TECHNICAL NOTE

Iodine determination in food by inductively coupledplasma mass spectrometry after digestionby microwave-induced combustion

Márcia F. Mesko & Paola A. Mello & Cezar A. Bizzi &Valderi L. Dressler & Guenter Knapp &

Érico M. M. Flores

Received: 26 February 2010 /Revised: 15 April 2010 /Accepted: 19 April 2010 /Published online: 13 May 2010# Springer-Verlag 2010

Abstract Iodine determination in food samples was per-formed by inductively coupled plasma mass spectrometry(ICP-MS) after digestion by microwave-induced combustion(MIC). Sample masses up to 500 mg of bovine liver, cornstarch, milk powder, or wheat flour were completelycombusted using the MIC system. Ammonium nitrate(6 mol l−1 solution, 50 μl) was used as an aid for ignitionand vessels were charged with 15 bar of O2. The use of H2O,0.9 mmol l−1 H2O2, 10 to 50 mmol l−1 (NH4)2CO3 and56 mmol l−1 tetramethylammonium hydroxide was investi-gated as absorbing solutions, as well as the suitability ofperforming a reflux step after the combustion process.

Digestion of food samples by pressurized microwave-assisted acid digestion, microwave-assisted extraction andconventional extraction of iodine in alkaline solution werealso evaluated. Iodine recoveries higher than 99% wereobtained using MIC and 50 mmol l−1 (NH4)2CO3 or56 mmol l−1 tetramethylammonium hydroxide as absorbingsolution and with 5 min for the reflux step. Accuracy wasevaluated using certified reference materials (bovine muscle,corn bran, and milk powder) and agreement better than 97%was obtained. The limit of quantification by MIC and furtherICP-MS determination was 0.002 µg g−1. Blanks werealways low and no memory effects were observed. Digestionby MIC allowed the processing of up to eight samples byeach run in 25 min with high efficiency of digestion (residualcarbon content lower than 1%) providing a suitable mediumfor further iodine determination by ICP-MS.

Keywords Iodine determination . Food samples .

Microwave-induced combustion . Inductively coupledplasma mass spectrometry

Introduction

Iodine plays an important role in human nutrition and it is wellknown as an essential trace element for growth, developmentand human essential functions. Iodine deficiency results innutritional disorders as the thyroid gland disease and fetuscongenital anomalies [1]. On the other hand, adversereactions have been associated with excessive iodine intakeas the increase of thyroid volume. Therefore, humans need acontrolled daily uptake of iodine to ensure an optimal leveland as, a consequence, many food products are enriched withiodine [1]. With regard to iodine determination, it is always

Awarded an ABC Poster Prize on the occasion of the 2010 WinterConference on Plasma Spectrochemistry held in Fort Myers, Florida,January 4–9, 2010.

Electronic supplementary material The online version of this article(doi:10.1007/s00216-010-3766-9) contains supplementary material,which is available to authorized users.

M. F. MeskoInstituto de Química e Geociências,Universidade Federal de Pelotas,96010-610 Pelotas, RS, Brazil

P. A. Mello :C. A. Bizzi :V. L. Dressler : É. M. M. FloresDepartamento de Química, Universidade Federal de Santa Maria,97105-900 Santa Maria, RS, Brazil

P. A. Mello :C. A. Bizzi :V. L. Dressler : É. M. M. Flores (*)Instituto Nacional de Ciência e Tecnologia de Bioanalítica,Campinas, SP, Brazile-mail: flores@quimica.ufsm.br

G. KnappInstitute for Analytical Chemistry, Micro- and Radiochemistry,Graz University of Technology,8010 Graz, Austria

Anal Bioanal Chem (2010) 398:1125–1131DOI 10.1007/s00216-010-3766-9

an analytical problem, especially at the relatively natural lowconcentration in food, which is in general below 1 µg g−1 [2, 3].

The analytical techniques that have been used includespectrophotometry [4, 5], ion selective electrode [6, 7], X-ray fluorescence spectrometry [8, 9], neutron activationanalysis [3, 10, 11], ion chromatography (IC) [12], gaschromatography mass spectrometry [13, 14], voltammetry(mainly cathodic stripping voltammetry) [15, 16], induc-tively coupled plasma optical emission spectrometry (ICPOES) [8, 17] and inductively coupled plasma massspectrometry (ICP-MS) [2, 10, 18–22]. For the determina-tion step ICP-MS is a very powerful technique and itcombines a suitable throughput and high sensitivity usefulfor low-level iodine determination in food.

