Permeable Reactive Barrier of Coated Sand by Iron Oxide for Treatment of Groundwater Contaminated with Cadmium and Copper Ions

In this research, coated sand iron-oxide (CSIO) has been used as a permeable reactive barrier (PRB) for removal of cadmium and copper ions from the contaminated groundwater. The prepared material has been manufactured by precipitation of nano-particles based this oxide type on the sand surfaces by impregnation process. Therefore, this technique can be considered the main objective of the present study. The description of sorption data for sorbate-sorbent under consideration by the Langmuir isotherm model was more valuable than the Freundlich model. The maximum adsorption capacity of CSIO reaches 1.9181 and 7.6425 mg/g for cadmium and copper respectively. COMSOL Multiphysics Version release 3.5 has a good ability in the simulation and prediction of the cadmium and copper transport through one-dimensional CSIO-PRB. The outcome of this investigation prove that the manufactured CSIO has significant capability in the delay of contaminants the migration through barrier packed with this material. The root means squared errors between predicted and measured data were not exceeded the 0.121; so, this means that there a good agreement between these data.


Introduction
Ground-and surface-water contamination by hazardous compounds considers an urgent issue because it has negative impacts on public health and environment. Heavy metals are classified as one of hazardous compounds because they are non-biodegradable and tend to accumulate in living organisms, causing various biological disorders. Both cadmium and copper ions are presented in the effluents of several industries. Cadmium is one of the dangerous pollutants, originating from metal plating, metallurgical alloying, mining, ceramics, and other industrial operations [5,11]. Copper mining and smelting, brass manufacture, use of Cu-based agri-chemicals and electroplating industries are the major sources for pollution with this element [6,19]. The reclamation of the contaminated groundwater quality for standards specified by USEPA and WHO within acceptable time framework can be considered a big task because it is difficult and more expensive. A Pump-and-treat method is a historical technique that utilized for remediating the groundwater contaminated with different types of elements and compounds by extraction the polluted water to the ground surface and, then, treated it by conventional methods like adsorption, flotation, precipitation, and others. The resulted water from the treatment process can be disposed to water bodies, sewer system or reinjected to the aquifer. Unfortunately, the collected data related to this technique certified that this system is not effective, costly and not able to reclamation the contaminated water with acceptable periods of time. So, the permeable reactive barrier (PRB) technique is introduced as an alternative method for in-situ treatment of the contaminated groundwater. This technique has many advantages;1) passive (not need any exterior source of energy), 2) underground (not effect on the activities on the ground surface), safe (no contact between workers and contaminants) and others. The PRB is based on the idea of constructing a trench filled with the reactive material perpendicular on the direction of the flow where water will pass the barrier and contaminants can be entrapped within the barrier by physicochemical and biological forces [5,19].
There are a significant number of studies that directed towards the use of different sorbents in the treatment of aqueous solutions polluted with heavy metals such as activated carbon, zeolite, waste foundry sand, sewage sludge, cement kiln dust, and others [9,17]. Precipitates of iron oxides have been used as promising sorbents for organic compounds metal elements [4,13]. The particles of iron oxide are fine powders and their separation from aqueous solution can be represented as a serious obstacle in the usage of this material for the treatment process. Several previous studies are suggested to achieve plantation for the iron oxide on the immobilized solid supporting bed like sand, silica, cement, granular activated carbon, biomass and others [3,14]. Accordingly, the importance of this work is manufacturing of iron-oxide nanoparticles-immobilized sand and use it as PRBs for capturing the elements of cadmium and copper from contaminated water based on the set of the experiments and theoretical simulations.

