Combined exposures to lead, cadmium and arsenic are common in both occupational and environmental situations [7] [8] . Since all three elements are nephrotoxic, additive or synergistic manifestations of renal cell injury are frequently reported depending on dose or exposure level. Toxic effects on renal tubule cells from combined lead and arsenic exposures have been reported following both moderate and low level dosing of experimental animals in vivo [9] [10] .

The clinical manifestations of such arsenic exposures on the kidney are consistent across studies and may involve effects on the blood vasculature [11] [12] leading to axotemia and/or tubular toxicity, and in some studies renal carcinogenesis. The renal proximal tubules are a major metabolic component of the kidney, which is responsible for reabsorption of proteins, glucose, and electrolytes from the urinary filtrate [13] . Renal proximal convoluted tubule cells (PCTCs) have high energy requirements in the form of ATP in order to meet their metabolic functions for the kidney and for this reason contain many mitochondria [14] that are highly sensitive to arsenical inhibition of respiration [15] . The high metabolic activity of the renal PCTCs coupled with their high mitochondrial content combine to make these cells and this portion of the renal nephrons major targets for arsenical toxicity. According to data provided by Correa-Rotter et al., 2014, in Mesoamerican Nephropathy, tubules and kidney interstitium are the kidney structures predominantly affected [5] .

In view of the toxic effects of As to human health, it becomes important to remove it from the environment. Some of the most common technologies for arsenic removal include oxidation, coagulation/precipitation, adsorption, ion exchange, electrocoagulation, and membranes [16] [17] . Among these removal methods, adsorption has proven to be the best method because of its versatility and affordability [18] . Additionally, adsorbent materials can be regenerated and reused and in this way the amount of generated waste material is considerably reduced [19] . The affordability of adsorbents makes them ideal to implement in third-world countries [20] and in small municipalities seeking to comply with standards. Some of the most widely used adsorbents for arsenic removal include activated alumina (AA), natural zeolites [21] , and granular ferric oxide [22] . Among the most widely used nanoparticles, magnetic nanoparticles, mainly zero-valent iron (nZVI), magnetite (Fe₃O₄) and maghemite (γ-Fe₂O₃) nanoparticles, have sparked an immense interest in research for engineering applications for treatment of polluted water or subsurface environments [23] [24] . Adsorption process is also occurred on the iron oxides (Fe₃O₄@SiO₂ and γ-Fe₂O₃) surface, while Fe₃O₄ possess reducing power [25] .

In recent years, there has been an increasing use of engineered magnetic nanoparticles for remediation and water treatments, leading to elevated public concerns. To this end, it is necessary to enhance the understanding of how these magnetic nanoparticles react with contaminants and interact with the surrounding environment during applications.

The large contaminants removal capacity and fast reaction rate of the magnetic nanoparticles are the major advantages. The total surface area of particles with the same mass increases tremendously (can add up to a thousand times), when the particle size decreases to nano-scale [26] . Due to the high surface-area-to-volume ratio, thus increase of active sites for the reaction, the mass required for treatment processes can be less than the case using the micro-sized counterparts. Various studies showed that the removal capacity and reactivity of nanoparticles are highly size dependent. [27] - [29] reported that the removal capacity of Fe₃O₄ nanoparticles (8 nm) was about seven times higher than that of coarse-grained counterparts (50 mm).

Using magnetic nanoparticles for contaminant removal has proven to be a highly efficient technology. Desorption is desirable for controlled applications; but this can lead to a disaster of pollutant migration and spreading out if desorption occurs when the spent magnetic nanoparticles are released into the environment in an uncontrolled manner [25] .

One of the most important characteristics of magnetic nanoparticles, showing advantage over other nanoparticles, is their magnetism. Abundant evidence showed that magnetic properties of particles alter when particle size reaches nanoscale [30] . In particular, the magnetic property of Fe₃O₄ and γ-Fe₂O₃ changed from ferromagnetic to superparamagnetic [31] . This property offers an encouraging option, satisfying requirements of high accessibility as well as reusability. Most water or wastewater treatment systems require a settling, filtration or centrifuge process to separate solids like sludge. However, magnetic nanoparticles can be separated and recovered with an external magnetic field due to the intrinsic magnetic characteristic of the nanoparticles, aiding in prominent nanoparticles recovery without a filtration process.

When considering the applicability of magnetic nanoparticles for treatment technologies, another major consideration is the reusability of magnetic nanoparticles. To maintain reactivity, physical and chemical properties of the nanoparticle products, the synthesis and manufacturing process of nanoparticles require meticulous control of conditions. The treatment cost of applying nanoparticles is relatively high, compared with traditional treatments. For instance, the Price of 1 kg of γ-Fe₂O₃ nanoparticles is ranging from USD 180 to 380, depending on the particle size, purity and methods of synthesis [32] ; Reuse of nanoparticles would substantially reduce the treatment cost. Regeneration is a common to regain the removal capacity of adsorbents, whilst feasibility of regeneration of magnetic nanoparticles is closely related to their removal mechanism.

