^{7,42,44,50–54} Those characters were similarly found in Cr@Fe₃O₄ and Cr@FeOOH. However, the area ratio of Cr₂O₃/CrO(OH) in Cr@FeOOH was higher than that in Cr@Fe₃O₄. The finding correlated with the enhancement of O²⁻_lattice species of Cr@FeOOH in O 1s spectrum. This result revealed the presence of precipitated Cr₂O₃ species on the α-FeOOH surface at pH 3.⁵⁵ It was noticeable that a high CrO(OH) content in Cr@Fe₃O₄ might proceed through Cr³⁺ migration on the Fe₃O₄ surface.

Fig. 6 Cr 2p XPS spectra of Cr@Fe₃O₄ and Cr@FeOOH.

Bulk analysis of α-FeOOH and Fe₃O₄ by XAS

XANES of Fe K-edge and L₃-edge. The Fe K-edge spectra of Fe₃O₄ and α-FeOOH compared with the iron oxide standards are shown in Fig. 7. The Fe K-edge energy of the pristine Fe₃O₄ was calculated as 7120.7 eV while the energy of Cr@Fe₃O₄ slightly shifted to a lower value of 7120.5 eV. This result indicated that ferrous specie (Fe²⁺) in Fe₃O₄ increased after Cr-adsorption. On the other hand, the Fe K-edge energies of the pristine α-FeOOH and Cr@FeOOH were similar (Fig. 7b). Thus, the bulk structure of α-FeOOH was not affected by Cr-adsorption process.

Fig. 7 Fe K-edge XANES spectra of bare and Cr-adsorbed (a) Fe₃O₄ and (b) α-FeOOH and (c) Fe L₃-edge XANES spectra of Cr@Fe₃O₄ and Cr@FeOOH comparing with the iron standards.

In principle, Fe L₃-edge has a better resolution for fine structures and manifests a lower intrinsic lifetime broadening than Fe K-edge. Thus, the local geometry of Fe₃O₄ was further investigated by Fe L₃-edge. The spectra of Fe₃O₄ and Cr@Fe₃O₄ are presented in Fig. 7c. The I_a/I_b intensity ratio in the Fe L₃-edge spectrum was used to determine the mixed valence state.⁵⁶ The Cr@Fe₃O₄ had I_a/I_b ratio of 0.59 which was higher than that of pristine Fe₃O₄ (0.53). The higher I_a/I_b intensity ratio corresponded to the lower Fe²⁺ content in the tetrahedral site of Fe₃O₄.⁵⁶

XANES of Cr K-edge. A comparison of Cr K-edge in the Cr-adsorbed samples with CrCl₃ (Cr³⁺) and K₂Cr₂O₇ (Cr⁶⁺) standards are shown in Fig. 8. The Cr⁶⁺ (4-fold Cr–O) character was exhibited by a strong pre-edge intensity at 5999 eV.⁵⁷ Besides, a weak pre-edge intensity was found in Cr³⁺(6-fold Cr–O). XANES profile of Cr@Fe₃O₄ explored a reduction of pre-edge intensity compared with the Cr⁶⁺ standard. Thus, the Fe₃O₄ could partially reduce Cr⁶⁺-species (HCrO₄⁻, CrO₄²⁻, Cr₂O₇²⁻) to Cr³⁺ on the surface under aqueous solution at pH 3. This behavior was also observed on Cr@FeOOH.⁵⁸ However, a decrease in pre-edge intensity of Cr@FeOOH was much lower than that of Cr@Fe₃O₄. Furthermore, Cr⁶⁺ was a major species in the solution after adsorption (S15†).

Fig. 8 Cr K-edge XANES spectra of Cr@Fe₃O₄ and Cr@FeOOH comparing with Cr-standards.

