Highly organized nanofiber formation from zero valent iron nanoparticles after cadmium water remediation

Abstract

Many studies have used nanoscale zero valent iron (nZVI) nanoparticles to remove redox-sensitive metals (e.g., As, Cr, U, Se, Ni, Cu) from aqueous systems by absorption or reduction processes. However, very few investigations present a detailed study of the product formed after the remediation process. In order to quantify the efficiency of nZVI particles as a possible cadmium remediation agent, we prepared nZVI by sodium borohydride reduction of an iron complex, FeCl₃·6H₂O, at room temperature and ambient pressure. Fe⁰ and nanocrystalline structures of iron oxides and oxyhydroxides were obtained with this method. We exposed the nZVI to 6 ppm of Cd²⁺ and characterized the products with X-ray diffraction, X-ray absorption and X-ray photoelectron spectroscopy. Inductively coupled plasma analysis showed that the nZVI remediation efficiency of cadmium ions was between 80% and 90% in aqueous media. All of the physical characterization results confirmed the presence of Fe⁰, α-Fe₂O₃ and FeOOH. High resolution transmission electron microscopy images showed nanofiber formation of a mixture of Fe⁰, oxyhydroxides and oxides iron formed after interacting with cadmium ions, possibly forming CdFe₂O₄. These results suggest that the FeOOH shell and other iron oxides in nZVI could enhance Cd²⁺ removal. This removal is observed to cause a change of the initial structure of nZVI to nanofibers due to possible formation of CdFe₂O₄ as a waste product.

---

A. Introduction

Cadmium is a metal used today by industry (e.g. electronic industry) that has been released to the environment through the combustion of fossil fuels, metal production, use of phosphate fertilizers, electroplating, and the manufacturing of batteries, pigments, and computer screens.¹ This metal tends to accumulate in plant and animal tissues, reaching dangerous levels in the food chain, even when released in small amounts.^2–4 Cadmium is toxic at high concentrations because it disrupts enzyme functions, replacing essential metals in pigments or producing reactive oxygen species.⁵ It is poorly retained by the soil and is readily absorbed by plants.^2,4,6 This heavy metal can be accumulated in leaves and cause phytotoxic effect on plants, such as chloroplast structure change and cell death.^7,8 In hazardous waste sites, the levels of cadmium in the soil has become about 4 parts per million, while in water the levels have reached 6 ppm.⁹

U.S. Environmental Protection Agency (EPA) recognize cadmium to be a probable human carcinogen (e.g. breast cancer) and has been classified as a Group B1 carcinogen.^9–11 Since cadmium represents a great threat to human health, considerable effort has been placed on the development of techniques to remove this contaminant from water.¹²

In the course of developing suitable options to treat heavy metal contaminants in wastewater, nanoscale zero valent iron (nZVI) particles have been found to be an alternative approach to reduce the concentration of several kinds of contaminants.^13,14 This approach provides a real alternative to remove cadmium by using particles at nanoscale. Cadmium remediation is of the utmost importance because Cd species have been detected in aquatic ecosystems and found to bioaccumulate in organisms in nanomolar to micromolar concentrations.² Iron nanoparticles are particularly attractive for remediation purposes due to their significant surface area to mass ratio, leading to a greater density of reactive site and heavy metal removal capacity.¹⁴ Because of their smaller size, nanoscale based iron materials are much more reactive than conventional iron powders and they can be easily suspended in slurry and pumped straight to the contaminated site.^15,16 Efficient nZVI remediations of groundwater contaminants have been shown in multiples studies.^12,17–20 However, in the literature there is a lack of comparable studies for different nZVI materials and deployment strategies.¹³

Very few investigations of nZVI materials present a detailed study of the products formed in the remediation process. Various adsorption and kinetic models to describe metal adsorption on nZVI and Cd–nZVI surfaces using SEM/EDX and XPS measurements have been studied^1,13 but, to our knowledge, the interactions between surface Fe(0) and Cd(II) and subsequent cadmium retention in nZVI particles has not been the subject of detailed study. The specific objectives of this work are to (1) analyze by inductively couple plasma measurements the efficiency of the nanoscale zero valent iron (nZVI) at low treatment dose in the removal of Cd(II); (2) characterize the nanoparticles of Fe⁰, Fe⁰–H₂O and Fe⁰–Cd using spectroscopy and microscopy techniques; and (3) establish a conceptual model of the Cd ions with the surface of nZVI. Results from this study provide important information of the products formed during the remediation process of cadmium. We also report an unexpected nanofiber formation and the potential synthesis of cadmium ferrite. This remediation product has been suggested as a photocatalyst material that has tremendous potential application for light harvesting.²¹ These results could be useful because by understanding the products of remediation processes, alternative applications may be suggested for them. This will decrease the amount of new toxic substances formed after water remediation processes.

