8/13/2019 1-s2.0-S003991400600662X-main

1/6

Talanta 72 (2007) 6065

Comparison of slurry sampling and microwave-assisted digestion forcalcium, magnesium, iron, copper and zinc determination in fish tissue

samples by flame atomic absorption spectrometry

Raquel Alonso Bugallo, Susana Ro Segade , Esperanza Fernandez Gomez

Departamento de Qumica Analtica y Alimentaria, Area de Qumica Analtica, Universidad d e Vigo,

Facultad de Ciencias de Ourense, As Lagoas s/n, 32004 Ourense, Spain

Received 27 March 2006; received in revised form 24 August 2006; accepted 27 September 2006

Available online 1 November 2006

Abstract

The development of a slurry sampling method for the determination of calcium, copper, iron, magnesium and zinc in fish tissue samples by flame

atomic absorption spectrometry is described. In comparison with microwave-assisted digestion, the proposed method is simple, requires short time

and eliminates total sample dissolution before analysis. Suspension medium was optimized for each analyte to obtain quantitative recoveries from

fish tissue samples without matrix interferences. Nevertheless, iron recoveries higher than 46% were not found. Treatment of samples slurried

in nitric acid by microwave irradiation for 1530 s at 75285 W permitted to achieve efficient recoveries for calcium, iron, magnesium and zinc.

Further improvement in the matrix effects for iron determination was accomplished by the use of an additional step of short microwave-assisted

suspension treatment. However, standard addition method was required for calcium and copper determination, being necessary hydrochloric

acid as suspension medium for the last one. Although copper could not be determined in the certified reference material using microwave-assisted

digestion, the accuracy of the slurry sampling method was verified for all the investigated analytes. Detection limits were 22.8 8.0, 0.884 0.092,

5.07 0.76, 35.5 0.7 and 1.17 0.04g g1 for calcium, copper, iron, magnesium and zinc, respectively. The standard deviations obtained using

slurry sampling method and microwave-assisted digestion were not significantly different, and the mean relative standard deviation of the over-all

method (n = 3) of the slurry sampling method for different concentration levels was below 12%. 2006 Elsevier B.V. All rights reserved.

Keywords: Slurry sampling; Microwave-assisted digestion; Calcium, magnesium, iron, copper and zinc determination; Fish tissue samples; Flame atomic absorption

spectrometry

1. Introduction

The most time consuming step in solid materials analysis is

often sample treatment.Several procedures have been developed

for elements determination by atomic spectrometric techniques,

in order to short the analysis time and to minimize the problems

associated with solid sample treatment (conventional wet acid

digestion and dry-ashing), such as sample contamination and

analyte loss. So, alkaline[13]and acid[311]digestion meth-

ods assisted by microwave[38,10,11]or ultrasound[24,8,9]

Corresponding author. Present address: Estacion de Viticultura y Enologa

de Galicia, Ponte San Clodio s/n, Leiro, 32427 Ourense, Spain.

Tel.: +34 988 488033; fax: +34 988 488191.

E-mail address: evegadoc2@cesga.es(S.R. Segade).

energy (involve complete or partial matrix solubilization) permit

to short the time required for sample treatment. Furthermore,

direct solid sampling offers several advantages, including the

minimization of the sample handling and, as a consequence, the

reduction of the risks previously mentioned[1215].

The slurry sampling technique has been extensively used

for the determination of trace elements in biological and envi-

ronmental solid samples by electrothermal atomic absorption

spectrometry (ETAAS)[4,6,8,1620].This approach combines

the advantages of both liquid and direct solid sampling. How-

ever, it is important to emphasize that slurry homogeneity must

be preserved during the time required for sample introduction

into the atomizer. Furthermore, the particle size affects accuracy

and precision. Anyway, literature has been also published in the

last years on the analysis of biological[8,21,22]and inorganic

[2326]solid samples by slurry sampling flame atomic absorp-

0039-9140/$ see front matter 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.talanta.2006.09.023

mailto:evegadoc2@cesga.eshttp://localhost/var/www/apps/conversion/tmp/scratch_6/dx.doi.org/10.1016/j.talanta.2006.09.023http://localhost/var/www/apps/conversion/tmp/scratch_6/dx.doi.org/10.1016/j.talanta.2006.09.023mailto:evegadoc2@cesga.es