XANES of O K-edge. The O K-edge XANES spectra and their derivative plots of the pristine and Cr-adsorbed Fe₃O₄ and α-FeOOH structures are exhibited in Fig. 9. The O K-edge energy of Cr@Fe₃O₄ (Fig. 9a) shifted to higher values compared to pristine Fe₃O₄. The result revealed a decrease in the electron density of oxygen atoms in the bulk structure of Cr@Fe₃O₄. This finding was probably due to an enhancement of oxygen bonding with a higher ionic strength species of Fe³⁺ rather than Fe²⁺. Hence, an increase of Fe³⁺–O component in the Cr@Fe₃O₄ structure was suggested. An increase of Fe³⁺ species in Cr@Fe₃O₄ was in good agreement with Fe K-edge XANES. In addition, on Fig. 9b, O K-edge energy of Cr@Fe₃O₄ nearly fitted to α-FeOOH. Thus, a part of FeO(OH) as well as CrO(OH) could be created on the Fe₃O₄ structure after Cr-adsorption. Meanwhile, the O K-edge energies of α-FeOOH and Cr@FeOOH were similar indicating no structural change in the bulk structure of α-FeOOH during Cr adsorption.

Fig. 9 O K-edge XANES spectra (a) and their derivatives (b) of bare and Cr-adsorbed Fe₃O₄ and α-FeOOH.

EXAFS analysis of Fe and Cr K-edge

EXAFS of Fe K-edge. The local structures of Fe₃O₄ and Cr@Fe₃O₄ were investigated by Fe K-edge EXAFS analysis. The EXAFS analysis is reported in Table 1 and S16.† The 4-fold and 2-fold Fe–O bonds with distances of 1.61 Å and 1.89 Å, respectively, were observed in bare Fe₃O₄. The bond length of 1.61 Å was shorter than the typical tetrahedral Fe–O bond length in Fe₃O₄ (1.88 Å) due to the existence of oxygen vacancies.⁵⁹ The second shell of the Fe₃O₄ structure consisted of Fe–O (3.14 Å) and Fe–Fe (3.47 Å) bonds. After Cr adsorption, the first shell represented 2-fold (1.97 Å) and 4-fold (2.40 Å) of Fe–O bonds. The bond distances fitted well with octahedral Fe–O in Fe₂O₃ or tetrahedral Fe–O in α-FeOOH. For α-FeOOH, the first shell consisted of tetrahedral Fe–O bonds with an average bond distance of 1.97 Å (Table 1). Moreover, 3-fold Fe–Fe bonds with bond distances of 3.00 Å and 3.30 Å were also displayed. Octahedral Fe–O (1.98 Å) and Fe–Fe (3.00 Å) bonds in Cr@FeOOH were similar to α-FeOOH. Thus, the bulk structure of α-FeOOH did not change after Cr adsorption.

Table 1 k² (χ) weight of Fe K-edge EXAFS of bare and Cr-adsorbed Fe₃O₄ and α-FeOOH

Sample	Bond	Distance (Å)	Coordination number (CN)	Debye–Waller factor (σ^2)
Fe3O4	Fe–O	1.61	3.9	0.054
	Fe–O	1.89	2.8	0.006
	Fe–O	3.14	12.0	0.048
	Fe–Fe	3.47	12.0	0.013
Cr@Fe3O4	Fe–O	1.97	2.2	0.012
	Fe–O	2.40	3.8	0.059
	Fe–Fe	2.76	0.9	0.004
	Fe–Fe	3.00	2.9	0.008
	Fe–Fe	3.49	3.2	0.013
α-FeOOH	Fe–O	1.97	4.3	0.012
	Fe–Fe	3.00	3.1	0.012
	Fe–Fe	3.30	3.0	0.013
	Fe–O	3.40	3.0	0.002
	Fe–Fe	3.89	12.0	0.035
Cr@FeOOH	Fe–O	1.98	6.0	0.013
	Fe–Fe	3.00	6.0	0.017
	Fe–O	3.32	6.0	0.003
	Fe–O	3.81	12.0	0.004

---

EXAFS of Cr K-edge. The Cr K-edge EXAFS spectrum of Cr@Fe₃O₄ is displayed in Fig. 10a. Cr–O bonds consisted of short Cr⁶⁺–O bonds (1.22 Å and 1.56 Å) and long Cr³⁺–O bonds (1.87 Å) (Table S17†).²⁹ Hence, mixed phases of Cr₂O₃/CrOOH (Cr³⁺) and CrO₃ (Cr⁶⁺) were formed on Cr@Fe₃O₄. Moreover, 2-fold Cr–Cr bonds of 2.41 Å and 3.62 Å correlated to binuclear Cr₂O₇²⁻ species which could be physisorbed on the adsorbent in an aqueous solution at pH 3.

