Talanta 80 (2009) 158–162

Contents lists available at ScienceDirect

Talanta

journa l homepage: www.e lsev ier .com/ locate / ta lanta

Separation and preconcentration of trace amounts of aluminum ions in surfacewater samples using different analytical techniques

Sumaira Khan 1, Tasneem G. Kazi ∗, Jameel A. Baig 1, Nida F. Kolachi 1, Hassan I. Afridi 1,Abdul Q. Shah 1, Ghulam A. Kandhro 1, Sham Kumar 1

Center of Excellence in Analytical Chemistry, University of Sindh, Jamshoro 76080, Pakistan

a r t i c l e i n f o

Article history:Received 24 April 2009Received in revised form 16 June 2009Accepted 17 June 2009Available online 3 July 2009

Keywords:AluminumWater sample8-HydroxyquinolineTriton X-114Cloud point extraction

a b s t r a c t

A separation/preconcentration of aluminum (III) (Al3+) has been developed to overcome the problemof high matrix species, which may interfere with the determination of trace quantity of Al3+ in natu-ral water samples. The separation of Al3+ in water samples was carried out from interfering cations bycomplexing them with 2-methyle 8-hyroxyquinoline (quinaldine) on activated silica. Whereas the sep-arated trace amounts of Al3+ was preconcentrated by cloud point extraction (CPE), as prior step to itsdetermination by spectrofluorimetry (SPF) and flame atomic absorption spectrometry (FAAS). The Al3+

react with 8-hydroxyquinoline (oxine) and then entrapped in non-ionic surfactant Triton X-114. Themain factors affecting CPE efficiency, such as pH of sample solution, concentration of oxine and TritonX-114, equilibration temperature and time period for shaking were investigated in detail. The validity ofseparation/preconcentration of Al3+ was checked by certified reference material of water (SRM-1643e).After optimization of the complexation and extraction conditions, a preconcentration factor of 20 wasobtained for Al3+ in 10 mL of natural water samples. The relative standard deviation for 6 replicates con-

taining 100 �g L−1 of Al3+ was 5.41 and 4.53% for SPF and FAAS, respectively. The proposed method has

ation 3+

1

anhmfirsrncao

((aw

0d

been applied for determin

. Introduction

Aluminum (Al3+) is a non-essential, toxic metal to which humansre frequently exposed. Normally it is very insoluble and in mosteutral natural waters its concentration is very low. In recent years,owever, a large amount of Al3+ has been released into the environ-ent through water acidification, waste discharge, and soils extract

rom acidic rain. The maximum permissible content of Al3+ in drink-ng water is only 200 �g L−1 [1]. Nowadays, much interest has beenaised by the toxicity and biological effect of Al3+ [2]. Some studiesuggested that Al3+ may be accumulated in the brain via differentoutes (drinking waters, food, and medicines) and interfere with the

ormal activities of nervous system [3,4]. This metal ion has beenonsidered as a possible cause of renal osteodystrophy, Parkinsonnd Alzheimer disease [4]. The determination of very low levelsf Al3+ has become increasingly very important in environmental

∗ Corresponding author. Tel.: +92 022 2771379; fax: +92 022 2771560.E-mail addresses: skhanzai@gmail.com (S. Khan), tgkazi@yahoo.com

T.G. Kazi), jab mughal@yahoo.com (J.A. Baig), nidafatima6@gmail.comN.F. Kolachi), hassanimranafridi@yahoo.com (H.I. Afridi),qshah07@yahoo.com (A.Q. Shah), gakandhro@yahoo.com (G.A. Kandhro),adhwashamkumar@yahoo.com (S. Kumar).1 Tel.: +92 022 2771379; fax: +92 022 2771560.

039-9140/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.talanta.2009.06.055

of trace amount of Al in natural water samples with satisfactory results.

© 2009 Elsevier B.V. All rights reserved.

and clinical chemistry since its negative role in the human life [5].Therefore, there is a strong need for Al3+ monitoring in natural waterresources [6].

