Determination of lead in milk and yoghurt samples by solid phase extraction using a novel aminothioazole-polymeric resin

Abstract

A preconcentration method was developed by using a new aminothioazole-containing

sulfonamide resin in solid phase extraction for the determination of lead and nickel in milk

and yoghurt samples. The optimisation of experimental conditions was performed. The

optimum parameters for Pb and Ni were found to be 3.5, 30 min and 2.5 mL for pH, contact

time, eluate volume, respectively. After preconcentration step, atomic absorption

spectrometry was used for the determinations. The enhancement of 350- and 50-fold in the

sensitivity of Pb and Ni were achieved by combination of the slotted tube atom trap-atomic

absorption spectrometry with the optimised preconcentration method, respectively. Limits of

detection were found to be 0.15 ng mL−1 for Pb and 0.75 ng mL−1 for Ni. The lead

concentrations in the studied samples were found to be in the range of 15–61 ng Pb mL−1 for

milk and 21–42 ng Pb g−1 for yoghurt samples.

O metodă de preconcentrare a fost dezvoltat prin utilizarea unui noua aminothioazole conținând rășină sulfonamidă în extracție în fază solidă pentru determinarea plumb și nichel înprobe de lapte și iaurt . A fost realizată optimizarea condițiilor experimentale. Parametrii optime pentru Pb și Ni s-au dovedit a fi 3,5, 30 de minute și 2,5 ml pentru pH, timp de contact, respectiv volum eluat,. După primul pas care a fost preconcentrarea ,a fost folosita spectrometrie de absorbție atomică pentru determinări. Îmbunătățirea 350- și 50 ori a sensibilității Pb si Ni au fost realizate prin combinarea atom tub-capcană atomică spectrometria de absorbție cu fante cu metoda de preconcentrare optimizat, respectiv. Limitele de detectare s-au dovedit a fi 0,15 ng ml-1 pentru Pb și 0,75 ng ml-1 pentru Ni. Concentrațiile de plumb în probele studiate au dovedit a fi în intervalul de 15-61 ng Pb ml-1 pentru lapte și 21-42 ng Pb g-1 pentru probele de iaurt.

Highlights

► A new modified resin, aminothioazole-containing sulfonamide resin was synthesised. ►

The synthesised resin was characterised and used for preconcentration. ► Lead and nickel

levels in milk and yoghurt samples were determined using the resin. ► The synthesised resin

has high adsorption capacity. ► The resin can also be examined to remove Pb and Ni from

polluted waters.

Keywords

Lead; 

Nickel; 

Preconcentration; 

Aminothioazole-containing sulfonamide resin; 

Milk; 

Yoghurt

1. Introduction

There are high interested in the determinations of Pb and Ni in environmental, food and

biological samples due to their extremely toxic for human and animal health. It is known

that lead is health-endangering element for human and its effects include blood enzyme

changes, anaemia, hyperactivity, and neurological disorders (Wang et al., 2009). Lead can

also cause pathophysiological changes in several organ systems including central nervous,

renal, hematopoietic and immune system (Mishra, 2009). Among these damages, one of the

most important issues is that lead can strongly affect intelligence development of children.

Further, inorganic lead compounds are classified as probably carcinogenic to humans (group

2A) by the International Agency for Research on Cancer (IARC, 2006 and Yaman, 2006).

Nickel is a nutritionally essential trace metal for, at least, several animal species, micro-

organisms and plants (Cempel & Nikel, 2006). It is also known that Ni can cause allergic

reactions and certain Ni compounds have carcinogenic effects (Yaman, 2006). Therefore, its

either deficiency or toxicity symptoms can be occurred when too little or too much Ni is

taken up, respectively. World Health Organization (WHO) established the provisional

tolerable weekly intakes (PTWI) of Pb as 25.0 μg kg−1 body weight (equivalent to

3.5 μg kg−1 of body weight per day) for all human groups on the basis that lead is a

cumulative poison, and Ni was updated by WHO, in 2005, as 12 μg Ni kg−1 of body weight

(Bilandzic et al., 2011, WHO, 1995 and WHO, 2005). As a result, allowable Pb concentrations

in beverages, particularly, step down to the lower values, in recent times. Regarding these

results, the allowable Pb concentration in milk samples was established as 20 μg L−1 by

authorised committee of Turkish (Turkish Food Codex, 2010). Among lead and nickel sources

in daily life, beverages including milk are very important sources. In case of lead, the major

protein fractions of milk alpha-casein interaction with Pb(II) through cysteine and sulfhydryle

are of strong binding sites as reported by Srinivas, Kaul, and Prakash (2007). On the other

hand, the most affected group from lead toxicity is children because of their incomplete

development. In view of this point, milk is another important food which should be checked

against the determination of toxic metal concentrations, constantly. Briefly, there is a high

interested to determine those metals in milk, yoghurt and similar food samples (Chakraborti

et al., 1987, Linsinger, 2005, Maas et al., 2011, Chakraborti et al., 1987 and Orun et al.,

2011).