Fig. 10 Fourier transforms of the k³ (χ) weighted in Cr K-edge EXAFS data collected from (a) Cr@Fe₃O₄ and (b) Cr@FeOOH.

The Cr K-edge EXAFS analysis of Cr@FeOOH are presented in Fig. 10b and Table S17.† Tetrahedral Cr–O bonds with the bond distances of 1.33 Å and 1.59 Å corresponded to Cr⁶⁺ in HCrO₄⁻. Moreover, the average bond distance of 2-fold Cr–Cr bonds of 3.17 Å was found in Cr@FeOOH which was different from Cr–Cr bonds in Cr@Fe₃O₄ confirming that the HCrO₄⁻ species was presented on α-FeOOH.

Removal mechanism of Cr⁶⁺ on Fe₃O₄ and α-FeOOH. In the solution pH 2.0–6.5, Cr⁶⁺ in an aqueous solution existed as a predominant form of HCrO₄⁻ (eqn (8)). This species moved and adsorbed on the surface of both adsorbents. Some Cr⁶⁺ was reduced to Cr³⁺ or precipitated on surfaces which confirmed by the existence of Cr₂O₃, CrO(OH)/Cr(OH)₃ and CrO₃ from XPS surface analysis. Compared with α-FeOOH, the lower Cr₂O₃/CrO(OH) ratio in Cr@Fe₃O₄ implied the Cr-ion migration into the bulk structure of Fe₃O₄. XANES and EXAFS as bulk analysis further revealed this behavior. The contained Fe²⁺ species inside Fe₃O₄ could reduce Cr⁶⁺ in HCrO₄⁻ to Cr³⁺ (eqn (9)) with a relatively slow reaction rate due to the limited electron transfer rate in the maghemite (Fe₂O₃) passivation layer.³⁰ Then, Cr³⁺ species migrated to bulk Fe₃O₄ and formed Fe_(1−x)Cr_xOOH containing Fe³⁺ and Cr³⁺ in between FeO lattice (eqn (10)). These dissolved octahedral Cr³⁺–O sites could prevent the bulk structure of Fe₃O₄ from collapsing.^60,61

H₂CrO₄ → HCrO₄⁻ + H⁺ (8)

2HCrO₄⁻ + 2Fe²⁺ + 10H⁺ → 2Fe³⁺ + 2Cr³⁺ + 6H₂O + O₂ (9)

(1−x)Fe³⁺ + (x)Cr³⁺ + 2H₂O → Fe_(1−x)Cr_xOOH(s) + 3H⁺ (10)

Additionally, the XPS and XAS results revealed the increase of FeO species in Fe₃O₄ after Cr-adsorption. Thus, the most likely form was a Fe_(3−x)Cr_xO₄ solid solution containing Fe²⁺ and Cr³⁺ in forms of FeO and Cr₂O₃.⁶² The presented species might be created through Cr⁶⁺ reduction mechanism by FeO in Fe₃O₄ (or FeO·Fe₂O₃) structure as expressed by eqn (11).

2FeO + 4HCrO₄⁻ + 4H⁺ → 2Fe_(3−x)Cr_xO₄ + 4H₂O + 3O₂ (11)

Meanwhile, α-FeOOH surface changed to positive [FeO–OH₂]²⁺ species at pH = 3.0 (eqn (12)). These [FeO–OH₂]²⁺ molecules could react with the presented negative HCrO₄⁻ species by an electrostatic force (physisorption) and created FeO–HCrO₄ species (eqn (13)).⁷ This adsorption behavior did not change the local geometry and oxidation state of iron atoms in α-FeOOH.