B. Experimental

B.1. Chemicals

The chemicals used were iron(III) chloride hexahydrate (97%, ACS reagent), sodium borohydride (≥98.5%, reagent grade), trichloroethylene (99.5+%, ACS reagent), Tween® 80 and humic acid (sodium salt) obtained from Sigma-Aldrich, FeCl₃·6H₂O (F2877 Sigma Aldrich), NaBH₄ (Alfa Aesar), ethanol 200 proof, 99.5% anhydride (Sigma Aldrich), Cd (CHOO₃)₂·2H₂O (Sigma Aldrich) and HCl (Alfa Aesar). All solutions were prepared with deionized water (18.2 MΩ cm, Nanopure Diamond, Barnstead).

B.2. Preparation of nanoscale zero valent iron

Nanoscale zero valent iron (nZVI) particles were produced by adding NaBH₄ as the reducing agent to a solution of FeCl₃·6H₂O. The resulting reductive reaction may be given as:²²

2Fe³⁺ + 6BH₄⁻ + 18H₂O → 2Fe⁰(s) + 6B(OH)₃ + 21H₂ (1)

To obtain 1.13 g of nZVI, a 0.6 M solution of FeCl₃·6H₂O (molar mass 270.30 g mol⁻¹) was prepared in 30 mL of ethanol (83% v/v) and nanopure water (17% v/v). The solution was purged with N₂ for 30 minutes to remove oxygen prior to the reaction; this was done to avoid the rapid oxidation of iron. Then, the solution was titrated by adding a total of 100 mL of 0.8 M NaBH₄ aqueous solution under a nitrogen atmosphere.

After 30 minutes of stirring with a magnetic bar at 700 rpm, the solution was filtered using a 0.22 μm filter paper (Millipore) under vacuum at 25 °C.

The sample was rinsed three times with 99% absolute ethanol. Various filtered samples, prepared as mentioned, were placed immediately in a vacuum desiccator.

B.3. Preparation of cadmium concentrations in synthetic wastewater

Cadmium solutions were prepared using Cd(CH₃COO)₂·2H₂O due to its high solubility in water and ethanol. To represent different samples of wastewater contaminated with cadmium, a solution of 6 ppm was prepared. This was followed by two dilutions of the 6 ppm solution (stock concentration) to prepare solutions of 3 ppm and 1 ppm. The initial pH of each solution was adjusted to 2 using 1.0 M HCl. Concentration units were presented as “ppm” in keeping with toxicity nomenclature (ppm = mg L⁻¹).

B.4. Determination of the reduction of the cadmium concentration in aqueous solution nZVI

All reactions were carried out in a dry box under argon atmosphere to prevent Fe oxidation. In some treatment processes with 1 ppm or more of Zn (which has chemical properties similar to Cd) the dose used was between 1.0 g L⁻¹ to 3.5 g L⁻¹ of nZVI.²³ Here, 200 mL of all the prepared cadmium solutions of 1, 3 and 6 ppm were treated with 3.0 g L⁻¹ of nZVI in periods of 1, 3, and 5 hours. One sample of each solution was taken before and after treatment to corroborate the accuracy in the preparation of the initial concentrations and to determine the final concentration of cadmium.

The pH of each cadmium solution was determined before and after adding the nZVI. To determine the effect of the nanoparticles in the amount of cadmium ions in the samples after certain reaction time, the samples were processed and analyzed using inductively coupled plasma (ICP-AES) Model: PerkinElmer 4300 DV with a detection limit of 0.002 ppm for cadmium.

B.5. Characterization of Fe⁰, Fe⁰/H₂O and Fe⁰/Cd particles

Characterization of nZVI is needed to understand their reaction mechanism and kinetics.¹⁸ In this study, the synthesized nanoparticles and nanofibers were characterized by X-ray Photoelectron Spectroscopy (XPS), X-ray Absorption (XAS), High Resolution Transmission Electron Microscopy (HRTEM) and X-ray diffraction (XRD).

HRTEM images were done with a Cs-corrected JEOL JEMARM200F electron microscope operated at 200 kV and equipped with a CEOS Cs corrector on the illumination system. XRD patterns were obtained on a PANalytical X'Pert Material Research Diffractometer operating with a Cu Kα radiation (λ = 1.54 Å) source. XPS binding energy spectra were obtained using a PHI spectrometer equipped with an Al Kα mono and polychromatic X-ray source operating at 15 kV, 350 W and pass energy of 58.70 eV. All the binding energies reported were corrected fixing the carbon 1s peak (C 1s) at 284.5 eV. X-ray absorption spectroscopy (XAS) was carried at National Synchrotron Light Source (NSLS) in Brookhaven National Laboratory (BNL) using beam line X18A. The iron K edge (7112 eV) was measured for the Cd–nZVI sample in transmission mode. The data acquisition setup consisted of three 12 in. ionization chamber detectors: incidence (I₀) transmittance (I_T) and reference (I_Ref) detector.

