7/28/2019 ,mmbnb

1/6

Abstract A simplified and fast sample pretreatmentmethod based on ultrasound-assisted solubilization ofmetals from plant tissue with ethylenediaminetetraaceticacid in alkaline medium is described. Powdered unknownand certified plant samples (particle size < 50 m) wereslurried in the solubilization medium and subjected tohigh intensity ultrasonication by a probe ultrasonic proces-sor (20 kHz, 100 W). Metal solubilization can be accom-plished within 3 min using a 30% vibrational amplitudeand 0.1 M EDTA at pH 10, the supernatant obtained uponcentrifugation being used for analysis. The method is ap-plied to several food plants with unknown metal contentsand certified plant samples such as CRM GBW07605 tealeaves, BCR CRM 61 aquatic moss and BCR CRM 482lichen, with good trueness and precision. Intensive treat-ments with concentrated acids involving total matrix de-composition can be avoided. Metal determination (Ca, Cd,Mg, Mn, Pb and Zn) in the alkaline extracts was carriedout by flame and electrothermal atomic absorption spec-trometry.

Introduction

The use of ultrasonic metal extractions has been shown re-cently to be a realistic alternative to traditional sample pre-treatment mostly based on intensive attacks involving con-centrated acids and pressure [1, 2]. Thus, ultrasonic ex-tractions have proved to be successful for achieving quan-titative recoveries from industrial, environmental and bio-

logical matrices [38], only a few applications of this tech-nique being so far reported. In most applications, sampleswere sonicated in an ultrasonic cleaner bath at variabletemperatures in the range of 2585C for a period of time

comprised between 5 min and 2 h and using single acids(e.g. HCl, HNO3, HF), acid mixtures (e.g. HCl + HNO3,HNO3 + HF) or acids mixed with an oxidizing agent (e.g.H

2O

2).

The authors have reported several applications forsolid-liquid extraction of Cd, Pb, Hg and Cu from a vari-ety of real and certified samples using high intensityprobe sonication, where the importance of having anacidic medium as extractant is highlighted [913]. In allthese cases, extraction is based on the following reactionaccelerated by ultrasound:

ML + nH+Mn+ + HnL

Position and rate of establishment of this exchange equi-librium depend on the composition and texture of the bio-logical matrix, type of bonding of the metal ions, type andconcentration of the acid and temperature.

A significantly high concentration of acid has been re-ported to be needed for quantitative extraction by severalworkers, mostly using bath sonication [7], which may beinconvenient for final determination by some analyticaltechniques. Basic extractants such as the ammonium sul-fate/ammonium hydroxide buffer combined with bath son-ication, however, have provided good recoveries for anionicspecies such as Cr(VI) from spiked filter, sand and soil [3,5], but are inadequate for extraction of metals encapsulatedwithin cells or strongly bound to the matrix.

The use of alkali digestions of biological tissues assistedby microwaves or ultrasound has been reported by severalworkers [14, 15]. All these procedures involve total solu-bilization with a tissue solubilizer such as tetramethylam-monium hydroxide (TMAH) for trace metal determinationby flame or electrothermal atomic absorption spectrome-try (FAAS, ETAAS) [14, 15], inductively coupled plasmaatomic emission spectrometry (ICP-AES) [16], electrother-mal vaporization ICP-mass spectrometry [17] and cold va-por atomic absorption spectrometry (CVAAS) [18]. In somecases, addition of ethylenediaminetetraacetic acid to im-prove metal solubilization has been reported [14, 16]. Metalsolubilization by chelating agents from solid environmen-tal samples (e.g. sediments) is a well known process which

A.V. Filgueiras I. Lavilla C. Bendicho

Ultrasound-assisted solubilization of trace and minor metals

from plant tissue using ethylenediaminetetraacetic acidin alkaline medium

Fresenius J Anal Chem (2001) 369 : 451456 Springer-Verlag 2001

Received: 31 July 2000 / Revised: 23 October 2000 / Accepted: 5 November 2000

ORIGINAL PAPER

A. V. Filgueiras I. Lavilla C. Bendicho ()Departamento de Qumica Analtica y Alimentaria,Universidad de Vigo, Facultad de Ciencias (Qumica),As Lagoas-Marcosende s/n, 36200 Vigo (Spain)e-mail: bendicho@uvigo.es

7/28/2019 ,mmbnb

2/6

strongly influences metal mobilization in the environmentand can be used in the laboratory to assess the bioavail-able metal fraction [19].