However, despite some analytical techniques allowingiodine determination with low limits of detection (LOD), as,e.g., ICP-MS, the main problem is still the sample preparationstep [18]. Current sample preparation methods are time-consuming and wet digestion using concentrated acids are notwidely applicable due to iodine losses by volatilization, as HIor I2, that result in nonquantitative recoveries [10, 18, 20–23].In addition, memory effects in the ICP-MS introductionsystem have been reported [18, 24, 25]. In this sense, samplepreparation approaches have been developed using sampledilution or dispersion in alkaline solution as ammonia,sodium hydroxide, tetramethylammonium hydroxide(TMAH) and water-soluble tertiary amines solution (CFA-C) [10, 18, 20–22]. Alternative procedures using oxidizingacid mixtures can be used, whereby the oxidizing potentialmust be high enough for oxidation of iodide into non-volatileiodate. However, these procedures are time-consuming;reagents generally present high blank values or could causeinterferences in the further determination step [22, 23, 26–28].

Combustionmethods are particularly efficient to decomposeorganic matrices due to the high temperature achieved. Organicsamples can be burnt in a closed vessel filled with oxygen usingSchöniger combustion flask and combustion bombs. Usingthese systems, iodine can be absorbed as iodide in an alkalinemedium [18–21, 28, 29]. Schöniger combustion flask hasbeen applied for nutritional and biological materials using0.5% v/v TMAH [18], CFA-C solution [10] and 0.1 moll−1 NaOH as absorbing solutions [20]. However, digestionusing this system presents some limitations for trace analysisbecause, in general, lower than 100 mg of sample can be burntusing a conventional vessel of 1 l. For combustion bombs, upto 1.5 g can be digested and absorbing solutions, similar tothose used for Schöniger flask, could be used for analyteabsorption [18, 21, 29]. Nevertheless, in spite of its suitabilityfor further iodine determination, low sample throughput canbe pointed out as the main drawback related to Schönigerflask and also combustion bomb; cleaning of vessels istroublesome and a reflux step that could speed and increasethe analyte absorption is not possible with these systems [29].

Microwave-induced combustion (MIC) system allowscombining some advantages of classical combustion tech-niques with those achieved using conventional closedsystems heated by microwave radiation [29]. This methodis performed in the same system conventionally used formicrowave-assisted wet digestion and the only instrumentalchange is a small quartz sample holder which is placed intothe quartz vessel containing a suitable absorbing solution.A reflux step can be applied after combustion, assuring aneffective washing of vessel walls and holder surface, whichis advantageous over Schöniger flask and combustionbombs, where cleaning of vessel must be performedmanually. Microwave-induced combustion has been appliedfor digestion of biological samples [30–32], carbon blackcontaining elastomers [33], coal [34] and for crude oil andcrude oil products [35–40], and for both metals and non-metals determinations.

The purpose of this work was to demonstrate the feasibilityof MIC for fast and efficient digestion of food samples forfurther iodine determination by ICP-MS, allowing accuratequantification of iodine at low level. Different types of foodsamples were evaluated. Sample mass and oxygen pressurewere investigated as well as the type and concentration ofabsorbing solution for MIC procedure. The suitability ofperforming a reflux step was studied using recovery assays fora bovine liver sample. In addition, microwave-assisted aciddigestion, microwave-assisted extraction and conventionalextraction both using alkaline solutions were carried out forcomparison with the results obtained usingMIC. Digests werealso analyzed by IC and accuracy was checked using certifiedreference materials (CRMs).

Experimental

Instrumentation

A microwave sample preparation system (Multiwave 3000,software version v1.27-Synt, Anton Paar, Graz, Austria)equipped with high-pressure quartz vessels (internal volumeof 80 ml, maximum operating temperature and pressure of280 °C and 80 bar, respectively) was used for MIC and alsofor microwave-assisted acid digestion and microwave-assisted alkaline extraction. Commercial combustion quartzholders (Anton Paar, part number 16427) were used for MIC.Conventional extraction in alkaline media was performed in aheating block (Tecnal, Piracicaba, Brazil) with digital temper-ature control (TE007D) equipped with polytetrafluorethylene(PTFE) vessels (50 ml of internal volume).