Data analysis was done with the Demeter Suite of Programs.²⁴ The Athena software was used for linear combination fittings of the X-ray Absorption Near Edge (XANES) region with theoretical XANES spectra of different iron species. The reference spectra were calculated with the Hephaestus program.

C. Results and discussion

C.1. Analysis of reduction process in cadmium concentration after treatment with Fe⁰ nanoparticles

Previous researchers have studied redox and adsorption processes as mechanisms of metal concentration reduction by nanoscale iron.¹⁵ However, there is very little research that presents in detail adsorption features that occur with certain metals (i.e. cadmium).¹³ Iron nanoparticles have standard a reduction potential (E°) of −0.447 V, which is lower than the reduction potential of cadmium (−0.40 V, 25 °C).^1,13 Similar with other studies, our data suggest that Cd²⁺ is sequestrated on nZVI by adsorption or surface complex formation with no apparent reduction of Cd²⁺.²⁵

To analyze the results and determine the concentration reduction of cadmium in the nZVI sample, we used ICP. According to the results, we observe a change in the decrease of cadmium concentration as treatment time elapsed. The maximum percentages of cadmium adsorption relative to the initial concentration (% C_Cd) for Cd²⁺ aqueous solutions of 1, 3 and 6 ppm were 93%, 91% and 89%, respectively (see Fig. 1). These percentages were obtained in a period of 5 hours, which indicates that longer interaction times between cadmium ions and nZVI promoted larger cadmium concentration reduction.

Fig. 1 Bar graph of the ratio between final and initial cadmium(II) ion concentration (C_Cd) versus time. The nZVI concentration (C_Fe) was 3 g L⁻¹ and cadmium concentrations (C₀) were 1, 3 and 6 ppm. Experimental conditions: pH = 2; 25 °C; argon atmosphere; magnetic bar stirrer at 700 r min⁻¹.

The correlation between the final concentration versus time and the coefficient of determination (R²) for the three concentrations was calculated. The correlations for the 6 ppm, 3 ppm, and 1 ppm Cd²⁺ samples gave values of 0.944, 0.989 and 0.992, respectively. These values indicate that the percentage of the variability of the final cadmium concentration may be proportional with respect to reaction time. Nevertheless, for the 6 ppm Cd²⁺ sample, there seems to be an induction period until the pH reaches a value higher than 4 when the rate of adsorption increases and reaches the same rate as the 3 and 1 ppm samples after 3 hours (Fig. 2).

Fig. 2 Correlation of the final cadmium concentration and solution pH vs. time using nZVI with initial pH of 2 and C_Fe = 3 g L⁻¹; initial Cd²⁺ concentrations of 1, 3, and 6 ppm; temperature of 25 °C and under argon atmosphere; magnetic bar stirrer was at a w of 700 r min⁻¹.

The samples showed changes of pH from 2 to 6 with increasing cadmium treatment time. This is due to the formation of hydroxides (OH⁻) on the surface of the nZVI by its interaction with water.²³ The maximum cadmium reduction efficiency was greater in 5 hours of treatment. Previous studies suggested that the rate and efficiency in metal reduction increases as the pH decreases.²⁵ When water has a basic initial pH, iron may precipitate as an oxyhydroxide. The presence of this elements cause a passivation on the surface of the iron nanoparticles. Thus, acidic pHs should be used for best results.²³

These studies prove that nZVI are a viable option in order to decrease cadmium contamination in affected sites. Since cadmium is being removed from the solution, it must mean that the nZVI plays a role in cadmium sequestering. Thus, it is expected that the used nZVI contains cadmium, which would in turn make the remediation product toxic. For this reason, we decided to analyze the structure of used nZVI in order to gain an understanding of the interactions between nZVI and cadmium. This new knowledge may serve to optimize the remediation process and to provide alternatives uses for the remediation product.

C.2. X-ray photoelectron spectroscopy analysis

In the XPS binding energy spectrum, we observed peaks for Fe 2p_3/2 and O1s of the prepared nanoparticles (Fig. 3A). The XPS spectrum also shows other peaks for oxygen, iron and adventitious carbon (from atmosphere contamination). The amount and extension of rinsing was determined following the presence or absence of B1s and Na1s peaks at the XPS spectrum. Rinsing in this synthesis procedure is a critical step, necessitating the use of absolute alcohol, because water promotes rapid oxidation when the samples are exposed to the atmosphere.²⁶ Cd–nZVI particles were analyzed by high resolution XPS to determine the elemental composition and the valence state of elements on the nZVI surface.

Fig. 3 X-ray photoelectron spectroscopy spectrum of nZVI particles (A) and spectra of the Fe 2p binding energy region for (B) freshly synthesized and (C) one year old, under dry air, of nZVI particles. (D) Spectra of the Cd 3d binding energy region for the samples of nZVI particles exposed to 6 ppm Cd²⁺ aqueous solution (3 g L⁻¹ of nZVI, contact time of 5 h).

