Chemical Engineering Journal 215–216 (2013) 315–321

Contents lists available at SciVerse ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier .com/locate /cej

Novel ion imprinted polymer coated multiwalled carbon nanotubes as a highselective sorbent for determination of gold ions in environmental samples

Homeira Ebrahimzadeh a,⇑, Elahe Moazzen a, Mostafa M. Amini a, Omid Sadeghi b

a Department of Chemistry, Shahid Beheshti University, GC, Tehran 1983963113, Iranb Department of Chemistry, Islamic Azad University, Shahr-e-Ray Branch, Tehran, Iran

h i g h l i g h t s

" Synthesis of a novel ion imprintedpolymer coated multiwalled carbonnanotube as a new sorbent.

" Preconcentration and determinationof trace amounts of gold with highselectivity and accuracy.

" Synthesis of nanosize particles.

1385-8947/$ - see front matter � 2012 Elsevier B.V. Ahttp://dx.doi.org/10.1016/j.cej.2012.11.031

⇑ Corresponding author. Tel.: +98 21 29902891; faxE-mail address: h-ebrahim@sbu.ac.ir (H. Ebrahimz

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 16 July 2012Received in revised form 29 October 2012Accepted 6 November 2012Available online 14 November 2012

Keywords:Multiwalled carbon nanotubesGoldIon imprinted polymerSolid-phase extraction

a b s t r a c t

In this work, for the first time an ion imprinted polymer (IIP) has been synthesized for gold ions precon-centration and determination. In order to increase extraction efficiency and improve the figures of merit,this polymer has been grafted on multiwalled carbon nanotube and was characterized by infrared (IR)spectroscopy, high resolution transmission electron microscopy (HRTEM), scanning electron microscopy(SEM), elemental analysis (CHN) and thermal analysis (TGA/DSC). The application of this sorbent hasbeen investigated for rapid extraction, preconcentration and determination of trace amounts of gold ionsin environmental samples. The optimum conditions, including pH of sample, eluent type and concentra-tion, also eluent and sample flow rates and the least amounts of eluent for desorption of gold ions wereobtained. The effects of various cationic interferences on the adsorption of gold ions, especially platinumand palladium ions, as the most important interferences, were evaluated. The analytical efficiency valueof gold ions was 98.3% in the optimum condition. The limit of detection (LOD) was less than0.041 ng mL�1 for gold extraction. The maximum capacity of this IIP modified nano-tubes were obtainedmore than 67 mg g�1. The relative standard deviation for this method was determined to be 1.14. Fur-thermore, the accuracy of this method was confirmed using various standard reference materials. Thismethod was successfully applied for determination of gold ions in natural samples.

� 2012 Elsevier B.V. All rights reserved.

1. Introduction

Gold is one of the most precious and important elements due toits usage in different areas such as jewelry, and industry. Gold pos-sesses unique properties that promote its widespread industrialapplications. So there have been so many techniques presentedfor its determination in environmental and mineral samples andwaste waters [1–3]. Among them flame atomic absorption spec-

ll rights reserved.

: +98 21 22403041.adeh).

trometry (FAAS) is a good and common technique to determinelow levels of gold ions due to its simplicity, easy operation andits fast response. But it has some difficulties such as lower levelsof gold ions than the limit of detection of FAAS and effects of thematrix components of the working media [4–7]. To overcome theselimitations on the determination of gold by FAAS, separation-enrichment techniques, including solid-phase extraction (SPE)[8–10], cloud point extraction [11,12], liquid–liquid extraction[13,14], coprecipitation [15], etc. have been used by the research-ers around the world. Among these methods SPE has attractedmore attention due to being inexpensive (does not need so much

316 H. Ebrahimzadeh et al. / Chemical Engineering Journal 215–216 (2013) 315–321

solvent), simplicity and ease of operate. But in the case of gold ionsdetermination, lack of selectivity especially toward platinum andpalladium, is the failure [16–18].