There are a variety of analytical techniques such as, atomicabsorption spectrometry [7], inductively coupled plasma-atomicemission spectrometry (ICP-AES) and spectrofluorimetry are usedfor the determination of Al3+ in environmental samples [8,9]. How-ever, all of the techniques are required enrichment methods for thedetermination of trace amounts of Al3+ [10].

Due to presence of interfering cations such as iron (Fe3+),chromium (Cr3+), copper (Cu2+), lead (Pb2+) and zinc (Zn2+) in envi-ronmental samples or the presence of Al3+ below the detectionlimit, becomes difficult for accurate determination of Al3+ by spec-troscopic techniques [11]. So, the separation and preconcentrationsteps for Al3+ contents are still necessary to applying simple and lessexpensive techniques such as spectrofluorometry (SPF) and flameatomic absorption spectrometry (FAAS) [12,13]. The widely usedtechniques for the separation and preconcentration of Al includeliquid–liquid extraction [14], ion exchange [15], solid-phase extrac-

tion (SPE) [16] and cloud point extraction (CPE) [17]. Separationand preconcentration based on SPE and CPE are important alterna-tives and have a lot of practical applications in the field of surfacechemistry [16,18]. The SPE and CPE are offered the most frequentlyused extraction methodologies, which are simple, cheap, most

nta 80

etaetcbpaettta

afceFpbo

r(iuaabfSsas

2

2

(PbsloeITdsgaaoah

2

Ut

S. Khan et al. / Tala

fficient and less toxic than other extraction methods [19]. Recently,he SPE technique has become increasingly popular, by using

suitable complexing reagent which initiates the separation orlimination of interference [20]. The CPE is based on the preconcen-ration of metal ions after the formation of sparingly water-solubleomplexes in the surfactant-rich phase prior to the determinationy different spectrophotometric techniques [21,22]. The CPE cou-led with spectrophotometry for the determination of Al3+ haslso been reported [23]. In addition, the cloud point strategies cannlarge the FAAS applications, because depending on preconcentra-ion factors obtained, its sensitivity can significantly be increases,hus making the method more advantageous when compared withhose bases on direct determination using electrothermal atomicbsorption spectrometry and ICP-AES techniques [24].

The complexing reagent 8-hydroxyquinoline (oxine) has beencknowledged as one of the most sensitive organic ligands usedor the determination of Al3+ [25,26]. It forms a highly fluorescentomplex, without showing any intrinsic fluorescence itself. In thextraction of Al3+ by complexing with oxine, the interference ofe3+, Cu2+, Cr3+, Zn2+ and other metals ions, which are made com-lex with oxine, can be eliminated partly by using a masking agent,ut it is inconvenient in operation and not very sensitive becausef the use of organic reagent [16].

The aim of present work was to develop and establish a sepa-ation/preconcentration of trace quantity of Al3+ in surface waterriver, canal and lake) samples. For separation of Al3+ from othernterfering cations, 2-methyl 8-hydroxyquinoline (quinaldine) wassed which is selective for complexation with different metal ionsdsorbed on activated silica, while for enrichment of separated Al3+,CPE was used to react with oxine and resulted complex entrappedy Triton X-114 prior to its determination by SPF technique, andor comparative purpose FAAS with nitrous oxide-acetylene flame.everal experimental variables affecting the method sensitivity andtability were investigated in detail. The proposed method has beenpplied for determination of trace amount of Al3+ in surface wateramples of different origin with acceptable results.