FeOOH + H⁺ → FeO–OH₂⁺ (12)

FeO–OH₂⁺ + HCrO₄⁻ → FeO–HCrO₄ + H₂O (13)

Conclusions

Fe₃O₄ (magnetite) and α-FeOOH (goethite) synthesized by the hydrothermal method were mainly in cubic and orthorhombic structures, respectively. The Cr adsorption isotherms of Fe₃O₄ and α-FeOOH fitted well with the Langmuir model, indicating a monolayer adsorption process. The maximum Cr⁶⁺ uptake of 2.87 mg g⁻¹ was found by Fe₃O₄. After Cr-adsorption, the Cr³⁺-doped forms of Fe_(1−x)Cr_xOOH and Fe_(3−x)Cr_xO₄ solid solution was observed in bulk Fe₃O₄ resulting in a high Cr-adsorption capacity. Besides, an electrostatic attraction of FeO–HCrO₄ occurred only on the α-FeOOH surface leading to a low Cr-adsorption capacity. The Cr-adsorption ability and behavior in an aqueous solution of the adsorbents were strongly influenced by an adsorbed species stabilized on their structures. The discovered Cr-solid solution in Fe₃O₄ could be a promising catalyst preparation through water treatment process.

Author contributions

Nichapha Senamart-data curation, formal analysis, writing-original draft, writing-review & editing; Krittanun Deekamwong-writing-review & editing; Jatuporn Wittayakun-supervision & editing; Sanchai Prayoonpokarach-review & editing; Narong Chanlek-data analysis; Yingyot Poo-arporn-review & editing; Suttipong Wannapaiboon-data curation, formal analysis, writing; Pinit Kidkhunthod-data analysis and Sirinuch Loiha-conceptualization, funding acquisition, supervision, writing-review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Science Achievement Scholarship of Thailand (SAST), Center for Innovation in Chemistry (PERCH-CIC). Materials Chemistry Research Center (MCRC), Thailand. Department of Chemistry, Faculty of Science, Khon Kaen University, Khon Kaen, Thailand. Research and Graduate Studies Department, KKU (Research Program, grant number RP64-6/001), fiscal year 2021. Synchrotron Light Research Institute (SLRI), Nakhon Ratchasima, Thailand (BL1.1W, BL2.2, BL3.2Ua, BL5.2 and BL5.3). Suranaree University of Technology (SUT), Thailand Science Research and Innovation (TSRI), and National Science, Research and Innovation Fund (NSRF) (90464/project code 90464).