In ion imprinted polymers (IIPs) preparation process, functionalmonomers are copolymerized with a cross-linking agent in thepresence of the desired ion, the imprint ion. Then the imprint ionis removed by a suitable optimized eluent and the IIP is gained.The IIP nanoparticles have been grafted on different supports suchas mesoporous silica [19], Fe3O4 nanoparticles [20], SiO2 [21–22]and TiO2 [23]. Furthermore, IIP nanoparticles have been graftedon multiwalled carbon nanotubes (MWCNTs) for several purposes.As the IIP particles could be synthesized in nanoscale, the surfacearea is increased presumably [24]. Moreover, other supports havesome limitations; as an example Fe3O4 nanoparticles are destroyedin acidic conditions, therefore, it cannot be eluted with acidic solu-tions and the another example is mesoporous silica (MCM) whichis destroyed in alkaline conditions, in contrast CNTs are stable inboth acidic and alkaline medias. In addition to such advantages,these synthetic particles have good mechanical stability (decreasedsusceptibility to swelling and shrinking). The high surface area andhigh chemical and mechanical stability as well as high thermal sta-bility and the efficiency gained by grafting the IIP on the nano-tubes, make MWCNTs preferable material for being a support forIIP. It needs to be mentioned that it is the first time that gold IIPis grafted on CNT and in this work advantages and disadvantagesof the system are investigated by comparing the figures of meritof bulk IIP.

In this work, a new strategy was used for extraction of gold ionsusing an IIP with high selectivity toward Pt(II) and Pd(II) ions. Tothe best of our knowledge it is the first report for employing anIIP for determination of gold ions. In order to increase the efficiencyof this IIP, it was grafted on multi-walled carbon nanotubes(MWCNTs). The maximum capacity, optimum factors related toeluent including type and concentration and least amount of elu-ent, interfering ions effects and adsorption conditions containingpH and analyte flow rate, was also investigated. Moreover, the ana-lytical performance of this nano-sized sorbents was compared tothe bulk ones and also other IIP grafted supports. Finally, the appli-cation of this sorbent was investigated by determination of goldions in several natural samples.

2. Experimental

2.1. Reagents and materials

All chemicals used were analytical reagent grade without anyfurther purification. A 1000 lg mL�1 standard solution of HAuCl4,

HCl, HNO3, Na3C3H5O(CO2)3, Na2HPO4, NaH2PO4, CH3COOH, thio-urea, oxalyl choloride, triethylamine, 3-aminopropyl triethoxysi-lane and 3-vinyl triethoxysilane were purchased from MerckCompany (Darmstadt, Germany). Ethylene glycol dimethacrylate(EGDMA) was obtained from Fluka (Buchs, Switzerland). 2,20-Azobis iso butyronitrile (AIBN) was obtained from Acros Organics(New Jersey, USA). Carboxilic acid modified MWCNTs, 10–40 nmin diameters, 1–25 lm in length, were purchased from NeutrinoCompany (Tehran, Iran). The solutions of Au(III) were obtainedby diluting the standard solution with buffer, and pH adjustmentswere performed with the appropriate buffer solutions. To adjustpH of solutions to pH of 1–4, a mixture of Na3C3H5O(CO2)3/HCl(trisodium citrate/hydrochloric acid) was used. A solution of CH3

COOH/NaCH3COO was used to adjust pH values to the range of4–6 while a buffer solution containing Na2HPO4/NaH2PO4 wasused for pH values to the range of 6–10. All the required solutionswere prepared using deionized water provided by a Milli-Q(Millipore, Bedford, MA, USA) purification system. Standard

material sample (SRM 2711 and SRM 2781) with a certified goldions content was obtained from the China National Analysis Center.The mine stones were obtained from gold mineral which certifiedconcentration of gold was reported by Geological Survey of Iran.