Materials
Quartz sand with initial porosity of 36% was brought from a local market and washed carefully with distilled water before the use. The Cd(NO 4 ) 2 .4H 2 O and Cu(NO 4 ) 2 .3H 2 O (manufactured by HIMEDIA, India) were dissolved in distilled water to prepare the aqueous solutions contaminated with 1000 mg/L of cadmium and copper ions respectively. This solution was used to prepare the required concentration of used metals; however, the pH of the solution was changed by the addition of 0.1 M of hydrochloric acid or sodium hydroxide. The heavy metals concentrations were measured using atomic absorption spectrophotometer (AAS, AA-7000 Shimadzu, Japan).

Preparation of Coated Sand and Equilibrium Experiments
To prepare the sand coated with nano-particles of iron oxide (CSIO), the method of modified precipitation was applied [15]. The amount of 1 g sand was added to 50 mL of a solution containing 2 g Fe(NO 3 ) 2 . To ensure the occurrence of precipitation, the NaOH was added to raise the pH of the mixture. This mixture must be agitated for the duration of three hours; then, it is dried for four hours at 105 • C and thereafter the unattached iron oxide can be removed by successive washing until the pH of the wash water stabilized within the range of (7-8). The coated solids must put in the oven for drying for twenty-four hours at 105 • C and stored until usage in the sorption tests. The coating process was assessed under the effects of solution pH and sand/ferric nitrate ratios based on the achieved maximum removal efficiency of cadmium ions. The required concentration of the solutions contaminated with metal ions was obtained by dilution of stock solution. The removal efficiency was calculated by applying the initial (C o ) and final (C e ) concentration before and after the sorption process respectively as follows [7,10]: The isotherm data were obtained by taking 50 mL with different concentrations (5-100) mg/L of Cd(II) and Cu(II) solutions and distributed into different conical flasks at pH =5. The 1 g of coated sand by iron oxide (CSIO) must distribute on the conical flasks which agitated by the shaker for three hours at room temperature to enhance the transport of metal ions from the liquid phase to solid phase until the reaching of equilibrium status. These sorption measurements are fitted with the Langmuir and Freundlich isotherm models that represented by Eqs. 2 and 3 respectively [1,7,8,16,18]: where was the amount of metal ions adsorbed at equilibrium (mg/g). The constants and 1/ are Freundlich constants which reflect the adsorption capacity and intensity respectively. Also, the b and are affinity constant (L/mg) and maximum adsorption capacity (mg/g) for Langmuir isotherm respectively.

Continuous Study
The diagram for column setup utilized for representation the contaminant transport is illustrated in Figure 1. This setup consists of three Perspex columns and each column has dimensions of height = 50 cm and diameter = 2.5 cm. Each column was supplied with ports P1, P2 and P3 at distances of 10, 20, and 30 cm as clear in this figure. The ports were assembled with stainless steel fittings and Viton stoppers can be for blockage them. By using the syringes, the water samples have been taken periodically from the column center. Each column was packed with 30 cm of CSIO and this bed must saturate with distilled water by feeding it from the bed bottom in the upward direction. Using a constant head tank with a flow-meter and four valves, the contaminated solution was introduced into the beds for the Column-1, -2 and -3. The flow rates of (5, 10, and 15 mL/min) with an initial concentration of (10, 20, and 30 mg/L) are tested in the present experimental program. The concentrations of Cd(II) and Cu(II) at locations are identical in ports P1, P2 and P3 and measured by AAS for a period not exceeded 13 days. The measurements are also included the monitoring of the accumulative aqueous solution volume with time and results proved that the flow rate values remained constant through each experiment.

Preparation of CSIO
The pH effect of aqueous solution associated with Fe(NO 3 ) 2 concentration and also the mass of sand on the removal efficiency for Cd(II) ions were evaluated as showen in Figure 2. This figure proved that the increase of Fe(NO 3 ) 2 solution pH from 5 to 12 can cause a significant increase in the removal efficiency by CSIO. The precipitation of ferric hydroxide on the solid particles surfaces was achieved by changing the pH to became > 3 (11). In addition, the removal efficiency be CSIO was approximately stabilized the pH 11; so, this value can be chosen for the coating of sand by iron oxide. Then, the mass of sand used in the coating process was changed within the range from 0.5 to 10 g added to 50 mL of solution contained on 2 g of Fe(NO 3 ) 2 . The removal efficiency of Cd(II) ions was increased with an increase in the sand dosage from 0.5 to 5 g (0.25 to 2.5 sand dosage /ferric nitrate). The washing after ending the process of preparation can be applied to remove the uncoated iron oxide. Therefore, the sand dosage of 5 g (2.5 sand dosage /ferric nitrate) was proposed for the achieving of the coating process.