Iodine determination was carried out in an inductivelycoupled plasma mass spectrometer (PerkinElmer-SCIEX,Model Elan DRC II, Thornhill, Canada) equipped with aconcentric nebulizer (Meinhard Associates, Golden, USA),

1126 M.F. Mesko et al.

a cyclonic spray chamber (Glass Expansion, Inc., WestMelbourne, Australia) and a quartz torch with a quartzinjector tube (2 mm i.d.). Instrumental performanceoptimization, including nebulizer gas flow rate, ion lensvoltage, and torch alignment, was carried out following theinstructions of the manufacturer [41]. The operationalconditions are shown as Electronic supplementary material.Argon (99.996%, White Martins–Praxair, São Paulo, SP,Brazil) was used for plasma generation, for nebulizationand as auxiliary gas. Digests obtained by MIC were alsoanalyzed by IC using a modular chromatographic system(Metrohm Ion Analysis, Herisau, Switzerland), with ananion-exchange column (Metrosep A Supp 5, polyvinylalcohol with quaternary ammonium groups, 150×4 mm i.d.), a guard column (Metrosep A Supp 4/5 Guard), achemical suppressor module and a conductivity detector. Asample loop of 100 μl was used. The mobile phase was3.2 mmol l−1 Na2CO3 and 1.0 mmol l−1 NaHCO3 at a flowrate of 0.7 ml min−1. A simultaneous inductively coupledplasma optical emission spectrometer (Spectro Ciros CCD,Spectro Analytical Instruments, Kleve, Germany) with axialview configuration was used for residual carbon content(RCC) determination according to previous works [34].

A cryogenic mill (Spex CertiPrep, Model 6750, Metu-chen, USA) and a hydraulic press (Manual Hydraulic Press15 Ton, Specac, Orpington, UK) was used for sampleconditioning. An infrared thermometer (Ultimax InfraredThermometer, Ircon, Niles, IL) equipped with a close-upVX-CL1 lens was used for measurement of the temperatureachieved during the combustion.

Reagents, samples, and sample preparation

Milli-Q water (18.2 MΩ cm, Millipore, Billerica, USA) andanalytical-grade reagents (Merck, Darmstadt, Germany)were used throughout. Concentrated nitric acid (Merck,Darmstadt, Germany) was purified using a sub-boilingdistillation system (Model DuoPur, Milestone, Sorisole,Italy). Iodide stock reference solution was prepared bydissolving KI salt (Merck, Darmstadt, Germany) in waterand standard solutions were prepared before use by serialdilution. For the determination by ICP-MS workingsolutions were prepared in 10 mmol l−1 (NH4)2CO3.Absorbing solutions were prepared by dilution of acommercial 25% TMAH solution (Sigma-Aldrich, St.Louis, USA), a solution of 50% w/w H2O2 (Vetec,Diadema, Brazil) and by dissolution of ammonium carbon-ate salt in water. Ammonium nitrate solution (6 mol l−1)used as igniter for MIC procedure was prepared bydissolving the solid reagent in water.

Small disks of paper (12 mg, 15 mm of diameter) withlow ash content (Black Ribbon Ashless, Schleicher andSchuell GmbH, Dassel, Germany) were used as aid for the

ignition process. Paper was previously cleaned by immer-sion in 10% (v/v) HNO3 and further in absolute ethanol by20 min in an ultrasonic bath, washed with water and driedin a class 100 laminar bench (CSLH-12, Veco, Brazil). Thecleaning procedure of vessels and holders was carried outwith 6 ml of 14 mol l−1 HNO3, in the microwave oven at1,400 W for 10 min and 0 W for 20 min (cooling step).

CRMs from National Institute of Standards and Tech-nology (NIST) of bovine muscle powder (NIST 8414), non-fat milk powder (NIST 1549), corn bran (NIST 8433), andwhole milk powder (NIST 8435) and from EuropeanCommission Institute for Reference Materials and Measure-ments—Community Bureau of Reference (BCR) of skimmilk powder (BCR150 and BCR 151) were used.

Samples of bovine liver, corn starch, milk powder, andwheat flour were purchased from a local market. Bovineliver was prepared by lyophilization, ground in a cryogenicmill, and dried at 60 °C for 1 h. Samples and CRMs wereprepared as pellets (100 to 500 mg) using a hydraulic pressset at 3 tons during 1 min for digestion by MIC.