. Experimental

.1. Instrumentation

A centrifuge of WIROWKA Laboratoryjna type WE-1, nr-6933speed range 0–6000 rpm, timer 0–60 min, 220/50 Hz, Mechanikahecyzyjna, Poland) used for centrifugation. The pH was measuredy pH meter (720-pH meter, Metrohm) and global positioningystem (iFinder GPS, Lowrance, Mexico) was used for samplingocations [27]. Fluorescence measurement of Al–oxine was maden a Shimadzu RF-510 (Kyoto, Japan) spectrofluorophotometerquipped with a 150 W Xenon lamp and using 1.00 cm quartz cells.nstrument excitation and emission slits were adjusted to 10 nm.he concentration of Al3+ in extracts was also determined by aouble beam PerkinElmer model A Analyst 700 atomic absorptionpectrometer (Norwalk, CT, USA) equipped with deuterium back-round correction. An aluminum hollow-cathode lamp was useds radiation source at wave length (nm) 309.3, slit width (nm) 1.3,nd lamp current (10 mA). The Al3+ was measured under optimizedperating conditions by FAAS with fuel (acetylene 0.45 kg cm−1)nd oxidant (nitrous oxide as oxidant = 1.6 kg cm−2) at burnereight (12.5 mm).

.2. Reagents

Ultrapure water obtained form ELGA labwater system (Bucks,K), was used throughout the work. The non-ionic surfactant Tri-

on X-114 was obtained from Sigma (St. Louis, MO, USA) and was

(2009) 158–162 159

used without further purification. Stock standard solution of Al3+ ata concentration of 1000 �g L−1 was obtained from the Fluka Kamica(Bush, Switzerland). Working standard solutions were obtained byappropriate dilution of the stock standard solutions before analysis.Concentrated nitric acid, hydrochloric acid were analytical reagentgrade from Merck (Darmstadt, Germany) and were checked forpossible trace Al3+ contamination by preparing blanks for each pro-cedure. The 8-hydroxy quinoline and 2-methyl 8-hydroxyquinolinewere obtained from (Merck), both reagents were prepared by dis-solving appropriate amount of these reagents in 10 mL ethanol(Merck) and diluting to 100 mL with 0.01 M acetic aid and were keptin refrigerator (4 ◦C) for one week. The 0.01 mol L−1 acetate bufferwas used to control the pH of the solutions. The pH of the sampleswas adjusted to the desired pH (3–8) by the addition of 0.1 mol L−1

HCl or NaOH solution in acetate buffer. For the accuracy of method-ology, certified reference material of water SRM-1643e (NationalInstitute of Standards and Technology (NIST), Giathersburg, MD,USA) was used.

For analysis of trace quantity of Al3+, contamination is a seriousproblem. Hence, the utmost care was taken in preparing standardsand samples, and the working area around the instruments waskept as clean as possible, especially for the determination of Al3+

by SPF technique. The glass and plastic wares were soaked in 10%nitric acid overnight and rinsed many times with deionized waterprior to use to avoid contamination.

2.3. Sampling

The surface water samples of different origin (river, canal andlake) were collected on alternate month in 2008 from 210 samplingsites of Jamshoro, Sindh (southern part of Pakistan) with the helpof Global positioning system (GPS). The understudy district posi-tioned between 25◦19′–26◦42′N and 67◦12′–68◦02′E. The samplingnetwork was designed to cover a wide range of determinates ofwhole district. From each sampling site, a fresh surface water sam-ples from canal (CS), river (RS), and lake (LS) were collected frommain stream of five to six different sampling points at a depth of20–30 cm. The collection of surface water samples was performedby using Van Dorn plastic bottles (1.5 L capacity) and was kept inwell stoppered polyethylene plastic bottles previously soaked in10% nitric acid for 24 h and rinsed with ultra pure water. All watersamples were filtered through a 0.45 �m pore size membrane filterto remove suspended particulate matter and were stored at 4 ◦C.

2.4. Solid-phase extraction

For separation of Al3+ from different interfering cationic speciespresent in matrices of water samples, a glass column with an innerdiameter of 20 mm and a length of 25 cm was filled up to a heightof about 20 cm with activated silica gel at 100 ◦C. Prior to use, thecolumn was preconditioned with buffer solution (acetate bufferpH 6.2), and passed 10 mL of 0.1 mol L−1 quinaldine solution pre-pared in ethyl alcohol-acetic acid at pH 6.2. A 25 mL of replicate sixsample of SRM and real water samples were passed through thiscolumn to separate Al3+ from other interfering cations at the flowrate of 0.5–5.0 mL min−1. The treated SRM and real water sampleswere then divided into two sub-samples of 10 mL each. After sev-eral separation experiments (n = 20), the column was rinsed with10% ethanol in 0.1 mol L−1 HNO3 and deionized water then used forfurther experiments.

