doelhert niobium

8/3/2019 Doelhert Niobium

1/6

Introduction

Lead (Pb) is one of the most common environmentalcontaminants due to its wide use in petroleum, mining, paintand pigments, ceramics and weapons industries. This metal iscommonly found in many places where humans may beexposed to it. It is considered an element that has toxic effectson human and animals and has no physiological function inliving organisms. 1,2 The harmful effects of Pb have been knownfor a long time. Indeed, this metal affects practically all vitalorgans and systems of the human body. 3

Since heavy metals are in trace levels in many parts of theenvironment, suitable analytical methods are required for

quantitative determinations.4

In Brazil, the NationalEnvironment Council ( Conselho Nacional do Meio Ambiente ,CONAMA) regulates the maximum levels of contaminantspermitted in environmental samples. The maximum Pbconcentration allowed in treated water and in seawater wherefishing is carried out is 10.0 g L 1.5

Atomic absorption spectrometry (AAS) is a technique thatoffers simple, fast, precise and low cost analysis. 6 However,flame atomic absorption spectrometry (FAAS) has limitations interms of detectability for low concentrations, such as thoserequired by CONAMA. Furthermore, the complexity of environmental samples is a factor which makes the quantitationof analytes difficult. For these reasons, extraction andpreconcentration techniques have been employed. Solid phaseextraction, 6 liquid liquid extraction, 7,8 coprecipitation, 9,10 andcloud point extraction 11 are commonly applied sample

preparation techniques. Solid phase extraction (SPE) is widelyemployed due to the low generation of residues, highenrichment factor, easy accommodation of the solid phase in aminicolumn coupled to a flow preconcentration system and thefact that it, generally, does not require the use of toxicsolvents. 12

Pyrzy n ska and Cheregi 13 proposed a methodology using SPEfor lead determination in natural waters after preconcentrationon the sorbent Cellex P and reported a detection limit of 1.8 g1. Goswami and Singh 14 used 1,8-dihydroxyanthraquinoneimmobilized on silica gel for the preconcentration anddetermination of lead by FAAS in tap and river water samplesand achieved a detection limit of 0.45 g L 1. Other sorbentmaterials have also been used for solid phase extraction, such as

2-aminothiazole modified silica gel,15

Chromosorb-102,16

Amberlite XAD-2 and Amberlite XAD-7. 17

In analytical chemistry, methods proposed for thedetermination of contaminants in environmental samples areusually optimized using a univariate approach where eachvariable is optimized individually. This requires a great numberof experiments and considerable amounts of reagents and time,and does not take into account the interaction betweenvariables. 18 To overcome these disadvantages, researches haveused procedures involving multivariate techniques. 1820 Amongthe multivariate procedures available, factorial experimentaldesign has been used for the evaluation of the interaction effectsof the selected factors on the analytical response. The factorsthat have been shown to be statistically significant can then beoptimized through a response surface methodology. TheDoehlert matrix is a non-factorial response surfacemethodology, in which two or more factors can besimultaneously optimized. The response is fitted to a second-order polynomial, and the optimum values for the factors can be

365ANALYTICAL SCIENCES MARCH 2008, VOL. 242008 The Japan Society for Analytical Chemistry

Application of Factorial Design and Doehlert Matrix for

Determination of Trace Lead in Environmental Samplesby On-line Column Preconcentration FAAS Using SilicaGel Chemically Modified with Niobium(V) Oxide

Kalya Cravo Di Pietro R OUX , Heloisa Frana M ALTEZ , Jeferson Schneider C ARLETTO ,Edmar M ARTENDAL , and Eduardo C ARASEK

Departamento de Qumica, Universidade Federal de Santa Catarina, Florianpolis, 88040-900, Brazil

In this study a new method for Pb determination in water using solid phase extraction coupled to a flow injection systemand flame atomic absorption spectrometry was developed. The sorbent used for Pb preconcentration and extraction wassilica gel chemically modified with niobium(V) oxide. Flow and chemical variables of the system were optimizedthrough a multivariate procedure. The factors selected were buffer type, eluent concentration, and sample and eluent flowrates. It was verified that the aforementioned factors as well as their interactions were statistically significant at the 95%confidence level. The effect of foreign ions was evaluated using a fractionary factorial experimental design. Thedetection limit was 0.35 g L1 and the precision was 1.6%. Results for recovery tests using different environmentalsamples were between 90 and 104%. Certified reference materials were analyzed in order to check the accuracy of theproposed method.

