Development of iron-based nanoparticles for Cr ( VI ) removal from drinking water

A great deal of research over recent decades has been motivated by the requirement to lower the concentration of chromium in drinking water. This study has been conducted to determine the feasibility of iron-based nanoparticles for chromium removal from contaminated water. Single Fe, Fe3O4 and binary Fe/Fe3O4 nanoparticles were grown at the 45-80 nm size range using the solar physical vapor deposition technique and tested as potential hexavalent chromium removing agents from aqueous solutions. Due to their higher electron donation ability compared to the Fe3O4 ones, single Fe nanoparticles exhibited the highest Cr(VI) removal capacity of more than 3 g/mg while maintaining a residual concentration 50 g/L, equal to the regulation limit for drinking water. In combination to their facile and fast magnetic separation, the applicability of the studied particles in water treatment facilities should be considered.


Introduction
Among the uses of magnetic nanoparticles in a variety of application fields, their incorporation in traditional environmental and health safety treatment methods is gradually earning much interest [1].Their ability to sufficiently remove heavy metals or other pollutants from drinking or waste water accompanied by the possibility of magnetically controlling their dispersion may introduce additional opportunities for increasing the yield and reducing the cost of such processes.
However, depending on the contaminant, its concentration levels and the relevant regulation, the required physical properties of nanoparticles may significantly vary.For instance, most of the reported studies for the removal of arsenic, lead, chromate and other heavy metals appear very efficient in terms of percentage reduction at relatively high concentrations [2] but only few of them provide reasonable results complying with the regulations for drinking water safety when tested at reliable concentrations met in natural water [3].Therefore, in order to decide and prepare the proper magnetic nanoparticle system for each case, specialized approaches should be designated and carried out considering parameters like the mechanism of contaminant capture, their affinity, the kinetic rate of the process and the effect of interfering factors usually existing during application.
The presence of chromate ions, also referred to as hexavalent chromium Cr(VI), in groundwater has only recently been discussed as an important issue related to drinking water in many regions worldwide due to the indications for harmful effects on human health upon regular consumption [4].Since the existing tolerance limit in potable water corresponding to total chromium (50 g/L) [5] practically underestimates the health risks, the establishment of a separate and very strict limit for the Cr(VI) form is anticipated soon.Thus, the development of optimized low-cost methods for its treatment is essential.In practice, Cr(VI) is usually removed from water either by chemical coagulation/filtration, ion exchange, reverse osmosis or adsorption [6].The experience from other contaminants (e.g.arsenic) indicates that the adsorption process, providing high efficiency and environmentally safe disposal, is the simplest and one of the cheapest methods.Moreover, the use of consumable adsorbents in the case of Cr(VI) removal can be further facilitated by the incorporation of a Cr(VI) to Cr(III) reduction stage.
Considering this, the increased specific surface area and surface reactivity of nanosized particles favor the enhancement of chromate reduction.In this work, zerovalent iron (ZVI) and magnetite (Fe 3 O 4 ) nanoparticles, were prepared by solar physical vapor deposition (Solar PVD) and tested as potential Cr(VI) adsorbents capable of reducing Cr(VI) and subsequently immobilize the resulting Cr(III) within their surface.

Preparation of nanoparticles
The studied Fe, Fe 3 O 4 and Fe/Fe 3 O 4 nanoparticles were synthesized by using the solar physical vapor deposition technique under inert argon atmosphere [7].In this process, the target material is placed in the focus of a 2 kW solar concentrator and the fume produced by evaporation is condensed on a cold finger or trapped on a nanoporous ceramic filter (Figure 1).The nanoparticles are then collected by brushing.The arrangement is constituted of a "heliostat" which tracks the sun and reflects the radiation on a parabolic concentrator mirror (diameter 2 m).The particles size and the growth rate can be controlled the gas pressure and the power supplied which is controlled by a flag located between the concentrator and the heliostat.The targets used for nanoparticles production were cold-pressed pellets prepared by Fe and/or Fe 3 O 4 powders purchased from Alfa Aesar (>99 %).In the studied samples, the chamber's dynamic pressure was set to 80 Torr by introducing an argon flow.