Due to the allowable low Pb and Ni concentrations in milk samples as well as their low

concentrations in the matrices described above, reliable and sensitive analytical methods

are required for the determinations of these elements. For this purpose, inductively coupled

plasma-atomic emission spectrometry (ICP-AES), electrothermal atomic absorption

spectrometry (ETAAS), voltammetry and inductively coupled plasma-mass spectrometry

(ICP-MS) have been mostly used. Although ETAAS is a more sensitive technique than the

flame atomic absorption spectrometry (FAAS), ETAAS has some disadvantages such as

slowness of procedure, greater proneness to interferences and high cost (Pereira and

Arruda, 2003 and Yaman, 2005). For example, a period of 2 or 3 min is sometimes necessary

for one measurement by ETAAS against only a few seconds are sufficient for FAAS. In

addition, FAAS is more economical, and does not require expert operators. On the other

hand, direct determinations of trace level of heavy metals in environmental and food

samples by FAAS are difficult due to low sensitivity of this instrument. Some isolation and/or

preconcentration procedures prior to the determination step can be used to overcome these

limitations (Kaya et al., 2008b and Yaman, 1998). Due to severe interference effects coming

from the matrix such as salinity, preconcentration/separation procedures for Pb and Ni

determinations in water samples have been used to minimise the matrix effects before

ETAAS and ICP-MS measurements (Chakraborti et al., 1987,Rivas et al., 2009 and Sekhar et

al., 2003). Among preconcentration methods, solid-phase extraction (SPE) has been found a

great attention in recent years (Kaya et al., 2008b and Sekhar et al., 2003). Low amount of

waste generated, high potential for the re-use of solid phase, low matrix effect, high

preconcentration factor and no need to use of toxic solvent are the advantages of SPE when

compared with the other separation and preconcentration methods (Karaaslan et al.,

2010, Senkal et al., 2007,Yaman, 1998, Yaman, 2000 and Yaman, 2001).

It can be said that some of the preconcentration methods are not practical to be used due to

very long preconcentration periods up to 24 h (Venkatesh et al., 2005 and Yavuz et al.,

2009). Rapid and simple preconcentration procedure without using any complexing reagent

can be achieved by using proper adsorbent resin. For this purpose, the use of polymer-

bonded ligands in preconcentration of metals has been the subject of many research articles

(Karaaslan et al., 2010 and Senkal et al., 2007). Finally, slotted tube atom trap (STAT) has

been used to enhance the sensitivity of FAAS for some elements (Kaya and Yaman,

2008 and Yaman, 2005).

In this study, an aminothioazole-containing sulfonamide resin derived from crosslinked

chlorosulfonated polystyrene (CCPS) was synthesised and characterised. Then, experimental

parameters were optimised for the separation and preconcentration of Pb and Ni at

ng mL−1 level by using batch method. The method is based on the SPE of Pb and Ni ions on

the synthesised resin without using any complexing reagent in preconcentration step.

Another purpose of this study was to figure out the lead and nickel contents of milk and

yoghurt samples consumed in Elazig city (with population of 320 000) of Turkey. Due to the

lack of relevant data in this region, the study was focused on the determinations of lead and

nickel in the mentioned beverages. For this purpose, the optimised method was applied to

the digested milk and yoghurt samples for determinations of their Pb and Ni contents. After

the preconcentration procedure, STAT-FAAS was used for improving in the sensitivity of

FAAS for Pb because any increase in the sensitivity of FAAS for Ni was not observed. To

check the reliability of the results, the certified reference material (BCR 151 Skim Milk

Powder) was analysed. Further, the recoveries of Pb and Ni from milk samples fortified with

these metals were found, and some milk samples were also analysed by ICP-MS for their Pb

and Ni contents to the validation the method.

2. Experimental

2.1. Apparatus and reagents

ATI UNICAM Model 929 flame atomic absorption spectrophotometer equipped with ATI

UNICAM hollow cathode lamp (HCl) was used for the determination of Pb and Ni (Unicam

Ltd., Cambridge, England). The optimum conditions for Pb and Ni by FAAS were applied as

follow: wavelength: 217.0 and 232.0 nm, respectively; HCL current: 7.5 mA for both

elements; acetylene and air flow rates: 0.5 and 4.0 L min−1 for both elements; slit width: 0.5

and 0.2 nm, respectively. The FT-IR spectroscopy was used for characterisation of the

synthesised resin. To assess the reliability of measurements, some samples were analysed

by Perkin–Elmer Elan 9000 ICP-MS (Perkin Elmer SCIEX, Concord, Ontario, Canada). The

operation conditions for this instrument were applied as recommended by the

manufacturers.

