Abstract
This work focused on the synthesis of iron oxide nanoparticles by the coprecipitation method with three basic solutions, namely, NH4OH, KOH, and NaOH. The synthesized iron oxides were characterized by various techniques such as XRD, MET, BET, and a SQUID magnetometer. The results showed nanosized particles of 13.2, 9.17, and 8.42 nm and different phases associated to maghemite and maghemite/hematite. The surface areas were 113, 94, and 84 m2/g and the magnetization strength were 58, 61, and 75 emu/g to iron oxides synthesized with NaOH, KOH, and NH4OH, respectively. The magnetic iron oxides obtained using NaOH were more efficient in the removal of lead and arsenic by adsorption than iron oxides obtained with KOH and NH4OH. However, the magnetic strength decreases using NaOH and KOH. The highest adsorption capacities attained for lead and arsenic removal were 16.6 and 14 mg/g, respectively, using NaOH-based iron oxides.
1. Introduction
Heavy metals constitute an important source of water pollution. In developing countries, water contaminated with heavy metals is caused by the excessive use of pesticides, chemical fertilizers in agriculture, and rapid industrialization. According to the World Health Organization (WHO), the toxicity of heavy metal pollution such as lead (Pb) and arsenic (As) affects not only human health but also our ecosystem [1]. Many countries have restricted levels of As and Pb in the water supply due to their harmful effects [1, 2]. The contaminated water needs to be treated and thus many researchers are interested in developing effective methods for contaminated water treatment. Several conventional physicochemical methods [3], including membrane filtration [4], chemical oxidation [5], chemical precipitation [6], ion exchange [7], coagulation [8], flocculation [9], and adsorption [10, 11], were applied for the treatment of heavy metals containing wastewater. Among these techniques, adsorption is considered as one of the most effective, fast, and simple operations for the efficient removal of heavy metals from aqueous solution. Recently, many researchers have focused on iron oxide magnetic nanoparticles as adsorbents because they exhibited high adsorption [12] and provided feasibility for separation from aqueous media by using low magnetic field gradients. It was shown that an ideal magnetic sorbent should have the following advantages: (i) strong magnetism to achieve fast magnetic separation; (ii) good dispersion, so as to improve the adsorption/desorption kinetics; (iii) large specific surface area, suitable porosity, (iv) good stability, and (v) environmentally friendly. The use of iron-based nanomaterials allows easy separation using a magnetic field, but also increases the adsorption capacity. Therefore, magnetic particles can be a very good option for the adsorption of various metal ions [3]. The goal of this work is to highlight the contribution of iron oxide nanomaterials particles sizes and porosity on the adsorption process by varying only the magnetization strength of iron oxide particles. The adsorption of metal ions by only magnetic iron oxide particles is not often well studied. Different magnetic iron oxides were synthesized and tested for the removal of arsenic and lead, which are highly toxic pollutants.
2. Material and Methods
2.1. Synthesis
Iron oxide nanoparticles were synthesized by the coprecipitation method using an aqueous solution of Fe2+ and Fe3+ with the stoichiometric ratio of 1 : 2. The synthesis was done according to the following operating procedure: a mixture of 50 ml of Fe2+ and Fe3+ previously prepared with adding in 50 ml of deionized water, 1.41 and 2.76 g of FeSO4·7H2O, and FeCl3·6H2O, respectively. One hundred and eighty milliliter of deionized water was added by stirring to finally obtain 300 ml. Different amounts corresponding to 4 M concentration solutions of ammonium hydroxide (NH4OH), sodium hydroxide (NaOH), and potassium hydroxide (KOH) were added dropwise while stirring. The black precipitates obtained between pH 10 and 12 were recovered, washed several times with deionized water until the pH was reduced to pH 7, and then dried at 70°C overnight. The samples were denoted, respectively, as IO-K, IO-Na, and IO-NH, where IO indicates iron oxide and K, Na, and NH indicate the basic solutions KOH, NaOH and NH4OH used, respectively to synthesize iron oxide nanoparticles.
2.2. Characterization
XRD was employed for analyzing the phase of the synthesized iron oxide nanomaterials. The XRD patterns were obtained using a Bruker D8 powder (XRD) instrument employing CuKα radiation (λ = 1.5418 Å). The iron oxide nanoparticles were further characterized using a transmission electron microscopy (TEM) instrument, using JOEL JEM-2100F model with a 200 keV. Nitrogen gas adsorption measurements were performed at −196°C by using a Micromeritics TriStar II plus instrument. Before adsorption measurements, all samples were outgassed at 200°C for at least 2 hr. The Brunauer–Emmett–Teller (BET) surface area was calculated from nitrogen adsorption isotherms in the relative pressure (p/p0) range of 0.05–0.20. The total pore volume was estimated from the amount adsorbed at a relative pressure of 0.98. The pore size distributions (PSDs) were calculated from the adsorption branches of the isotherms by using the Barrett–Joyner–Halenda (BJH) method. For magnetic measurements, a vibrating sample magnetometer (VSM, Lakeshore, model 7400 series) was used at room temperature. The magnetic response of the iron oxide nanoparticles was also studied using a squid magnetometer (MPMS, Evercool, 7T Squid instrument supplied by Quantum Design). The neutral charge of the iron oxide synthesized surfaces was verified by point of zero charge (PZC). For this method, 0.1 g of iron oxide nanoparticles was added in 20 ml of deionized water under nine different initial pH conditions (from 2 to 11) and after 48 hr at 25°C, the pH was measured again. The pH-PZC corresponds to the range in which the final pHf remains more constant at the different initial pHi applied.