The Fe 2p_3/2 peak in the XPS spectrum for elemental iron commonly arises at 707 eV.²⁷ The Fe 2p_3/2 peak located at 711 eV was ascribed to ferric iron.

Since the major contribution on the Fe 2p_3/2 envelope from a fresh iron nanoparticles sample corresponds to Fe(III) (see Fig. 3B), the particles should have a core of metallic iron surrounded by a thick outer shell of iron oxides/hydroxides. Probably this oxygenated iron layer was rapidly formed after the particles had been exposed to the atmosphere. The FWHM for the Fe^III 2p_3/2 peak in the XPS spectrum was large (4.0 eV) which also suggests multiple oxidized species of iron on the surface.

This interpretation is compatible with the XPS spectrum for the oxygen that presented great amounts of metal oxides (530.0 eV, 41%) and metal hydroxides (531.5 eV, 57%) species, as determined by deconvolution treatment on the O1s binding energy peak (data not shown). The O1s peak also indicates that water was practically absent on the surface, since any significant contribution (less than 2%) was found for the O1s peak at binding energies higher than 533 eV. Examining the ratio of peak areas Fe/OH and OH/O²⁻ it was established that the covering oxide layer may be composed of iron hydroxides or oxyhydroxides. This is an important observation because the mechanism for aqueous metal contamination adsorption is regulated by the presence of different oxygenated groups on the particle surface. A number of mechanisms have been proposed, depending also of the adsorption conditions such as pH and the kind of metal ion that is being fixed on the surface.^26–28 The binding energy difference between Fe 2p_3/2 and Fe 2p_1/2 was 13.6 eV, making the assignment of the principal iron specie as Fe^III in Fe₂O₃ and FeOOH.²⁸

In the XPS spectrum, metallic iron was only identified in fresh samples (see Fig. 3B). After approximately 12 months aging, little or any presence of the zero valence iron could be determined (see Fig. 3C).

The arrows in the spectra indicate the Fe 2p satellites contribution, which are characteristic of Fe(III) state. The spectrum also suggests small amounts of Fe(II) for freshly prepared nanoparticles, since a small shoulder can be seen in Fig. 3B at 729 eV. This is because iron reacts not only with the contaminants, which are generally present at low concentrations, but also with dissolved oxygen and with water as shown below.²⁹

2Fe⁰(s) + O₂(g) + 4H⁺ → 2Fe²⁺(aq) + 2H₂O(l) (2)

Fe⁰(s) + 2H₂O → Fe²⁺(aq) + H₂(g) + 2OH⁻(aq) (3)

Through a high resolution XPS, Cd 3d_3/2 and 3d_5/2 peaks can be observed (see Fig. 3D). After completing the deconvolution of the peaks, a Cd²⁺ signal can be detected, especially at 405 eV which is in accordance with the values reported for in the literature.¹³ The XPS results confirm the ICP data; the reduction of cadmium concentration on the test solutions is due to nZVI. Data suggest that Cd²⁺ is sequestrated within nZVI by adsorption or surface complex formation with no apparent reduction on the nZVI surface.

C.3. XAS analysis

XPS studies confirmed the presence of both cadmium and oxygen containing iron species on the nZVI surface. Due to the limitations of the surface sensitive XPS technique, XAS experiments were performed because they can provide a better understanding of the bulk composition of the various oxygen containing iron species.^30,31 The linear combination fit was done in the 7092 eV and 7142 eV range of the sample spectrum and the theoretical spectra. The fit was done with reference spectra of Fe⁰, Fe₂O₃, ferrihydrite, FeO(OH) and γ-FeO(OH). 25 different combinations of the 5 reference spectra were performed and the model with the least residual error was chosen.

Fig. 4 shows that a good fit was achieved with the reference spectra used and Table 1 summarizes the percent composition results. The predominant oxygen containing iron species in the Cd–nZVI sample were ferrihydrite and α-Fe₂O₃ which is consistent with the XPS analysis. This implies that ferrihydrite and α-Fe₂O₃ can be found both in the surface of the sample and in the bulk. The linear combination fitting detected the presence of metallic iron in the Cd–nZVI sample whereas the XPS shows almost no metallic iron signals. The apparent contradiction between XPS and XAS can be explained by the fact that the metallic iron contained in the samples is covered by and oxidized iron layer. The surface sensitive XPS technique cannot detect it but the bulk sensitive XAS technique can. This suggests that the structure of the Cd–nZVI may contain an Fe⁰ core surrounded by a shell made up of ferrihydrite and α-Fe₂O₃. This hypothesis is confirmed by HRTEM studies discussed later. The presence of α-Fe₂O₃ throughout the bulk is particularly interesting because studies show that CdFe₂O₄ is formed under similar conditions as those in this study when α-Fe₂O₃ is in the presence of Cd²⁺. In order to confirm the formation of CdFe₂O₄, XRD and HRTEM studies were performed.

