Anal. Chem. 1998, 70, 1021-1025
Determination of Total Iodine in Nutritional and Biological Samples by ICP-MS Following Their Combustion within an Oxygen Stream Yves Ge´linas, Antoaneta Krushevska, and Ramon M. Barnes*
Department of Chemistry, Lederle Graduate Research Center, University of Massachusetts, Box 34510, Amherst, Massachusetts 01003-4510
A mineralization and determination method for total iodine in nutritional and biological samples is described. Combustion of the sample in an oxygen stream is followed by collection of the combustion products in a 5% watersoluble tertiary amine solution. Iodine is determined by inductively coupled plasma mass spectrometry. The accuracy and precision of the quantitative iodine analysis using standard addition is better than (10%. A semiquantitative analysis of four standard reference materials is evaluated. Owing to the presence of low-level iodine contaminant in the blank solution, the determination limit of the method is (10 µg kg-1. Good agreement with certified iodine values is obtained for six reference materials. The use of the tertiary amine matrix solution permits the simultaneous determination of iodine and other trace metals of biological and toxicological importance, including Mn, Co, Ni, Cu, Zn, Rb, Cd, and Pb. Iodine is an essential trace element that is ubiquitous in nature. In human nutrition, iodine is an integral part of the thyroid hormones that play an important role in controlling the rate of basic metabolism and in reproduction.1,2 Levels of iodine in tissues and body fluids are affected by dietary intakes of the element. Low soil levels of iodine are indirectly responsible for inadequate dietary intakes by humans and animals, which are readily overcome by supplementation.3 Nonetheless, iodine deficiency disorders remain a major health problem in many regions of the world. Thus, a need exists to develop precise methods of quantification for the low levels of iodine usually found in food products and biological tissues (i.e., mg kg-1 and µg kg-1 levels). Until recently, all methods available were tedious, expensive, or needed extensive sample preparation: spectrophotometry,4,5 X-ray fluorescence analysis,6,7 neutron activation analysis,8-14 gas (1) Casey, C. E.; Smith, A.; Zhang, P. In Handbook of Milk Composition; Jensen, R. G., Ed.: Academic Press: Boston, 1995; pp 622-7. (2) Hetzel, B. S.; Maberly, G. F. In Trace Elements in Human and Animal Nutrition, 5th ed.; Mertz, W., Ed.: Academic Press: Boston, 1986; pp 13964. (3) Anke, M.; Groppel, B.; Mu ¨ ller, M.; Scholz, E.; Kra¨mer, K. Fresenius J. Anal. Chem. 1995, 352, 97-101. (4) Joerin, M. M. Analyst 1975, 100, 7-14. (5) Moxon, R. E. D.; Dixon, E. J. Analyst 1980, 105, 344-9. (6) Crecelius, E. A. Anal. Chem. 1975, 47, 2034-7. (7) Jonckheer, M. H.; Velkeniers, B.; Vanhaelst, L.; Van Blerk, M. Nucl. Med. Commun. 1992, 13, 2-10. S0003-2700(97)00974-8 CCC: $15.00 Published on Web 01/28/1998
© 1998 American Chemical Society
chromatography,15 and cathodic stripping voltametry.16,17 Inductively coupled plasma mass spectrometry (ICP-MS) provides high sensitivity and sample throughput for iodine determinations. Reports of the iodine determination by ICP-MS in biological fluids or milk samples without extensive sample preparation have demonstrated the possibility of achieving sensitive quantification with limited amounts of material.18-26 Sample preparation usually consisted of a simple dilution or dispersion in an alkaline solution (e.g., 0.5% v/v ammonia,23 0.05 mol L-1 sodium hydroxide,24 or 2.5% tetramethylammonium hydroxide24,25) to avoid a selective evaporation of iodine as HI or I2. This causes nonquantitative recoveries and serious memory effects in the ICP-MS introduction system.23 The determination of iodine in nutritional materials introduces an additional complication owing to the possible loss of volatile iodide when solid samples are being digested with acid.23 Alternative mineralization protocols must then be used. Wet ashing procedures with perchloric acid to convert the volatile iodide into nonvolatile iodate have been used, but the iodine analysis must be completed on the same day because of the fast reconversion (8) Chou, F. I.; Lin, H. D.; Wei, J. C.; Wang, A. Y.; Lo, J. G. Nucl. Med. Biol. 1993, 20, 631-8. (9) Landsberger, D.; Wu, D. J. Radioanal. Nucl. Chem. 1993, 167, 219-24. (10) Susetyo, W.; Lahagu, F.; Das, H. A. J. Radioanal. Nucl. Chem. 1992, 164, 373-8. (11) Sato, T.; Kato, T. J. Radioanal. Chem. 1982, 68, 175-9. (12) Dermelj, M.; Slejkovec, Z.; Byrne, A. R.; Stegnar, P.; Stibilj, V.; Rossbach, M. Fresenius J. Anal. Chem. 1990, 338, 559-66. (13) Rao, R. R.; Chatt, A. Anal. Chem. 1991, 63, 1298-303. (14) Rao, R. R.; Holzbecher, J.; Chatt, A. Fresenius J. Anal. Chem. 1995, 352, 53-7. (15) Gaebler, H. E.; Heumann, K. G. Fresenius J. Anal. Chem. 1993, 345, 53-9. (16) Wong, G.; Zhang, L. S. Talanta 1992, 39, 335-40. (17) Yang, S. X.; Fu, S. J.; Wang, M. L. Anal. Chem. 1991, 63, 2970-3. (18) Durrant, S. F.; Ward, N. I. Micronutr. Anal. 1989, 5, 111-26. (19) Allain, P.; Mauras, Y.; Douge´, C.; Jaunault, L.; Delaporte, T.; Beaugrand, C. Analyst 1990, 115, 813-5. (20) Baumann, H. Fresenius J. Anal. Chem. 1990, 338, 809-12. (21) Cox, R. J.; Pickford, C. J.; Thompson, M. J. Anal. At. Spectrom. 1992, 7, 635-9. (22) Ward, N. I.; Durrant, S. F.; Gray, A. J. Anal. At. Spectrom. 1992, 7, 113943. (23) Vanhoe, H.; Van Allemeersch, F.; Versieck, J.; Dams, R. Analyst 1993, 118, 1015-9. (24) Schramel, P.; Hasse, S. Mikrochim. Acta 1994, 116, 205-9. (25) Stu ¨ rup, S.; Bu ¨ chert, A. Fresenius J. Anal. Chem. 1996, 354, 323-6. (26) Michalke, B.; Schramel, P.; Hasse, S. Mikrochim. Acta 1996, 122, 67-76.
Analytical Chemistry, Vol. 70, No. 5, March 1, 1998 1021
of iodate into the volatile iodine upon storage.27-30 The use of oxidants such as K2Cr2O7, H2O2, and NaOCl was not demonstrated probably owing to the incomplete conversion of iodide into iodate or to the fast reduction of iodate from the solution.23 A peroxydisulfate oxidation method was recently introduced, but its detection limit is not sufficiently low for most nutritional and biological materials.31 Other methods involving the oxidation of iodide followed by the generation of volatile iodine (I2) for the determination of total iodine in the gas phase by ICP-MS or ICPAES were also described for biological and aqueous samples.32-34 These approaches all work to a certain extent and were developed to solubilize and convert the iodine to a form that is appropriate for the instrumentation being used. However, none of them allows easy multielemental determination including iodine with ICP-MS instruments. Schramel and Hasse reported quantitative iodine recovery in a wide variety of biological materials using a simple Scho¨niger flask combustion device and sodium hydroxide or tetramethylammonium hydroxide as the collecting solution.24 However, their limit of determination (50 µg kg-1) was limited by the small maximum sample size combusted (20-25 mg) and by the contamination introduced by the filter paper in which the samples were wrapped. Although their proposal was not substantiated by experimental results, they suggested a lower determination limit could be obtained when the Trace-O-Mat oxygen combustion apparatus was used. This equipment mineralizes as much as 1 g of sample (depending on the matrix) and eliminates the need for wrapping the sample in filter paper.35 This report presents the results obtained for the determination of the total iodine content of nutritional and biological materials using the Trace-O-Mat oxygen combustion instrument followed by ICP-MS analysis. It also evaluates the collection of the incinerated sample with a solution of water-soluble tertiary amines at pH 8. This solution allows for multielemental determination of iodine and other trace metals using ICP-MS.36,37 EXPERIMENTAL SECTION Instrumentation. The Trace-O-Mat instrument (Anton Paar, Graz, Austria) is a partially mechanized oxygen combustion device consisting of a combustion unit made of quartz that is divided into three functional zones: a burning chamber, an overlying cooling unit, and a test tube with sampler holder at the bottom.35 Briefly, it allows for an organic sample of 0.1-1 g, pressed into a pellet by the use of a molybdenum alloy pressing tool and placed on a quartz sample holder, to be incinerated within a stream of oxygen flowing at an optimized rate. Nonvolatile elements remain in the burning chamber as a residue while the volatile products of combustion are condensed against the walls of the overlying coldfinger and cooling jacket filled with liquid nitrogen prior to the incineration. Once the combustion is completed, the sample (27) Fasset, J. D.; Murphy, T. J. Anal. Chem. 1990, 62, 386-9. (28) Holak, W. Anal. Chem. 1987, 59, 2218-21. (29) Schindlmeier, W.; Heumann, K. G. Fresenius J. Anal. Chem. 1985, 320, 745-8. (30) Larsen, E. H.; Ludwigsen, M. B. J. Anal. At. Spectrom. 