Digestion by microwave-induced combustion

Sample pellets (100 to 500 mg) were placed together the filterpaper on the quartz holder and 50 μl of 6 mol l−1 NH4NO3

solution were added on the paper. Then, sample holder wasintroduced into the quartz vessels previously charged with6 ml of absorbing solution (ammonium carbonate solutionsfrom 10 to 50 mmol l−1, 56 mmol l−1 TMAH, 0.9 mmol l−1

H2O2 or water). After closing and capping of the rotor,vessels were pressurized with oxygen (10 to 25 bar) usingthe valve originally designed for pressure release afterconventional wet digestion. Microwave program for MICwithout reflux was 1,400 W for 1 min (combustion only) and0 W for 20 min (cooling step). For MIC with reflux,microwave program was 1,400 W for 5 min (samplecombustion and reflux of absorbing solution) and 0 W for20 min (cooling step). After digestion, solutions weretransferred to polypropylene vessels, diluted with water to30 ml and iodine was further analyzed by ICP-MS. Thevolume and concentration of the NH4NO3 and the mass ofpaper for ignition process were used according to previouswork for digestion of biological materials [31].

Microwave-assisted acid digestion, microwave-assistedalkaline extraction, and conventional alkaline extraction

A high-pressure microwave-assisted acid digestion proce-dure was also carried out based on previous works [18].The procedure was performed using 250 mg of sample and6 ml of concentrated HNO3. The microwave heatingprogram was (1) 1,400 W for 50 min (ramp of 10 min)and (2) 0 W for 20 min (cooling step). After cooling,

Iodine determination in food by inductively coupled plasma mass spectrometry after digestion by microwave-induced combustion 1127

digests were diluted with water to 30 ml in a polypropylenevessel and iodine was determined by ICP-MS. Based onpreviously reported procedures using TMAH for iodineextraction [2, 18, 22], a microwave-assisted extractionprocedure in alkaline medium was also evaluated. Samples(200 mg) were placed in the quartz vessel and 6 ml of0.11 mol l−1 solution of TMAH were added. Heatingprogram was (1) 1,400 W for 50 min (ramp of 10 min) and(2) 0 W for 20 min (cooling step). Temperature was set upto 90 °C. In addition, a conventional extraction procedurein alkaline medium was performed using 200 mg of sampleand 5 ml of 0.11 mol l−1 TMAH solution by 3 h at 90 °C ina heating block following the reported procedures [2, 18,22]. After cooling, extracts form extraction procedures werediluted with water to 30 ml in a polypropylene vessel andthen centrifuged at 3,000 rpm for 10 min prior todeterminations by ICP-MS.

Results and discussion

Optimization of operational conditionsfor microwave-induced combustion

An initial study was carried out to evaluate the maximumsample mass that could be burnt in MIC procedure and alsothe oxygen pressure necessary for complete combustionunder safe conditions. Pellets of bovine liver and milkpowder samples from 100 to 500 mg were digested using15 bar of O2. This pressure was selected based on theprevious studies for digestion of biological materials byMIC [31]. For digestion of 500 mg with 15 bar of O2, thecombustion always occurred (n=10) for all samples andthis value was chosen as initial pressure for MIC. Themaximum pressure achieved in the system was 38±3 barand 27±3 bar for bovine liver and for milk powder,respectively, that represents about 34% to 48% of themaximum pressure allowed for the system. This result isimportant because if a procedure could be effective fordigestion of relatively high sample mass it could beadvantageous for low-level iodine determination. There-fore, the possibility to digest 500 mg of sample using MICcan be pointed out as an important feature in comparisonwith other procedures as, e.g., Schöniger flask.

Digestion of 500 mg of milk powder was evaluatedusing 10 to 25 bar of oxygen as initial pressure. It wasobserved incomplete combustion with carbon deposits onsample holder and vessel surfaces using 10 bar of oxygen.Using higher pressures (15, 20, and 25 bar of O2)combustion always occurred (n=10) and no residues wereobserved. Then, 15 bar of oxygen was considered enoughfor digestion of 500 mg using MIC and this condition wasselected to further studies. For 500 mg of milk powder,

using 15 bar as initial oxygen pressure, quick ignition ofsample was observed (3 to 9 s) and the combustion timewas about 30 s. Using the selected conditions, thetemperature for combustion of all samples was higher than1,400 °C and a white and bright radiation during the wholecombustion process was also an indicative about the hightemperature achieved. However, despite the high tempera-ture, no damages were observed in the holders and vessels.