Fig. 4 XANES region of the XAS measurements of the Cd–nZVI particles with the theoretical linear combination peak with reference spectra of Fe⁰, Fe₂O₃, ferrihydrite and FeO(OH).

Table 1 Composition of the Cd–nZVI sample derived from linear combination fitting of the XANES sample spectrum with reference spectra of Fe⁰, Fe₂O₃, ferrihydrite, FeO(OH) and γ-FeO(OH)

Iron species	Composition%
Fe^0	23%
α-Fe2O3	17%
Ferrihydrite	47%
FeO(OH)	13%
γ-FeO(OH)	0%

---

C.4. X-ray diffraction patterns

Three XRD patterns with broad peaks characteristic of small particles are shown in Fig. 5. In Fig. 5A, a nZVI sample pattern with two peaks at 44.9° and 65.1°, which indicate the presence of body-centered cubic crystal Fe⁰, is presented. Both peaks are in good agreement with those reported in the literature.^32–35 In the next pattern (Fig. 5B), nZVI were placed in water for 5 hours as a blank to compare the oxidation states without the presence of cadmium. The intensity of the characteristic peak for Fe⁰ near 44.9° decreases in this pattern. A possible explanation to this change is a strong interaction of the iron with the oxygen either from the atmosphere at the time of preparation or the hydroxides (OH–) group from the water. Preliminary results indicate that these iron nanoparticles can prevent oxidation under atmospheric conditions for extended periods.^18,21,22 Characteristic peaks of hematite (α-Fe₂O₃) were identified.³⁶ Xiaoqin Li et al. also proves the formation of Fe₂O₃ and FeOOH in a previous work.^37,30

Fig. 5 X-ray diffraction pattern of nZVI particles synthesized using the sodium borohydride reduction method. (A) Pure nZVI particles freshly synthesized; (B) nZVI particles exposed to nanopure water (3 g L⁻¹ of nZVI, contact time of 5 h) and (C) nZVI particles exposed to 6 ppm Cd²⁺ aqueous solution (3 g L⁻¹ of nZVI, contact time of 5 h).

In this work, we obtained similar peaks of α-Fe₂O₃ and FeOOH visible at 30.2°, 38.0°, 43.3°, 49.1°, 57.2° and 62.8°; and 27.1°, 35.6°, 46.9°, 52.9° and 60.6°, respectively.

In Fig. 5C, an XRD pattern of nZVI at 3 g L⁻¹ placed for 5 hours with 6 ppm of Cd²⁺ is presented. This pattern has several peaks that required a comprehensive analysis. Previous studies show that CdFe₂O₄ may occur in similar chemical environments as the one presented in this work. Fig. 5C due to the low amount of cadmium used during the preparation of the sample. A pattern analysis was completed and characteristic peaks were identified for Fe⁰: 44.7° and 65.0°; α-Fe₂O₃: 36.3°, 38.2°, 43.2°, 49.1° and 62.7°; FeOOH: 27.1°; and CdFe₂O₄: 30.1°, 35.6°, 46.9°, 52.6°, 57.1° and 60.6°.^38,21 Diffraction lines of Fe₂O₃ in the XRD pattern of nZVI exposed to Cd indicates that possible formation of CdFe₂O₄ is partial.³⁹ This analysis was done taking in consideration that possible CdFe₂O₄ formation can occur at different temperatures based on previous works.^39–41

An important observation is the nZVI peak in 44.7° has a higher intensity in Fig. 5C than in Fig. 5B. The change in intensity may be explained as a result of cadmium interacting with oxygen. This reaction prevents iron to have enough oxygen available to fully react with it, preserving metallic iron integrity. In general, intensity changes of all peaks are observed in both samples (Fig. 5B and C), suggesting different physical arranges after the cadmium exposure to nZVI. These results can be explained by a strong interaction between the Fe(III), O and Cd(II) detected in the XPS results, probably by the formation of cadmium ferrite.

The mean crystallite size (L) of nZVI samples can be calculated by using Scherrer's equation:⁴²

L = 0.9λ/(B_2θ cos(2θ)) (4)

where λ is the wavelength of the X-ray (1.54056 Å), 2θ is the angle at the maximum of the peak, and B_2θ is the width of the peak at half-height.

The pure iron standard had a Fe⁰ particle size of 7.7 nm, the Fe⁰–H₂O sample, where we observe Fe₂O₃, FeOOH and small quantities of Fe⁰, had a particle size of 16.6 nm. The samples with Fe⁰–Cd, where FeOOH, Fe₂O₃, Fe⁰ are clearly observed and possible formation of CdFe₂O₄ was identified, had a particle size of 12.4 nm. The observed decrease in the average particle size of the samples with the possible formation of CdFe₂O₄ with respect to samples of Fe₂O₃ and iron oxides is referred to the strong chemical affinity of Cd²⁺ to the tetrahedral A site and the metastable cation distribution in the nanoscale range of the ferrites particles.⁴⁶ These chemical characteristics can help explain that there is a dependence of particle size with the stoichiometry of the cations.