References

1. A. Chiu, X. L. Shi, W. K. P. Lee, R. Hill, T. P. Wakeman, A. Katz, B. Xu, N. S. Dalal, J. D. Robertson, C. Chen, N. Chiu and L. Donehower, J. Environ. Sci. Health, Part C: Environ. Carcinog. Ecotoxicol. Rev., 2010, 28, 188–230.
2. M. Narayani and K. V. Shetty, Crit. Rev. Environ. Sci. Technol., 2012, 43, 955–1009.
3. R. Rakhunde, L. Deshpande and H. D. Juneja, Crit. Rev. Environ. Sci. Technol., 2011, 42, 776–810.
4. E. J. Tomaszewski, S. Lee, J. Rudolph, H. Xu and M. Ginder-Vogel, Chem. Geol., 2017, 464, 101–109.
5. L. Li, Y. Li, L. Cao and C. Yang, Carbohydr. Polym., 2015, 125, 206–213.
6. P. Lazo, J. Int. Environ. Appl. Sci., 2009, 4, 207–213.
7. S. Wu, J. Lu, Z. Ding, N. Li, F. Fu and B. Tang, RSC Adv., 2016, 6, 82118–82130.
8. S. R. Chowdhury, E. K. Yanful and A. R. Pratt, J. Hazard. Mater., 2012, 235–236, 246–256.
9. X. Lv, J. Xu, G. Jiang, J. Tang and X. Xu, J. Colloid Interface Sci., 2012, 369, 460–469.
10. Y. L. Wei, Y. C. Lee and H. F. Hsieh, Chemosphere, 2005, 61, 1051–1060.
11. A. H. Meena and Y. Arai, Geochem. Trans., 2016, 17, 1.
12. M. Gheju and I. Balcu, J. Hazard. Mater., 2011, 196, 131–138.
13. A. Maleki, B. Hayati, M. Naghizadeh and S. W. Joo, J. Ind. Eng. Chem., 2015, 28, 211–216.
14. M. Gheju, Water, Air, Soil Pollut., 2011, 222, 103–148.
15. S. M. El-Sheikh, A. B. Azzam, R. A. Geioushy, F. M. El Dars and B. A. Salah, J. Alloys Compd., 2021, 857, 157513.
16. R. A. Geioushy, S. M. El-Sheikh, A. B. Azzam, B. A. Salah and F. M. El-Dars, J. Hazard. Mater., 2020, 381, 120955.
17. F. Di Natale, A. Erto, A. Lancia and D. Musmarra, J. Hazard. Mater., 2015, 281, 47–55.
18. S. M. Lee, L. Lalchhingpuii and D. Tiwari, Chem. Eng. J., 2016, 296, 35–44.
19. Q. Liu, Q. Liu, W. Ma, W. Liu, X. Cai and J. Yao, Colloids Surf., A, 2016, 511, 8–16.
20. E. I. Basaldella, P. G. Vázquez, F. Iucolano and D. Caputo, J. Colloid Interface Sci., 2007, 313, 574–578.
21. L. N. Døssing, K. Dideriksen, S. L. S. Stipp and R. Frei, Chem. Geol., 2011, 285, 157–166.
22. D. L. Sedlak and P. G. Chan, Geochim. Cosmochim. Acta, 1997, 61, 2185–2192.
23. A. E. Gleason, R. Jeanloz and M. Kunz, Am. Mineral., 2008, 93, 1882–1885.
24. H. Abdel-Samad and P. R. Watson, Appl. Surf. Sci., 1997, 108, 371–377.
25. T. Mineno and M. Okazaki, Soil Sci. Plant Nutr., 2004, 50, 1043–1046.
26. V. D. Nguyen, J. Kynicky, P. Ambrozova and V. Adam, Materials, 2017, 10, 783.
27. M. Kumari, C. U. Pittman Jr and D. Mohan, J. Colloid Interface Sci., 2014, 442, 120–132.
28. S. Rajput, C. U. Pittman and D. Mohan, J. Colloid Interface Sci., 2016, 468, 334–346.
29. A. H. Meena and Y. Arai, Geochem. Trans., 2016, 17, 1.
30. J. Chang, H. Wang, J. Zhang, Q. Xue and H. Chen, Colloids Surf., A, 2021, 611, 125784.
31. N. Senamart, J. Watcharakitti, K. Atthawilai, K. Phedluen, P. Nasomjai and S. Loiha, Suranaree J. Sci. Technol., 2018, 25, 445–454.
32. P. Ou, G. Xu, Z. Ren, X. Hou and G. Han, Mater. Lett., 2008, 62, 914–917.
33. U. Pal, S. Uribe Madrid and F. Sanchez De-jesus, Adv. Nano. Res., 2014, 2, 187–198.