8/13/2019 1-s2.0-S003991400600662X-main

2/6

R.A. Bugallo et al. / Talanta 72 (2007) 6 065 61

tion spectrometry (FAAS) in spite of the continuous aspiration

of the slurry, which affects atomization efficiency. In this con-

text, some investigations have been realized in order to improve

the transport of the solid particles to the flame and, therefore, the

atomization efficiency of the analytes from the slurry. The four

principal proposals were the use of special nebulizers[21,27],

the addition of wetting agents [28], the acid extraction of analyte

into the liquid phase[21]or the use of flow injection systems

for the introduction of slurried solid samples into the flame

[22,23,25,26].

Most of the methods for determining elements in biolog-

ical samples by FAAS are based on microwave-assisted acid

digestion. In this paper, a method has been developed for the

determination of calcium, copper, iron, magnesium and zinc in

fish tissue samples by slurry sampling FAAS, using a conven-

tional nebulizer. Very few works have been reported in relation

to the analysis of fish tissue samples by the combination of both

techniques[8].The main aims proposed were to short the time

required for sampletreatment, to avoidthe problems related with

sample decomposition and to achieve quantification limits suit-able for the determination of element concentration present in

fish from high sea. The effect of the suspension medium and

a novel microwave-assisted slurry treatment was studied, being

tested not only the most commonly used acid medium, but also

alkaline medium and a complexing agent. Moreover, the influ-

ence of the addition of glycerol as wetting agent and the slurry

concentration was also evaluated. The results so obtained for all

thetunasamplesanalyzedwerecomparedwiththoseobtainedby

means of microwave-assisted acid digestion in order to empha-

size the advantages of the proposed method.

2. Experimental

2.1. Instrumentation

The slurry treatment was performed with a Moulinex

(Barcelona, Spain) 900 W microwave oven. A Selecta

(Barcelona, Spain) ModelAgimatic-N magnetic stirrer was used

to homogenize the slurry. A Kubota (Tokyo, Japan) Model 5100

centrifuge was employed for the separation of the liquid phase

from the slurried samples. The solid materials were also acid

digested in a Parr (Moline, IL, USA) Model 4782 medium-

pressure reactor heated by the above-mentioned microwave

oven. A Perkin-Elmer (Norwalk, CT, USA) Model 2380 atomic

absorption spectrometer equipped with an acetylene-air flamewas used for element determination. Cathodeon hollow-cathode

lamps were used as the radiation source. The instrumental

parameters used were those recommended by the manufacturer.

2.2. Reagents, samples and reference material

All chemicals used were of analytical reagent grade (Merck,

Darmstadt, Germany) and deionized water was used throughout

the experiments. The stock standard solutions (1000 mg l1)

were prepared from pure metal (copper, iron, magnesium and

zinc) or from high purity calcium carbonate salt (calcium).

Working standard solutions were prepared daily by appro-

priately diluting the stock standard solutions. Diluted nitric

acid, diluted hydrochloric acid, disodium ethylenediaminete-

traacetate solution, sodium hydroxide solution and diluted

hydrogen peroxide were used for the slurry preparation by

adding glycerol as dispersing agent. Concentrated nitric acid

was also used in combination with hydrogen peroxide for solid

samples digestion.

Thecomestible part of thetuna samples (IIII) was oven dried

at 50 C for 48 h to constant weight and ground. The powdered

sample with particle size less than 70m was selected by siev-

ing (nylon sieves) for analysis and stored at room temperature in

polyethylene bottles in a desiccator. A certified reference mate-

rial was used to validate the proposed method. The NCS ZC

80006 (prawn) was obtained from the China National Analysis

Center for Iron and Steel.

2.3. Microwave-assisted acid digestion procedure

A microwave-assisted wet decomposition of fish tissue sam-

ples was performed according to a slightly modified procedure[29]. About 0.1 g of tuna sample were digested with 4 ml of con-

centrated nitric acid in a Parr reactor by heating in a microwave

oven at 510 W for 60 s. After cooling to room temperature in an

ice bath, the reactor was opened and the digested sample was

after treated with 2 ml of concentrated hydrogen peroxide in the

closed reactor by microwave irradiation at 510W for 120 s. The

resulting solution was partially evaporated and quantitatively

transferred into a 10 ml calibrated flask, where it was diluted

to volume with deionized water. Blanks were prepared with the

same reagents undergoing a similar treatment. Finally, all solu-

tions were stored in polyethylene bottles at 4 C.