2.5. Cloud point extraction

For Al3+ preconcentration, aliquots of 10 mL of standard solu-tions containing Al3+ (in the range of 10–200 �g L−1), replicate sixsamples of SRM, surface water samples, treated with quinaldine to

1 nta 80 (2009) 158–162

rbia

0Xp3taba

0oSciaflps

3

iisfimstraw

3

tmaZpitc

3

wobhaiwf

3

i>H

dard, SRM and real sample solution containing Al3+ in the presenceof 0.1% (w/v) Triton X-114 and added appropriate buffer solu-tion to kept pH 6.5 and concentrations of oxine solutions in therange of (1.37–10.9) × 10−5 (0.1–0.8 mL from 1.37 × 10−3 mol L−1

60 S. Khan et al. / Tala

emove interfering species were simultaneously preconcentratedy CPE. The same sub-samples of SRM and water without remov-

ng interfering cations were taken in graduated centrifuge tubesnd subjected to preconcentration for comparison purposes.

In 10 mL of each standard SRM, and real samples, added.2–0.8 mL of oxine solution (6.90 × 10−3 mol L−1), 2 mL of Triton-114 (0.05–0.2%, v/v) and 2 mL of different buffers to adjust aH range of 4–8. The tubes were placed in an ultrasonic bath at0–70 ◦C for 10–30 min. After different time intervals, the separa-ion of the two phases was achieved and then centrifuged for 5 mint 3500 rpm (1852 × g). The contents of tubes were cooled in an ice-ath, the surfactant-rich phase became viscous and the supernatantqueous phase was decanted.

To decrease the viscosity of one of the duplicate extracts, added.5 mL of ethanol–water mixture (1:1) and fluorescence intensityf obtained Al3+ chelate in the micellar phase was measured byPF at �excitation 375 ± 3 nm and �emission at 510 ± 3 nm against aorresponding reagent blank. While simultaneously, the viscos-ty of surfactant-rich phase of other sub-sample was decreased bydding acidic-ethyl alcohol (0.1 mol L−1 HNO3) and introduced intoame by conventional aspiration. A blank submitted to the samerocedure was measured parallel to the calibration solutions oftandards, SRM and real samples.

. Results and discussion

To overcome the problem of high matrices concentration, whichs thought to interfere with the determination of low levels of Al3+

n environmental samples. The Al3+ was separated from interferingpecies in real water sample by adsorbing on quinaldine modi-ed silica, then separated Al3+ was react with oxine in a micellaredium and form a hydrophobic complex (Al–oxine), which is sub-

equently trapped in the non-ionic surfactant and separated fromhe aqueous phase. The oxine is strong chelators for Al3+ and givesise to fluorescent complex [20,5]. To check the accuracy of the SPEnd CPE, the results obtained by spectrofluorimetry were comparedith those obtained by FAAS.

.1. Optimization of SPE

As reported earlier that Al3+ was not reacted with quinaldine dueo stearic hindrance of the methyl group at the 2-position of aro-

atic ring [28]. The quinaldine adsorbed on activated silica used asbsorbent for SPE of interfering species (Fe3+, Cr3+, Cu2+, Pb2+ andn2+), which can quantitatively eliminate the interfering cations atH 5.5 with high adsorption capacity. The separation of Al3+ from

nterfering ions is necessary for determination of its trace quan-ity, particularly to measure the fluorescence intensity of Al–oxineomplex using SPF technique.