In this work, the applicability of a chelating agent (i.e.

ethylenediaminetetraacetic acid) in alkaline medium forsolubilization of Ca, Cd, Fe, Mg, Mn, Pb and Zn fromplant tissue following high intensity ultrasonication isdemonstrated. The extraction method is assessed withseveral real and certified food and environmental plants,and metal determination in the extracts is carried out byflame and electrothermal atomic absorption spectrometry.Microwave-assisted digestion is used for comparison.

Experimental

Apparatus. A Perkin-Elmer model 373 atomic absorption spec-trometer (Norwalk, CT, USA) equipped with a 10 cm burner head

was used for determination of Ca, Fe, Mg, Mn and Zn. A Perkin-Elmer (berlingen, Germany) Model 4110 ZL atomic absorptionspectrometer was used for determination of Pb and Cd in combi-nation with a Perkin-Elmer AS-72 autosampler and a transverseheated graphite atomizer (THGA). Longitudinal Zeeman effectwas used for background correction.

Perkin-Elmer (Cd, Fe, Pb, Mn and Zn) and Cathodeon (Ca andMg) hollow cathode lamps were used as radiation sources. The in-strumental parameters are shown in Table 1. Thermal programs for

determination of Pb and Cd by ETAAS are shown in Table 2 andwere previously optimized.

A Retsch micro-mill model MM 2000 equipped with agateballs was used for grinding the plant samples. Powdered sampleswere subsequently passed over the 50 m mesh-size before ultra-

sonic extraction. Microwave-assisted digestion of unknown plantsamples was carried out with a CEM microwave sample prepara-tion system model MDS-2000 (Matthews, NC, USA). This systemwas equipped with a pressure monitoring option and allowed tooperate at a power of up to 630 50 W (100% power) program-mable in 1% increments. Advanced composite vessels (CEM) wereused for digestion.

Reagents. All reagents used were of analytical reagent-grade. Ul-trapure water from a Milli-Q water purifier was used throughout.65% m/m HNO3 and 48% m/m HF were used for acid digestion(Carlo Erba, Milano, Italy). Citric acid, sodium acetate, sodium di-hydrogen phosphate, sodium carbonate and ammonium chloride(Merck, Darmstadt, Germany) were used to make buffer solutions(0.1 M). The pH and composition of the buffer solutions was thefollowing: buffer solutions for pH 1, pH 2 and pH 6: 3.4 g of di-hydrogen sodium dihydrogen phosphate dissolved in 200 mL ofdeionized water and adding HCl up to pH 1, pH 2, pH 6, pH 7 andpH 8, respectively; buffer solution for pH 3: 4.2 g of citric acid dis-solved in 200 mL of deionized water and adding NaOH solutionup to pH 3; buffer solution for pH 4 and 5: 2.0 g of sodium acetatedissolved in 250 mL of deionized water and adding HCl up to pH 4and pH 5, respectively; buffer solution for pH 9: 2.0 g of am-monium chloride dissolved in 200 mL of deionized water andadding NaOH solution up to pH 9; buffer solution for pH 10: 2.1 gof sodium carbonate dissolved in 200 mL of deionized water andadding HCl solution up to pH 10; buffer solution for pH 11: 3.4 gof sodium dihydrogen phosphate dissolved in 200 mL of deionizedwater and adding NaOH solution up to pH 11.

Ethylenediaminotetraacetic acid (di- sodium salt) (Merck) wasused in alkaline medium (sodium hydroxide, Merck) for metal sol-ubilization.