C.5. High resolution transmission electron microscopy

HRTEM images of synthesized iron nanoparticles exhibit spherical shapes and chain aggregates (see Fig. 6A and B) as shown by others.⁴³ These images show that the particles have a diameter range between 25 nm and 64 nm. The aggregates may be due to the magnetic properties of Fe⁰, which promotes iron linear orientation.^44–46

Fig. 6 Transmission electron microscopy images of nZVI particles after suspension in pure water. Experimental conditions: nZVI particles concentration of 3 g L⁻¹; pH adjusted to 2 with dilute solution of HCl, temperature at 25 °C, argon atmosphere; 5 hours; magnetic bar stirrer at 700 r min⁻¹.

The pure iron standard had an Fe⁰ particle size of 7.7 nm, the Fe⁰–H₂O sample, where we observe Fe₂O₃, FeOOH and small quantities of Fe⁰, had a particle size of 16.6 nm. The samples with Fe⁰–Cd, where FeOOH, Fe₂O₃, Fe⁰ are clearly observed and possible formation of CdFe₂O₄ is identified, had a particle size of 12.4 nm. The observed decrease in the average particle size of the samples with the possible formation of CdFe₂O₄ with respect to samples of Fe₂O₃ and iron oxides is referred to the strong chemical affinity of Cd²⁺ to the tetrahedral A site and the metastable cation distribution in the nanoscale range of the ferrites particles.⁴⁶ These chemical characteristics can help explain that there is a dependence of particle size with the stoichiometry of the cations.

The HRTEM images (see Fig. 6C and D) show a core–shell structure where the core is Fe and the shell may be iron oxides such as FeOOH, α-Fe₂O₃ (see Fig. 6B). The HRTEM images of the Cd²⁺ treated nZVI show nanofiber formation (see Fig. 7), while the HRTEM images of untreated nZVI samples do not present nanofibers, only clusters and nanoparticles.

Fig. 7 High resolution transmission electron microscopy of iron nanofibers formed during the cadmium ion remediation process. The contact time was 5 h, the nZVI concentration (C_Fe) was 3 g L⁻¹; and Cd(CH₃COO)₂·2H₂O concentration (C₀) was 6 ppm. Experimental conditions: pH = 2; 25 °C; argon atmosphere; magnetic bar stirrer at 700 r min⁻¹.

The presence of two nano-constituents in the core–shell structure may impart dual properties for contaminant removal: the metallic iron acts as the electron source and exerts a reducing character, while the oxide shell facilitates the sorption of contaminants via electrostatic interactions and forming complexes with the surface. At the same time, the oxide shell allows electron passage from the metal core.⁴⁷

A structural model of Cd-nZVI fibers is illustrated in Fig. 8. We proposed this conceptual model of Fe⁰(nZVI), FeOOH/α-Fe₂O₃ and CdFe₂O₄/α-Fe₂O₃/FeOOH nanofibers considering the results of XRD patterns, XPS, XAS and HRTEM images. Fig. 8A shows the Fe⁰ nanoparticles, Fig. 8B shows the Fe⁰–H₂O where we observe Fe₂O₃ and FeOOH surrounding small quantities of Fe⁰ and Fig. 7C shows Fe⁰–Cd where we observe CdFe₂O₄, FeOOH, Fe₂O₃ and Fe⁰. In the case of the formation of the structure of Fig. 8C, cadmium ions act as the limiting reagent, where the onset of the reaction of Cd, Fe⁰, Fe₂O₃ and FeOOH allowing the possible formation of cadmium ferrite. Henceforth, the structure changes detected by HRTEM are consistent with the presence of Cd ions in the sample as presented in XPS and XRD studies.

Fig. 8 Conceptual model of cadmium adsorption process on nZVI nanostructures with cadmium–iron oxides on the surface.

The oxyhydroxide iron (FeOOH) has a crystal structure containing tunnel-shaped cavities that run parallel to the c-axis. This sites are bound by double rows of fused octahedral in which cadmium ions probably reside.⁴¹

This is important for understanding that the presence of cadmium ions with FeO and other iron oxides (e.g. Fe(OH)₂) is one of the possible interaction that promotes the formation of cadmium ferrite, therefore we propose:

2Fe²⁺ + Cd²⁺ + 4OH⁻ → CdFe₂O₄ + 2H₂ (5)

The fiber formation as product of nZVI in the presence of cadmium ions possibly is due to the rearrangement of the nanomaterial structure as a consequence of the adsorption process. Fig. 8 shows a schematic of the possible surface reaction.