The STAT described in previous studies (Kaya and Yaman, 2008 and Yaman, 2005) was used

to enhance the Pb sensitivity by FAAS. The pH was measured with (SCHOTT Lab Star pH,

Mainz, Germany) pH meter. In the preconcentration procedure, magnetic stirrers (Velp

Scientific, Milano, Italy) and Hettich EBA III centrifuge (Tutlingen, Germany) were used.

All glass apparatus (Pyrex®) were kept permanently in full of 1.0 mol L−1 nitric acid when not

in use. Concentrated HNO3 (Merck, Darmstadt, Germany) was used for decomposition of

adsorbed Pb and Ni on the synthesised resin surface. Concentrated hydrogen peroxide

(Merck) and perchloric acid (Merck) together with nitric acid were used for digestion of milk

sample. The diluted standard solutions were prepared from 1000 mg L−1 of stock standard

solution of lead and nickel (Merck). Milk powder (BCR 151) as standard reference material

was used to check validation of results.

The buffer solutions in the pH range of 1.5–6.0 ± 0.1 were prepared by using

0.1 mol L−1 citric acid plus 0.1 mol L−1 HCl/0.1 mol L−1 NaOH solutions. The synthesised resin

was ground and sieved by using a 325-mesh sieve.

ATI UNICAM Modelul 929 Spectrofotometru de absorbție atomică in flacara ATI UNICAM Modelul 929 echipat cu lampă cu catod tubular (HCI) ATI UNICAM a fost utilizat pentru determinarea Pb și Ni (Unicam Ltd., Cambridge, Anglia). Condițiile optime pentru Pb și Ni prin FAAS au fost aplicate după cum urmează: lungime de undă: 217,0 și respective 232,0 nm,;HCL curent: 7.5 mA pentru ambele elemente; acetilena și rata debitului

de aer : 0.5 și 4.0 L min-1 pentru ambele elemente; Lățimea fantei: 0,5 și respective 0,2 nm,.Spectroscopia FT-IR a fost utilizat pentru caracterizarea rășinii sintetizate. Pentru a evalua fiabilitatea măsurătorilor, unele probe au fost analizate prin Perkin-Elmer Elan 9000 ICP-MS (Perkin Elmer SCIEX, Concord, Ontario, Canada). Au fost aplicate condițiile de funcționare pentru acest instrument, conform recomandărilor de către producători. STAT-ul descris în studiile anterioare (Kaya și Yaman, 2008 și Yaman, 2005) a fost utilizat pentru a spori sensibilitatea Pb de FAAS. PH-ul a fost măsurată cu (SCHOTT Lab Steaua pH, Mainz, Germania) pH-metru. În cadrul procedurii de preconcentrare, au fost utilizate agitatoare magnetice (Velp științific, Milano, Italia) și centrifuga Hettich EBA III (Tutlingen, Germania).Toate aparat de sticlă (Pyrex®) au fost ținute în permanență în acid azotic de 1,0 mol L-1 acid azotic când nu au fost utilizate.  HNO3 concentrat (Merck, Darmstadt, Germania) a fost utilizat pentru descompunerea Pb si Niabsorbit pe suprafața rășinii sintetizate.Peroxid de hidrogen concentrat (Merck) și acid percloric (Merck), împreună cu acid azotic au fost utilizate pentru digestia eșantionelor de lapte. Soluțiile standard diluate au fost preparate din 1000 mg L-1 de soluție etalon stoc de plumb și nichel (Merck). Lapte praf (BCR 151) ca material standard de referință a fost folosit pentru a verifica validarea rezultatelor. Soluțiile tampon cu un pH cuprins de 1.5-6.0 ± 0,1 s-au preparat folosind 0,1 mol L-1 acid citric plus 0,1 mol L-1 HCl / 0,1 mol L-1 soluții de NaOH. Rășina sintetizat a fost măcinat și cernut folosind o sită de 325.

2.2. Collection and digestion of milk samples

The raw and processed cow’s milk samples were obtained from local farms, shops and

individual farmers. The samples were collected in polyethylene containers which were

washed in 1% nitric acid. A 10-mL sample of animal milk was placed in a flask and a mixture

of 5-mL of concentrated nitric acid and 5-mL of concentrated hydrogen peroxide was added.