Characterization
The crystal structure of the nanoparticles was determined using transmission electron microscopy (TEM) and X-ray diffraction (XRD).For the TEM analysis, a drop of the colloidal solution in water was deposited onto a thin carbon film supported by a 400 mesh copper grid.
Measurements were carried out with a JEOL 100Cx TEM microscope working at an acceleration voltage of 100 kV.XRD patterns were taken with a Rigaku Ultima+ powder diffractometer using Cu-K  radiation.Magnetic characterization of the nanoparticles by vibrating sample magnetometry (VSM) provided the necessary information for the possibility of their magnetic recovery from an aqueous dispersion.Hysteresis loops at room temperature were recorded using an Oxford Instruments 1.2H/CF/HT VSM.In addition, magnetic separation experiments were performed by counting the rate of particles recovery from a suspension when applying a 15 T/m external field by means of permanent magnets in the sides of a 20 mL test tube.Chemical analysis of the nanoparticles was performed via graphite furnace atomic absorption spectrophotometry (Perkin Elmer Analyst 800) after diluting a weighted quantity in HCl.Similarly, residual iron concentrations dissolved after chromate removal were measured.

Chromate removal
The applicability of the nanoparticles as Cr(VI) adsorbents was evaluated by batch adsorption tests.Experiments were held by adding a quantity of nanoparticles (100-200 mg/L) in a volume of Cr(VI)containing distilled water (250-1000 g Cr(VI)/L) and shaking for a sufficient period of time (24 h) at room temperature.The equilibrium pH was adjusted to 7. Determination of the residual chromate was performed by the diphenylcarbazide spectrophotometric method while total chromium concentration was comparatively measured by graphite furnace atomic absorption spectrophotometry.The plotted isotherms represent the variation of Cr(VI) uptake per particles mass as a function of residual Cr(VI) fitted by a Langmuir function.

Morphology and structure
The identification of the samples' crystal structure through XRD mostly indicated the preservation of the 08007-p.2target's composition during evaporation/condensation whereas there is evidence for the surface oxidation of Fe (to FeO and Fe 3 O 4 ) after exposure to the atmosphere (Figure 2).On the other hand, single magnetite nanoparticles are found to be very stable.The quantitative Rietveld analysis of the diagrams estimated the composition for single Fe nanoparticles 80 %wt Fe, 5 %wt FeO and 15 %wt Fe 3 O 4 while the co-evaporation of Fe and Fe O 4 resulted in a mixture of nanoparticles with a total composition of 20 %wt Fe, 25 %wt FeO and 55 %wt Fe 3 O 4 .
The TEM observations for the samples are summarized in the images of Figure 3. Using Solar PVD for the growth of iron-based nanoparticles, a truncatedoctahedral shape was always favored.It is important to note that single Fe 3 O 4 particles formed long linear chains which indicate the effect of dipole-dipole magnetic interactions without serious agglomeration despite their relatively large diameter and the absence of surfactants.However, the mean diameter was different for Fe and Fe 3 O 4 nanoparticles: iron nanoparticles were grown at an average size of 45 nm with a standard deviation around 15 % while Fe 3 O 4 nanoparticles showed a diameter of 80 nm with a standard deviation around 10 %.Similar sizes for the two kinds of particles were met in the sample consisting of Fe/Fe 3 O 4 particles being prepared by the coevaporation of the two phases.The fraction of smaller Fe particles appears more susceptible to complete (Fe 3 O 4 ) or partial (FeO) oxidation and thus should be responsible for the decrease of total metal iron content suggested by XRD analysis.