The interactions of Cd²⁺ and Fe³⁺ particularly, possibly promote the CdFe₂O₄ formation which is in accordance with what is reported in the literature.³⁹ Our results showed the fiber formation as previous studies in the literature with CdFe₂O₄ particles synthesized by coprecipitation method.⁴¹ These studies showed the aggregation of fine particles of CdFe₂O₄ having two kinds of shapes, fibrous and granular in HRTEM. Due to the fact that XPS and XAS analysis showed that we had the necessary precursors to synthesize CdFe₂O₄ in our system and because of the formation of the nanofibrous structures, we can conclude that our remediation process has CdFe₂O₄ as a product.

D. Conclusions

nZVI were synthesized using chemical reduction technique and exposed to a Cd–water environment. Accordingly to results obtained in the XPS, XRD and HRTEM techniques, ICP study confirm that cadmium has been removed from water and integrated to the iron oxide after the nZVI treatment. A comprehensive characterization confirmed the production of highly organized nanofibers composed of iron oxides and cadmium ferrite after adding nZVI contact with the cadmium solution. Cd ions were removed from the solution at ca. 90% in 5 h by a pH dependent process. Optimum operational pH was identified at a range between 2 and 4. As general overview of the results, hydroxides groups are important to assure the nZVI particles interaction with Cd ions, yet there is a compromise were pH higher than 4 induce a surface passivation that exacerbates when this pH increased. The unintended formation of CdFe₂O₄ nanofibers as a remediation product presents an opportunity to reuse the remediation products for applications pertaining to light harvesting.

Acknowledgements

The authors acknowledge the use of facilities of the Materials Characterization Center of the University of Puerto Rico. We would like to thank Dr Lisandro Cunci (UPR) and Josue Ortiz for the XRD and XPS analysis. This work had financial support of NASA-URC Grant no. NNX10AQ17A. Financial support of the NSF-NSEC Center for Hierarchical Manufacturing, Grant no. CHM-CMMI-0531171, is also gratefully acknowledged. The authors of this work would like to thank the NSF PREM Grant no. DMR-0934218, Title: Oxide and Metal Nanoparticles. The Interface between life sciences and physical sciences. The authors would also like to acknowledge The Welch Foundation Agency Project # AX-1615, “Controlling the Shape and Particles Using Wet Chemistry Methods and Its Application to Synthesis of Hollow Bimetallic Nanostructures” and to NSF for its support through Grants DMR-1103730, “Alloys at the Nanoscale: The Case of Nanoparticles Second Phase”. Use of the National Synchrotron Light Source, Brookhaven National Laboratory, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract no. DE-AC02-98CH10886.