The rapid expansion of nanotechnology promises to have great benefits for society, yet there is increasing concern that human and environmental exposure to engineered nanomaterials may result in significant adverse effects. That is why the field of nanotoxicology-dealing with effects and potential risks of particulate structures <100 nm in size-has emerged, growing significantly over the past decade from long-standing foundations of well es- tablished knowledge on the toxicology of fibrous and non-fibrous particles and the interactions of viruses with cells.

Discoveries of nanoparticle-specific concepts of toxicology related to their small size and large specific surface area go back to the early parts of the past century, although the distinctive biological effects and kinetics of nanoparticles were not recognized until the last decade of the past century [33] .

Studies also illustrated that nanoparticles have the ability to enter organisms during ingestion or inhalation [34] and can translocate within the body to various organs and tissues [35] where the nanoparticles have the possibility to exert the reactivity being toxicology effects. Although some studies have also addressed the toxicological effects of nanoparticles on animal cells [36] [37] .

Adsorption of magnetic nanoparticles on the bacteria membrane surface, and the reactions internally in cytoplasm of bacteria were reported to be the causes of bactericidal effect [38] - [40] . However, the mechanism of toxicity is still unclear and biocompatibility varies depending on numerous parameters, such as nanoparticle size and shape, surface property, applied nanoparticle concentration, type of cell and nanomaterial.

The specific gene alteration and hierarchical clustering revealed that SPIONs-COOH altered genes associated with cell proliferative responses due to their reactive oxygen species (ROS) properties. It was also found that the cell type can have quite a significant role in the definition of suitable pathways for detoxification of NPs, which has deep implications for the safe and high yield design of NPs for biomedical applications and will require serious consideration in the future [41] .

The small size facilitates uptake into cells and transcytosis across epithelial and endothelial cells into the blood and lymph circulation to reach potentially sensitive target sites such as bone marrow, lymph nodes, spleen, and heart. Access to the central nervous system and ganglia via translocation along axons and dendrites of neurons has also been observed. The greater surface area per mass compared with larger-sized particles of the same chemistry renders nanoparticles more active biologically. This activity includes a potential for inflammatory and pro-oxidant, but also antioxidant, activity, which can explain early findings showing mixed results in terms of toxicity of nanoparticles, to environmentally relevant species [42] .

The research findings on the toxic effects of magnetic nanoparticles have generally focused on “clean” particles. However, during environmental applications of magnetic nanoparticles, various contaminants could interact with the particles and alter their physiochemical properties. Magnetic nanoparticles may act as contaminant carriers, if the “used” magnetic nanoparticles are internalized by cells. A synergistic toxic effect could be a consequence in this ternary system, cells-contaminants-magnetic nanoparticles, where the bioavailability of loaded magnetic nanoparticles and its affinity towards bacterial cells have yet to be studied. One should also note that the type of bacteria plays a critical role on bactericidal effects, since different bacteria have a different tolerance on oxidative stress and exposure to nanoparticles [43] .

For instance, autotrophic bacteria can utilize chemical energy which can probably withstand the high reactivity of nanoparticles [44] [45] . The strain of bacteria, gram positive or negative, also exhibits varying resistances to magnetic nanoparticles where electrostatic interactions may be taken into account [46] [47] . Other microbes like fungi have a more robust structure and are less sensitive to the environmental disturbance than bacteria [46] , these are also abundant in natural environments. In addition, the bactericidal ability of magnetic nanoparticles can be utilized in a beneficial way as a disinfecting and sedimentation agent [46] [48] .

The research data on the environmental effects and bioavailability of magnetic nanoparticles is still limited, especially on phytotoxicity, although the toxicological effects of several other types of nanoparticles, such as fullerenes, carbon nanotubes, titanium dioxide and silver, have been widely studied [49] - [52] . Since plants are essential constituents in ecosystems, and probably a pathway for magnetic nanoparticles transport and bioaccumulation along food chain, the recent findings on phytotoxicity of magnetic nanoparticles are discussed.

Various experimental methods including in vitro study with isolated plant cells, a whole plant, or seeds as well as various experimental setups such as water, soft agar medium, and nature soil have been applied [50] . The evidence on the phytotoxicity of nanoparticles is yet conclusive because it likely depends on the plant species, composition, concentration, size and surface properties of the nanoparticles, and experimental methods.