After evaporation for 15 min at 100 °C, 2-mL of concentrated perchloric acid was also added

and digested at 150 °C with stirring, until a clear digest was obtained (approximately 1 h).

After evaporation of reuses up to 5-mL, the clear digest was diluted to 125 mL with water.

The blank digests were carried out in the same way. Skim Milk Powder (BCR 151) was

digested as described elsewhere (Kaya, Akdeniz, & Yaman, 2008a).

2.3. Collection and digestion of yoghurt samples

The yoghurt samples produced from raw and processed milk samples were obtained from

local farms and shops. The samples were collected in polyethylene containers which were

washed in 1% nitric acid. A digestion procedure modified from Kaya et al. (2008a) was

applied to yoghurt samples. A 10-g sample of yoghurt sample was placed in a flask and a

mixture of 5-mL of concentrated nitric acid and 5-mL of concentrated hydrogen peroxide

was added. After evaporation for 15 min at 100 °C, 2-mL of concentrated perchloric acid was

also added and digested at 150 °C with stirring, until a clear digest was obtained

(approximately 1 h). After evaporation of reuses up to 5-mL, the clear digest was diluted to

125 mL with water. The blank digests were carried out in the same way.

Colectarea și digestia probelor iaurt Probele de iaurt, produse din probele de lapte crud și prelucrate au fost obținute de la ferme și magazine locale. Probele au fost colectate în recipiente de polietilenă care au fost spălate în 1% acid azotic. O procedură de dizolvare modificat de la Kaya și colab. (2008a) a fost aplicata la probele de iaurt. O probă de 10 g de probă iaurt a fost plasată într-un balon și a fost adăugat un amestec de 5-ml de acid azotic concentrat și 5 ml de peroxid de hidrogen concentrat. După evaporare la 100 ° C timp de 15 minute, 2-mL de acid percloric concentrat a fost de asemenea adăugat și dizolvat la 150 ° C cu agitare, până când o dizolare clară a fost obținută (aproximativ 1 h). După evaporarea reutilizări până la 5 ml, limpede digest a fost diluată la 125 ml cu apă. Produsele de dizolvare goale au fost realizate în același mod.

2.4. Preparation of the aminothioazole-containing resin

The polymer beads were prepared by the suspension polymerisation of a mixture of styrene

(54 mL, 0.48 mol) and DVB (55% grade, 10 mL, 0.038 mol) in toluene (60 mL), using gum-

Arabic as stabiliser, according to a previously described procedure (Yavuz et al., 2009). The

beads were sieved and the 420–590 μm size fractions were used for further reactions.

The beaded polymer was chlorosulfonated using chlorosulfonic acid as described elsewhere

(Senkal et al., 2007). The degree of chlorosulfonation was determined by analysis of the

liberation of chloride ions. For these purpose, a polymer (0.2 g.) sample was added to 10%

NaOH (20 ml) and boiled for 4 h. After filtration and neutralisation with HNO3 (5 M), the

chlorine content was determined by the mercuric-thiocyanate method. This gave a final

chlorosulfonation degree of 4 mmol g−1.

Chlorosulfonated resin (10 g) was added portion wise to a stirred solution of 6 g (0.060 mol)

of 2-aminothiazole in 50 ml of 2-methyl-1-pyrrolidone (NMP) at 0 °C. The mixture was

shaken with a continuous shaker for 24 h at room temperature. The reaction content was

poured into water (0.5 L), filtered and washed with excess amounts of water and acetone,

respectively. The resin was dried under vacuum at room temperature for 24 h. The yield was

12 g. The reaction mechanism was given in Fig. 1.

Fig. 1. 

The scheme of the reaction mechanism in the synthesis of resin.

Figure options

2.5. Batch preconcentration procedure

The adsorbent resin of 30 mg was added to 60 mL of the model solution containing

20 ng mL−1 of Pb and 40 ng mL−1 of Ni in flask including the matrix components as mg L−1; Ca,

100; Mg, 50; and Fe, 0.5, to represent natural water. This solution containing matrix

components was described as the “model solution” in the subsequent parts. After pH

adjustment to near the desired value using diluted HCl and NaOH (optimum pH = 3.5 ± 0.1),

10.0 mL of buffer solution was added to this mixture. Then, the pH of this mixture was again

adjusted to the studied pH, if necessary. The mixture was mechanically stirred for 30 min at

room temperature and centrifuged. The residue was dried at 50 °C in an oven. After

transferring the residue to a glass beaker by scraping, concentrated HNO3 of 5.0 mL was

added, and then the mixture was evaporated to near the dryness. After cooling, 3.0 mL of

1.0 mol L−1 HNO3 was added, mixed, and centrifuged to take liberated metal to the solution.