(Received January 16, 2007; Accepted May 29, 2007; Published March 10, 2008)

To whom correspondence should be addressed.E-mail: [email protected]


2/6

estimated through this mathematical relationship. 21

The factorial design and Doehlert matrix have been used tooptimize preconcentration procedures for the determination of nickel in food samples 22 and lead in river waters 23 using,respectively, FAAS and inductively coupled plasma opticalemission spectrometry (ICP OES) as detection techniques.

In this paper, a two-level full factorial experimental designand Doehlert matrix were the multivariate tools used foroptimization of the factors affecting the preconcentrationsystem for Pb determination in aqueous samples using FAAS.This methodology was based on sorption of Pb(II) on thesorbent silica gel chemically modified with niobium(V) oxide.

Experimental

InstrumentationA Varian SpectrAA 50 (Victria, Australia) flame atomic

absorption spectrometer equipped with a deuterium lamp as a

corrector for non-atomic absorption was used. The lead hollowcathode lamp was manufactured by Hitachi (Mitorika, Ibaraki,Japan). Operational parameters of the spectrometer were:wavelength, 217 nm; lamp current, 5 mA; slit width, 1 nm. Theflame consisted of an oxidizing air acetylene flame. Theacetylene was atomic absorption-grade (White Martins, SoPaulo, Brazil).

A Mettler Toledo 320 pH meter was used to set the sampleand working solutions at the desired pH.

Two peristaltic pumps manufactured by Ismatec, MS-Reglo(Glattbrugg, Switzerland) with eight channels fitted withTygon tubes were used to pump the solutions through the flowsystem.

Reagents and solutionsThe working solutions were prepared with deionized water

taken from a Milli-Q purification system (Millipore , Bedford,MA, USA). All reagents were analytical grade. Nitric acid65% (v/v) (Merck, Darmstadt, Germany) was bi-distilled belowits boiling point in a quartz distiller (Krner Analysentechnik,Rosenheim, Germany). The laboratory glassware was cleanedas follows: immersion in a 2% (v/v) Extran (Merck) solutionfor 24 h; washing thoroughly with distilled water; immersionfor 48 h in a 20% (v/v) nitric acid solution (Vetec, Rio deJaneiro, Brazil); ultrasound bath for 1 h and finally washingwith deionized water.

Sodium citrate buffer solution was prepared by weighing 2.94

g of sodium citrate and diluting with deionized water. The pHwas adjusted to 7 with 0.1 mol L 1 nitric acid and the finalvolume was made up to 100 mL. Tris(hydroxymethyl)-aminomethane (HOCH 2)3CNH 2 0.1 mol L 1 was prepared byweighing 1.2 g and diluting with water. The pH 7 was reachedwith 0.1 mol L 1 nitric acid and the final volume made up to 100mL.

Lead standard solutions were prepared from a 1000 mg L 1

stock solution (SPEX, Edison, NJ, USA).Silica gel chemically modified with niobium(V) oxide

(Nb 2O5-SiO 2) was synthesized as described in the literature. 24,25

To evaluate the effect of foreign ions on Pb sorption, wetested the following ions: Na +, Ca 2+, K+, Mg 2+, Zn 2+, Cu 2+ andFe3+. Na + was obtained as NaCl from Reagen (Rio de Janeiro,Brazil); Ca 2+ as CaCO 3 and Mg 2+ as MgSO 4 from Vetec (Rio deJaneiro, Brazil); K + as KCl from Isofar (Rio de Janeiro, Brazil);and standard solutions of Zn 2+, Cu 2+ and Fe 3+ from SPEX(Edison).