Influence of absorbing solution and reflux step for iodinerecovery using MIC procedure

It is well known that iodine can be lost as HI or I2 using wetdigestion procedures and then a suitable solution should beused to absorb gaseous products delivered during combus-tion [22]. In the present work, a systematic study was carriedout in order to evaluate the suitability of absorbing solutionfor iodine. Spikes were performed with sample pellets of300 mg (bovine liver). Absorbing solutions were selectedbased on previous works using Schöniger combustion flaskand combustion bombs [29]. The use of 10 to 50 mmol l−1

(NH4)2CO3 [34], 56 mmol l−1 TMAH [18], 0.9 mmol l−1

H2O2 [29] solutions and water as absorbing solutions wasevaluated. Previous works using MIC have shown that theuse of a reflux step allows assuring quantitative recoveriesonce it results in a better washing of internal parts ofcombustion vessel [29]. When MIC is used without refluxstep, the microwave power is applied only by 1 min which isenough for sample ignition and combustion. If a reflux stepis used, microwave radiation can be applied, for example, by5 min, and after sample combustion (about 30 s) theabsorbing solution is immediately refluxed promoting betteranalyte absorption. In this study, for both procedures, MICwithout or with reflux, an additional step for cooling of20 min (without microwave irradiation) was used, makingthe total time of the procedure 21 or 25 min, respectively.

Spike recoveries for this study are shown in Fig. 1. Lowrecoveries were obtained without reflux step and they werefrom 62% to 72% for all the evaluated solutions. Previousworks using Schöniger flask and combustion bombs havereported longer waiting times after combustion for halogens(usually 30 to 60 min) in order to obtain a completeabsorption [29]. In the present study, as incompleterecoveries were obtained for MIC without reflux step andwaiting for 20 min, a longer waiting time (30 min) wasevaluated. However, even for this time, recoveries were nothigher than 85% and relative standard deviation (RSD) fordigestion ranged from 9% to 15%.

As expected, recoveries were improved using the refluxstep (5 min of microwave irradiation and 20 min for cooling).For water, they were not higher than 87% and using0.9 mmol l−1 H2O2 solution recoveries were 95.9±3.2%.When TMAH at 56 mmol l−1 was used as absorbing solution,

1128 M.F. Mesko et al.

recoveries were 99.3±3.7%. Solutions of 10 and 25 mmol l−1

(NH4)2CO3 resulted in recoveries between 92% and 95%.Quantitative recoveries were also obtained for 50 mmol l−1

(NH4)2CO3 solution (99.7±2.8%). For all the experimentsusing the reflux step, the RSD was lower than 3.9% (Fig. 1).

Based on the results obtained for spike recoveries, it wasconsidered that 56 mmol l−1 TMAH or 50 mmol l−1

(NH4)2CO3 solutions can be both used for MIC procedure.At this point, ammonium carbonate can be considered asimpler, less toxic, and less expensive medium in compar-ison to TMAH and it was chosen for MIC procedure. Blanksfor both solutions were lower than 5 ng l−1. When TMAHwas used, the temperature for reflux was set at 90 °C becausethis solution could decompose in higher temperatures [42].In fact, if temperature for MIC was not limited for 90 °C, thetemperature of solution was about 180 °C during the refluxstep, the pH of solution after procedure was about 3, andonly 85% of iodine recovery was obtained. Using the refluxstep at 90 °C for TMAH, pH after combustion was 7 andrecoveries close to 100% were obtained. The pH for(NH4)2CO3 was lower (<3) using 10 and 25 mmol l−1

solutions and, for 50 mmol l−1 solution, pH was about 7 forMIC followed by reflux. When water or H2O2 solutions wereused, the final pH ranged from 1 to 4 and, probably, theacidity of the digests can be the reason of the low recoveries.Although the use of high alkaline solutions could causecorrosion in quartz surface of vessels this problem was notobserved in this study after at least 150 combustions.

Determination of iodine after MIC and other samplepreparation procedures

A bovine liver sample and CRMs of milk powder BCR 150and BCR 151 were digested using MIC, high-pressuremicrowave-assisted acid digestion, microwave-assisted al-kaline extraction, and conventional alkaline extraction, in

order to evaluate the performance of these sample prepa-ration procedures against the performance of MIC. Solutionof TMAH has proven to be a suitable medium for use withICP-MS (extraction procedures commonly use solutionsfrom 1% to 25%, at 90 °C, by 3 h) [2, 18, 22]. In this way,for this study, alkaline extractions were performed using a0.11 mol l−1 TMAH solution (1% v/v). Ammoniumcarbonate (50 mmol l−1) was used as absorbing solutionfor MIC, with reflux step.