The clear solution was measured by using STAT-FAAS for Pb and FAAS for Ni. The blank

solutions were carried out in the same way.One-way Analysis of Variance (ANOVA) was used

for the statistical evaluation of the results.

3. Results and discussion

3.1. The characterisation of new polymeric resin

The major disadvantage of using polymeric resins to remove or preconcentrate toxic metals

from solutions is their hydrolysis by acid and bases in the regeneration step. Sulfonamides

have chemical properties differently from carboxylic amides in that they have reasonable

stability towards acid and base hydrolysis. Therefore, there is no such disadvantage in the

synthesised resin in this study because the sulphonamide group in the resin has resistance

to hydrolysis by acid and bases.

The resin was characterised by FT-IR spectroscopy (Fig. 2). In the sulfonamide resin, N–H

stretching vibration occurs at 3400 cm−1 and S O stretching vibration occurs at 1320 and

1144 cm−1, respectively. The other bands of thiazole group were observed at 1451 cm−1 for

(C N) and 668 cm−1 for (C–S) stretching frequencies.After regeneration process by using

concentrated HNO3 as described in section 2.5, these peaks were observed at the same

values (Fig. 2).

Fig. 2. 

FT-IR spectra of aminothioazole containing sulfonamide resin before and after preconcentration procedure.

The line that has the lower transmittance value is before preconcentration procedure.

Figure options

3.2. Optimisation of the method

All of the parameters that may affect the recoveries in the preconcentration step of the

analytical scheme were optimised. These parameters were investigated by using “Model

Solutions” described above. In the optimisation procedure, other parameters were kept at

optimum values while one parameter was changing for optimisation. The studies on

investigation of each parameter were repeated three times. The pH of the solution and the

contact time are of the most important factors controlling the extractability of metal ions.

The effects of these parameters on recoveries of lead and nickel were examined by batch

method.

3.2.1. Effect of pH on recovery

It is known that hydrogen-ion concentration plays an important role in the preconcentration

of metals by adsorption on the resin because it significantly affects the binding of metal. The

pH values of the “model solutions” described above were adjusted to appropriate pH in the

range of 1.5–6, and stirred for 30 min, separately. The obtained results were given in Fig. 3.

The recoveries of Pb and Ni were found to be 90% and 93%, respectively, in the range of pH

3.0–4.0. It was observed that matrix components were separated from lead and nickel in the

analyte solutions in these pH ranges. Therefore, the pH of solution was selected as 3.5 ± 0.1

in all subsequent studies.

Fig. 3. 

Effect of pH on the recovery.

Figure options

Surface of the synthesised resin could be positively charged in strong acid solutions

(pH ⩽ 2.5) due to the protonation of N atoms, which cannot bind with metal cations.

Because hydroxide compounds of metals in matrix solution can precipitate at high pH

values, lower pH values are considered to be more appropriate for preconcentration step.

3.2.2. Effect of the preconcentration period on recovery

In the optimisation of this parameter, different preconcentration periods (contact times)

were applied to the “model solutions”. The results showed that the preconcentration period

of 30 min was sufficient for maximum recoveries (up to 90% and 93% for Pb and Ni,

respectively). The obtained contact time is shorter than that of other SPE methods reported

(up to 24 h) in literature (Yavuz et al., 2009). The fast extraction rate indicates that the

sorbent is highly suitable for the preconcentration of trace lead and nickel in aqueous

solution. Further, the shorter extraction period indicates that Pb and Ni ions have a good

accessibility through the chelating sites on the newly synthesised resin, and the binding

constants between the lead and nickel ions and aminothioazole immobilised on the

sulfonamide resin surface are sufficiently high. As a result, 30 min was selected as optimum

contact time and applied throughout the study.

3.2.3. Effect of the buffer volume on recovery

Buffer volume was also optimised to find out the effect of buffer volume on recovery. The

results showed that no obvious variation took place in the extraction yields when more

volume than 10.0 mL of buffer solution was added to 60.0 mL of the sample volume. Hence,

10.0 mL aliquot of buffer solution was added into the mixture in all subsequent experiments,

and this volume was proportionally increased with the sample volume.