On-line preconcentration systemThe on-line flow system used for the development of the

proposed method is illustrated in Fig. 1. The flow systemconsists of two peristaltic pumps equipped with Tygon tubes,four three-way solenoid valves and a minicolumn filled with100 mg of the sorbent Nb 2O5-SiO 2. The flow system wascoupled to the FAAS. During preconcentration (A), valve V1 isopen and the other valves remain closed; the sample or workingsolutions are pumped through the minicolumn and the effluentis discarded. After this first step, water is pumped for 5 s in thesame way as in the preconcentration procedure to eliminate anysample matrix that might be present in the dead volume of theminicolumn.

In the elution step (B), V1 is closed and valves V2, V3 and V4are open. Thus, the eluent percolates through the minicolumn inthe opposite direction to that of the preconcentration step. Theeluate is carried directly to the nebulization system of theFAAS. 18,25

Optimization strategy for the proposed systemThe optimization of the parameters affecting Pb sorption by

the sorbent Nb 2O5-SiO 2 was performed using a two-level fullfactorial experimental design involving four factors and a finaloptimization using a Doehlert matrix design. All experimentswere carried out in duplicate, using 15.0 mL of a 100.0 g L 1

Pb solution. The four variables were: buffer type, eluentconcentration, sample flow rate and eluent flow rate. Theanalytical response was taken as peak area. The experimentaldata were processed using the Statistica 6.0 computer program.

Environmental samples and certified reference materialsQuantification of Pb was carried out using an external

calibration procedure. The analytical curve was obtained in therange of 5.0 to 120.0 g L 1, submitting each solution to theoptimized preconcentration procedure.

A calibration curve with concentrations from 1.0 to 7.0 mg L 1

prepared in 0.5 mol L 1 HNO 3 (the same concentration as the

366 ANALYTICAL SCIENCES MARCH 2008, VOL. 24

Fig. 1 Diagram of the on-line preconcentration system used in thisstudy. (A) Adsorption process and (B) desorption process. V, Valve;L, open; D, closed; MC, minicolumn containing adsorbent; R, sampleor eluent back stream; hatched circle, valve on; white circle, valve off.


3/6

eluent) was constructed without preconcentration for estimationof the enrichment factor.

Water samples of acid waste material from a coal mine(Cricima-SC, Brazil) were collected in suitable polypropyleneflasks and kept at 4C until analysis. Seawater and well watersamples were collected in Florianpolis-SC, Brazil.

The following certified reference materials were used forvalidation of the proposed methodology: NIST 1643e(freshwater) from the National Institute of Standards &Technology (Gaithersburg, USA) and BCR 397 (human hair)from the Community Bureau of Reference, Brussels, Belgium.For the environmental samples, recovery tests were performed.

For the preparation of the certified reference material (hair),samples of approximately 250 mg were weighed directly inPTFE flasks. Subsequently, 3.0 mL of HNO 3 and 1.0 mL of H2O2 were added to each sample in the PTFE flasks. An MLS-1200 microwave (Milestone, Sorisole, Italy) was used formicrowave-assisted acid-digestion with the following program:2 min at 250 W, 2 min at 0 W, 6 min at 250 W, 5 min at 400 W

and 5 min at 650 W, followed by 5 min of ventilation. Thedigested samples were transferred to 50 mL polypropylenetubes (Sarstedt, Nqmbrecht, Germany) and kept underrefrigeration until use.

Results and Discussion

Preliminary tests were performed to investigate Pb(II)adsorption on the sorbent Nb 2O5-SiO 2 under different pHconditions. For a pH lower than 5, the amount of Pb(II)adsorbed was extremely low, while between 5 to 9 no

significant difference was observed. Thus, pH 7 was selected,since the pH for natural water samples is usually around 7. Thesample volume used for the optimization was 15.0 mL with a100.0 g L 1 Pb solution.