2.3. Heavy Metals Adsorption Experiments
To measure the removal of heavy metals by the synthesized iron oxide nanoparticles, batch adsorption experiments were conducted by mixing 0.1 g of iron oxide with 25 ml of lead solutions and 50 ml of arsenic solutions at room temperature. Metal solutions of 5–50 mg/L were prepared by dissolving Pb(NO3)2 and AsNaO2 in deionized water and adjusted the pH to 6.7 and 8 corresponding to the optimum pH (maximal adsorption conditions of lead and arsenic) for lead and arsenic metal ions, respectively. The iron oxide nanoparticles were added in each solutions, shaken in a mechanical shaker at 150 rpm for 24 hr to achieve equilibrium, and then filtered through a 0.45 μm syringe filter. The residual concentrations of lead or arsenic ions in the aqueous filtrated were measured using atomic absorption spectrophotometer (spectr AA20) to assess the heavy metal adsorption capacity. The amount of heavy metals adsorbed per unit weight of iron oxide nanoparticles (Qe, mg/g) is calculated using the following equation:where C0 and Ce represent the initial and equilibrium concentrations (mg/L) of the metal ions (Pb2+ or As3+) in the solution, V is the aqueous solution volume (L), and m is the mass (g) of the iron oxide materials. The adsorption isotherms of Pb2+ and As3+ ions on iron oxide nanomaterials were fitted to linear Langmuir and Freundlich models, which are the most frequently used models for describing sorption isotherms. The Langmuir isotherm model is defined as Equation (2) [13], while the Freundlich isotherm model is given in Equation (3) [14]:where Qe is the mass of metals adsorbed per mass of iron oxide (mg/g) at equilibrium, while Ce represents the equilibrium concentration (mg/L) of the metals. The Langmuir constants are a and b referred to the adsorption capacity and adsorption rate, respectively. The adsorption constants for the Freundlich model are k and n.
3. Results and Discussion
3.1. Characterization
XRD patterns of synthesized iron oxide nanoparticles are shown in Figure 1. IO-NH sample shows intense peaks attributed to maghemite at 2θ = 30.24°, 35.63°, 43.28°, 53.73°, 57.27°, and 62.92° corresponding to the (220), (311), (400), (422), (511), and (440) family of planes, respectively, in agreement with JCPDS No. 00-039-1346. The maghemite crystalline structure was also observed using NH4OH as basic solution [15]. The diffraction pattern corresponding to the samples IO-K and IO-Na exhibited same peaks associated to mixture of maghemite/hematite phase. The additional hematite phase is shown by the additional peaks at 2θ = 33.15, 40.86, 49.46, and 54.05 corresponding to the (104), (113), (024), and (116), respectively, identified in American Mineralogist Structure Database (AMSD) with the code 143 for hematite. The basic agents KOH and NaOH used for the synthesis of iron oxide showed the same crystalline phase comparatively to NH4OH. The choice of the basic solution influenced the nucleation and the pH of the medium. The mechanism of iron oxide formation involved the step of hydroxylation of Fe3+ and Fe2+ ions. The subsequent addition of the basic solution at room temperature (pH < 3) led quasi instantaneously to a highly hydrated phase called ferrihydrite [16, 17]. Its transformation proceeded via different pathways depending on the acidity of the medium and resulted to different phases. The addition of strong bases such as KOH and NaOH, comparatively to the weak base NH4OH, rapidly modified the acidity of the medium, thus the ferrihydrite was transformed into very small particles of hematite, followed by subsequent nucleation to produce maghemite structure. This transformation proceeded only by dehydration in situ and local rearrangement mechanism [16]. Hematite was therefore formed from ferrihydrite by an internal dehydration-rearrangement mechanism. The addition of the weak base NH4OH modified slowly the acidity of the medium and led to a progressive enlargement of the nucleation. The stoichiometric ratio of Fe2+/Fe3+ = 0.5 is known to produce magnetite [17]. The transformation of magnetite into maghemite is due to the oxidation of magnetite by the air. The morphology of the samples IO-K, IO-Na, and IO-NH synthesized was studied as well as the particle sizes distribution was determined, as shown in Figure 2. The samples contain aggregates of nanoparticles whose dimensions were determined with TEM images and statistical analysis. The particle sizes of these samples were well described by a log-normal distribution (Figure 2(d)–2(f)) and showed values around 13.2, 9.17, and 8.42 nm for IO-NH, IO-K, and IO-Na, respectively. The change of iron oxide particle sizes was observed in previous work [18].