2.2. Apparatus

Gold concentrations were determined by an AA-680 Shimadzu(Kyoto, Japan) flame atomic absorption spectrometer (FAAS) inan air-acetylene flame, according to the user’s manual, providedby the manufacturer. A gold hollow cathode lamp was used asthe radiation source with a wavelength of 242.8 nm. All pH mea-surements were performed at 25 ± 1 �C with a digital WTW Metr-ohm 827 Ion Analyzer (Herisau, Switzerland), equipped with acombined glass-calomel electrode. Flow rate investigating was per-formed using a peristaltic pump obtained from Leybold (Germany).To facilitate regulation of the flow rate during extraction, anadjustable vacuum gauge and controller were used, obtained fromAnalytichem International (Harber City, CA). The elemental analy-sis was performed with a Thermo Finnigan Flash-1112EA microan-alyzer (Italy). Thermal analysis (TGA–DTA) was carried out on aBahr STA-503 instrument under air atmosphere with heating rateof 10 �C/min. IR spectra were recorded on a Bruker IFS-66 FT-IRSpectrophotometer. High resolution transmission electron micros-copy (HRTEM) was performed on a Philips CM200. Scanningelectron microscopy was performed on a Vega-TeScan.

2.3. Ion imprinted polymer preparation

2.3.1. Preparation of vinyl functionalized MWCNTsVinyl functionalized MWCNTs were prepared according to the

earlier method [25] with some modifications. In this approach,1.0 g of COOH–MWCNTs was suspended in 50 mL of dried CH2Cl2.Afterward, 5.0 mL of oxalyl chloride was added to the solution andthe mixture was stirred for 24 h. Then the solvent was removedunder reduced pressure and the residue was suspended in 50 mLTHF and triethylamine mixture (1:3 V:V). 5 mL 3-aminopropyltri-ethoxy silane was added to the mixture subsequently. After 2 hthe particles were separated by centrifugation. Then the particleswere placed in HCl solution (pH = 4) for 1 h. The product was des-ignated as SiO2@MWCNT. After this step, to prepare vinyl function-alized MWCNTs, 1.0 g of SiO2@MWCNT was suspended in 50 mL oftoluene. Afterward 1.0 g of 3-vinyletriethoxy silane was added tothe solution and the mixture was stirred for 24 h. The solid-phasewas separated from the solvent and washed 3 times with 50 mL ofethanol. The vinyl functionalization of MWCNTs was confirmed byIR spectroscopy and elemental analysis. Elemental analysis showed0.48 mmol g�1 vinyl coated on this sorbent (C = 1.72%, H = 0.23%).

2.3.2. Preparation of N-allyl-N-(pyridin-2-yl)pyridin-2-amineN-allyl-N-(pyridin-2-yl)pyridin-2-amine (a-bipy) was prepared

according to following steps. 0.17 g dipyridile amine and0.079 mL vinyl chloride were added to 50 mL solution of triethylamine and methanol (1:4 V/V) at room temperature. After 2 h thesolvents were removed under reduced pressure and the solid prod-uct was characterized by 1H NMR spectroscopy.

2.3.3. Preparation of gold ions imprinted nano-particlesFor preparation of gold(III) complex of the ligand N-allyl-N-

(pyridin-2-yl)pyridin-2-amine (Au(a-bipy)2Cl3), 0.16 g (0.5 mmol) ofHAuCl4 was added to a methanolic solution containing 1 mmol ofN-allyl-N-(pyridin-2-yl)pyridin-2-amine. The colorless solutionturned to purple. In order to synthesize the gold IIP, in a typicalpolymerization, in a two-necked glass reactor equipped with acondenser, a mechanical stirrer and a gas inlet to maintain a nitro-gen atmosphere, 1 mmol of the complex (Au(a-bipy)2Cl3), 1.0 g of

H. Ebrahimzadeh et al. / Chemical Engineering Journal 215–216 (2013) 315–321 317

vinyl functionalized MWCNT and 1.1 mL ethylene glycol dimethac-rylate (EGDMA) as a cross linking agent were suspended in 100 mLof methanol. The mixture was heated to 60 �C and then 0.08 g ofAIBN, as an initiator was added to it. The product was separatedby centrifugation after 48 h and the Au(III) ions were removed bya solution containing 0.1 mol L�1 of HNO3 and 0.1 mol L�1 of thio-urea. The removal of the gold ions was followed by FAAS. The FAASdata showed that the removal was completed after seven timeselution. In order to confirm the removal of Au(III) ions, the amountof gold ions was determined by FAAS after using piranha solution(H2SO4 + H2O2). Piranha solution containing concentrated H2SO4

and 30% solution of H2O2 with (3:1 v/v) dissolves organic parts ofthese particles and release gold ions in solution. The formation ofthis IIP was confirmed by IR spectroscopy, thermal analysis, SEMand also HRTEM. A schematic diagram for entire process of theIIP synthesis is shown in Fig. 1.