Factorial designThe factors considered for the preconcentration procedure

were: buffer type, eluent concentration (HNO 3), sample flowrate and eluent flow rate. A total of 16 experiments werecarried out, in duplicate, to determine the significance of eachvariable, as well as the influences of their interactions on thepreconcentration system. Table 1 shows the minimum andmaximum levels obtained for the selected factors. Themaximum and minimum conditions for each variable wereselected based on other studies using the same sorbentmaterial. 18,25

The combination used for each experiment and the analyticalresponse obtained (as integrated absorbance) are summarized inTable 2.

A Pareto chart (Fig. 2) was plotted to check the influence of

367ANALYTICAL SCIENCES MARCH 2008, VOL. 24

Table 2 Matrix of f ull factorial desi gn and the analytical response for each experiment for preconcentration of Pb(II)

1 1 1 1 1 0.373 0.4142 1 1 1 1 0.419 0.4153 1 1 1 1 0.378 0.3784 1 1 1 1 0.323 0.3565 1 1 1 1 0.475 0.4776 1 1 1 1 0.456 0.477 1 1 1 1 0.398 0.3848 1 1 1 1 0.381 0.3819 1 1 1 1 0.196 0.18

10 1 1 1 1 0.188 0.19511 1 1 1 1 0.149 0.156

12 1 1 1 1 0.126 0.1113 1 1 1 1 0.146 0.11814 1 1 1 1 0.126 0.13515 1 1 1 1 0.084 0.10716 1 1 1 1 0.068 0.044

Inte grated absorbanceSample flow rate/ mL min 1

Eluent flow rate/ mL min 1

Eluent concentration/ mol L 1

Experiment Buffer typeReplicate 1 Replicate 2

Table 1 Factors and levels used in the factorial desi gn

Buffer type Sodi um citrate TRIS a

Sample flow rate/mL min 1 3.3 6.3Eluent concentration/mol L 1 0.5 2.5Eluent flow rate/mL min 1 3.3 5.3

Minim um () Maxim um (+)Factor

a. Tris(hydroxymethyl)aminomethane.

Fig. 2 Pareto chart of standardized effects for the variables of thepreconcentration system using integrated absorbance as analyticalresponse.


4/6

the factors and their interactions in the system using analysis of variance (ANOVA) and p-values ( p > 0.05). As can be seen

from the Pareto chart, the factors buffer type, sample flow rate,eluent flow rate and the interaction among them are statisticallysignificant at the 95% confidence level. The very high negativevalue obtained for the effect of buffer type (51.90) indicatesthat the analytical response increases considerably when citratebuffer is used. Thus, citrate buffer was used for the subsequentexperiments. The negative values for sample (11.44) andeluent (2.63) flow rates indicate an improvement in theanalytical response when the level changes from maximum tominimum. Eluent concentration was not statistically significantand, therefore, its concentration was fixed at 0.5 mol L 1.

Final optimization using Doehlert designIt was observed from the results of the factorial design that the

factors sample flow rate and eluent flow rate required a finaloptimization. Thus, a Doehlert matrix involving these variableswas constructed. Seven experiments were carried out induplicate, and the levels and responses are summarized in Table 3.

The surface obtained through application of the Doehlert

matrix is shown in Fig. 3. The analytical response was taken asintegrated absorbance. The response surface can be describedby a quadratic equation (Eq. (1)).

Abs=0.100+0.115flow sample +0.038flow eluent0.011flow 2sample 0.001flow sample flow eluent0.003flow 2eluent (1)

Abs/ flow sample =0.1150.022flow sample 0.001flow eluent (2)

Abs/ flow eluent =0.0380.001flow sample 0.006flow eluent (3)

The critical point of the quadratic function is obtained bymaking the partial derivative equations (Eqs. (2) and (3)) equalto zero. The application of the Lagrange criterion 19,20 indicateswhether this critical point is a maximum, minimum or a saddlepoint. The critical points obtained were 5.5 and 5.0 mL min 1

for eluent and sample flow rates, respectively. The applicationof Lagrange criterion indicated that the critical point is themaximum point of the response surface.