34. G. Du, Z. Li, L. Liao, R. Hanson, S. Leick, N. Hoeppner and W. T. Jiang, J. Hazard. Mater., 2012, 221–222, 118–123.
35. M. Kazemi, M. Jahanshahi and M. Peyravi, J. Hazard. Mater., 2018, 344, 12–22.
36. O. Ali and S. K. Mohamed, Turk. J. Chem., 2017, 41, 967–986.
37. F. Geng, Z. Zhao, J. Geng, H. Cong and H.-M. Cheng, Mater. Lett., 2007, 61, 4794–4796.
38. A. R. Amani-Ghadim, S. Alizadeh, F. Khodam and Z. Rezvani, J. Mol. Catal. A: Chem., 2015, 408, 60–68.
39. S. Alibeigi and M. R. Vaezi, Chem. Eng. Technol., 2008, 31, 1591–1596.
40. Y. F. Shen, J. Tang, Z. H. Nie, Y. D. Wang, Y. Ren and L. Zuo, Sep. Purif. Technol., 2009, 68, 312–319.
41. N. Guo, X. Lv, Q. Li, T. Ren, H. Song and Q. Yang, Microporous Mesoporous Mater., 2020, 299, 110101.
42. Z. Liu, G. Chen, L. Xu, F. Hu and X. Duan, ChemistrySelect, 2019, 4, 13817–13827.
43. M. Bhaumik, A. Maity, V. V. Srinivasu and M. S. Onyango, J. Hazard. Mater., 2011, 190, 381–390.
44. S. U. N. Zhenya, Z. H. U. Chunshui, H. Jiangbo, G. Wenqi, C. Hesheng and M. U. Shanbin, Acta Geol. Sin., 2006, 80, 597–603.
45. X. Q. Zhang, Y. Guo and W.-C. Li, RSC Adv., 2015, 5, 25896–25903.
46. M. Ahmaruzzaman, Adv. Colloid Interface Sci., 2011, 166, 36–59.
47. G. Vijayakumar, R. Tamilarasan and M. Dharmendirakumar, J. Mater. Environ. Sci., 2012, 3, 157–170.
48. M. Hasanpour and M. Hatami, Adv. Colloid Interface Sci., 2020, 284, 102247.
49. Y. H. Li, Z. Di, J. Ding, D. Wu, Z. Luan and Y. Zhu, Water Res., 2005, 39, 605–609.
50. A. P. Grosvenor, B. A. Kobe, M. C. Biesinger and N. S. McIntyre, Surf. Interface Anal., 2004, 36, 1564–1574.
51. X. Zhang, W. G. Sloof, A. Hovestad, E. P. M. van Westing, H. Terryn and J. H. W. de Wit, Surf. Coat. Technol., 2005, 197, 168–176.
52. Z. Wen, J. Ke, J. Xu, S. Guo, Y. Zhang and R. Chen, Chem. Eng. J., 2018, 343, 416–426.
53. Z. Wen, Y. Zhang, S. Guo and R. Chen, J. Colloid Interface Sci., 2017, 486, 211–218.
54. Z. Wen, Y. Zhang, G. Cheng, Y. Wang and R. Chen, Chemosphere, 2019, 218, 1002–1013.
55. S. M. Walker, M. C. Marcano, W. M. Bender and U. Becker, Chem. Geol., 2016, 429, 60–74.
56. B. P. von der Heyden, A. N. Roychoudhury, T. Tyliszczak and S. C. B. Myneni, Am. Mineral., 2017, 102, 674–685.
57. S. Ould-Chikh, O. Proux, P. Afanasiev, L. Khrouz, N. Hedhili Mohamed, H. Anjum Dalaver, M. Harb, C. Geantet, J. M. Basset and E. Puzenat, ChemSusChem, 2014, 7, 1361–1371.
58. C. J. Keturakis, M. Zhu, E. K. Gibson, M. Daturi, F. Tao, A. I. Frenkel and I. E. Wachs, ACS Catal., 2016, 6, 4786–4798.
59. F. Pinakidou, M. Katsikini, K. Simeonidis, E. Kaprara, E. C. Paloura and M. Mitrakas, Appl. Surf. Sci., 2016, 360, 1080–1086.
60. S.-f. Niu, Y. Liu, X.-h. Xu and Z.-h. Lou, J. Zhejiang Univ., Sci., B, 2005, 6, 1022–1027.
61. T. Kendelewicz, P. Liu, C. S. Doyle and G. E. Brown, Surf. Sci., 2000, 469, 144–163.
62. X. Wang, N. Chen and L. Zhang, Environ. Sci.: Nano, 2019, 6, 2185–2194.

---

Footnote

† Electronic supplementary information (ESI) available. See https://doi.org/10.1039/d2ra03676b

---