Determination of Cu , Ni and Pb in Aqueous Medium by FAAS After Pre-Concentration on 2-Aminothiazole Modified Silica Gel

Este trabalho descreve a síntese e caracterização da sílica gel modificada com grupos 2-aminotiazol (SiAT), e os resultados de um estudo de adsorção e pré-concentração (em batelada e em fluxo, utilizando-se um sistema por injeção em fluxo acoplado com um espectrômetro de absorção atômica) de Cu(II), Ni(II) e Pb(II) em meio aquoso. A capacidade máxima de adsorção da SiAT determinada para os íons metálicos estudados em mmol g foram: Cu(II)= 1,18, Ni(II)= 1,15 and Pb(II)= 1,10. Os resultados obtidos nos experimentos em fluxo, mostraram uma recuperação de praticamente 100% dos íons metálicos adsorvidos em uma mini-coluna empacotada com 100 mg de SiAT, utilizando-se 100 μL de HCl 2 mol L como eluente. A sorção-desorção quantitativa dos íons Cu(II), Ni(II) e Pb(II), permitiu a aplicação do sistema em fluxo na pré-concentração e posterior quantificação desses íons metálicos em amostras de águas naturais digeridas e não digeridas por oxidação com radiação UV.


Introduction
The toxic effects of heavy metals (e.g., Cd, Cr, Hg, Pb) have been demonstrated in certain "at-risk" humans. 1,2ince the main sources of these metals for humans are water and food.It is of paramount importance to control the metal content in natural water.For the direct determination of heavy metals in aquatic environments a number of sensitive instrumental methods (e.g., atomic spectroscopy) are available, which, however, can suffer from systematic interferences by water constituents.For this reason, a preconcentration step is often necessary before quantifying these analytes. 3][9][10][11][12][13] In such combined method the enrichment of analyte and removal of interfering matrix constituents, can considerably improve the method of analysis extending the limit of detection to lower concentration.
This paper describes the preparation of silica gel chemically modified with 2-aminothiazole aiming to find an efficient material for separation and chemical analysis of the metal ions present in natural water samples.Primarily the material was tested with a synthetic aqueous solution containing some metal ions and further it was used to analyse real samples.

Reagents, analytical solutions and samples
All solutions were prepared with high purity chemical and distilled-deionized water (Milli-Q system, Millipore).Hydrochloric and nitric acids used were Suprapur (Merck).
Samples collected from natural water (Tietê River, Paranapanema River; Paraná River, SP/Brazil.) were immediately filtered through a 0.45 µm membrane.After acidification (pH < 2.0) with 6.0 mol L -1 HNO 3 solution, they were stored in highly purified polyethylene containers.The river water samples were digested by oxidizing UV-photolysis in the presence of H 2 O 2 10 mol L -1 using a low-pressure Hg-lamp which was installed into a closed quartz vessel. 14eparation of modified silica gel 15 Silica gel (Merck) with specific surface area of 500 m 2 g -1 and average pore diameter of 0.6 nm, was activated at 420 K under vacuum (0.133 Pa).About 50 g of this silica was immersed in 200 mL of purified dimethylformamide and 12 mL of phosphorus oxychloride was added.The mixture was refluxed under nitrogen atmosphere for 36 h, filtered, washed with dimethylformamide, ethanol and heated under vacuum in order to eliminate all the solvent.The resulting solid was immersed in 135 mL of purified dimethylformamide and 18 g of 2-aminothiazole was added.The mixture was stirred for 36 h at 380 K under nitrogen atmosphere.The resulting modified silica was filtered off, washed with dimethylformamide, ethanol and heated for 8 h at 348 K under vacuum (0.133 Pa).
The equations 1 and 2 describe the preparation of the material: The quantity of 2-aminothiazole attached to the silica surface was determined by the nitrogen analysis using the Kjeldhal method. 13The specific surface area was determined by the BET method, starting from the nitrogen adsorption isotherms at 77 K in automated Micromeritics ASAP 2010 equipment.

Infrared spectra
Diffuse reflectance infrared Fourier Transform (DRIFT) spectra were obtained with a Nicolet spectrometer model Nexus 670 with Smart Collector, in Kubelka-Munk units, with 1mg of material mixed with 500 mg of KBr.