2.4. Slurry preparation procedure

A 3300 mg portion of tuna sample was weighed into a 50 ml

polyethylene tube and 1050 ml of 0.13mol dm3 hydrochlo-

ric acid, 0.071.5 mol dm3 nitric acid, 0.051 mol dm3

sodium hydroxide, 0.053% (m/v)disodium ethylenediaminete-

traacetate or 15% (m/v) hydrogen peroxide containing 0.13%

(v/v) glycerol were added. The slurry was homogenized by

magnetic stirring for 5 min. When nitric acid was employed

as suspension medium, the slurry was microwave irradiated at

75400 W for 560 s and, before aspirating, again homogenized

by magnetic stirring. The supernatant liquid was separated from

the slurried tuna sample by centrifugation at 2500 rpm for 5 minin order to check extraction efficiency of thesuspension medium.

Blanks were prepared with the same reagents undergoing a sim-

ilar treatment.

3. Results and discussion

3.1. Optimization of suspension medium

The atomization efficiency for the elements studied from

slurried tuna samples was evaluated by comparison of the

absorbance values with those obtained from an aqueous stan-

dard. The tuna sample suspended in deionized water and the

8/13/2019 1-s2.0-S003991400600662X-main

3/6

62 R.A. Bugallo et al. / Talanta 72 (2007) 6065

Table 1

Initial and optimum parameters for calcium, copper, iron, magnesium and zinc determination using slurry sampling

Parameter Ca Cu Fe Mg Zn

Initial

Suspension medium HNO3 HNO3 HNO3 HNO3 HNO3Medium concentration (mol dm3) 0.15 0.15 0.15 0.15 0.15

Suspension medium volume (cm3) 10 10 10 50 10

Glycerol concentration (%, v/v) 0.3 0.3 0.3 0.3 0.3Sample amount (mg) 100 150 150 50 100

Optimum

Suspension medium HNO3 HCl HNO3 HNO3 HNO3Medium concentration (mol dm3) 0.15 0.10 0.15 0.15 0.15

Suspension medium volume (cm3) 10 10 10 50 10

Microwave time (s) 15 30 15 15

Microwave power (W) 123 285 75 75

Glycerol concentration (%, v/v) 0.3 0.3 0.3 0.3 0.3

Sample amount (mg) 40 150 150 3 100

aqueous standard solution were prepared to contain the same

analyte concentration. The absorbance signals obtained for all

elements from slurried tuna sample were less than those cor-responding to the aqueous standard, except for magnesium.

The effect of nebulization flow-rate and air/acetylene flow-rates

relation on the atomization efficiency from the slurry was inves-

tigated. Maximum atomization efficiency from both the slurry

and the aqueous standard was 0.1, 0.02, 0.2, 0.9 and 0.08 for

calcium, copper, iron, magnesium and zinc, respectively. The

instrumental conditions selected were those corresponding to

maximum sensitivity for all elements from aqueous standard

solutions.

Some investigators have reported that partial extraction of

several analytes into the liquid phase improves the precision

and accuracy of slurry sampling ETAAS[30].Moreover, slurry

sampling couldbe replaced by supernatant liquid sampling when

the analytes were quantitatively extracted in the suspension

medium. Further advantages were the elimination of stirring sys-

tems and stabilizing agents since a homogeneous distribution of

solid particles was not necessary[31].

In this sense, suspension medium was evaluated in order to

establish: (i) the percentage of analyte recovered by the slurry

sampling technique in relation to the analyte concentration deter-

mined using the microwave-assisted acid digestion procedure;

(ii) the extraction efficiency of analyte from the slurry into the

liquid phase; (iii) the presence of matrix effects. The initial

parameters are shown in Table 1. The suspension medium inves-

tigated was hydrochloric acid (as acid medium), sodium hydrox-ide (as alkaline medium) and disodium ethylenediaminetetraac-

etate (as complexing medium). All determinations were carried

out by triplicate using a tuna sample.

