As(III) Removal by Dynamic Adsorption onto Amberlite XAD7 Functionalized with Crown Ether and Doped with Fe(III) Ions

MIHAELA CIOPEC1, ADINA NEGREA1, NARCIS DUTEANU1, CORNELIU MIRCEA DAVIDESCU1, IOSIF HULKA2, MARILENA MOTOC3, PETRU NEGREA1*, OANA GRAD2, DELIA BERCEANU VADUVA3 1University Politehnica Timisoara, Faculty of Industrial Chemistry and Environmental Engineering, 2 Victoriei Sq., 300006, Timisoara, Romania 2Research Institute for Renewable Energy of the Politehnica University of Timisoara, 138 Muzicescu Str., Timisoara, Romania 3Victor Babes University of Medicine and Pharmacy Timisoara, 2 Eftimie Murgu Sq., 300041, Timisoara, Romania

Arsenic is 20 th element present in the Earth's crust, in form of 245 mineral species, considering its distribution, predominantly ores containing sulphides, nickel, copper, cobalt, lead etc. [1].
Arsenic it is a toxic metalloid that is known to be a contaminant into groundwater's from many parts of world, like South-East Asia area [2,3], but also in the West of Romania [4,5]. Consumption of arsenic contaminated water leads to a series of health problems.
Acute toxicity of arsenic occurs due to accidental ingestion, consumption of contaminated waters or inhalation of arsenic compounds with suicidal intent. First symptoms of poisoning appear when a dose less than 5 mg of arsenic compounds is ingested [6], even if the arsenic forms are of different toxicity. Thus, for inorganic arsenic the highest toxicity form is the methylate (MMA and DMA) [7,8] and the lethal dose for an adult being considered between 1 and 3 mg·Kg -1 body weight [9]. Also, arsenic ingested in the inorganic form is adsorbed in gastrointestinal tract and turns into MMA and DMA forms [9][10][11][12].
Studies have demonstrated that arsenic removal can be realized through various technologies [13] such as: coagulation with salts [14], ion exchange, which is becoming more popular due to the simplicity and ease of operation and due to the fact that not result sludge. A particular technique used for arsenic removal is represented by the adsorption onto the iron oxide particles, knowing the As (III) affinity for iron ions [15,16].
The fact that, in general, the classical methods used for arsenic removal from aqueous solutions are expensive methods, but also lead at production of some secondary products, we aim to implement an environmentally friendly technology for arsenic removal from water. It is possible to * email: petru.negrea@upt.ro obtain that by using adsorption as removal technology, but in order to get higher efficiencies it is necessarily to use materials with advanced adsorbent properties. Such materials can be obtained by chemical synthesis [17,18], or by functionalization of organic solid supports (natural or synthetic polymer) or inorganic solid support (silica or florisil) with pendant groups (N, P or S) in order to improve the adsorption properties [19][20][21][22][23].
Based on the operation mode, adsorption can be generally classified into static adsorption and dynamic adsorption. Static adsorption, also called batch adsorption, occurs in a closed system containing a desired amount of adsorbent kept in contact with a certain volume of contaminated solution, while dynamic adsorption usually occurs in an open system where contaminated solution continuously passes through a column packed with adsorbent.
In some studies, materials with highly efficient adsorbent properties, specific to arsenic removal are presented. These materials point out significant adsorption capacities (> 100 mg As·g -1 of material), but such materials cannot be used to remove trace arsenic from underground waters (below 100µg·L -1 ) and also the residual concentration in these cases is is higher than10 µg·L -1 , which does not agree with WHO requirements for arsenic content, especially in drinking water (maximum 10µg·L -1 ) [18].
Crown ethers present a considerable potential for use as high selectivity reagents for metals due to their ability to form stable complexes with their ions [37][38][39], but also metal ions are able to penetrate the circular cavity of the crown ether forming rings with three-dimensional structure [40,41].
Aim of present study it is to test a new material obtained by functionalization of a commercial polymer-Amberlite XAD7 with Di-benzo 18-crown-6 (DB18C6) ether, doped with Fe (III) ions for arsenic removal from real water samples obtained from Western area of Romania. All adsorption experiments were carried out using adsorption in a fixed bed column.
Also, the influence of competing ions (nitrite, nitrate, ammonium, phosphate sulphates and bicarbonates) present in water sample onto the As(III) adsorption process was studied.