3.2.4. Effect of eluent acid concentration and volume on recovery

Acid solutions are widely used for the elution of metal ions from a sorbent due to the

protonation at a chelating site of the sorbent. Among acids, nitric acid replaces the metal

ions from binding sites, and hence is commonly used. Moreover, this acid does not interfere

in the subsequent determination step by FAAS. As a result, nitric acid was chosen as a

striping or back-extraction solution. In addition, suitable acid concentration is another

important factor for elution. Dilute acid solutions (2 and 4 mol L−1) were not found to be

enough to complete protonation. On the other hand, high concentration may contaminate

the sample and cause the problem in determination step. So, the back-extraction procedure

was carried out in two steps. In the first step, concentrated nitric acid was used, and then

this acid was evaporated to near the dryness. In the second step, diluted nitric acid solutions

with various concentrations between 0.5–1.5 mL were used. Related with this step, it was

found that higher recoveries than 90% for Pb and Ni were observed by using

1.0 mol L−1 HNO3 in compared to 0.5 and 1.5 mol L−1 HNO3. The efficiency of the eluent

volume of 1.0 mol L−1 HNO3 was examined by taking its different volumes (1.5–3.0 mL). It

was observed that 2.5 mL of 1.0 mol L−1 HNO3 is sufficient for obtaining the maximum

recoveries (90% for Pb and 93% for Ni). In the first step, concentrated HCl was also used

instead of HNO3. Then, 3.0 mL of 1.0 mol L−1 HCl was examined to elution procedure. The

recoveries with 82% and 75% were observed for Pb and Ni, respectively. Consequently,

2.5 mL of 1.0 mol L−1 HNO3 in the second step was chosen after using concentrated nitric

acid in the first step to complete desorption of the adsorbed lead and nickel because

maximum recoveries were obtained.

3.2.5. Effect of initial volume of sample on recovery

An important parameter in SPE is the breakthrough volume (initial analyte volume) because

enrichment factor is depending on the volumes of analyte and volume of eluate. In other

words, the breakthrough volume is the maximum solution volume in which the analyte can

be retained within the acceptable recovery range. The optimum initial volume was

experimentally evaluated by using the model solution containing a constant concentration of

20 ng mL−1 of Pb and 40 ng mL−1 of Ni in different volumes of solutions varying from 25 to

125 mL. It was found that the recoveries were higher than 90% for all studied volumes in

ranges of 25–125 mL. So, the initial volume of 125 mL was used for all real samples.

3.3. Analytical performance

The calibration plots were observed to be linear at the concentration range of 1.0–20.0 ng Pb

mL−1 and 6.0–40.0 ng Ni mL−1, by taking of 60 mL initial solution to final volume of 2.5 mL,

due to using 60 ml of model solution in all optimisation steps. In addition, the calibration

curves were also obtained in the concentration ranges of 0.5–10.0 ng Pb mL−1 and 3–30 ng Ni

mL−1 by taking a 125-mL of initial solution to eluate volume of 2.5 mL because Pb and Ni

concentrations in the studied samples may be lower than 1.0 and 6.0 ng mL−1, respectively.

Hence, the enrichment factor of 50-times was achieved by using the optimised

preconcentration method. Due to a 7-times sensitivity improvement for Pb by using STAT

(Kaya & Yaman, 2008), a 350-times improving in sensitivity was achieved by using the total

analytical scheme. The obtained calibration plots were linear in the concentration ranges

described above, and the equations of the plots were as follows:

Turn MathJaxon

(0.5–10.0 ng Pb mL−1 after preconcentration-STAT-FAAS)

Turn MathJaxon

(3.0–30.0 ng Ni mL−1 after preconcentration-FAAS)

Relative standard deviations (RSD) were found to be 9% for 4.0 ng Pb mL−1 and 10% for

20 ng Ni mL−1using 10 replicates of preconcentration procedures. In the statistical method,

the limit of detection (LOD) is defined as three times of the mean values of standard

deviation (SD) of blanks. As for limit of quantification (LOQ), it is defined as ten times of the

mean SD values of blanks. In this study, the levels of Pb and Ni in blank samples were found

to be 0.15 and 0.6 ng mL−1 with the SDs of 0.05 and 0.25 ng mL−1, respectively. Therefore,

the LODs defined as three times of the SD of the reagent blank were found to be 0.15 and

0.75 ng mL−1, respectively. Furthermore, LOQs were found to be 0.5 and 2.5 ng mL−1 for lead

and nickel, respectively. The validation of the method was checked by examining the

certified reference material BCR 151 Skim Milk Powder. The results are given in Table 1. It

can be seen that the recovery values were 97% for Pb. For assessing the method

performance, the measured values of the CRMs are compared with the certified values

(from Table 1) following a procedure described by Linsinger (2005). The absolute difference

(Δm) between the mean measured value (cm) and the certified value (ccrm) was calculated as

Δm=1.992-2.002=-0.01Turn MathJaxon

The uncertainty of Δm is uΔ, was calculated from the uncertainty of the certified value (ucrm)

and the uncertainty of the measurement (um) result (from Table 1) as

Turn MathJaxon

The expanded uncertainty UΔ, corresponding to a confidence level of approximately 95%,

was obtained by multiplication of uΔ by a coverage factor (k) that is usually equal to 2 as

U Δ=2× u Δ=2x0.030=0.060.Turn MathJaxon

Table 1.