Magnetism
The response of the magnetic nanoparticles under a constant field is the major motivation for their application in water treatment.Therefore, the magnetic characterization may provide an estimation for the potential success of magnetophoresis.Figure 4   According to saturation magnetization values and the geometrical characteristics of the nanoparticles, the minimum particle size that can be removed by the applied field of 15 T/m could be estimated by setting the applied magnetic force equal to the Brownian forces in the dispersion [8].Such calculation assumes the absence of interactions between the nanoparticles.The solution gives e.g. a critical diameter of 140 nm for the sample mainly consisting of zero valent iron, indicating that separation of the 45 nm particles is not feasible.However, the preliminary experiments using dispersion of 100 mg/L, showed not only the complete but also the fast separation in all the samples.In particular, the time of observed separation for the Fe, Fe/Fe 3 O 4 and Fe 3 O 4 nanoparticles was 21±3, 26±2 and 25±2 s.To clarify this divergence, the presence of dipole-dipole interactions incorporating a reversible aggregation-assisted separation should be considered.Such mechanism occurs when the Bjerrum length, the distance between particles with parallel dipoles at which the attractive magnetic energy becomes equal to the thermal energy, is larger than the particle diameter [9].In the studied samples not only this condition is satisfied but the relatively large particle size and the increased concentration contribute in the acceleration of separation.The measured separation rates were found in agreement to the prediction of complete separation time t S by the empirical equation for the reversible aggregation mechanism [9]: where t 0 is a time constant set at 66 s, D is the particle' s diameter, M s the saturation magnetization, k B the Boltzmann's constant,  0 the magnetic permeability, T the temperature,  particle density and C the dispersion concentration.More specifically, the separation time was calculated in the range 25-30 s for the three samples.

Cr(VI) uptake
The efficiency of the samples in the removal of Cr(VI) is described by the quantity of Cr(VI) captured by a mass of the nanoparticles for a specific residual concentration.The chromate removal isotherms measured at pH 7 and concentrations below 1 mg/L are also indicative for the applicability under reliable conditions, showing the potential of the nanoparticles to reduce Cr(VI) below the regulation limit.As shown in Figure 5, zero-valent iron nanoparticles are much more efficient compared to the magnetite ones, as expected according to the electron donation ability per iron atom even in the case of FeO formation in the surface.This is an indication that Cr(VI) removal mechanism is actually based on the reduction reaction which produces the insoluble Cr(III) state:

Conclusions
Magnetic iron-based nanoparticles produced by Solar PVD were found adequate to remove hexavalent chromium below the current regulation limit for drinking water.Since reduction to the trivalent state is the occurring mechanism, the efficiency is defined by the composition of nanoparticles and the oxidation state of iron atoms setting ZVI nanoparticles the most advantageous for the treatment of chromium contaminated water.In addition, the relative large size (45-80 nm) and the increased magnetization of the studied samples enable the easy application of a magnetic separation procedure.

Fig. 1 .
Fig. 1.Solar physical vapor deposition apparatus for the growth of nanoparticles.
presents the hysteresis loops for the single and binary Fe and Fe 3 O 4 nanoparticles.Samples showed relatively high saturation magnetization values approaching reference values for the corresponding bulk materials.The highest value was measured for the ZVI nanoparticles (200 emu/g) whereas Fe/Fe 3 O 4 nanoparticles reached 86 emu/g and Fe 3 O 4 90 emu/g.In all the cases magnetization mostly scales with the Fe and Fe 3 O 4 content calculated by XRD refinement.

Fig. 5 .
Fig. 5. Chromate removal isotherms at pH 7 for the Fe, Fe/Fe 3 O 4 and Fe 3 O 4 nanoparticles.Dotted lines correspond to the Langmuir model's fitting of the points.
chromate onto Fe 0 and Fe 3 O 4 particles limits further extend of removal activity.On the other hand, the partial inhibition of passivation may be the reason for the improved efficiency of the binary Fe/Fe 3 O 4 nanoparticle system compared to the single Fe 3 O 4 nanoparticles, since that cannot be explained by the relatively low iron's content.A possible electron interaction in a supposed nanoscale galvanic element between Fe 3 O 4 and Fe actually increases the lifetime of Fe 3 O 4 nanoparticles in reducing chromate oxy-ions.Importantly, dissolved iron concentration in the treated water was found always below 0.1 mg/L.