2.3.4. Preparation of gold ions imprinted bulk particlesThe bulk particles were synthesized similar to above procedure

in the absence of vinyl functionalized MWCNT. Briefly, 1 mmol of

Fig. 1. A schematic diagram fo

the gold complex (Au(a-bipy)2Cl3) was dissolved in 100 mL ofmethanol and then the solution was heated to 60 �C. Afterward0.08 g of AIBN and 1.1 mL of EGDMA were added to the mixture.The product was filtered after 48 h and the Au(III) ions were re-moved by a solution containing 0.1 mol L�1 of HNO3 and0.1 mol L�1 of thiourea. The FAAS results show that the removalof gold ions was completed after nine times. The formation of thisIIP was confirmed by IR spectroscopy and thermal analysis.

2.4. Procedure

2.4.1. Column preparationA glass column (120 mm in length and 20 mm in diameter) with

a porous disk was packed with 200 mg of IIP@MWCNTs or bulk IIPparticles with a bed height of approximately 2 mm. The columnends were blocked with polypropylene filters to prevent loss ofmaterial during sample loading. To remove organic and inorganiccontaminants prior to extraction, each glass column was firstwashed with 5 mL of a 1 mol L�1 solution of HCl, followed by5 mL of absolute ethanol and 20 mL of distilled water.

r synthesis IIP@MWCNT.

318 H. Ebrahimzadeh et al. / Chemical Engineering Journal 215–216 (2013) 315–321

2.4.2. Preconcentration methodSample solutions containing certain amount of gold ions at a pH

of 4 were prepared. The column was conditioned by passingthrough a buffer solution with a pH of 4. The gold solutions werethen passed through the columns at optimum flow rate(7 mL min�1). The retained Au(III) ions were eluted with 12 mL ofa solution containing 0.1 mol L�1 thiourea and 0.1 mol L�1 HNO3

with the flow rate of 4 mL min�1. Gold ions content was deter-mined by FAAS. Five measurements were made for each sample,and the results were subsequently averaged.

Fig. 2. HRTEM micrograph of synthesized IIP@MWCNT.

2.4.3. Real sample pretreatmentSome real samples were used such as distilled water, tap water

(Tehran, Iran), sea water (Caspian Sea), jewelry wastewater (Teh-ran, Iran) and river water (Karaj River). The samples were storedin cleaned polyethylene bottles and were filtered through nylon fil-ters (Millipore) before the analysis. After adjusting the samples pHto the optimum pH, the solutions were passed through the pro-posed column.

Two standard materials (SRM 2711 and SRM 2781) with a cer-tified amount of gold have been used. These samples were di-gested in an 8 mL mixture of 5% aqua regia with the assistanceof a microwave digestion system. Digestion was carried out for2 min at 250 W, 2 min at 0 W, 6 min at 250 W, 5 min at 400 Wand 8 min at 550 W, and the mixture was then vented for8 min and the residue from this digestion was diluted with deion-ized water. The mine stone samples were digested as explainedabove.