Mean Pb and Ni concentrations (ng mL−1) determined for milk and yoghurt samples in this study. The

values are mean ± SD, n = 3.Sample type Pb Ni

Raw milk from Farm 1 15 ± 2 bdl⁎

Raw milk from Farm 2 19 ± 2 bdl⁎

Raw milk from Farm 3 27 ± 3 bdl⁎

Raw milk from Farm 4 35 ± 4 bdl⁎

Packed Milk 5 29 ± 3 bdl⁎

Packed Milk 6 42 ± 5 bdl⁎

Packed Milk 7 37 ± 4 bdl⁎

Packed Milk 8 61 ± 6 bdl⁎

Yoghurt 1 22 ± 4(ng g−1) bdl⁎

Yoghurt 2 26 ± 5(ng g−1) bdl⁎

Yoghurt 3 21 ± 4(ng g−1) bdl⁎

Yoghurt 4 32 ± 5(ng g−1) bdl⁎

Yoghurt 5 27 ± 4(ng g−1) bdl⁎

Yoghurt 6 42 ± 6(ng g−1) bdl⁎

Raw milk from farm 1 + 10 ng mL−1 Pb and Ni

24 ± 3Recovery: 93%

9.0 ± 1.0Recovery: 90%

CRM milk powder (BCR 151) Certified: 2.002 ± 0.026 ng/gFound: 1.992 ± 0.015Recovery: 99%

–

Raw milk from Farm 1 This method: 15 ± 2⁎⁎

ICP-MS: 14 ± 2Recovery: 93%

This method: bdl⁎

ICP-MS: 8 ± 0.7Recovery: –

⁎

bdl: Below detection limit.

⁎⁎

p < 0.05 With respect to the ICP-MS results.

Table options

After comparing Δm (−0.01) with UΔ (0.06) (because of Δm < UΔ), it was concluded that there is

no significant difference between the measurement result and the certified value, at a

confidence level of about 95%.

In addition, the accuracy of the method was also studied by examining the recoveries of Pb

and Ni from milk samples fortified with these metals. The results in Table 1 show that, at

least, up to 95% of the lead and nickel added to these samples were recovered. Finally,

some samples were also analysed by ICP-MS to determine the accuracy of results. It was

found that the differences between the results obtained by the developed preconcentration

method and ICP-MS were not meaningful (Table 1). In other words, the differences were

originated from unknown sources. For this purpose, One-way Analysis of Variance (ANOVA)

is conducted to test the equality of group means. One of the pairwise comparisons test,

Tukey HSD, was carried out to find which of group means are different from each other.

SPSS (version: 13) statistical program was used for all statistical computations. Statistical

findings are given in Table 1.

3.4. Effect of interfering ions on recovery

In view of the high selectivity provided by FAAS, the interference effects of some cations and

anions were examined using the preconcentration procedure. The hypothesis was initially

that the interferents might decrease the extraction efficiency of the analyte. The interfering

cations and anions were selected depending on their major natural abundance in milk and

yoghurt. A binary mixture of 125 mL of Pb and Ni, and each interfering ion, i.e., Na+, K+, Mg2+,

Ca2+, Cl− and SO42− in three concentration levels (100, 200 and 500 mg L−1) was extracted by

the batch method. Nitrate or sodium forms were used as ionic salts of compounds. From the

results obtained, the overall percent recoveries of Pb and Ni were found to be higher than

90%. These results indicated that the ions mentioned above do not interfere the Pb and Ni

determination at the optimised conditions.