As can be observed in Table 1, the results obtained byMIC and by alkaline extraction (microwave-assisted andconventional procedures) for bovine liver were in goodagreement (no statistical difference, 95% confidence level).When microwave-assisted acid digestion was used, resultswere lower than those obtained by other procedures (about67% in relation to MIC) that are probably due to iodinelosses in acidic medium. Results for CRMs also showedthat lower concentrations were obtained when using aciddigestion. Despite some procedures reporting the suitabilityof acid digestion as sample preparation method for iodinedetermination, they usually require the use of perchloricand sulfuric acid combined to nitric acid to assure anoxidizing system to convert iodide into non-volatile iodate[18, 28]. In this work, the use of these acids was avoided inview of security and simplicity for the proposed procedureand mainly due to possible interferences in determinationstep [23, 24]. Results for MIC, microwave-assisted andconventional alkaline extraction were in agreement with thecertified values for CRMs. Higher RSDs were observed forboth extraction methods that can be related to the lowerdigestion efficiency once only a leaching is observed. Inaddition, the effect of particle size in the effectiveness ofthe leaching procedure has been also reported [22] and thiscould contribute for incomplete recoveries. After extraction,a solid dark residue from sample matrix remained in vesselsand after dilution the sample required a centrifugation step

(NH4)2CO3, mmol l-1

0

15

30

45

60

75

90

105

H2O H2O2 TMAH Carb 10 Carb 25 Carb 50

without reflux with reflux

H2O 0.9 mmol l-1

H2O2 56 mmol l

-1

TMAH

10 25 50

Absorbing solution Io

din

e re

cover

y, %

Fig. 1 Iodine recoveries inbovine liver for differentabsorbing solutions using MICwith and without reflux step.Determinations by ICP-MS(mean ± standard deviation,n=4)

Iodine determination in food by inductively coupled plasma mass spectrometry after digestion by microwave-induced combustion 1129

prior to analysis. Additionally, a more diluted solution ofTMAH (56 mmol l−1, used as absorbing solution for MIC)was also evaluated for extraction but it was not effectiveand only about 60% of the certified value was found.Results for MIC showed RSDs lower than 4.6%. Based onthe main characteristics of the evaluated procedures (see acomparison in Electronic supplementary material) MICprovides a fast digestion, only 25 min to prepare eightsamples, with lower RSD and minimal reagent consump-tion. In addition, MIC procedure was effective to digesthigher amounts of sample (500 mg) in comparison to otherprocedures then allowing lower limit of detection (LOD)and limit of quantification (LOQ). With regard to LOD, itwas 0.0007 µg g−1 for MIC, 0.0075 µg g−1 for microwave-assisted alkaline extraction and 0.0090 µg g−1 for conven-tional alkaline extraction. Values for LOQ were 0.002,0.023, and 0.027 µg g−1 for MIC, microwave-assistedalkaline extraction and conventional alkaline extraction,respectively. At this point, LOD by MIC was improvedabout ten times of the correspondent LOD found by usingthe other procedures evaluated in this work.

Determination of iodine in food samples by ICP-MSafter MIC

Based on the results previously discussed and on the suitabilityof MIC to perform digestion of food samples, different kindsof food and CRMs were digested by MIC. Bovine liver, cornstarch, milk powder, and wheat flour commercial samples andCRMs were digested using MIC followed by the reflux stepand 50 mmol l−1 (NH4)2CO3 as absorbing solution.

Table 2 shows the feasibility of using MIC and ICP-MSfor iodine determination in different food samples. It wasobserved that agreement for CRMs was better than 97%,even at low contents, as observed for corn bran (NIST 8433)and bovine muscle (NIST 8414). Milk powder sample andmilk powder CRMs presented higher contents among allsamples analyzed. In general, RSD was about 2% to 3% andfor lower concentrations (<50 ng g−1) they were about 15%.In addition, high digestion efficiency was observed for MIC(RCC was lower than 1% for all samples) which can be

attributed to the high temperature achieved during combus-tion, assuring the complete destruction of all organic matrix.To evaluate the suitability of the proposed procedures,digests of samples shown in Table 2 were also analyzed byIC and no statistical differences (95% confidence level) wereobserved in comparison with the values obtained by ICP-MS(for BCR 151 and NIST 1549). On this aspect, it is importantto point out that ICP-MS allows a better limit of detectionand a fast analysis in comparison to analysis by IC (LOD forIC was 3.0 µg g−1). The possibility of analyzing the digestsby ICP-MS or IC is an important aspect when consideringequipment availability in quality control laboratories.