Stock standard solutions of Ca, Mg, Mn and Zn (1000 g mL1)were obtained by dissolving the pure metals (Merck) or CaCO3(Aldrich, Milwaukee, WI, USA) for Ca with 1 + 1 (v/v) HCl. Stan-dard solutions for Pb and Cd (1000 g mL1) were made by dissolv-ing the nitrate salts (Merck) in 3% v/v HNO3. Calibration standardswere made by appropriately diluting the stock solutions in 0.1 MEDTA. All glasware and plasticware were rinsed with 5% v/v HNO3prior to use. The mixture of 750 g mL1 Pd (Merck) + 500 g mL1

Mg(NO3)2 (Merck) was used as a matrix modifier for ETAAS de-terminations. 5 L of matrix modifier and 5 L of sample/standardwere injected into the graphite furnace.

Solubilization procedure. A portion of 0.1 g of sample (particle sizeless than 50 m) was accurately weighed into polypropylene cen-trifuge tubes (50 mL volume) and 5 mL of extractant solution were

452

Table 1 Instrumental param-eters for FAAS and ETAASdeterminations

Flame atomic absorption spectrometry

Instrumental parameter Ca Fe Mg Mn Zn

Wavelength /nm 422.7 248.3 285.2 279.5 213.9Lamp current /mA 4 30 4 20 15Slit width /nm 0.7 0.2 0.7 0.2 0.7

Electrothermal atomic absorption spectrometry

Instrumental parameter Pb Cd

Wavelength /nm 283.3 228.8Spectral bandpass /nm 0.7 0.7Background correction Longitudinal Zeeman Longitudinal ZeemanLamp current /mA 10 10Pipetting sequence Diluent, matrix modifier, Diluent, matrix modifier,

sample / standard sample / standard

Table 2 Thermal programs for determination of Pb and Cd byETAAS

Stage Tempera- Hold Ramp/s Gas flow-rate/ ture/C time/s mL min1

Drying 110 30 1 250Drying 130 30 15 250Pyrolysis 300 20 10 250Pyrolysisa 600 (Cd) 30 10 250

900 (Pb)Atomizationa, b 1400 (Cd) 5 0 0

1700 (Pb)Cleaning 2450 3 1 250

a Thermal programs for lead and cadmium differed only in the py-rolysis and atomization temperaturesb Read was set up in the atomization stage

7/28/2019 ,mmbnb

3/6

added. Then, the sample was sonicated at room temperature for3 min at 30% of the nominal amplitude of the converter. After son-ication, samples were centrifuged for 4 min at 4500 rpm, and theclear supernatant taken for analysis. Blanks were run in the sameway. EDTA-matched calibration standards were used for FAASdeterminations whereas aqueous standards containing 0.2% v/vHNO3 were used for ETAAS determinations.

The target samples employed for this study were several foodplants with unknown metal concentration (clove, nutmeg, hot pa-prika, rosemary, orange blossom, cinnamon, basil, black tea, black

pepper and tarragon) and three certified reference materials, i.e.CRM GBW07605 tea leaves, BCR CRM 61 aquatic moss and BCRCRM 482 lichen. Both certified and unknown plant samples werepreviously ground and the fraction with particle size less than 50 mwas subjected to the metal solubilization procedure. For non certi-fied samples, microwave-assisted digestion was used according tothe procedure established elsewhere [20].

Results and discussion

Effect of pH on metal solubilization

In order to study the effect of pH on the ultrasound-as-sisted solubilization of the different metals, the cinnamon

sample was used as target sample. 0.1 g of the powderedsample were subjected to ultrasonic treatment with 5 mLof the corresponding buffer solution. Figure 1 shows thatthe solubilization of Ca, Mg, Mn and Zn decreases with de-creasing acidity. For Mg and Mn, absorbance remains con-stant up to pH 2. Absorbance of Ca decreases from pH 1.As can be observed, solubilization is negligible for Zn inthe 411 pH interval. For Mg, a decreasing absorbance is

seen in the 28 pH interval, but a 70% solubilization takesplace at pH 8 in comparison with that observed at pH 1.In the 910 pH interval, Mg solubilization is only about37%. Solubilization of Ca reaches a minimum at pH 8,where only 5% of Ca is solubilized in comparison with theamount solubilized at pH 1. In the latter case, solubiliza-tion increases slightly in the pH interval 911.

In conclusion, acidity is necessary in order to obtain agood metal recovery when applying ultrasonic treatment forsolubilization. This finding is in agreement with previouswork showing efficient metal solid-liquid extraction in acidmedia containing an acid concentration of at least 1% v/v(pH 1) [10, 13].