Fig. 1 shows that quantitative calcium recovery was achieved

using either disodium ethylenediaminetetraacetate concentra-

tions equal to or less than 0.5% (m/v) or hydrochloric acid con-

centrations comprised between 0.5 and 1 mol dm3. Taking into

account that both the greatest slopes of standard addition lines

corresponded to the use of disodium ethylenediaminetetraac-

etate (0.007580.0113 l mg1) and absorbance values obtained

for reagent blank increased with increasing complexing agent

concentration from 0.5% (m/v), the suspensionmedium selected

Fig. 1. Effect of %disodium ethylenediaminetetraacetate (m/v) (), hydrochlo-

ric acid (mol dm3) () and sodium hydroxide (mol dm3) () on recovery of

calcium, copper, iron, magnesium and zinc.

8/13/2019 1-s2.0-S003991400600662X-main

4/6

R.A. Bugallo et al. / Talanta 72 (2007) 6 065 63

was 0.05% (m/v) disodium ethylenediaminetetraacetate. Fur-

thermore, the influence of pH value of disodium ethylenedi-

aminetetraacetate on calcium recovery was also investigated.

So, an increase in pH value caused calcium amount recov-

ered to decrease, obtaining a calcium recovery of 68% for a

pH value of 9. This effect was verified by the use of ammo-

nium chloride/ammonium hydroxide buffer solution or sodium

hydroxide concentrations higher than 0.05 mol dm3 as suspen-

sion medium, which gave calcium recoveries of 41 and 74%,

respectively.

The effect of disodium ethylenediaminetetraacetate and

sodium hydroxide concentrations on copper recovery was not

significant. Moreover, pH value of disodium ethylenediaminete-

traacetatesolutionhad no influence on copper amount recovered.

Fig. 1shows quantitative copper recoveries only for hydrochlo-

ric acid concentrations ranging from 0.1 to 1 mol dm3. As the

maximum slope of standard addition line was attained using

0.1 mol dm3 hydrochloric acid (0.131 l mg1), this suspension

medium was chosen.

The maximum iron recovery (42%) corresponded tohydrochloric acid concentrations comprised between 0.1 and

1moldm3, 0.05 mol dm3 sodium hydroxide or 0.5% (m/v)

disodium ethylenediaminetetraacetate, as can be observed in

Fig. 1. The alkalinization of the complexing agent or the use

of sodium hydroxide concentrations higher than 0.05 mol dm3

caused an important reduction of iron amount recovered.

The slopes of external calibration and standard addition

lines obtained for iron were not statistically different (t-test;

p = 0.05) for the sample slurried in 0.51 mol dm3 hydrochlo-

ric acid or 0.053% (m/v) disodium ethylenediaminetetraacetate

(0.03830.0400 l mg1). Therefore, 0.5 mol dm3 hydrochloric

acid could be selected as optimum suspension medium.Fig. 1shows that magnesium recovery depended on the sus-

pension medium used. Quantitative recoveries were obtained

for hydrochloric acid concentrations comprised between 0.1 and

1moldm3, followed by those (67%) corresponding to the use

of disodium ethylenediaminetetraacetate concentrations ranging

from 0.05 to 2% (m/v). Thelast ones were maintained practically

constant when the solution was alkalinized. However, sodium

hydroxide concentrations within the range 0.050.1 mol dm3

or ammonium chloride/ammonium hydroxide buffer solution

permitted to obtain magnesium recoveries of 68%. On the other

hand, insignificant absorbance signals corresponding to reagent

blank and no significant difference (t-test; p = 0.05) between

the slopes of external calibration and standard addition lines(0.647 l mg1) were found when low hydrochloric acid concen-

trations were used as suspension medium. Thus, the suspension

medium chosen was 0.1mol dm3 hydrochloric acid.

The influence of suspension medium on zinc recovery is

shown in Fig. 1. So, the use of disodium ethylenediaminete-

traacetate concentrations equal to or higher than 0.5% (m/v)

or hydrochloric acid concentrations equal to or less than

1moldm3 permitted to achieve quantitative recoveries for

zinc. The alkalinization of disodium ethylenediaminetetraac-

etate solution caused the diminution of zinc amount recov-

ered, which was verified by the suspension of the sample in

either sodium hydroxide or ammonium chloride/ammonium

hydroxide buffer solution. The slope of standard addition line

obtained for both acid and complexing suspension medium was

0.162 l mg1. Taking into account that low absorbance values

were obtained for reagent blank when 0.1 mol dm3 hydrochlo-

ric acid was employed, this suspension medium was selected.