3. Results and discussion

3.1. Characterization of the sorbent

Formation of IIP@MWCNTs was confirmed by HRTEM micro-graph, IR spectroscopy and TGA/DTA analysis. Modification ofnano-sized MWCNTs surface with vinyl groups was carried outthrough the earlier report using carboxyl MWCNTs according tothe sol–gel method [25]. Formation of the IIP was carried out ina polymerization reaction including vinyl functionalized MWCNTsas a monomer with N-allyl-N-(pyridin-2-yl)pyridin-2-amine asanother monomer in presence of an initiator like AIBN in a solu-tion containing EGDMA as crosslinking agent. A schematic dia-gram of gold IIP synthesis is shown in Fig. 1. Formation of thispolymer on the surface of MWCNT was confirmed by IR spectros-copy, TEM and thermal analysis. The IR spectrum showed bandsat 3010, 2940, 1735, 1551 and 1610 cm�1 which assigned to C–H aromatic, C–H aliphatic, C@O, C@C and C@N, respectively. TheHRTEM confirmed the presence of IIP on MWCNTs (Fig. 2). As itcan be seen in this figure, the MWCNT walls and presence ofpolymer on the nano-tubes can be clearly seen. For more investi-gation, the SEM micrograph of IIP@MWCNT was compared toSiO2@MWCNT in Fig. 3. Finally, the thermal stability of this com-posite was investigated by TG/DTA analysis. The thermal analysisof this sorbent showed that its stable up to 240 �C and approxi-mately 93% of this composite is polymer (Fig. 4). Bulk particleswere synthesized by the same procedure in the absence ofMWCNTs. Formation of these particles was confirmed by IR spec-troscopy and comparison of its thermal analysis withIIP@MWCNTs composite (Fig. 4). Comparison of TG/DTA curvesshows that the patterns are the same with two differences. Asthe bulk particles have no support, the losing weight is 100%and also the peaks are broad. As at nano-scale the polymer layersare thin, it is burnt more homogeneous and the peaks are sharperthan bulk sphere.

3.2. Optimization studies

3.2.1. Effect of pHThe effect of pH on the adsorption of Au(III) ions was tested by

passing 20 mL of 5 ppm solution of HAuCl4 through the column un-der different pH conditions from pH = 1 to pH = 10. It was figuredout that in pH of 3–5 the IIP particles show the highest adsorption(Fig. 5). The electrostatic interactions cause the best adsorptionefficiency in pH of 3–5, since in these pHs the major species forgold ions is AuCl�4 and for the ligand is pyridinium+ form. Expect-edly, in pHs above 5, the ligands exist in neutral form and conse-quently, in these pH ranges the adsorption efficiency is reduced.In lower pHs, the gold ions convert to HAuCl4 and cause adsorptionreduction. Hence, the sorption of gold ions increases with pH andreaches to a peak at the pH of 3–5 and then decreases with pH in-crease. As it is the functional groups (ligands) which are prorogatedin the acidic pH and cause the removal of the ions and the func-tional groups are the same in both of the sorbents, the bulk parti-cles showed the same pH pattern and optimum pH (Fig. 5).

3.2.2. Effect of type, concentration and volume of eluentTo elute the gold ions from this synthesized gold IIP particles, a

series of selected eluent solutions including different concentra-tions of HCl, H2SO4 or HNO3 and also each solution mixed with0.1 mol L�1 of thiourea were used. It was found out that the solu-tion containing 0.1 mol L�1 of HNO3 and 0.1 mol L�1 solution ofthiourea results an effective elution of gold from both nano andbulk IIP particles (Fig. 6). Thiourea helps to remove the ions owingto the strong interaction between its sulfur and the gold ionstrapped in the polymer. The effect of eluent volume on the recov-ery of gold ions was also examined. It was figured out that a quan-titative recovery can be obtained with 8 mL of the eluent solution.For bulk particles, this value is 12 mL. As the results show, the vol-ume of eluent for elution of the gold ions from bulk IIP particles, ishigher than the one grafted on MWCTs. This is due to the fact thateluting the ions from the bulk particles is more difficult, as the cen-ter parts of the bulk particles are difficult to reach for eluent, butthe ions is removed more easily from the thin layers of polymeron MWCNTs. Being the same conditions, optimum eluent was

Fig. 3. SEM micrograph of: (a) SiO2@MWCNT and (b) IIP@MWCNT.

Fig. 4. TGA/DSC curves of: (a) IIP@MWCNT and (b) IIP bulk nano-particles.

Fig. 5. Effect of sample pH on the recovery of Au(III) ions on synthesized IIP.

Fig. 6. Effect of the type and concentration of eluent for desorption of Au(III) ionsfrom synthesized IIP.

H. Ebrahimzadeh et al. / Chemical Engineering Journal 215–216 (2013) 315–321 319

selected 12 mL of a 0.1 mol L�1 solution of thiourea in 0.1 mol L�1

HNO3.