Experimental part Materials and methods
In order to establish the adsorption mechanism of As(III) ions presents in natural ground water using a dynamic regime through fixed bed column, filled with an adsorbent material obtained by functionalization of Amberlite XAD7 with DB18C6 and doped with Fe (III), the content of residual arsenic was determined as function of the volume of water passed over the material and the time required for breakthrough column, as well as a number of parameters specific to the drinking water quality (pH, nitrite, nitrate, ammonium, phosphate and bicarbonate content).
The extractant and the support remain in contact for 24 hours, after which the modified support was filtered and dried in oven (Nitech oven) at 50°C for 24 h. In order to include iron ions into the modified material, his was mixed with 25 mL of FeCl 3 (Sigma -Aldrich, Merck) solution (100 mg·L -1 ) and kept in contact for 24 h, then filtered and dried in the oven for 24 h.
All adsorption experiments using real water were performed in dynamic mode using a glass column with diameter of 2.5 cm and length of 30 cm, loaded with 20 g of prepared adsorbent material. The height of the material layer inside of the column was 10 cm.
Real water containing 80 µg·L -1 As(III) with pH 7.3 was introduced into the column using a HEIDOLF peristaltic pump with a flow-rate of 10 mL·min -1 . Analysis of As (III) content in water were determined by using inductively coupled plasma mass spectrometry ICP-MS BRUKER AuroraM90.
To study the influence of competing ions from real water upon arsenic adsorption process the content of nitrites, nitrates, ammonium, phosphate and bicarbonates was determined by atomic absorption spectrometry in UV-Vis using the VARIAN Cary 50 spectrometer.

Results and discussions Column adsorption experiments
In order to characterize the As(III) adsorption process in dynamic regime on column, the maximum adsorption capacity and the time need it to breakdown column were determined.
Evolution of residual concentration as function to volume sequence and implicitly time needed for analysis are presented in figure 1.
From figure 1 can observe that by increasing the volume of water passed over the adsorbent material, and implicit the time, the residual concentration of As (III) decreases below the detection limit of the device (<0.01µg·L -1 ) until the column is breakdown (after ~ 400 mL of water and 4 0 min). The adsorption capacity of the material begins to decrease after a volume of 700 mL water, respectively after a period of 70 min.