Iodine determination by ICP-MS

Special attention should be given in view of memory effectsthat have been reported when determining iodine by ICP-MS[24]. It has been attributed to the volatility of HI and I2mainly in spray chamber and nebulizer tubing walls [24]. Inthis study, memory effects were observed in the initialdeterminations by ICP-MS if reference solutions or digestswere prepared in water or acidic medium. This problem wassolved by preparing all the solutions in 10 mmol l−1

Table 2 Results for iodine in food samples after MIC digestion (mean ±standard deviation, n=4) and ICP-MS determination

Sample Certified value (μgg−1) Determined value (μgg−1)

NIST 8433 0.026±0.006 0.027±0.004

BCR 151 5.35±0.14 5.27±0.10

BCR 150 1.29±0.09 1.27±0.04

NIST 8414 0.035±0.012 0.034±0.005

NIST 8435 2.3±0.4 2.36±0.11

NIST 1549 3.38±0.02 3.29±0.09

corn starch – 0.007±0.001

wheat flour – 0.013±0.002

milk powder – 2.66±0.06

bovine liver – 0.33±0.02

Five hundred milligrams of sample, 6 ml of 50 mmol l−1 (NH4)2CO3

as absorbing solution, 15 bar of oxygen, 5 min at 1,400 W

Table 1 Results for iodine determination by ICP-MS for CRMs and bovine liver sample after digestion by MIC, microwave-assisted aciddigestion, microwave-assisted extraction and conventional extraction with TMAH (mean ± standard deviation, n=4)

Sample Iodine content (μgg−1)

Certified value MIC Microwave-assistedacid digestion

Microwave-assistedextraction with TMAH

Conventional extractionwith TMAH

BCR 150 1.29±0.09 1.27±0.04 0.91±0.06 1.33±0.11 1.26±0.18

BCR 151 5.35±0.14 5.27±0.10 4.12±0.21 5.59±0.43 5.12±0.37

Bovine liver – 0.33±0.02 0.22±0.06 0.31±0.06 0.30±0.05

1130 M.F. Mesko et al.

(NH4)2CO3 and also using this medium for washing thesample introduction system (tubing, spray chamber, nebuliz-er). Digests obtained using TMAH were analyzed usingcalibration curves also prepared in TMAH and a dilutedsolution of this alkaline medium was used for washing. Thisalternative was used because superestimated values wereobtained when solutions containing TMAH were analyzedwith calibration performed in (NH4)2CO3 solution.

Conclusions

In a general way, MIC was suitable for iodine determina-tion in a set of food samples with different iodine content.Using MIC, quantitative recovery was obtained using both50 mmol l−1 (NH4)2CO3 and 56 mmol l−1 TMAH solutionsas absorbing medium with a reflux step of 5 min aftercombustion. Results obtained by MIC were in agreementwith those obtained using alkaline extraction procedures.MIC procedure showed to be efficient for digestion ofrelatively high sample mass (up to 500 mg) under safeconditions. In addition, MIC avoided the use of concen-trated alkaline solutions or excessive amounts of reagents,reducing blank values and residues generation according tothe green chemistry recommendations. Using MIC, up toeight samples could be processed simultaneously in lesstime (only 25 min) than other procedures. Digests obtainedby MIC were suitable for ICP-MS determination and alsofor IC, presenting RCC lower than 1%. The proposedprocedure combines relatively high sample throughput, lowreagents consumption and good performance for digestionof high samples mass. The use of MIC and ICP-MSallowed the further determination of low iodine content infood samples (LOQ of 0.002 µg g−1) making the proceduresuitable for routine analysis.

Acknowledgments The authors are grateful to INCT-Bio/CNPq(Process Nr. 573672/2008-3), CAPES, and FAPERGS for support-ing this study. The authors also thank to Dr. Marco Aurélio ZezziArruda (UNICAMP) for lending the infrared thermometer.

References

1. Assessment of iodine disease disorders and monitoring theirelimination, 108 p. (2007) World Health Organization—WHO,Geneva. http://www.who.int. Accessed 13 Oct. 2009

2. Rädlinger G, Heumann KG (1998) Anal Chem 70:2221–22243. Isaac-Olive K, Acharya R, Chatt A (2008) Talanta 77:827–8324. Joerin M (1975) Analyst 100:7–115. Wu D, Deng H, Wang W, Xiao H (2007) Anal Chim Acta

601:183–1886. Butler ECV, Gershey RM (1984) Anal Chim Acta 164:153–1617. Kozlowski E, Górecki T, Siefikowska-Zyskowska E (1991)

Fresenius J Anal Chem 340:773–7768. Varga I (2007) Microchem J 85:127–131

9. Crecelius EA (1975) Anal Chem 47:2034–203510. Gélinas Y, Iyengar GV, Barnes RM (1998) Fresenius J Anal Chem

362:483–48811. NormanBR, Iyengar GV (1994) Fresenius J Anal Chem 348:430–43212. Malongo TK, Patris S, Macours P, Cotton F, Nsangu J,