Thus, as a result of all the optimizations, the followingworking conditions were selected: citrate buffer, 0.5 mol L 1

HNO 3 as the eluent, and sample and eluent flow rates of 5.0 and5.5 mL min 1, respectively.

Effect of foreign ions on the Pb preconcentrationThe effects of Ca 2+, Mg 2+, Na +, K+, Zn 2+, Cu 2+ and Fe 3+ ions

were assessed to check whether these ions interfere with theadsorption of 15.0 mL of a solution containing 100.0 g L 1 Pb,under the optimized conditions. Silica gel chemically modifiedwith niobium(V) oxide is characterized as being a cationic

exchanger. Therefore, competition between concomitantcations and the analyte will be dependent on their relativeconcentrations, the relative affinity for the sorbent surface, andthe number of available active sites at the sorbent surface.

Usually, the effect of concomitants is determined using aunivariate approach, adding to the working solution a knownconcentration of only one cation at a time. In this way, theeffect of each cation is evaluated individually. This type of study is tedious and time-consuming and is not representative of a real situation, in which a sample may contain several ions atthe same time. For this reason, in this study, interference wasinvestigated using a 2 73 fractionary factorial experimentaldesign including a central point, resulting in 17 experiments.

Table 4 summarizes the levels for each ion considered as afactor for this study and Table 5 shows the matrix of theexperimental design and the analytical response, obtained usingintegrated absorbance.

Figure 4 shows the main and 2-way interaction effects of thefactors on the analytical response. As can be observed, none of


Table 3 Doehlert matrix for the final optimization of sample andeluent flow rates

1 4.3 3.3 0.2392 3.3 4.1 0.2583 3.3 5.6 0.2464 4.3 6.3 0.2465 5.3 5.6 0.2576 5.3 4.1 0.2737 4.3 4.8 0.267

a. Avera ge of two replicates.

Eluent flow rate/ mL min 1

Sample flow rate/ mL min 1

Integratedabsorbance a /s

Experiment

Fig. 3 Response surface for optimization of sample and eluent flowrates: sample volume, 15.0 mL; sample concentration, 100.0 g L1;sample pH, 7.0; sorbent mass, 100.0 mg; eluent concentration, 0.5mol L 1 HNO 3; citrate buffer, 0.1 mol L 1.

Table 4 Levels and factors used for eval uation of interference of selected ions thro ugh fractionary factorial experimental desi gn

Ca2+

0 2.5 5.0Mg2+ 0 1.5 3.0Na+ 0 2.5 5.0K+ 0 1.5 3.0Zn2+ 0 0.5 1.0Cu2+ 0 0.5 1.0Fe3+ 0 0.5 1.0

Minim um ()/ mg L1

Center point/ mg L1

Maxim um (+)/ mg L1

Factor


5/6

these main and interaction effects is significant with regard tothe analytical response at a 95% confidence level. In otherwords, none of the selected ions interfere with lead adsorptionat these concentration levels.

Effect of sample volumeUsing the optimized conditions, a study to determine the

sample volume to be preconcentrated was performed. Theanalyte mass was maintained constant and the sample volumewas varied and, therefore, the analyte concentration changed.On increasing the sample volume, it was observed that theanalytical signal was almost constant over the volume rangestudied (20 to 120 mL). Thus, the detection of very lowconcentrations of the analyte can be improved by increasing thesample volume. However, this procedure can result in someloss of sample throughput. As a compromise between analyticalfrequency and sensitivity, 20 mL of sample volume wasselected for subsequent studies.