Adsorption of the metal ions by SiAT
Adsorption of MX n(aq) by SiAT from a solution can be described by the equilibrium: were mSiAT represents the mass of the modified silica gel and MXn the concentration of metal ions in the aqueous solution.
The time required to this reaction achieves the equilibrium condition was previously determined immersing 100 mg of SiAT in 50 ml of 5x10 -3 mol L -1 of the metal solution and shaking.At different time intervals, an aliquot of the supernatant solution was separated and the concentration of metal ions was determined by complexometric titration using Na 2 EDTA.
The quantity of the adsorbed metal, N f , was calculated applying the equation 4: (4)   where N i represents the initial mole number of the metal ion in the solution phase, N s the mole number of the metal ion in equilibrium with the solid phase and m is the mass of the adsorbent.J. Braz.Chem.Soc.

Effect of pH
The effect of pH on the adsorption of metal ion on SiAT was evaluated by varying the pH over range of 1.0-5.0 with addition of 0.10 mol L -1 HNO 3 .

Isotherms of adsorption
The adsorption capacity of SiAT was determined at 298 K using the batch technique.In 50 mL of aqueous solutions of the metal ions (concentrations between 2x10 -5 and 5x10 -3 mol L -1 ), about 100 mg of SiAT were added and the resulting mixture was shaken for 10 min.The solid phase was separated by centrifugation and the metal ion was determined in the supernatant solution by complexometric titration.

Apparatus
A SHIMADZU Model AA-6800 atomic absorption spectrometer equipped with a copper, nickel and lead hollow cathode lamps and deuterium lamp for background correction was used for the determination of metals.The operating conditions adjusted in the spectrometer were carried out according to the standard guidelines of the manufacturers.The atomic absorption signal was measured as a peak height mode against an analytical curve.
Ismatec Model IPC-8 peristaltic pump furnished with tygon pumping tubes (i.d.0.5 mm) and homemade injectorcommutator of perspex were used in the flow injection system.

Flow injection system
The flow diagram of the system with the SiAT packed mini-column is shown in Figure 1, consisted of a peristaltic pump, an injection valve (homemade injector-commutator) and a mini-column (glass tubing, 5.5 cm length and 3 mm i.d.) packed with 100 mg of SiAT.Standard solutions or water samples in the 40 -200 µg L -1 Cu(II), Ni(II) and Pb(II) concentration range were passed through the column at 1.0 mL min -1 for 5 min for the sorption of metal ions in the column.The metal ions were eluted with 100 µL of 2.0 mol L -1 HCl solution, added through the injection valve at 1.0 mL min -1 .The metal ions eluted were transported to the burner of spectrometer and the signal was recorded as a peak height.

Characteristics of material
The infrared spectra confirms the presence of 2aminothiazole groups bounded to the silica surface.The absorption band observed at 2784 cm -1 is due to the C-H stretching of aromatic olefins.The other two bands of interest observed at 1471 and 1395 cm -1 , can be assigned to the δ C=N and δ N-H , both dislocated from low frequencies due to strong absorption at 1630 cm -1 of δ Si-O of the remaining groups of silica gel. 15he chemical analysis of SiAT yield a 1.24±0.07mmol g -1 of the functional groups attached to silica surface and 354±0.6 m 2 g -1 of the specific surface area.The attached functional groups were very stable under the various cycles of adsorption-elution of the metal ions by the adsorbent in a column.Losses of the adsorption capacity for individual metal ions were not detectable during the experiments.

Adsorption isotherms
An important aspect of this material is the time necessary for the adsorption process to achieve an equilibrium condition. 9,13Figure 2 shows the plot of N f against time for Cu (II) as an example.The system achieves  equilibrium condition very fast, about 10 min, because in the present case, the functionalization occurs on the matrix surface having sufficiently large pores (0.6 nm).
The influence of the pH value on pre-concentration by SiAT is plotted in Figure 3, showing that the exchange of metal ions strongly decreases with the acidity of the solution.Obviously, a pH value 5.0 is optimal for the adsorption of metal ions.Exchange experiments at pH values higher than 5.0 were not carried out considering the risk of hydrolysis. 16e adsorption capacity for each metal ion, determined from saturation condition of the isotherms shown in Figure 4, in mmol g -1 was, Cu = 1.18,Ni = 1.15 and Pb = 1.10.The fraction of surface-attached ligand bonded to metal ion is given by equation 5: where N o is the amount of organofunctional groups (mmol) per mass of SiAT (g).The values of φ max.for each metal were, Cu = 0.95, Ni = 0.93 and Pb = 0.90.As φ max.approximately the same for all metal ions, it is possible to infer that the complexes formed is presumably of the metal:ligand = 1:1 type.This assumption is reasonable taking into account that the coordination of metal could be done in the following ways: (a) at the nitrogen atoms; (b) at the thiazole sulphur atom.