The concentrations found in the tuna slurry are in accordance

with those found in the supernatant liquid for all analytes deter-

mined, achieving recoveries higher than 96% in the last one in

relation to the concentrations determinedin the slurried samples.

3.2. Microwave-assisted suspension treatment

The low recoveries obtained for iron and the presence of

matrix interferences for calcium, copper and zinc led to the

optimization of a microwave-assisted treatment of slurried

tuna sample in order to achieve greater matrix decomposition

degree. Nitric acid is adequate for microwave-assisted treat-

ments, because of its capability to absorb efficiently microwave

energy and it is one of the most widely used reagents for bio-

logical samples treatment because of its oxidant properties.

The effect of nitric acid concentration on calcium, copper,

iron, magnesium and zinc recovery was investigated and the

results can be seen inFig. 2.Calcium, copper and magnesium

recovery remained practically constant for nitric acid concen-

trations between 0.07 and 0.7 mol dm3. Although magnesium

recovery was quantitative for a nitric acid concentration of

1.5 mol dm3, the absorbance values obtained for reagent blank

were also high. On the other hand, iron recovery decreased

slightly and zinc recovery increased markedly when nitric acid

concentration was increased from 0.07 to 0.15 mol dm3 and

both recoveries kept constant for higher nitric acid concen-

trations. Thus, the slopes of standard addition lines obtainedusing nitric acid as suspension medium were 0.00935, 0.139,

0.0280, 0.668 and 0.195 l mg1 for calcium, copper, iron, mag-

nesium and zinc, respectively. The slopes of external calibration

and standard addition lines were significantly different (t-test;

p = 0.05) for calcium, copper and iron. 0.15 mol dm3 (1%,

v/v) nitric acid was selected for further experiments. The same

concentration was previously used by other authors for metals

determination in human scalp hair by slurry sampling FAAS

[21].

Fig. 2. Effect of nitric acid concentration on recovery of: calcium (), copper

(), iron (), magnesium () and zinc ().

8/13/2019 1-s2.0-S003991400600662X-main

5/6

64 R.A. Bugallo et al. / Talanta 72 (2007) 6065

Microwave heating times between 5 and 60 s were opti-

mized at 75 W power. Calcium and iron recoveries increased

with increasing microwave time up to 15 and 30 s, respectively.

However, the effect of microwave time on copper, magnesium

and zinc recoveries was not significant. Then, in another exper-

iment the influence of different microwave powers (75400 W)

on analyte recovery revealed that maximum recoveries were

obtained for calcium (88.1%), copper (55.1%), iron (94.5%),

magnesium (85.6%) and zinc (97.9%) when microwave powers

ranging 123285, 75, 285400, 75170 and 75170 W, respec-

tively, were employed. The microwave-assisted slurry treatment

permitted to avoid matrix interferences for iron because of both

slopes of external calibration and standard addition lines were in

good agreement (t-test;p = 0.05). Furthermore, the slope of stan-

dard addition line for copper increased when microwave energy

was applied to the slurry heating.

The effect of the addition of hydrogen peroxide to nitric acid

suspension medium was also investigated for calcium, copper,

iron, magnesium and zinc. However, recoveries and slopes of

standard addition lines remained practically constant for all oneswhen hydrogen peroxide in the range of concentrations from 1

to 5% (m/v) was added.

After selecting the optimum conditions for each element

(Table 1), glycerol concentration (0.13%, v/v) and sample

amount slurried (3300 mg) were evaluated. Thus, the effect

of wetting agent concentration on recovery and slope of stan-

dard addition line for all elements studied was not significant.

While the sample amount slurried negatively affected to mag-

nesium and zinc recoveries for values higher than 6 and 100 mg,

respectively, or to calcium and zinc sensitivity for values higher

than 45 and 100 mg, respectively. On the other hand, an analyte

extraction percentage into the liquid phase of the slurry close to100% was reached for all elements.