3.2.3. Sample and eluent flow rates and eluent volumeAs analysis time and the recovery of adsorption are affected by

flow rate, it is an important parameter and needs to be optimizedcarefully. Thus, the effect of the flow rates of both the sample andeluent solutions over IIP@MWCNTs and bulk particles were stud-ied. With the aid of a peristaltic pump, 100 mL solutions containing1 lg mL�1 gold ions were adjusted to a pH of 4 and passed throughthe column at flow rates ranging from 1 to 15 mL min�1. As shown

in Fig. 7, increasing the sample flow rate from 1 to 10 mL min�1

had no effect on the gold ions extraction efficiency on nanoparti-cles. Furthermore, in the same study, 8 mL of eluent was passedover the column to desorb the retained gold ions at flow rates of1–15 mL min�1. As the results show, the optimum flow rate of

Fig. 7. Effect of sample and eluent flow rate on the recovery of Au(III) ions.

320 H. Ebrahimzadeh et al. / Chemical Engineering Journal 215–216 (2013) 315–321

eluent for IIP@MWCNTs was found to be 6 mL min�1. For bulkpolymer particles, these values are 7 and 4 mL min�1 for sampleand eluent flow rates, respectively (Fig. 7). So in further experi-ments, optimum sample and eluent flow rate were selected 7and 4 mL min�1, respectively. Being more efficient in higher sam-ple and eluent flow rates, is a result of thin layers of polymer onMWCNTs which increase the liquid and surface.

3.3. Influence of interfering ions

The effect of variety of cations found in natural samples on thedetermination of gold ions was studied. In the form of their chlo-ride salts, various concentrations of Na+, K+, Cs+, Mg2+, Ca2+, Cd2+,Fe2+, Mn2+, Pb2+ and Cr3+ were added to gold-containing solutionslisted in Table 1. Also the impact of the most interfering ions suchas Pt2+ and Pd2+ has been investigated (Table 1). As shown in Ta-ble 1, the vast majority of transition metals even Pt2+ and Pd2+ doesnot have any interference and the method is selective toward goldextraction at a pH = 4 . Furthermore, extraction is not affected byhigh concentrations of alkaline and alkaline earth metals. More-over, there is no meaningful difference between nano and bulk par-ticles selectivity. This high selectivity for both balk and nanoparticles could be attributed to selective sites which their sizeand shapes are fitted to the gold ions specifically.

Table 1The tolerance limit of various ions on the determination of gold.

Interferingions

Tolerable concentration ratioX/Au

IIP@MWCNTsR%

Bulk IIPR%

Na+ 5000 98.3 92.4K+ 5000 98.9 93.7Cs+ 5000 99.2 92.6Ca2+ 2500 98.6 93.5Mg2+ 2500 98.5 91.8Fe2+ 2500 97.8 92.7Cd2+ 2500 98.7 94.4Mn2+ 2500 98.4 93.8Pb2+ 2500 99.1 91.9Cr3+ 2500 97.5 93.6Zn2+ 1000 98.8 94.6Ni2+ 1000 96.4 93.9Ag+ 1000 95.8 90.7Pd2+ 50 94.8 87.4Pt2+ 50 95.1 85.9Hg2+ 100 97.6 84.6

3.4. Adsorption capacity

The capacity of these synthesized IIPs with respect to Au(III)was studied by passing 250 mL of 50 ppm solution of Au(III) atpH = 4 through the columns at optimum flow rate (7 mL min�1).The retained Au(III) ions were eluted with 12 mL of the solutioncontaining 0.1 mol L�1 thiourea and 0.1 mol L�1 HNO3 with theflow rate of 4 mL min�1. The adsorbed gold content in eluent wasdetermined by FAAS. The capacities of these IIPs were found tobe 14 and 67 mg/g for bulk and nano particles, respectively. Thisincrease in capacity is due to the higher surface area of theIIP@MWCNTs than the bulk IIP particles.

3.5. Dilution and kinetic effect

The breakthrough volume of sample solutions was investigatedby dissolving 1 mg of gold in 100, 200, 500 and 1000 mL of distilledwater and the SPE protocol was followed. Gold was quantitativelyretained from 500 and 100 mL of sample solution for nano andbulk particles, respectively. The gold ions recoveries were greaterthan 98.3 and 92.7 for nano and bulk particles, respectively andthe loaded gold ions were desorbed with 8 and 12 mL eluent fromnano and bulk particles, respectively. Subsequently, they were sub-jected to distillation until the volume reached to 2 mL. As a result,the enrichment factors were found to be 224 and 42 using nanoand bulk IIP particles as sorbent, respectively.