Modelling of fixed bed adsorption
In order to determine the arsenic adsorption mechanism, it is necessarily to make the kinetic study of the adsorption process in a dynamic regime by using several simplified mathematical models. Thus can be achieved by fitting the experimental data with Bohart-Adams, Yoon-Nelson and Thomas simplified models [42]. Bohart-Adams model is used for description of the initial part of the breakdown curve of the column, by using the following mathematical expression: (1) where: C 0 is the influent concentration, µg·L -1 C t is the effluent concentration, µg·L -1 t is time, min k BA is the kinetic constant of the Bohart-Adam model, L µg -1 . min -1 F is the linear velocity calculated by dividing the flow rate by the column section area, Z is the bed depth of column, cm N 0 is the saturation concentration, µg·L -1 Curve associated with Bohart-Adam model, consisting in linear dependence of ln(Ct/C0) versus time is depicted in figure 2. Analyzing data presented in figure 2 can evaluate From obtained results can evaluate if the column adsorption process is due to an external mass transfer during the first part of the process. The k BA and N O constants are presented in table 1 and the obtained regression coefficient is 0.9854. In order to model the column breakdown time, two models can be used, namely Yoon-Nelson and Thomas. Yoon-Nelson model is used to model a monocomponent system and does not require data on the characteristic of adsorbate, the type of adsorbent and the adsorbent physical properties of the adsorption bed.
The equation regarding a single component system is expressed in its linearized form as: (2) where: C t is the solution concentration at time t, µg·L -1 C 0 is initial solution concentration, µg L -1 k YN is the rate constant, min -1 τ is the time required for 50% adsorbate breakthrough, min.
The parameters τ and k YN can be obtained from the intercept and slope of the plot linear dependence graph ln[C t /(C 0 -C t )] function time ( fig. 3).
Based on the experimental data modelling, it was found that the process folds relatively well, the correlation coefficient R 2 being 0.9508, slightly lower than in the Bohart-Adams model. Also, the parameters τ and k YN are calculated and are found in table 1.
If the adsorption process is a process follows a secondorder reversible reaction kinetics, then obtained experimental data can be modelled using Thomas model. The Thomas model is frequently applied to estimate the adsorptive capacity of an adsorbent and predict breakthrough curves [43,44].
The Thomas model in its linearized form is given by: (3) where: C 0 is the solution concentration in the influent, µg·L -1 C t is the solution concentration at time t in the effluent, µg·L -1 k Th is the Thomas rate constant, L·min -1 . mg -1 q th is the equilibrium compounds uptake per g of the resin, µg·g -1 m is the mass of adsorbent resin, g Q is he flow rate, mL·min -1 .
A linear plot of ln [(C 0 /C t -1] against time (t) was employed to determine values of k Th and Q Th from the intercept and slope of the plot ( fig. 4). It can be estimated that the adsorption process is due to an external and internal diffusion and the resistance forces being extremely small [45]. Based on this model,  . 1), namely 72.5 min. Also, theoretically adsorption capacity value is 3.012 µg·g -1 value that is approaching the experimental value from adsorption capacity 2.18 µg·g -1 .
Influence of competing ions present in water onto As(III) adsorption process Present study investigates the effect of the presence of some competing ions (NO 3 -, NO 2 -, NH 4 -, PO 4 3-, SO 4 2and CO 3 2-) in real water, but also the pH of water onto As(III) adsorption process on produced material obtained by functionalization of Amberlite XAD7 with DB18C6 and doped with Fe (III) ions (experimental data are depicted in table 2).
From experimental data presented in table 2 can observe that in addition to the adsorption of the As (III), the use of such adsorbent material exhibit adsorption capabilities for other ions present in real water. It was also noted that the adsorbent material present affinity for the phosphate anions, reducing them from 12.5 mg·L -1 to 3.6 mg·L -1 and from bicarbonate ions reducing them from 424 mg·L -1 to 225 mg·L -1 . Also, can be affirmed that presence of phosphate anions and bicarbonate ions in water have a positive effect upon As(III) adsorption process [46]. Based on experimental data can conclude that the adsorbent material present adsorbent capacities for As (III) ions and for the competing ions present into the real waters. The pH of water decreases from 7.3 to 6.9 but not significantly.

Conclusions
Present paper aimed to study As(III) removal from a real water collected from Western area of Romania by adsorption on a new material obtained by the functionalization of Amberlite XAD7 polymer with Di-benzo 18-crown-6 (DB18C6) crown ether and then doped with iron ions. The studies were performed in dynamic regime using a glass column. Also, was studied the influence of competing ions (nitrite, nitrate, ammonium, phosphate sulphates and bicarbonates) present in water sample during the As(III) adsorption process.
To establish the adsorption mechanism, the experimental data obtained in this study were analyzed using three mathematical models: Bohart-Adams, Yoon-Nelson and Thomas models. Modelling was performed by linear regression analysis of experimental data. Regression coefficients R 2 were calculated, which are an indicator of the model, providing information about how well the process studied folds on one of the three models. Each model allows the determination of some parameters characteristic of the adsorption process.
Thus, the Thomas model allows the prediction of maximum adsorption capacity (q Th ) and Thomas rate constant (k Th ), the Yoon-Nelson model simulates breakthrough time (t) and Yoon-Nelson rate constant (k YN ) and the Adams-Bohart model predicts the saturation constant (N 0 ). In conclusion, the material can be used for treatment of water with arsenic content provided that the maximum adsorption capacity is increased and finding a solution to reduce the occurrence of preferential drainage through the column.