3.3. Validation of the method

The proposed method was validated by the analysis of one

biological reference material using the optimum experimental

conditionsdescribedin Table 1. External calibrationwith a series

of aqueous standards was applied to iron, magnesium and zinc

determination. However, standard addition method was applied

to calcium and copper determination because of the slope of

external calibration line was significantly different with that

corresponding to standard addition line (t-test; p = 0.05). The

sample was analyzed 10 times. The results obtained for cal-cium, copper, iron, magnesium and zinc concentration in the

certified reference material by slurry sampling method and the

certified values are shown inTable 2.It can be observed that the

concentrations found of each ones were in good agreement with

the certified values (t-test;p = 0.05).

3.4. Analysis of fish tissue samples

Three tuna tissue samples (IIII) were analyzed by triplicate

using both microwave-assisted digestion and slurry sampling

methodologies. The results obtained for calcium, copper, iron,

magnesium and zinc determination are shown in Table 3. Recov-

Table 2

Calcium, copper, iron, magnesium and zinc concentration (g g1) in the certi-

fied reference material using microwave-assisted digestion and slurry sampling

methods

Element Certified valuea Obtained valuea,b,c Obtained valuea,b,d

Ca 3040 60 3100 50 3008 25

Cu 4.66 0.23 4.58 0.19

8/13/2019 1-s2.0-S003991400600662X-main

6/6

R.A. Bugallo et al. / Talanta 72 (2007) 6 065 65

(t-test;p = 0.05). Furthermore, the standard deviations obtained

for calcium, iron, magnesium and zinc using both method-

ologies were not significantly different (F-test; p = 0.05) and

mean relative standard deviations of the over-all method (n = 3)

for different concentration levels were 5.3, 6.1, 1.3 and 4.9%

for calcium, iron, magnesium and zinc, respectively, by slurry

sampling. Calcium, iron, magnesium and zinc recoveries were

greater than 95%. Finally, 10 tuna tissue samples were ana-

lyzed using slurry sampling and standard addition method in

order to check the absence of matrix interferences on iron,

magnesium and zinc determination by comparing the slopes

associated with standard addition lines and with external cal-

ibration lines. The differences between both slopes were only

significant (t-test;p = 0.05) for iron determination in some tuna

samples (0.02510.0389 l mg1). As expected, external calibra-

tion with aqueous standards can be used for magnesium and zinc

determination. The slopes of standard addition lines for calcium

varied between 0.00881 and 0.0186 l mg1.

The detection limits based on the amount necessary to yield

a net signal equal to three times the standard deviation of theblank were 22.8 8.0, 0.884 0.092, 5.070.76, 35.50.7

and 1.17 0.04g g1 for calcium, copper, iron, magnesium

and zinc determination, respectively. It should be noted that the

detection limits obtained by the proposed method for copper

and zinc are 34 times less than those reported for other slurry

sampling FAAS approaches [21]. Therefore, the slurry sampling

method proposed in this work permits to spread the application

field to the analysis of fish samples from high sea containing

lower copper and zinc concentrations than those found in human

scalp hair samples.

4. Conclusions

Slurry sampling was compared with microwave-assisted acid

digestion for the determination of calcium, copper, iron, magne-

sium and zinc in fish tissue samples by flame atomic absorption

spectrometry. The main advantages of using slurry sampling

are the elimination of a tedious and time-consuming step of

sample dissolution and quantitative extraction of all analytes

studied into the liquid phase of the slurry. Unlikethe microwave-

assisted digestion method, in which sample amount is limited,

long times are required for cooling the digestion vessel after the

microwave irradiation and concentrated acids are used, involv-

ing high blank values, sampling of the liquid phase present in

slurried samples, after microwave irradiation except for copper,does not require stabilizing agents nor finely ground mate-

rial without nebulizer blockage. The accuracy of both methods

was checked using a certified reference material. The results

obtained were in good agreement with the certified values for

all chosen analytes when slurry sampling method was used.

The precision (R.S.D.) was less than 5.3, 11.1, 6.1, 1.3 and

4.9% for calcium, copper, iron, magnesium and zinc at concen-

tration ranging of 233713, 3.7312.0, 21.058.8, 12241708

and 16.237.2g g1, respectively. Furthermore, the results

obtained using slurry technique were not significantly differ-

ent with those corresponding to microwave-assisted digestion

for calcium, iron, magnesium and zinc in the fish samples ana-

lyzed.

References

[1] D. Pozebon,V.L. Dressler,A.J. Curtius,J. Anal. Atom. Spectrom.13 (1998)

1101.