Although diverging results of the phytotoxicology of the magnetic nanoparticles were reported, a comparatively conclusive finding showed that the presence of soil or sand reduced the impacts of magnetic nanoparticles on plants. The magnetic nanoparticles tend to attach or deposit on soil or sand grains, leading to a reduction of availability of magnetic nanoparticles [53] [54] . Several recent research papers reported the uptake of magnetic nanoparticles from aqueous medium by different plant species, such as pumpkin [54] .

The magnetic nanoparticles can also lead to little growth of leaves and death of older leaves (grown before dosing of nanoparticles). On the contrary, It was also reported that the magnetic nanoparticles could be taken up and translocated in plants, without any apparent visual impact, and could even enhanced the plant growth at low concentration [55] .

Taken together with the toxicology effects on bacteria, surface modifiers can reduce degree of toxicity on both plants and bacteria, probably due to reduction of attachment, and thus bioavailability. Moreover, the attachment of magnetic nanoparticles to soil medium resulted in lower the toxic impacts, suggesting that magnetic nanoparticles at low concentration could be used for environmental engineering applications without pernicious effects on plants [55] .

The aim of the research was to develop a nature-friendly selective treatment, applying updated technology and at a reasonable cost, intended to be employed in the way out of medium flow wells.

A Fe₃O₄@SiO₂ composite was used as remover of arsenic species. This composite material was prepared using co-precipitation of Fe(II) and Fe(III) with NH₄OH(ac) onto SiO₂, using a method similar to the one proposed by Petcharoen and Sirivat [56] with some modifications. All processes were carried out under N₂(g) atmosphere at room temperature. In a typical procedure, SiO₂ was dispersed in a solution containing Fe(II) and Fe(III) cations under inert atmosphere. After that, Fe₃O₄ nanoparticles were deposited over the SiO₂ substrate using concentrated NH₄OH(ac) (equivalent to 28.0% w/w NH₃ in water) as precipitating agent. The resulting material was washed with water until neutral pH. The Fe₃O₄@SiO₂ composite obtained had a support-magnetite mass ratio of 2:1. A similar procedure developed without the SiO₂ was done to obtain single Fe₃O₄ nanoparticles. In the latter case, the single Fe₃O₄ nanoparticles were stabilized with carboxylic acid, dispersing the Fe₃O₄ nanoparticles in a solution of the carboxylic acid in methanol and then removing the excess of the acid by washing with methanol.

XRD data of single Fe₃O₄ nanoparticles and Fe₃O₄@SiO₂ were acquired with a PANalytical Empyrean diffractometer with Cokα1 radiation. Each pattern was obtained in the 2θ range of 15.0091˚ - 84.9881˚ with steps of 0.0130˚. The scan step times were 13.7700 s for the single Fe₃O₄ nanoparticles sample and 37.9950 s for the Fe₃O₄@SiO₂ composite. TEM micrographs were obtained for the single Fe₃O₄ nanoparticles using a JEOL JEM-2100 microscope with an accelerating voltage of 100 kV.

A scalable lab system using filtering down through porous materials Fe₃O₄@SiO₂ was implemented in order to study the performance of the removers using continuous flow. Filters with Fe₃O₄@SiO₂ composite material were produced, wrapped in semipermeable fabric allowing water flow, containing 1 g of composite material each filter. Several tests were needed to determine the appropriate number of filters per tested arsenic solution volume.

Then, activated carbon bed filters were assembled to the system, in order to reduce turbidity of filtered water coming through the composite material filters. The activated carbon bed filters were placed in the system at a ratio of 8:1 with respect to the composite material filters. Following, the water flows through a 0.45 µm filter, to finally come out through a magnetized tube that can remove all magnetic nanoparticles that may have been liberated.

A solution prepared in the laboratory with arsenic reagent grade with a concentration of (43.7 ± 2.1 µg/L) was used to carry out performance measurements of the filters mentioned above. Following this trial, the performance of the filtration system was tested using the water entering the treatment system at Bebedero, Cañas, Guanacaste in Costa Rica.

Total As in water samples was analyzed using Perkin Elmer AAnalyst-800 atomic absorption spectrometer with a FIAS-100 hydride generation system. All measurements were carried out using a method with a detection limit of 0.71 μg/L and a quantification limit of 2.38 μg/L. Alternatively, simple field test measurements were done using a test kit Lovibond^®, previously validated using the Atomic Absorption Spectroscopy analysis method.

Obtaining the smallest particle size possible is essential for a successful synthesis of nanoparticles. Once prepared for visualization through the electronic transmission microscope, the samples were analyzed to determine average size. The Fe₃O₄ magnetite nanoparticles obtained showed irregular shape with an average size of 15 nm, calculated by means of TEM. Figure 1 shows a TEM micrograph of bare particles.