Analytical featuresThe analytical data of interest for the method are summarized

in Table 6. The method precision was determined by

submitting 20 mL of a 100 g L1

sample solution to theoptimized procedure. Very good precision (RSD < 2%) wasobtained. An analytical curve was obtained to assesssensitivity, linear range, enrichment factor, detection andquantitation limits. The detection limit was calculated as threetimes the standard deviation of a blank solution divided by theslope. The enrichment factor was obtained through the ratiobetween the slopes of analytical curves with and withoutpreconcentration.

Analysis of samples and method validationThe proposed flow system was used for Pb determination in

water samples from a coal mine located in Cricima, Brazil, aswell as in seawater and well water, both from Florianpolis,Brazil. Recovery tests were performed and two certifiedreference materials were analyzed in order to check theaccuracy of the method. The results are given in Tables 7 and8, respectively.

Good recoveries between 90 104% were obtained for the

369ANALYTICAL SCIENCES MARCH 2008, VOL. 24

Table 5 Fractionary factorial desi gn matrix used to check theinfluence of selected ions on analytical response

1 1 1 1 1 1 1 1 0.51132 1 1 1 1 1 1 1 0.49473 1 1 1 1 1 1 1 0.49894 1 1 1 1 1 1 1 0.50635 1 1 1 1 1 1 1 0.50676 1 1 1 1 1 1 1 0.50857 1 1 1 1 1 1 1 0.49148 1 1 1 1 1 1 1 0.49949 1 1 1 1 1 1 1 0.5050

10 1 1 1 1 1 1 1 0.518511 1 1 1 1 1 1 1 0.509412 1 1 1 1 1 1 1 0.507913 1 1 1 1 1 1 1 0.520114 1 1 1 1 1 1 1 0.520015 1 1 1 1 1 1 1 0.5264

16 1 1 1 1 1 1 1 0.525517 0 0 0 0 0 0 0 0.5241

Ca2+ Mg2+ Na + K+ Zn2+ Cu2+ Fe3+Integratedabsorbance

Experiment

Sample vol ume, 15.0 mL; sample concentration, 100.0 g L1 Pb.

Fig. 4 Pareto chart of standardized effects of foreign ions on leadadsorption in the proposed preconcentration system.

Table 6 Relevant analytical data

Linear ran gea / g L1 5.0 120.0Correlation coefficient ( R) 0.9982Relative standard deviation (RSD, %) 1.6

(100 g L1, n = 7)

Sensitivity ( )/L g1

0.00411Limit of detection (LOD)/ g L1 0.35Limit of q uantification (LOQ)/ g L1 1.20Enrichment factor (EF) 46.4Analytical freq uency/samples h 1 12

a. St udied linear ran ge.

Table 7 Pb(II) concentration obtained for water samples d uring the recovery tests

Mine water 0 ND

30.0 28.7 1.2 96.040.0 41.5 2.0 104.0

Seawater 0 ND 5.0 4.5 1.0 90.0

10.0 9.6 1.1 96.4Well water 0 ND

5.0 4.6 0.7 92.010.0 9.9 0.9 99.0

ND = not detected.a. Avera ge sd, n = 3.

Sample Added/ g L1 Founda / g L1 Recovery, %

Table 8 Concentrations obtained for Pb(II) in certified reference

materials analyzed usin g the preconcentration proced ure heredeveloped

NIST 1643e 19.63 0.21 18.71 0.91BCR397 (h uman hair) 33.0 1.2 30.2 2.0

n = 5. St udent t -test for 95% confidence level.

Sample Certified/ g L1 Found/ g L1


6/6

environmental samples, demonstrating that the method can besuccessfully applied to these samples. For the certifiedreference materials, there was good agreement between foundand certified values.