[2] R.G.L. Silva, S.N. Willie, R.E. Sturgeon, R.E. Santelli, S.M. Sella, Analyst

124 (1999) 1843.

[3] L.M. Costa, S.T. Gouveia, J.A. Nobrega, Anal. Sci. 18 (2002) 313.

[4] E.C. Lima, F. Barbosa Jr., F.J. Krug, M.M. Silva, M.G.R. Vale, J. Anal.

Atom. Spectrom. 15 (2000) 995.

[5] P.R.M. Correia, E. Oliveira, P.V. Oliveira, Anal. Chim. Acta 405 (2000)

205.

[6] B. Perez Cid, C. Silva, C. Boia, Anal. Bioanal. Chem. 374 (2002) 477.

[7] J. Sastre, A. Sahuquillo, M. Vidal, G. Rauret, Anal. Chim. Acta 462 (2002)

59.[8] P. Bermejo Barrera, A. Moreda Pineiro, A. Bermejo Barrera, Talanta 57

(2002) 969.

[9] M.C. Yebra, A. Moreno Cid, R.M. Cespon, S. Cancela, Atom. Spectrosc.

24 (2003) 31.

[10] K. Eilola, P. Peramaki, Analyst 128 (2003) 194.

[11] E.C. Lima, J.L. Brasil, A.H.D.P. Santos, Anal. Chim. Acta 484 (2003)

233.

[12] E. Lucker, K. Failing, T. Schmidt, Fresen. J. Anal. Chem. 366 (2000) 137.

[13] E.M.M. Flores, A.B. da Costa, J.S. Barin, V.L. Dressler, J.N.G. Paniz, A.F.

Martins, Spectrochim. Acta 56B (2001) 1875.

[14] M.G.R. Vale, M.M. Silva, B. Welz, R. Nowka, J. Anal. Atom. Spectrom.

17 (2002) 38.

[15] E.M.M. Flores, J.N.G. Paniz, A.P.F. Saidelles, E.I. Muller, A.B. da Costa,

J. Anal. Atom. Spectrom. 18 (2003) 769.

[16] E.C.Lima, F.J. Krug, A.T. Ferreira, F. Barbosa Jr., J. Anal. Atom. Spectrom.14 (1999) 269.

[17] M. Gonzalez, M. Gallego, M. Valcarcel, Talanta 48 (1999) 1051.

[18] P. Bermejo Barrera, A. Moreda Pineiro, J. Moreda Pineiro, T. Kauppila, A.

Bermejo Barrera, Atom. Spectrosc. 21 (2000) 5.

[19] S.J. Huang, S.J. Jiang, Analyst 125 (2000) 1491.

[20] E.C. Lima, F. Barbosa, F.J. Krug, J. Fresenius, Anal. Chem. 369 (2001)

496.

[21] P. Bermejo Barrera, A. Moreda Pineiro, J. Moreda Pineiro, A. Bermejo

Barrera, Fresen. J. Anal. Chem. 360 (1998) 707.

[22] P. Jacob, H. Berndt, J. Anal. Atom. Spectrom. 17 (2002) 1615.

[23] M.A. Bautista, C. Perez Sirvent, I. Lopez Garca, M. Hernandez Cordoba,

Fresen. J. Anal. Chem. 350 (1994) 359.

[24] J.H. Brown, J.E. Vaz, Z. Benzo, M. Velosa, Analyst 120 (1995) 1215.

[25] M.D. de Almeida, K.C. Leandro, C.V. da Costa, R.E. Santelli, M. de la

Guardia, J. Anal. Atom. Spectrom. 12 (1997) 1235.[26] J.A. Sweileh, Anal. Chim. Acta 448 (2001) 151.

[27] N. Mohamed, R.C. Fry, Anal. Chem. 53 (1981) 450.

[28] J. Stupar, R. Ajlec, Analyst 107 (1982) 144.

[29] S. Ro Segade, C. Bendicho, Ecotoxicol. Environ. Saf. 42 (1999) 245.

[30] N.J. Miller-Ihli, J. Anal. Atom. Spectrom. 9 (1994) 1129.

[31] I. Lavilla, J.L. Capelo, C. Bendicho, Fresen. J. Anal. Chem. 363 (1999)

283.