.1.1. Effect of pH on sorptionThe influence of pH on the retention of interfering elements

as determined. The pH ranges of (4–8) of standards solutionf interfering ions, SRM and real water samples were adjustedy buffer solution with either 0.1 mol L−1 nitric acid or sodiumydroxide. These solutions were then passed through the columnt 0.5–5.0 mL min−1. It was found that the maximum retention ofnterfering ions were achieved at pH 5.0–6.0 and begins to decrease

hen the solution is over pH 6.5. Therefore, the pH 5.5 was usedor separation of Al3+ ions from interfering cations.

.1.2. Effect of flow rate on sorptionThe retention of interfering ions on the quinaldine modified sil-

ca was studied at different flow rates. The separation of Al was95% at optimum flow rate of the sample solution up to 2 mL min−1.owever, at a flow rate above 5 mL min−1, there was a decrease

Fig. 1. Effect of pH on the spectrofluorimetric responses: 100 �g L−1 Al,6.85 × 10−5 mol L−1 8-HQ, 0.1% (w/v) Triton X-114, temp. 40 ◦C, stirring time 20 min.

in the sorption percentage of interfering ions, because ligand for-mation between interfering ions and quinaldine is slow at roomtemperature. The retention of interfering ions was confirmed byFAAS (not reported) and for further work a 1.0 mL min−1 flow ratewas selected. The SPE is a highly selective and sensitive method forsimple and quick elimination of interfering of matrixes elements,to determined fluorescence intensity of Al–oxine chelate.

3.2. Optimization of CPE

3.2.1. pH effectsThe pH plays a distinctive role on metal-chelate formation and

subsequent extraction, this factor is proved to be a main parameterfor CPE method. The replicate six standard, SRM and real sam-ple solution of Al3+ (100 �g L−1) were adjusted to the pH range of(pH 3–8), each desired pH value was obtained by the addition of0.1 mol L−1 HNO3 and/or 0.1 mol L−1 NaOH solution in the presenceof acetate buffer solution. Fig. 1 shows the effect of pH on the extrac-tion recovery of Al-oxinate which were calculated on the basis ofcertified value of Al in SRM and in the surfactant-rich solution afterCPE using SPF. As can be seen that quantitative extraction (>96%)was obtained for Al3+ in the pH range of 5–7 and starts to decreaseafter pH 7.0. Hence, a pH of 6.5 was chosen for optimum CPE of Al3+.Based on the obtained results, there may be no need for adjustingthe pH in the case of understudy water samples, because their pHvalues fall near to the required range (6.8–7.5), while it was alsoobserved that after pH 7, abrupt decline in % recovery.

3.2.2. Effect of oxine concentrationThe % recovery of Al3+ as a function of the concentration of

the chelating agent, oxine is shown in Fig. 2, where 10 mL of stan-

Fig. 2. Effect of oxine concentration on the %recovery: 100 �g L−1 Al, 0.1% (w/v)Triton X-114, pH 6.5, temp. 40 ◦C, stirring time 20 min.

S. Khan et al. / Talanta 80

F8

oeostCi

3

eptswwwcAopseotbsF

3

cTs1rtttem

TV

E

CW

V

ig. 3. Effect of Triton X-114 on the %recovery: 100 �g L−1 Al, 6.85 × 10−5 mol L−1

-HQ, pH 6.5, temp. 40 ◦C, stirring time 20 min.

xine solution) were subjected to cloud point preconcentration. Thextraction efficiency of Al3+ increases up to 6.85 × 10−5 mol L−1 ofxine, then reaching a plateau, indicated that this amount of oxine isufficient for total complexation and the extraction. The concentra-ions above this value have no significant effect on the efficiency ofPE. The stoichiometry of the Al–oxine ternary complex, as reported

n literature is 1:3 ratios [29].