(a)

(b)

(c)

(d)

(e)

(f)
Figure 3 shows the nitrogen adsorption isotherms at −196°C and pore sizes distribution of iron oxide nanoparticles obtained with different basics solutions. Figure 3(a) shows the volume of adsorbed N2 at a relative pressure (p/p0) close to 0.1 increases slightly for all samples. However, the N2 adsorption capacity at the relative pressure close to unity increases from 197 to 210 cm3 STP/g. All the samples showed isotherms type II according to the IUPAC classification [19]. This type of isotherm is generally characteristic of nonporous materials. Figure 3(b) shows the diameter pores size and revealed a range mesoporous sizes; smaller diameter (<10 nm) and large diameter (>10 nm). We attributed these to the intraparticles porosity. As we showed above, the strong bases KOH and NaOH induce small particle size leading to very small intra-particles porosity (d < 10 nm) then the subsequent enlargement of the nucleation leads to large size with larger intraparticles porosity. The use of the weak base NH4OH generated larger particles with wider intraparticles porosity, in addition to the contribution of the evaporation of ammonia. The surface areas obtained were 113, 94, and 84 m2/g for IO-Na, IO-K, and IO-NH samples, respectively, according to the size of particles obtained.

(a)

(b)
The magnetic properties of the synthesized iron oxide nanoparticles were evaluated using a vibrating sample magnetometer (VSM) to investigate the saturation magnetization of iron oxide nanoparticles. Figure 4 shows a magnetization curve and determines the relationship between magnetization strength and external magnetic field intensity. The measurements were carried out at room temperature. Figure 4 shows the hysteresis-free magnetization curve indicating that all the samples are superparamagnetic. The use of different basic solutions for the synthesis of magnetic iron oxides had no effect on superparamagnetic feature of the particles obtained or do not destroy the superparamagnetism. The stationary point of the magnetization curve is the saturation magnetization strength and Figure 4 reveals that the magnetization strength are 58, 61, and 75 emu/g for the samples IO-Na, IO-K, and IO-NH, respectively. We observed that for the small particle sizes, magnetization strength is low contrary to large particles. In fact, the increase of magnetization increased the aggregation of particles. This effect has been frequently reported. The basic solutions caused difference in the H–M plot because of the particles size generated. These results indicated that iron-based materials resulting from NH4OH as basic solution can be recovered easily with magnetic field than those from NaOH and KOH as basic solutions.