Influence of the EDTA concentration

Metal solubilization can be promoted by adding to the al-kaline medium a strong complexing agent such as EDTA.The logarithms of the formation constants for the com-plexes Zn-EDTA, Mn-EDTA, Mg-EDTA and Ca-EDTA,Cd-EDTA, Fe-EDTA and Pb-EDTA are 16.5, 14.0, 8.7 and10.7, 16.5, 25.1 and 18.0, respectively [21]. For an effi-cient complexation with EDTA a pH value > 10 should beemployed. At pH 10, the logarithms of the conditional for-mation constants (K) with EDTA are 13.6, 13.4, 7.4, 9.3,15.5, 14 and 14.8 for Zn, Mn, Mg, Ca, Cd, Fe and Pb, re-spectively [21]. The influence of EDTA concentration onmetal solubilization was studied for Ca, Fe, Mg, Mn and Znin the concentration range of 0.010.1 M at pH 10 (Fig.2).In all cases, a remarkable increase in the solubilization ef-ficiency is observed. Incomplete solubilization is observedfor Fe which likely may be attributed to contaminationduring drying and packing operations from stainless steel.Metallic impurities would need oxidant conditions so thatthey could be solubilized. Incomplete Fe solubilizationcould also arise from the slow complexation kinetics ofFe(III) with EDTA.

Experiments carried out with CRMs showed that solu-bilization of Pb and Cd is also significantly enhanced whenadding EDTA to the alkaline medium. The latter elementshave also been efficiently extracted from slurried biologi-cal samples following ultrasonication in acidic media [10,13].

Influence of sonication parameters

Ultrasound effects are enhanced when increasing the vi-brational amplitude of the probe so that effective cavitationin the liquid medium takes place [1]. For all metals studied,

453

Fig.1 Effect of pH on metal solubilization following ultrasonicextraction

7/28/2019 ,mmbnb

4/6

good recoveries are reached with at least a 20% vibrationalamplitude. The only exception is Fe, which displays in-creasing solubilization with increasing amplitude in theinterval studied (1070%). Nevertheless, incomplete solu-bilization is observed for the latter metal even using themaximum amplitude allowable.

Under optimum vibrational amplitude, pH and EDTAconcentration, a time of at least 3 min was seen to beneeded for achieving quantitative solubilization being sim-ilar to that needed for solid-liquid extraction in acid media.A longer sonication time seems to increase Fe solubiliza-tion, but results are far from being quantitative.

The mass-to-volume ratio also influences metal solubi-lization. For a 5 mL volume, quantitative solubilization isobserved for Zn, Mn, Mg and Ca for a sample mass lessthan 0.15 g. In contrast, a decreasing Fe solubilization oc-curs in the range of masses studied (0.050.2 g). Particlesize has been demonstrated to be among the most criticalparameters influencing ultrasound-assisted extraction pro-cesses [1013]. Sample fractionation by sieving to yieldparticle size fractions in the range of less than 25 m,

2550 m, between 50100 m and 100150 m, showedthat a particle size less than 50 m should be employedfor efficient solubilization.

Figures of merit

Limits of detection (LODs) defined as 3 of a blank

EDTA solution were 2, 0.1, 1.3 and 0.7 g g1

in plant tis-sue (dry weight), for Ca, Mg, Mn and Zn, respectively.Characteristic masses for determination of Cd and Pb

by ETAAS were 2.3 and 44 pg which are in agreement withthose provided by the manufacturer (1.3 and 30 pg, respec-tively). LODs of Cd and Pb in plant tissue (dry weight)were 8 and 93 ng g1 when metals are solubilized from a100 mg sample mass treated with a 5 mL extractant vol-ume. Between-batch precision values, expressed as relativestandard deviation (RSD), were usually below 2% for Ca,Mg, Mn and Zn and in the 510% range for Cd and Pb.Solubilization blank expressed as equivalent metal con-centration (g g1) was only significant for Zn (6.3 g g1)

when no dilution of the extractant solution was performed.Nevertheless, a 1:10 dilution of the extractant solutionwas needed for most samples, and therefore the blank so-lution was much lower for this metal. For Mn, the undi-luted blank was below the LOD. For other metals whichare present at high level in the samples analyzed, dilutionfactors as high as 1:500 were needed, and therefore, solu-bilization blanks were below the corresponding LOD.