.2.3. Effect of Triton X-114 concentrationIn present work Triton X-114 was chosen because of its higher

xtraction efficiency and lower cloud point temperature as com-ared with other reported surfactants [30]. The low cloud pointemperature avoids back extraction during centrifugation. Fig. 3hows the variation in extraction efficiency of Al–oxine complexithin the concentration range of 0.05–0.2% (w/v) of Triton X-114,hile other factors were at optimum levels. The 60–70% recoveryas observed at 0.05% of Triton X-114, while the extraction effi-

iency reaches a maximum in the concentration range of 0.1–0.15%.concentration of 0.1% (w/v) of Triton X-114 was chosen as the

ptimum surfactant concentration in order to achieve the highestossible extraction recovery of Al3+ from standards, SRM and realamples. Whereas at lower than 0.1% of Triton X-114, the extractionfficiency of complexes is low probably because of the inadequacyf the assemblies to entrap the hydrophobic complex quantita-ively, on the contrary at higher concentrations, the signals decreaseecause of the increment in the volumes and the viscosity of theurfactant phase (ethanol–water for SPF and ethanol–nitric acid forAAS) to reducing the viscosity of the surfactant.

.2.4. Effect of temperature and timeA sufficiently long reaction time was allowed for the Al–oxine

omplex formation to precede up to 300 min at room temperature.he effects of equilibration temperature and time on the analyticalignal were studied in the range of 30–60 ◦C in ultrasonic bath for0–60 min, respectively. In the present study low temperature wasequired for optimum extraction efficiency of Al3+. It was observed

3+

hat the optimum % recovery of Al was achieved at equilibrationemperature of >50 ◦C and a time of 20 min. At higher temperaturehe fluorescent intensity of Al–oxine complex was lowered, mayffects the viscosity of medium and the number of collisions ofolecules of the fluorophore with solvent molecules. So, 40 ◦C was

able 1alidation of CPE of Al3+ in certified sample of water SRM 1643e using spectrofluorimetry

lement (Al3+) Certified value of SRM 1643e (water) SPF

PE 141.8 ± 8.6 139ithout CPE 137

alue in parenthesis = % recovery.a tcertical = 2.28 at 95% confidence limit (n = 6) (5.41) (4.53).

(2009) 158–162 161

chosen for further experiments. We preferred to stop the reactionat 60 min, as a longer reaction time did not increase fluorescenceintensity and % recovery of Al3+.

3.3. Interferences

The interferences studied were those relating to the preconcen-tration step, i.e. cations may react with oxine or species that mayreact with Al3+ and decrease the extraction efficiency and also inter-ference in determination of fluorescence intensity using SPF. Theinterference from these ions can be eliminated by SPE using quinal-dine that form chelate with these ions except Al3+. To perform thisstudy, 10 mL solution containing 100 �g L−1 of Al3+ and interferingions in different interferent to analyte ratios were subjected to thecomplete procedure. The determination of Al3+ in the presence ofZn (50 �g L−1) produce positive error and broad band in emissionspectrum was observed, while Fe3+, Cr3+, Pb2+ and Cu2+ (1000, 20,20 and 100 �g L−1), respectively, can interfere with the determina-tion of Al3+, to produce negative effects (Table 1). Our results arenot consistent with other study to determine simultaneously Al3+

and Zn2+ using oxine as chelating reagent [17].

3.4. Validity and applicability

The calibration graph using the preconcentration system for Al3+

was linear with a correlation coefficient of 0.9997 at 10–200 �g L−1.The surfactant-rich phase was diluted with 0.5 mL of water–ethanoland the fluorescence intensity of the chelate was measured. In allcases, linear relationships between the fluorescence measured andthe concentration of the Al in the standards, SRM and real waterwere attained. Simultaneous same sub samples diluted with acidicalcohol were subjecting to FAAS for comparison purposes.

The reproducibility was calculated as % relative standard devia-tion for 6 replicates containing 100 �g L−1 of Al3+, obtained 5.41 and4.53% for SPF and FAAS, respectively. The limit of detection (LOD)calculated (0.5 �g L−1) as three times the standard deviation of theblank signals.

The obtained LOD was sufficiently low as compared to reportedwork [5] and to be valuable for detecting Al3+ in different typesof surface water samples. Quantitatively accurate results wereobtained by adopting matrix-matched calibration of certified stan-dards and certified reference material of water, thus avoiding theuse of more time-consuming and laborious standard additions intoeach sample.