Sorption Isotherms
Langmuir and Freundlich's models are fitted with sorption data for the interaction of cadmium and copper ions with CSIO for q e (mg/g) with C e (mg/L) as shown in Figure 3. The fitted constants with statistical measures have been calculated from nonlinear regression -Microsoft Excel (2016) [1,7,8,16,18] as listed in Table 1. This figure in conjugation with a determination coefficient (R 2 ) and the sum of squared errors (SSE) elucidated that the model of Langmuir is more representative of sorption data.

Breakthrough Curves
The breakthrough curves for the propagation of cadmium and copper fronts measured at certain locations along the CSIO packed columns are plotted in Figures 4 to 6 under the different values of initial concentration and flow rate. A significant decrease in the concentrations of contaminants can be recognized due to the change of depth from 10 to 30 cm for all magnitudes of initial concentration and flow rate. Remaining of the polluted water for the long duration within the bed may be the cause for this result; however, this was improved the sorption process. With time, the barrier functionality decreases significantly because of the decrease of retardation factor due to saturate of the bed with sorbed metals. For initial concentration at a certain location, plotted results signified that the migration of the metal front relates directly with the flow velocity; so, the higher flow rate means faster transport for contaminant and this accompanies with decreasing of time required for achieving the state of equilibrium. Also, the increase of initial metal concentration will enhance the saturation of the bed with this contaminant more quickly due to the increase of the mass transfer driving force. The shapes of breakthrough curves are to be sharper due to the increase of metal concentration [12]. To represent the experimental measurements mathematically, the advection -dispersion equation (Eq.4) can be solved numerically by the finite element method through the application of COMSOL Multiphysics 3.5a (2008). This equation plots the contaminant transport in the column packed with CSIO porous medium and it can be written as follows: where C is the contaminant concentration, D z is the dispersion coefficient, V z is the flow velocity. Because the sorption of solute onto CSIO is more represented by the Langmuir model, the retardation factor (R) is expressed as follows: R = 1 ( where n is the porosity and ρ b is the bulk density. To complete the modeling process, the dispersivity must be calculated for the medium under consideration. Experimental measurements for the coefficient of longitudinal dispersion (D L ) versus velocity (V) through CSIO can be described by the following linear relationship [9]: D 9 = 6.49 V ( 0.5325 R 2 =0.996 … (7) By similarity of this equation with the following formula, dispersivity (α) can be evaluated: The outputs of solution for physical problem are illustrated in Figure 1

Conclusions
Coated sand by iron oxide (CSIO) that prepared by plantation of nano-sized-iron oxide onto immobilized-sand through simple impregnation process proved its ability in the remediation of contaminated water. Mainly, CSIO was applied in the permeable reactive barrier for removal of cadmium and copper ions from aqueous solutions in the batch and continuous tests. The sorption data were described in better form by the Langmuir isotherm model in comparison with the Freundlich model where the coefficient of determination (R 2 ) was more than 0.98. The maximum adsorption capacities for cadmium and copper onto CISO reached to 1.9181 and 7.6425 mg/g respectively. The results of COMSOL Multiphysics 3.5a certified that the barrier is efficient in the delay of contaminant migration due to a decrease of both flow rate and inlet concentration as well as the increase of bed depth. However, a good agreement between the predicted and experimental results was observed with root mean squared errors not exceeded 0.121.