Figure 2 shows the XRD measurements developed for the samples Fe₃O₄@SiO₂ [2(a)] and single Fe₃O₄ [2(b)]. The XRD pattern for Fe₃O₄@SiO₂ presents a broad signal between 17˚ and 34˚ indicating the presence of amorphous silica as well as another group of signals that clearly matches with the ones presented for single bare Fe₃O₄ nanoparticles, this evidence the actual incorporation of Fe₃O₄ onto the porous support. The presence of the magnetite gives to the materials important magnetic characteristics. This could be determined qualitatively in the samples using a magnet, one that led in all cases a migration of the particles in contact with the magnet through a glass.

In order to quantify the efficiency of the nanoparticles on the support material and to scale up the results to a field device, the determination of the efficiency of the composite material (Fe₃O₄@SiO₂) was carried out in a prototype laboratory consisting of a continuous flow system. Filter testing was performed starting with 25 filters per time, down to a minimum of 5 filters, concluding that this last number of filters of the composite material may have acted efficiently in the prototype. Work was carried out with continuous flow at a ratio of 0.015 L/min, tested with a laboratory solution of arsenic at 43.7 ± 2.1 mg/L. The resulting concentrations of arsenic in each

volume of sample filtered through the composite material were found to be lower than the detection limit of the atomic absorption equipment (pointed green line).The graphics presented on Figure 3 was generated.

It was found that the filters could treat an important amount of water with arsenic. In view of their high performance, it was decided to estimate the projected quantity of water that could be treated with the same filters. The yield volume of the composite material of the prototype was calculated, continuing the trend line. A straight line equation was used for calculation. Figure 4 shows the trend from 9 L of arsenic solution prepared in the laboratory and passed through the filters. The productivity of the filters was estimated in 50 L.

Once filter efficiency was determined under laboratory conditions, the filtration system was tested using raw untreated water from the Northern zone of the country, contaminated with arsenic at a concentration of 13.3 µg/L. This area, called Bebedero de Cañas, is characterized by having high pH of 8.1 and total hardness of 83 mg/L, which is considered as medium hardness. It is important to determine the performance of the nano-parti- cles under such conditions, since a previous treatment may be necessary in order to pass water through the filters and these accomplishing the task of selective arsenic absorption. Performance of the filters is shown in Figure 5.

As can be observed, while the water presents high pH and high hardness, there is interaction with the nano- particle charges of the composite material, which causes that after being adsorbed, the arsenic is released, which can be observed in the ascending behaviour of the graphics. In addition, cleaning filters play an important role in arsenic retention by the system; however, the filters become rapidly saturated with interaction of the charges, as shown in Figure 5.

For the above reasons, three tests were performed: test 1-water pH in Bebedero was changed from 8.1 to 5.6; test 2-pH was modified and the sample was put through a resin for cationic exchange; and test 3-the sample received the resin treatment to reduce hardness only. All this was performed to determine the influence of high water pH, high water hardness or both, on the efficiency and arsenic removal capability of the filtration system. Figure 6 shows the behaviour of the treatments.

As shown in Figure 6, only by lowering the initial pH of the water entering the system to 5.6, the required arsenic removal efficiency is attained.

High performance material was obtained, with average nanoparticle sizes of 15 nm; proven high efficiency; low cost and easily reproducible, because of the type of materials used. Appropriate material characterization was achieved; its magnetic nature was identified and nanoparticles were collected in the magnetized tube at the end, thus avoiding nanoparticle release into the environment in order to respect basic concepts of nanotoxicology.

By depositing nanoparticles on a support of porous silica (Fe₃O₄@SiO₂), an important stability of this composite material is obtained, thus increasing the yield of the reaction of adsorption of arsenic in the treated water.

Added to the above advantages is the possibility of recovery of the material, which allows substantial reduction of the overall cost of treatment. Regeneration of the composite material and capacity of adsorbent removal is made possible when using absorption-based treatments.

Performance of the composite material using arsenic solutions, in both laboratory solutions and in water for use in populations currently affected was validated by means of the filtration system assembled as a prototype laboratory. Although the maximum amount of admissible arsenic in water is surpassed by only 3 µg/L in the population under study, the cost of treatment is very high.

With the proposed system, the number of filters used can be increased, depending on the amount of arsenic to be treated in each particular case, and the volume of water in each system, and its parts may be changed independently, depending on the specific wear.

It was found that the impact of the pH is very high on the performance and water treatment under this system. According to the specific nature of each type of water to be treated it is essential?as is the case of the Bebedero System in Cañas?to lower the pH from 8.1 to 5.6. The results showed that when passing through the cleaning filters, the pH increases again.

The addition to the filtration system of an end tube to collect magnetized nanoparticles makes this kind of treatment an environmentally friendly option because it prevents the intrinsic nanotoxicity of these particles. Because of their small size, nanoparticles can enter directly into the cell, causing undetermined damage.