Kauffmann J (2008) Talanta 76:540–54713. D’Ulivo A, Pagliano E, Onor M, Pitzalis E, Zamboni R (2009)

Anal Chem 81:6399–640614. Shin H, Oh-Shin Y, Kim J, Ryu J (1996) J Chromatogr A

732:327–33315. Wong GTF, Zhang L (1992) Talanta 39:355–36016. Yang S, Fu S, Wang M (1991) Anal Chem 63:2970–297317. Naozuka J, Veiga MAMS, Oliveira PV, Oliveira E (2003) J Anal

At Spectrom 18:917–92118. Knapp G, Maichin B, Fecher P, Hasse S, Schramel P (1998)

Fresenius J Anal Chem 362:508–51319. Gélinas Y, Krushevska A, Barnes RM (1998) Anal Chem

70:1021–102520. Schramel P, Hasse S (1994) Mikrochim Acta 116:205–20921. Nóbrega JA, Gélinas Y, Krushevska A, Barnes RM (1997) J Anal

At Spectrom 12:1243–124622. Fecher PA, Goldmann I, Nagengast A (1998) J Anal At Spectrom

13:977–98223. Matusiewicz H (2003) In: Mester Z, Sturgeon R (eds) Compre-

hensive analytical chemistry—sample preparation for trace ele-ment analysis. Elsevier, Amsterdam, pp 193–233

24. Al-Ammar A, Reitznerová E, Barnes RM (2001) Fresenius J AnalChem 370:479–482

25. Vanhoe H, Allemeersch F, Versieck J, Dams R (1993) Analyst118:1015–1019

26. Larsen EH, Ludwigsen MB (1997) J Anal At Spectrom 12:435–43927. Inhat M (2003) In: Mester Z, Sturgeon R (eds) Comprehensive

analytical chemistry—sample preparation for trace element anal-ysis. Elsevier, Amsterdam, pp 765–856

28. Vtorushina EA, Saprykin AI, Knapp G (2009) J Anal Chem64:129–135

29. Flores EMM, Barin JS, Mesko MF, Knapp G (2007) SpectrochimActa Part B 62:1051–1064

30. Flores EMM, Barin JS, Paniz JNG, Medeiros JA, Knapp G (2004)Anal Chem 76:3525–3529

31. Mesko MF, Moraes DP, Barin JS, Dressler VL, Knapp G, FloresEMM (2006) Microchem J 82:183–188

32. Duarte FA, Pereira JSF, Barin JS, Mesko MF, Dressler VL, FloresEMM, Knapp G (2009) J Anal At Spectrom 24:224–227

33. Moraes DP, Mesko MF, Mello PA, Paniz JNG, Dressler VL, KnappG, Flores EMM (2007) Spectrochim Acta Part B 62:1065–1071

34. Flores EMM, Mesko MF, Moraes DP, Pereira JSF, Mello PA,Barin JS, Knapp G (2008) Anal Chem 80:1865–1870

35. Mello PA, Pereira JSF, Moraes DP, Dressler VL, Flores EMM,Knapp G (2009) J Anal At Spectrom 24:911–916

36. Pereira JSF, Mello PA, Moraes DP, Duarte FA, Dressler VL, KnappG, Flores EMM (2009) Spectrochim Acta Part B 64:554–558

37. Pereira JSF, Mello PA, Duarte FA, Santos MFP, Guimarães RCL,Knapp G, Dressler VL, Flores EMM (2009) Energy Fuels23:6015–6019

38. Mello PA, Giesbrecht CK, Alencar MS, Moreira EM, Paniz JNG,Dressler VL, Flores EMM (2008) Anal Lett 41:1623–1632

39. Pereira JSF, Diehl LO, Duarte FA, Santos MFP, Guimarães RCL,Dressler VL, Flores EMM (2008) J Chromatogr A 1213:249–252

40. Pereira JSF, Moraes DP, Antes FG, Diehl LO, Santos MFP,Guimarães RCL, Fonseca TCO, Dressler VL, Flores EMM (2010)Microchem J. doi:10.1016/j.microc.2009.12.016

41. PerkinElmer-SCIEX, Elan Version 3.0, Software Guide, 1006920A, 2003, Thornhill, Canada

42. Nóbrega JA, Santos MC, Souza RA, Cadore S, Barnes RM, TatroM (2006) Spectrochim Acta Part B 61:465–495

Iodine determination in food by inductively coupled plasma mass spectrometry after digestion by microwave-induced combustion 1131