Analytical results

Several food plants were analyzed for determination of Ca,Mg, Mn and Zn using the ultrasound-assisted solubiliza-tion method with EDTA at pH 10. Microwave-assisted di-gestion in closed vessels was used for comparison. Rele-vant parameters of the calibration curves for the metals de-termined are shown in Table 3.

As can be observed (Table 4), differences (%) betweenthe average concentrations obtained with the two methodsbeing usually within the 15% interval.

On the other hand, a good agreement is also observedbetween the certified and found concentration values for

454

Table 3 Parameters of the calibration curves for determinationsby FAAS and ETAAS

Element Slopea Zero Correlationintercept coefficient

Ca 0.0497 +0.0007 0.9998Mg 0.6681 0.0008 0.9999Mn 0.0792 +0.0001 0.9999Zn 0.2012 0.0002 0.9999Cd 0.0470 0.0017 0.9998Pb 0.00527 0.0082 0.9986

a Units for the slope were mL g1 (Ca, Mg, Mn and Zn) andmL ng1 (Cd and Pb)

Fig.2 Effect of EDTA concentration on metal solubilization at pH10 following ultrasonic extraction

7/28/2019 ,mmbnb

5/6

CRM GBW07605 tea leaves, BCR CRM 61 aquatic mossand BCR CRM 482 lichen for metals determined by FAAS(Ca, Mg, Mn and Zn) and ETAAS (Pb and Cd) (Table 5).Integrated absorbance was taken as the measurement sig-

nal for Cd and Pb determinations. Fe was not determinedas a result of the insufficient solubilization displayed bythis element.

Conclusions

Metals encapsulated within plant cells can be efficientlyreleased by ultrasonication combined with a chelating agentin alkaline medium. Metal determination in the extractscan be performed by flame and electrothermal atomic ab-sorption spectrometry with good trueness and precision.

The method proposed can be an attractive alternative so asto avoid intensive treatments with acids which are usuallytedious and time-consuming. Unlike solubilization proce-dures based on TMHA, potential interferences are sepa-

rated with the EDTA-solubilization procedure, and addi-tionally, some drawbacks inherent with the use of the lat-ter compound such as high blank values and odor createdduring the ETAAS ashing and atomization stages are over-come. A possible drawback concerning the use of EDTAas solubilizing agent combined with ultrasonic treatmentcould be the contamination risk as a result of metal leach-ing from the extraction vessel.

Acknowledgements This work has been financially supported bythe Ministerio Espaol de Ciencia y Tecnologa in the frameworkof project PB981081 and IN960101.

455

Table 4 Analytical results forsolubilization of Ca, Mg, Mnand Zn from food plants usinghigh intensity ultrasonicationin an EDTA-alkaline medium

a Average value standard de-viation (N = 3)

Element Plant sample EDTA-solubilization Microwave digestion Difference(g g1)a (g g1)a (%)

Zn Clove 14.0 0.1 14.5 0.6 3Nutmeg 18.8 0.2 19.6 0.6 4Hot paprika 27.3 0.2 33.9 0.4 19Rosemary 21.8 0.1 27.6 1.4 21

Mn Orange blossom 22.6 0.1 19.9 0.2 +14

Cinnamon 502 11 495 10 +1Nutmeg 36.5 0.7 42.7 0.1 15Hot paprika 27.4 0.3 29.8 1.1 8Rosemary 50.9 0.2 41.6 0.6 +22Black tea 798 2 735 13 +9

Mg Basil 8352 29 7458 153 +12Orange blossom 2417 7 2244 16 +8Cinnamon 1235 7 1435 19 14Clove 3418 11 3309 65 +3Hot paprika 2658 17 2927 73 9Black tea 2550 10 2458 44 +4Black pepper 2285 7 1971 19 +16

Ca Basil 21.8 0.1 21.4 0.4 +2

Cinnamon 7062 88 8082 132 13Clove 10.7 0.1 10.0 0.1 +7Tarragon 7306 7 8768 181 17Black pepper 2314 62 2383 46 3