The Student’s t-test showed that the results obtained by SPFand FAAS were not significantly different at 95% confidence level(Table 1). After optimization, the preconcentration factor (CF) of 20was achieved. The CF obtained in this work was comparable withthose obtained in other CPE methods used for extraction and pre-concentration of Al3+ [20]. It is obvious that this CF and LOD can beimproved by using larger volume of initial solution or small-volumecells and diluting the surfactant-rich phase by small amounts of

ethanol. The method was successfully applied to the determinationof Al3+ in different natural water samples.

The obtained results showed variations among the concentra-tion of Al3+ in surface water samples of three origins (Table 2).The concentration of Al3+ in RS, CS and LS water samples was

(SPF) and flame atomic absorption spectrometry (FAAS) (�g L−1).

x ∓ ts/√

n FAAS x ∓ ts/√

n Paired t testa tExperiment

± 7.5 (98.0%) 140 ± 6.4 (98.7%) 0.78(96.6%) 138 (97.3%)

162 S. Khan et al. / Talanta 80 (2009) 158–162

Table 2Ranges of pH and Al3+ concentration in the surface and ground water samples of district Jamshoro, Sindh, and Pakistan.

Parameter Unit WHO values Canal water River water Lake water

n = 120 n = 30 n = 30

Mean

p 7.27A 80

orocphisArwci

4

daosmutAsgeAiHvAa

R

[

[[

[

[[

[[[

[[[

[

[[

[

[[[30] D.R.C. McLachlan, C. Bergeron, J.E. Smith, D. Boomer, S.L. Rifat, Neurology 46

(1996) 401.[31] M.B. Arain, T.G. Kazi, M.K. Jamali, N. Jalbani, H.I. Afridi, A. Shah, Chemosphere

Min Max

H – 6.5–8.5 7.1 7.6l3+ �g L−1 200 90 150

bserved in the ranges of 10.0–150, 90–150 and 750–1340 �g L−1,espectively. The concentration of Al3+ in three studied origins wasbtained in increasing order: RS < CS < LS. It was observed that con-entration of Al3+ in RS and CS water samples are within WHOermissible level for drinking water [1], while in LS water samplesigh content of Al3+ was found. This can be attributed to reduction

n precipitation, surface wastage runoff with rain water into under-tudy Lake and increasing rate of evaporation during summer [31].ll this provides evidence that anthropogenic and geological envi-

onment play a key role in the distribution of Al3+ in understudyater bodies. Number of epidemiological studies showed an asso-

iation between Al in drinking water and Alzheimer’s disease, anmportant form of senile dementia in man [29,32].

. Conclusions

The separation and preconcentration methodologies for theetermination of trace quantities of Al3+ in water samples, offerssimple, sensitive, inexpensive and non-polluting alternative to

ther separation/preconcentration techniques using toxic organicolvents. To the best of our knowledge, modified silica with 2-ethyl 8-hydroxyquinoline as a solid-phase extractor is first time

sed by us to separate interference elements from environmen-al sample matrixes, to determine the fluorescence intensity ofl–oxine complex. The Al3+ is not reacting with quinaldine due totearic hindrance effect of methyl group. The proposed CPE methodives low LOD, good RSD, less time-consuming, achievement of highnrichment factors and solvent-free extraction of trace quantity ofl3+ in natural water matrixes. The surfactant-rich phase can be eas-

ly diluted with ethyl alcohol–water and ethyl alcohol–0.1 mol L−1

NO3 systems prior to its determination by SPF and FAAS. Due toersatility, it can be applied to the monitoring of trace quantity ofl3+ in various samples of environmental, toxicological and clinicalnalysis.

eferences

[1] WHO, Aluminium, World Health Organization, International Programme onChemical Safety, Geneva, 1997 (Environmental Health Criteria 194).