Notes and references

1. H. K. Boparai, M. Joseph and D. M. O'Carroll, J. Hazard. Mater., 2011, 186, 458.
2. S. Clemens, Biochimie, 2006, 88, 1707.
3. J. Szkoda, M. Durkalec, R. Kolacz, S. Opalinski and J. Zmudzki, Med. Weter., 2012, 68, 689.
4. H. A. Castillo-Michel, N. Hernandez, A. Martinez-Martinez, J. G. Parsons, J. R. Peralta-Videa and J. L. Gardea-Torresdey, Plant Physiol. Biochem., 2009, 47, 608.
5. P. Babula, V. Adam, R. Opatrilova, J. Zehnalek, L. Havel and R. Kizek, Environ. Chem. Lett., 2008, 6, 189.
6. N. Vrinceanu, D. M. Motelica, M. Dumitru, E. Gament, A. Manea, V. Tanase and M. Preda, Sci. Pap. Univ. Pardubice, Ser. A, 2010, LIII, 73.
7. M. P. Benavides, S. M. Gallego and M. L. Tomaro, Braz. J. Plant Physiol., 2005, 17, 21.
8. L. Wan and H. Zhang, Plant Signaling Behav., 2012, 7, 1.
9. J. A. McElroy, M. M. Shafer, A. Trentham-Dietz, J. M. Hampton and P. A. Newcomb, J. Natl. Cancer Inst., 2006, 98, 869.
10. J. Huff, R. M. Lunn, M. P. Waalkes, L. Tomatis and P. F. Infante, Int. J. Occup. Environ. Health, 2007, 13, 202.
11. U.S. Environmental Protection Agency, National Center for Environmental Assessment, Office of Research and Development, Washington, DC, 1999, http://www.epa.gov/iris/subst/0141.htm.
12. R. A. Crane and T. B. Scott, J. Hazard. Mater., 2012, 211, 112.
13. H. K. Boparai, M. Joseph and D. M. O'Carroll, Environ. Sci. Pollut. Res., 2013, 20, 6210.
14. B. I. Kharisov, H. V. Rasika Dias, O. V. Kharissova, V. Manuel Jimenez-Perez, B. Olvera Perez and B. Munoz Flores, RSC Adv., 2012, 2, 9325.
15. A. S. Adeleye, A. A. Keller, R. J. Miller and H. S. Lenihan, J. Nanopart. Res., 2013, 15, 18.
16. H. J. Shipley, K. E. Engates and A. M. Guettner, J. Nanopart. Res., 2011, 13, 2387.
17. S. M. Ponder, J. G. Darab and T. E. Mallouk, Environ. Sci. Technol., 2000, 34, 2564.
18. Y. P. Sun, X. Q. Li, J. S. Cao, W. X. Zhang and H. P. Wang, Adv. Colloid Interface Sci., 2006, 120, 47.
19. E. J. Bishop, D. E. Fowler, J. M. Skluzacek, E. Seibel and T. E. Mallouk, Environ. Sci. Technol., 2010, 44, 9069.
20. B. Schrick, B. W. Hydutsky, J. L. Blough and T. E. Mallouk, Chem. Mater., 2004, 16, 2187.
21. K. N. Harish, H. S. Bhojya Naik, P. N. Prashanth Kumar, R. Vishwanath and G. S. Yashvanth Kumar, Arch. Appl. Sci. Res., 2013, 5, 42.
22. C. B. Wang and W. X. Zhang, Environ. Sci. Technol., 1997, 31, 2154.
23. S. Y. Chen, W. H. Chen and C. J. Shih, Water Sci. Technol., 2008, 58, 1947.
24. B. Ravel and M. Newville, J. Synchrotron Radiat., 2005, 12, 537.
25. Y. L. Zhang, Y. T. Li, C. M. Dai, X. F. Zhou and W. X. Zhang, Chem. Eng. J., 2014, 244, 218.
26. C. Uzum, T. Shahwan, A. E. Eroglu, I. Lieberwirth, T. B. Scott and K. R. Hallam, Chem. Eng. J., 2008, 144, 213.
27. P. Ghods, O. B. Isgor, J. R. Brown, F. Bensebaa and D. Kingston, Appl. Surf. Sci., 2011, 257, 4669.
28. C. Uzum, T. Shahwan, A. E. Eroglu, K. R. Hallam, T. B. Scott and I. Lieberwirth, Appl. Clay Sci., 2009, 43, 172.
29. S. M. Ponder, J. G. Darab, J. Bucher, D. Caulder, I. Craig, L. Davis, N. Edelstein, W. Lukens, H. Nitsche, L. F. Rao, D. K. Shuh and T. E. Mallouk, Chem. Mater., 2001, 13, 479.
30. W. B. Kim, S. H. Choi and J. S. Lee, J. Phys. Chem. B, 2000, 104, 8670.
31. A. J. Berry, H. S. C. O'Neill, K. D. Jayasuriya, S. J. Campbell and G. J. Foran, Am. Mineral., 2003, 88, 967.
32. G. H. Lee, S. H. Kim, B. J. Choi, S. H. Huh, Y. Chang, B. Kim, J. Park and S. J. Oh, J. Korean Phys. Soc., 2004, 45, 1019.
33. X. B. Cao, W. H. Wang, X. H. Zhang, L. Y. Li, Y. H. Cheng, H. Liu, S. C. Du and R. K. Zheng, Nanoscale Res. Lett., 2013, 8, 423.
34. H. B. Liu, T. H. Chen, D. Y. Chang, D. Chen, Y. Liu, H. P. He, P. Yuan and R. Frost, Mater. Chem. Phys., 2012, 133, 205.
35. M. F. R. Fouda, M. B. ElKholy, S. A. Mostafa, A. I. Hussien, M. A. Wahba and M. F. El-Shahat, Adv. Mater. Lett., 2013, 4, 347.
36. Z. H. Ban, Y. A. Barnakov, V. O. Golub and C. J. O'Connor, J. Mater. Chem., 2005, 15, 4660.
37. L. Li, M. H. Fan, R. C. Brown, J. H. Van Leeuwen, J. J. Wang, W. H. Wang, Y. H. Song and P. Y. Zhang, Crit. Rev. Environ. Sci. Technol., 2006, 36, 405.
38. M. Chakrabarti, D. Sanyal and A. Chakrabarti, J. Phys.: Condens. Matter, 2007, 19, 236210.
39. S. M. Ismail, S. Labib and S. S. Attallah, J. Ceram., 2013, 2013, 8.
40. M. Yehia, S. Labib and S. M. Ismail, Phys. B, 2014, 446, 49.
41. M. Yokoyama, T. Sato, E. Ohta and T. Sato, J. Appl. Phys., 1996, 80, 1015.
42. A. L. Patterson, Phys. Rev., 1939, 56, 978.
43. A. Liu and W.-x. Zhang, Analyst, 2014, 139, 4512–4518.
44. J. Feng and T. T. Lim, Chemosphere, 2007, 66, 1765.
45. A. T. Ngo and M. P. Pileni, J. Phys. Chem. B, 2001, 105, 53.
46. W. Liang, C. Dai, X. Zhou and Y. Zhang, PLoS One, 2014, 9, 1.
47. W. L. Yan, A. A. Herzing, C. J. Kiely and W. X. Zhang, J. Contam. Hydrol., 2010, 118, 96.

---