Effect of pre-concentration and elution flow rates of metal ions
The pre-concentration and elution flow rates were varied from 0.5 to 5 mL min -1 , while keeping the amount of injected Cu (II) ions at 100 µg L -1 .As shown in Figure 5 (graph a) an optimum peak absorbance was obtained at flow rate about 1.5 mL min -1 .In the same Figure (graph b), it was observed that the values of absorbance increased with increasing elution flow rate, remaining constant from 2.0 mL min -1 .Considering the best values of absorbance, the pre-concentration and elution flow rates selected were 1.0 and 2.0 mL min -1 , respectively.As a compromise between loading capacities and eluent volumes, different column lengths and inner diameters were tested and the dimensions selected were 5.5 cm and 3 mm, respectively.

Analytical performance
A linear calibration curve was obtained for the 40 -200 µg L -1 range of Cu(II) using sampling times of 3, 4, and 5 min (volumes injected in triplicate).Accordingly, the calibration slopes increased, but not proportionally, with the increase of sampling time, which means that the retention efficiency in not constant.The experimental pre-concentration factor was calculated as the ratio of slope of the calibration graphs obtained using the pre-concentration system proposed (A = -0.0046+ 0.0022[Cu(II)]; A = 0.0003 + 0.0011[Ni(II)]; A = -0.0051+ 0.0009[Pb(II)] ) and without pre-concentration (A = -0.0012+ 4.2x10 -5 [Cu(II)]; A = -4x10 -5 + 2.2x10 -5 [Ni(II)]; A = -1.1x10 - + 1,9x10 -5 [Pb(II)] ), using sampling time of 5 min.The pre-concentration factors calculated in this conditions were 52, 50 and 47, for Cu(II), Ni(II) and Pb(II), respectively.Differences between the preconcentration factor and the added and eluted volumes can occur due to loss of efficiency with increasing sample volume.As can be seen in Table 1, the pre-concentration efficiency declines from 99%-95% with the increase of sample volumes from 3 to 5 mL through the column.The pre-concentration factor under these conditions increase from 29 to 48.The precision expressed as RSD for twelve independent determinations, was 4% for 100 µg L -1 of metal ions solutions.The detection limits calculated for Cu(II), Ni(II) and Pb(II) in respect of blanks of elements 17 , were 0.30, 0.80 and 2.80 µg L -1 , respectively.The sampling frequency was fifteen samples per hour.
The reproducibility of the pre-concentration column expressed as SD, tested by sampling 100 µg L -1 Cu(II), Ni(II) and Pb(II) solutions for 5 min through twelve independently prepared mini-column, was about 98 ± 4 µg L -1 .

Determination of metal ions in river water samples
The proposed method was successfully applied recovery tests of Cu, Ni and Pb using river water samples spiked with 150 µg L -1 of Cu(II), Ni(II) and Pb(II).The results obtained with the proposed method are shown in Table 2. Accordingly, recoveries within 95%-99% for digested sample and 70%-75% for samples without previous digestion were found.This difference has been attributed to formation of inert complexes of metal ions with aquatic humic substances (HS), being less accessible for the functional groups of the collector. 18After digestion (by oxidizing UV-photolysis in the presence of H 2 O 2 ), the metal ions bound to HS as inert complexes can quantitatively react with functional groups of SiAT.

Conclusions
The proposed method offers a good alternative for copper, nickel and lead determination at trace levels in aqueous medium.The high stability of the mini-column packed with SiAT is evidenced by its utilization during five hours per day for one month.The proposed method allows on-line copper, nickel and lead determination by FAAS of natural water samples, however, requires a previous sample digestion, for instance by UV photolysis, necessary for elimination of the interference of aquatic organic matter, particularly aquatic humic substances, that can form inert complexes with metal ions.

Figure 2 .
Figure 2. Plots of Nf versus time from Cu(II) aqueous solution at 298 K.

Figure 4 .
Figure 4. Adsorption of metal ions from aqueous solution at 298 K.

Figure 3 .
Figure 3. Influence of pH on the adsorption of metal ions from aqueous solutions at 298 K.

Table 2 .
Determination of Cu, Ni and Pb in different water samples (spiked with 150 µg L -1 Cu(II)), Ni(II) and Pb(II) by FAAS with preconcentration on-line on a mini-column packed SiAT, with and without UV-Photolysis digestion.Experimental conditions: eluent -100 µL of 2 mol L -1 HCl solution; sampling time = 5 min with flow rate = 1.0 mL min -1 , n = 3)