3.6. Figures of merit

In order to determine the LOD of this method, adsorption proce-dure was performed using 500 mL blank solution (n = 10) underoptimal experimental conditions. The LOD value for gold ions onnano and bulk particles were 0.037 and 0.041 ng mL�1, respec-tively. The results were obtained from the relationship expressingCLOD = kSb/m where k = 3 [15].

The precision of the method under the optimal conditions (vol-ume = 100 mL, concentration: 1 mg L�1) was determined by per-forming ten replicates. The gold recoveries were found to be98.3 ± 1.14 and 92.7 ± 2.31 using nano and bulk IIP particles,respectively.

3.7. Validation of method

This technique was validated by applying the presented methodto several reference materials containing certified gold ions con-tent and the results shows this IIP is suitable for natural samples(Table 2).

In order to investigate the applicability and effect of differentmatrices on this method, a number of natural samples were ana-lyzed. Specified solutions with known amounts of gold were pre-pared and tested by the method. In all cases, quantitative goldrecovery was obtained (Table 2).

3.8. Comparison with bulk ones

In order to compare the efficiency of this nano-sized sorbentparticles with bulk ones, important factors such as LOD, maximumcapacity, flow rate, enrichment factor and selectivity of the sor-bents were compared as shown in Table 3. As it can be seen in Ta-ble 3, the analytical performance, figures of merit and also recoveryis improved, by converting from Bulk sphere to nanosphere, bygrafting the IIP on the MWCNT. The increase in preconcentrationfactor and maximum capacity is due to higher surface area ofIIP@MWCNTs than the bulk polymer.

Table 2Determination of gold in real samples.

Real sample (ng mL�1) Added (ng mL�1) Found (ng mL�1) Recovery (%)

Nano Bulk Nano Bulk

Distilled water Nd1 5.0 4.95 ± 0.11 4.72 ± 0.19 99.0 ± 2.2 94.0 ± 3.8Tap water Nd 5.0 4.90 ± 0.09 4.67 ± 0.17 98.0 ± 1.8 93.0 ± 3.4Waste water 516 100 613.00 ± 6.58 582.00 ± 12.52 98.0 ± 1.1 94.0 ± 2.1River water Nd 5.0 4.90 ± 0.13 4.78 ± 0.23 98.0 ± 2.6 95 ± 4.6Sea water Nd 5.0 4.95 ± 0.09 4.81 ± 0.21 99.0 ± 1.8 96.0 ± 4.2Certified (SRM 2711) 130 – 122.20 ± 2.47 108.70 ± 5.92 94.0 ± 1.9 83.0 ± 4.5Certified (SRM 2781) 70 – 65.30 ± 0.94 61.80 ± 1.53 93.0 ± 1.3 88.0 ± 2.1Mine stone #1 0.600 – 0.61 ± 0.03 0.58 ± 0.04 101.0 ± 4.7 96.0 ± 6.6Mine stone #2 1.100 – 1.07 ± 0.05 1.04 ± 0.07 97.0 ± 4.6 94.0 ± 6.3

1st: Not determined.

Table 3Comparison of IIP performance in bulk and nanosphere (IIP@MWCNTs).

Solid phase SampleFlow RatemL min�1

EnrichmentFactor

Recovery%

Maximumcapacity

Limit ofdetection(ng mL�1)

RSD(%)

Bulk 7 42 92.7 14 0.041 1.14IIP@MWCNTs 9 244 98.3 67 0.037 2.31

H. Ebrahimzadeh et al. / Chemical Engineering Journal 215–216 (2013) 315–321 321

3.9. Reusability of the column

The reusability of these columns was investigated by subse-quent sorption and desorption cycles. It was determined that thecolumn containing IIP@MWCNTs is stable up to 9 adsorption–elution cycles and for bulk particles this value is seven withoutany noticeable drop off in the recovery of gold.