3.5. Applications to real samples

All of the optimum values determined were applied in the preconcentration procedure for

milk and yoghurt samples prepared/digested as described in experimental section. After

applying the optimised method, the results obtained were given in Table 1. The values given

are the mean of three different portions of the same sample. Concentrations of Pb in the

milk samples were found to be in range of 15–35 ng mL−1 for raw milk and 29–61 ng mL−1 for

commercially packed milk. Nickel concentrations in all studied milk samples could not be

detected due to its levels lower than the LOQ. The obtained lead concentrations except two

for raw milk are higher than the maximum allowable levels (20 ng Pb mL−1) restricted by

interested authority in our country (Turkish Food Codex, 2010). Lead and Ni concentrations

in yoghurt samples were found to be in the range of 21–42 ng mL−1 and bdl,

respectively. Venkatesh et al. (2005) applied a solid phase extraction method by using 2,3-

hydroxypyridine loaded Amberlite XAD-2 for preconcentration of trace metal ions in water,

milk and vitamin samples. Shan, Tie, and Xie (1988) determined trace elements in water,

sea-water and biological samples by ICP-AES after preconcentration with ammonium

pyrrolidinedithiocarbamate precipitation. The observed optimum conditions in this study are

found better than in those studies regarding time, pH and sensitivity. Related with reported

values for Pb and Ni concentrations in animal milk and yoghurt in the literature, Ayar, Sert,

and Akin (2009) found 103 ng Pb mL−1 for Cow’s milk and 93 ng Pb mL−1 for Cow’s yoghurt,

by using ICPAES (Ayar et al., 2009). Qin, Wang, Li, Tong, and Tong (2009) and Bilandzic et al.

(2011) determined lead concentrations in ranges of 12–35 ng mL−1 for different kinds of milk

and 36–59 ng mL−1 for Cow’s milk by using GFAAS, respectively (Qin et al.,

2009 and Bilandzic et al., 2011). Giri, Singh, Jha, and Tripathi (2011) found 480 ng Ni mL−1 in

Cow’s milk taken from U mining area while lead was below detection limit, by using AAS (Giri

et al., 2011). Al-Othman (2010)) reported Pb concentrations in fresh milk and yoghurt

samples as 6.0 and 25.0 ng mL−1using ICPAES, respectively (Al-Othman, 2010). Enb, Abou

Donia, Abd-Rabou, Abou-Arab, and El-Senaity (2009) found Pb and Ni concentrations in

ranges of 40–62 and 2–3 ng mL−1 in milk and 39–60 and 1–2 ng mL−1 in yoghurt samples,

respectively (Enb, Abou Donia, Abd-Rabou, Abou-Arab, and El-Senaity, 2009). The other

reported levels of Pb and Ni are as follows: 11–244 ng Pb mL−1 in Cow’s milk (De Castro et

al., 2010), 43 ng Pb mL−1 and 20 ng Ni mL−1 in Cow’s milk (Kazi et al., 2009), 18 ng Pb mL−1 in

Cow’s milk (Yaman, 1998), 9–11 ng Pb mL−1 in Cow’s milk using ICP-MS (Frazzoli & Bocca,

2008) and 19–126 ng Pb mL−1 in yoghurt (Kaya et al., 2008a). From those results, Pb and Ni

levels in milk and yoghurt samples change in the range of below detection limit (bdl)-244 μg

Pb L−1 and 1–480 μg Ni L−1 for milk samples, and 19–126 μg Pb kg−1 and 1–2 μg Ni kg−1 for

yoghurt samples, respectively. The observed Pb and Ni concentrations in this study are in

the ranges of those results.

4. Conclusions

The optimised preconcentration method in this study has some advantages than the other

reported methods in this field in view of sensitivity and no need to use of ligand at the

preconcentration step because the ligand is imprinted to the resin at the beginning of the

analytical procedure. The developed method is applicable for the determination of ultratrace

amounts of Pb and Ni in a variety of milk and yoghurt samples with low LOD, high accuracy

(recovery >95%) and high precision (RSD <9%). The optimised method together with the

STAT has high sensitivity, selectivity and reliability for Pb. The sensitivity of FAAS was

increased up to 50-times by using the optimised SPE conditions. A 350-fold improvement in

the sensitivity of Pb by FAAS was achieved by the combination of STAT and preconcentration

method because the STAT method can reduce the potential interferences by dilution. So, the

ng mL−1 levels of Pb and Ni in both milk and yoghurt samples could be determined by this

method. It can be concluded that the developed preconcentration method combined with

STAT-FAAS can be compete with ET-AAS method taking into consideration the sensitivity,

interferences and cost. The analysis procedure of this method is not complex compared to

the previous similar works taking into consideration LOD, operation conditions such as

column procedure, reagent species and resin amount or the required time. This method can

also be applied to more complex samples such as soil and biological samples. The obtained

Pb levels in most of the milk samples were higher than the restricted concentrations by the

authorised committees. Although Ni concentrations in the studied milk and yoghurt samples

were found to be lower than the LOQ and the maximum allowable levels, the developed

method can be applied to nickel analysis in different types of samples.

Finally, it was concluded that the synthesised resin can be examined to remediation of

polluted environmental matrices such as soil and water by using more than one use. The IR

spectra have supported this conclusion.