[

Min Max Mean Min Max Mean

7.0 7.5 7.2 7.10 8.20 7.410 150 90 750 1340 1000

[2] M. Venturini-Soriano, G. Berthon, J. Inorg. Biochem. 71 (1998) 135.[3] G.L. Klein, Curr. Opin. Pharmacol. 5 (2005) 637.[4] M. Soylak, U. Sahin, A. Ulgen, L. Elci, M. Dogan, Anal. Sci. 13 (1997) 287.[5] M. Buratti, C. Valla, O. Pellegrino, F.M. Rubino, A. Colombi, Anal. Biochem. 353

(2006) 63.[6] J. Tria, E.C.V. Butler, P.R. Haddad, A.R. Bowie, Anal. Chim. Acta 588 (2007)

153.[7] L. Zhu, K. Li, S. Zhou, Q. Hu, G. Yang, Z. Huang, J. Chen, Asian J. Chem. 20 (2008)

4433.[8] B. Bocca, A. Alimonti, F. Petrucci, N. Violante, G. Sancesario, G. Forte, O. Seno-

fonte, Spectrochim. Acta B 59 (2004) 559.[9] A. Sanz-Medel, A.B. Soldado Cabezuelo, R. Milacic, T. Bantan Polak, Coordin.

Chem. Rev. 228 (2002) 373.10] I. Narin, M. Tuzen, M. Soylak, Talanta 63 (2004) 411.

[11] E.K. Paleologos, C.D. Stalikas, S.M. Tzouwara-Karayanni, M.I. Karayannis, Anal.Chim. Acta 436 (2001) 49.

12] E.K. Paleologos, C.D. Stalikas, M.I. Karayannis, Analyst 126 (2001) 389.13] E.K. Paleologos, C.D. Stalikas, S.M. Tzouwara-Karayanni, G.A. Pilidis, M.I.

Karayannis, J. Anal. At. Spectrom. 15 (2000) 287.14] J. Komarek, R. Cervenka, T. Ruzicka, V. Kuban, J. Pharm. Biomed. Anal. 45 (2007)

504.15] S.B. Erdemoglu, K. Pyrzyniska, S. Gucer, Anal. Chim. Acta 411 (2000) 81.16] M.B. Luo, S.P. Bi, J. Inorg. Biochem. 97 (2003) 173.

[17] A.B. Tabrizi, Food Chem. 100 (2007) 1698.18] S. Rubio, D. Perez-Bendito, Trends Anal. Chem. 22 (2003) 470.19] C.D. Stalikas, Trends Anal. Chem. 21 (2002) 343.20] P. Vanloot, J.L. Boudenne, L. Vassalo, M. Sergent, B. Coulomb, Talanta 73 (2007)

237.21] J.L. Manzoori, A.B. Tabrizi, Anal. Chim. Acta 470 (2002) 215.22] M. Tuzen, M. Soylak, J. Hazard. Mater. 154 (2008) 519.23] L. Sombra, M.O. Luconi, M.F. Silva, R.A. Olsina, L. Fernandez, Analyst 126 (2001)

1172.24] M. Bahram, T. Madrakian, E. Bozorgzadeh, A. Afkhami, Talanta 72 (2007)

408.25] F. Papadimitrakopoulos, X. Zhang, K.A. Higginson, J. IEEE 4 (1998) 49.26] A. Alonso, M.J. Almendral, M.J. Porras, Y. Curto, C. García de María, Anal. Chim.

Acta 447 (2001) 211.27] J.A. Baig, T.G. Kazi, M.B. Arain, H.I. Afridi, G.A. Kandhro, R.A. Sarfraz, M.K. Jamal,

A.Q. Shah, J. Hazard. Mater. 166 (2009) 662.28] A. Asrar, T.G. Kazi, G.H. Kazi, J. Phys. Chem. 4 (1985) 21.29] A. Ohashi, A. Tsuguchi, H. Imura, K. Ohashi, Anal. Sci. 20 (2004) 1091.

70 (2007) 1845.32] V. Rondeau, D. Commenges, H.G. Jacqmin, J.F. Dartigues, Am. J. Epidemiol. 1521

(2000) 59.