4. Conclusion

Selectivity and recovery of a novel IIP based on dipyridile ligandas a high selective sorbent for extraction and preconcentration ofgold ions from environmental solutions have been investigated.In order to increase the efficiency of the sorbent and improve thefigures of merit, the IIP was synthesized on MWCNTs and com-pared to the bulk one. The high selectivity of this sorbent for Au(III)ions toward transition metals and platinum group ions and highpreconcentration factor as well as higher maximum capacity andgood recoveries make it an especial and efficient sorbent for so-lid-phase extraction of gold ions. Furthermore, various analyticalfactors such as extraction flow rate, enrichment factor, and LOD,

were investigated and compared to bulk ones in order to investi-gate the effects of transferring from bulk to nanosphere.

References

[1] D.A. Beardsley, G.B. Briscoe, J. Rika, M. Williams, Talanta 14 (1967) 879–886.[2] X. Shukun, S. Lijing, F. Zhaolun, Anal. Chim. Acta 245 (1991) 7–11.[3] A.S. Amin, Spectrochim. Acta Part A 77 (2010) 1054–1058.[4] D. Afzali, A. Mostafavi, M. Mirzaei, J. Hazard. Mater. 181 (2010) 957–961.[5] P. Liang, E. Zhao, Q. Ding, D. Du, Spectrochim. Acta Part B 63 (2008) 714–717.[6] H.B. Senturk, A. Gundogdu, V.N. Bulutb, C. Durana, M. Soylak, L. Elcid, M.

Tufekci, J. Hazard. Mater. 149 (2007) 317–323.[7] M. Tuzena, K.O. Saygia, M. Soylak, J. Hazard. Mater. 156 (2008) 591–595.[8] Q. Hu, X. Yang, Zh. Huang, J. Chen, G. Yang, J. Chromatogr. A 1094 (2005) 77–

82.[9] L. Zhang, Zh. Li, Zh. Hu, Ch. Chang, Spectrochim. Acta Part A 79 (2011) 1234–

1239.[10] M. Bagheri, M.H. Mashhadizadeh, S. Razee, Talanta 60 (2003) 839–844.[11] S. Chen, X. Zhu, Miner. Eng. 23 (2010) 1152–1154.[12] L. Tavakoli, Y. Yamini, H. Ebrahimzadeh, A. Nezhadali, Sh. Shariati, F.

Nourmohammadian, J. Hazard. Mater. 152 (2008) 737–743.[13] M.S. El-Shahawi, A.S. Bashammakh, S.O. Bahaffi, Talanta 72 (2007) 1494–1499.[14] M. Domingueza, E. Anticoa, L. Beyerb, A. Aguirrec, S. Garcia-Grandac, V.

Salvado, Polyhedron 21 (2002) 1429–1437.[15] M. Soylak, M. Tuzen, J. Hazard. Mater. 152 (2008) 656–661.[16] S. Daniel, R.S. Praveen, T.P. Rao, Anal. Chim. Acta 570 (2006) 79–87.[17] S. Daniel, P.P. Rao, T.P. Rao, Anal. Chim. Acta 536 (2005) 197–206.[18] S. Daniel, P.E.J. Babu, T.P. Rao, Talanta 65 (2005) 441–452.[19] Y. Liu, Z. Liu, J. Gao, J. Dai, J. Han, Y. Wang, J. Xie, Y. Yan, J. Hazard. Mater. 186

(2011) 197–205.[20] Y. Zhan, X. Luo, Y. Huang, X. Tu, Sh. Luo, Ind. Eng. Chem. Res. 50 (2011) 6355–

6361.[21] F. An, B. Gao, X. Feng, J. Appl. Polym. Sci. 112 (2009) 2241–2246.[22] X. Chang, N. Jiang, H. Zheng, Q. He, Zh. Hu, Y. Zhai, Y. Cui, Talanta 71 (2007) 38–

43.[23] Q. Li, H. Su, T. Tan, Biochem. Eng. J. 38 (2008) 212–218.[24] H. Ganjali, M.R. Ganjali, T. Alizadeh, P. Norouzi, Int. J. Electrochem. Sci. 6 (2011)

6085–6093.[25] Ch. Zheng, M. Feng, Y. Du, H. Zhan, Carbon 47 (2009) 2889–2897.