3.2. Heavy Metal Removal
Figure 5 shows the heavy metals removal capacities of lead and arsenic on the three iron oxides synthesized. The capacities were 11.2, 13.1, and 16.6 mg/g of lead on IO-NH, IO-K, and IO-Na, samples respectively while those of arsenic were 8, 13, and 14 mg/g, respectively. The sample IO-Na showed the highest removal capacity for both heavy metals even though the magnetization strength is the lowest. As shown in Figure 3, the surface area of IO-Na is high comparatively to IO-NH and IO-K samples. It was demonstrated that the removal of heavy metal depends also on the size of iron oxide particles [20]. The amount of heavy metal adsorption removal increased with decreasing particle size due to less aggregation and more sites exposed to adsorption [20]. We observed in TEM the smallest particles in IO-Na, less aggregation than IO-NH, IO-K samples in agreement with its high adsorption capacity. The low magnetization strength of IO-Na sample had no effect on the adsorption capacity. The adsorption process is not governed by the magnetic structure. The highest heavy metal adsorption capability of IO-Na sample can be attributed also to the small intraparticles porosity size observed (<10 nm) in IO-Na sample comparatively to IO-NH, IO-K. Adsorption is known to be favorable in very small pore size [21]. The addition effect is the electrostatic attraction between the iron oxide surface and the charge of metal ion. Figure 6 shows the PZC corresponding to the value of pH at which the charge detected on iron oxide particles surface becomes neutral. The PZC is helpful for the evaluation of charges on the surface [22]. PZC was found to be 5.9, 6.2, and 6.4 for IO-K, IO-Na, and IO-NH, respectively. At the optimal pH of this study (pH 6 and 8), all the iron oxides had negative charge (PZC < pH optimal), while lead (Pb) in cationic form and arsenic (As) in anionic form. The attraction of the opposite charges was an additional contribution to the sample intraparticle porosity to increase adsorption efficiency. These two contributions provided favorable conditions for heavy metal adsorption.


The difference in adsorption capacity of metal ions is due to the ions speciation. The negative surface charge of IO-Na generated a strong electrostatic attraction for cationic lead than for arsenic in anion form. Nevertheless, the iron oxide nanoparticles synthesized showed considerably higher heavy metal adsorption capability than the other types of iron oxide materials previously reported in the literature in Table 1.
The advantage of iron oxide nanoparticles is that they can be quickly recovered from the water using an external magnetic field even iron oxide nanoparticles easily undergo aggregation in aqueous solution systems which leads to decrease their specific surface area and adsorption capacity.
3.3. Isotherms
The heavy metals adsorption isotherms were simulated with the widely used Langmuir and Freundlich equations, and the corresponding parameters are summarized in Table 2, while the data are plotted in Figure 7. Adsorption isotherms of Pb (Figure 7(a)) and As (Figure 7(b)) on the sample IO-NH were better fitted with the Langmuir model, with a coefficient values of 0.97, 0.99 than the Freundlich model which showed values of 0.90 and 0.92. For the sample IO-K, Langmuir model was more suitable than the Freundlich model, although the difference in the coefficient values was very significant. In the case the sample IO-Na, the resulting correlation coefficients also showed that Langmuir model was better fitted than Freundlich model. Previous works have shown the Langmuir model as a suitable isotherm reflecting adsorption of heavy metal such Pb and As [30, 31].

(a)

(b)
Renewability and reusability were important factors for an adsorbent. Thus, NaOH solution was used for the regeneration of the used adsorbents. A near-complete desorption of heavy metals from iron oxide particles by NaOH solution was occurred immediately. The adsorption stability of the iron oxide nanoparticles was investigated for the reusability tests, as shown in Figure 8. It was observed that adsorption efficiency of Pb by regenerated iron oxide adsorbents decreased after four cycles (i.e., 90.5% to 74.1%, 92.4% to 70.8%, and 95.9% to 70.3%) for IO-NH, IO-K, and IO-Na, respectively. The adsorption efficiency of As by regenerated adsorbents also decreased after four cycles (i.e., 90.2 to 79.2, 90.3 to 77, and 91.2 to 75) for IO-NH, IO-K, and IO-Na, respectively. The regenerated adsorbents showed the excellent reusability and stability and the potential use to treatment of wastewater containing heavy metals.

(a)

(b)
4. Conclusion
Magnetic iron oxides were easily synthesized with different basic solution (NH4OH, NaOH, KOH), and their properties were investigated using various characterization techniques. All the iron oxides obtained were superparamagnetic and in nanosize. Iron oxide synthesized from NaOH exhibited the highest adsorption of lead and arsenic because of high surface area, intraparticles porosity, and very high electrostatic attraction. Iron oxide synthesized from NH4OH showed high magnetization strength but the magnetic properties had no effect on the adsorption process. The adsorption process was in agreement with Langmuir isotherm comparatively to Freundlich. The regenerated adsorbents showed the excellent reusability and stability. This work opens a way to the synthesis of new materials-based iron oxides from NaOH to increase the adsorption of micropollutants including heavy metals.
Data Availability
Characterization and analytical data supporting the findings of this study are available from the corresponding author upon request. Table and figure data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This work was carried out with the support from the University Felix Houphouet-Boigny University.