Table 5 Determination of Ca,Cd, Mg, Mn, Pb and Zn in cer-tified reference materials fol-lowing ultrasound-assisted sol-ubilization with EDTA in alka-line medium

a Average value confidenceinterval (p = 0.05)b Average value standard de-viation (N = 3)c Metal concentration expressedas mg g1

Plant sample Element Certified value Found value Difference(g g1)a (g g1)b (%)

CRM GBW 07605 Cac 4.3 2 4.1 0.1 4.6Tea leaves Mgc 1.7 0.1 1.8 0.04 +5.9

Mn 1240 40 1195 9 3.6Zn 26.3 0.9 28.5 0.5 +8.3

BCR CRM 61 Mn 3771 78 3612 52 4.2Aquatic moss Zn 566 13 581 21 +2.6

Cd 1.07 0.08 0.98 0.1 8.4Pb 64.4 13 59.6 4 7.4

BCR CRM 482 Zn 100.6 2.2 106 3 +5.5Lichen Cd 0.56 0.02 0.52 0.05 7.6

Pb 40.9 1.4 37.8 3 7.5

7/28/2019 ,mmbnb

6/6

References

1.Bendicho C, Lavilla I (2000) Extraction: Ultrasound Extrac-tions. Encyclopedia of Separation Science, Wilson ID, AdlardER, Cooke M, Poole CF (eds), Vol 3, Academic Press, London,pp 14481454

2.Bendicho C, Lavilla I (2000) Applications of Ultrasound-As-sisted Metal Extractions. Encyclopedia of Separation Science,Wilson ID, Adlard ER, Cooke M, Poole CF (eds), Vol 9, Aca-

demic Press, London, pp 442144263.Ashley K (1998) Trends Anal Chem 17:3663724.Ndungu K, Djane NK, Malcus F, Mathiasson L (1999) Ana-

lyst 124:136713725.Wang J, Ashley K, Kennedy ER, Neumeister C (1977) Analyst

122:130713126.Minami H, Honjyo T, Atsuya I (1996) Spectrochim Acta, Part

B 51:2112207.El Azouzi H, Cervera ML, de la Guardia M (1998) J Anal

Atom Spectrom 13: 5335388.McKiernan JW, JT Creed, Brockhoff CA, Caruso JA, Loren-

zana RM (1999) J Anal Atom Spectrom 14:6076139.Lavilla I, Capelo JL, Bendicho C (1999) Fresenius J Anal Chem

363:28328810.Capelo JL, Lavilla I, Bendicho C (1998) J Anal Atom Spectrom

13:12851290

11.Rio-Segade S, Bendicho C (1999) J Anal Atom Spectrom 14:19071912

12.Capelo JL, Filgueiras AV, Lavilla I, Bendicho C (1999) Ta-lanta 50:905911

13.Amoedo L, Capelo JL, Lavilla I, Bendicho C (1999) J AnalAtom Spectrom 14:12211226

14.Zhou CY, Wong MK, Koh LL, Wee YC (1996) Talanta 43:10611068

15.Silva RGL, Willie SN, Sturgeon RE, Santelli RE, Sella SM(1999) Analyst 124:18431846

16.Uchida T, Isoyama H, Yamada K, Oguchi K, Nakagawa G(1992) Anal Chim Acta 256:27728417.Pozebon D, Dressler VL, Curtius AJ (1998) J Anal Atom Spec-

trom 13:1101110518.Tao G, Willie SN, Sturgeon RE (1998) Analyst 123:1215121819.Quevauviller PH, Rauret G, Ure A, Bacon J, Muntau H (1997)

The Certification of the EDTA-and Acetic-Extractable Contents(Mass Fractions) of Cd, Cr, Cu, Ni, Pb and Zn in sewage sludgeAmmended Soils, Report EUR 17127 EN, Brussels, p 85

20.Lavilla I, Filgueiras AV, Bendicho C (1999) J Agric Food Chem47: 50725077

21.Ringbom A (1979) Formacin de complejos en Qumica Ana-ltica, 1st edition, Alhambra, Madrid, Spain, pp 387388, 416

456