Conclusions

In this paper, we describe a new method for Pb determination attrace levels using solid phase extraction for the samplepreparation and silica gel chemically modified with niobium(V)oxide as the sorbent. This sorbent presented excellent stability,having a longer lifetime than that necessary to carry out thisstudy. The use of a multivariate procedure for optimization of the parameters affecting the system, as well as the interferencestudy, was found to be very efficient, requiring a reducednumber of experiments. The instrumentation and methoddescribed provided a fast and convenient way to preconcentrateand separate Pb(II) from aqueous samples in a simple analytical

cycle.

Acknowledgements

The authors thank Conselho Nacional de DesenvolvimentoCientifico e Tecnolgico (CNPq) for financial support.

References

1. S. E. Manahan, Environmental Chemistry , 6th ed., 1994 ,CRC Press, Boca Raton.

2. D. A. Gidlow, Occup. Med. , 2004 , 54 , 76.3. L. H. Sann, T. G. Cosso, I. Romieu, and M. H. Avila,

Salud Pblica Mex , 1998 , 40 , 359.4. A. Ceccarini, I. Cecchini, and R. Fuoco, Microchem. J. ,

2005 , 79 , 21.5. CONAMA: http://www.mma.gov.br/port/conama/res/res05/

res35705.pdf.6. D. Budziak, E. L. da Silva, S. D. de Campos, and E.

Carasek, Microchim. Acta , 2003 , 141 , 169.7. H. Chen, J. Jin, and Y. Wang, Anal. Chim. Acta , 1997 , 353 ,

181.8. Y. K. Agrawal and K. R. Sharma, Talanta , 2005 , 67 , 112.9. M. Soylak and N. D. Erdogan, J. Hazard. Mater. B , 2006 ,

137 , 1035.10. M. Soylak, S. Saracoglu, U. Divrikli, and L. Elci, Talanta ,

2005 , 66 , 1098.11. J. Cheng and K. C. Teo, Anal. Chim. Acta , 2001 , 450 , 215.12. V. Camel, Spectrochim. Acta, Part B , 2003 , 58 , 1177.13. K. Pyrzy n ska and M. Cheregi, Water Res. , 2000 , 34 , 4215.14. A. Goswami and A. K. Singh, Talanta , 2002 , 58 , 669.15. I. L. de Alcntara, P. S. Roldan, M. A. L. Margionte, G. R.

Castro, C. C. F. Padilha, A. O. Florentino, and P. M.Padilha, J. Braz. Chem. Soc. , 2004 , 15 , 366.

16. S. Saracoglu and L. Elci, Anal. Chim. Acta , 2002 , 452 , 77.17. P. K. Tewari and A. K. Singh, Talanta , 2002 , 56 , 735.

18. R. L. Dutra, H. F. Maltez, and E. Carasek, Talanta , 2006 ,69 , 488.19. S. L. C. Ferreira, H. C. dos Santos, M. S. Fernandes, and

M. S. de Carvalho, J. Anal. At. Spectrom. , 2002 , 17 , 115.20. M. Zougagh, P. C. Rudner, A. G. de Torres, and M. C.

Pavn, J. Anal. At. Spectrom. , 2000 , 15 , 1589.21. S. L. C. Ferreira, W. N. L. dos Santos, C. M. Quintella, B.

B. Neto, and J. M. Bosque-Sendra, Talanta , 2004 , 63 ,1061.

22. V. A. Lemos, P. X. Baliza, J. S. Santos, L. S. Nunes, A. A.de Jesus, and M. E Rocha, Talanta , 2005 , 66 , 174.

23. M. S. Bispo, E. S. da Boa Morte, M. G. A. Korn, L. S. G.Teixeira, M. Korn, and A. C. S. Costa, Spectrochim. Acta,Part B , 2005 , 60 , 653.

24. S. Denofre, Y. Gushikem, S. S. Castro, and Y. Kawano, J.Chem. Soc. Faraday Trans. , 1993 , 89 , 1957.

25. E. L. Silva, E. M. Ganzarolli, and E. Carasek, Talanta ,2004 , 62 , 727.


doelhert niobium

Documents