1997, 12, 435-9. (31) Gu, F.; Marchetti, A. A.; Straume, T. Analyst 1997, 122, 535-7. (32) Kerl, W.; Becker, J. S.; Dietze, H.-J.; Dannecker, W. J. Anal. At. Spectrom. 1996, 11, 723-6. (33) Nakahara, T.; Mori, T. J. Anal. At. Spectrom. 1994, 9, 159-63. (34) Schnetger, B.; Muramatsu, Y. Analyst 1996, 121, 1627-31. (35) Knapp, G.; Raptis, S. E.; Kaiser, G.; To ¨lg, G.; Schramel, P.; Schreiber, B Fresenius J. Anal. Chem. 1981, 308, 97-103.
1022 Analytical Chemistry, Vol. 70, No. 5, March 1, 1998
Table 1. ICP-MS Operating Conditions and Measurement Parameters ICP-MS plasma conditions
Elan 250
Elan 5000a
rf frequency rf forward power (kW) argon flow rate (L min-1) outer intermediate nebulizer
27.16 1.1
40 1.0
11.4 1.4 0.91-1.00
15.0 0.8 0.92-1.04
measurement parameters
Elan 250 and Elan 5000a
resolution (m/z at 10% peak height) scanning mode replicate time (ms) sweeps per reading readings per replicate no. of replicates points per spectral peak
0.8 (normal) peak hopping 2000 20 1 10 1
holder is turned upside down to lay at the bottom of the underlying test tube that is filled with 2 mL of a reagent, usually nitric acid. Two infrared radiators, which are also used to ignite the sample, are then brought into the lower position to boil the reagent under reflux for 30 min, thus collecting both volatile and nonvolatile elements in the dissolved phase. The total preparation time is ∼45 min per sample. Two ICP-MS instruments (Perkin-Elmer Sciex Elan 250 and Elan 5000a, Norwalk, CT) were used. Both instruments were equipped with a standard torch, cross-flow nebulizer, and Ni sampler and skimmer cones. The limit of detection was determined for both instruments, and the quantitative analyses were carried out on the Elan 5000a under the plasma conditions and measurement parameters listed in Table 1. The optimal nebulizer flow rate was lower (0.92-0.95 L min-1) than with water when water-soluble tertiary amines were used as the matrix and collecting solution. To avoid memory effects resulting from the selective volatilization of iodide in acid solution, the complete introduction system (i.e., blanks, standard solutions, samples, rinsing solution, and waste container) was kept alkaline using 0.05 mol L-1 NaOH or 5% amine solution (see Reagents, below). This resulted in an iodine washout time similar to that of most of the other trace elements. Measurements. To find the best compromise between accuracy, precision, and sample throughput, external calibration and standard addition protocols of the instrument software were compared. Standard addition had to be used to obtain a relative standard deviation of less than 10% for samples with a very low iodine content. Although rhodium24 and tellurium20,23 have been used as internal reference elements for iodine determination in alkaline solutions, we found that internal standardization did not improve precision or accuracy significantly, providing the standards were matrix-matched. This agrees with the findings of Stu¨rup and Bu¨chert, who also reported no improvement in accuracy or precision when using europium as an internal reference.25 The performance of the multielement TotalQuant II program (Perkin-Elmer),37-39 designed as a semiquantitative screening or survey tool, was also evaluated. This program (36) No´brega, J.; Ge´linas, Y.; Krushevska, A.; Barnes, R. M. J. Anal. At. Spectrom. 1997, 12, 1239-42.
performs completely automated mass spectrum interpretation, correcting for common isobaric interferences and using precalibrated intensities per concentration unit to evaluation the concentration of a given element. Reagents. The match cords used to ignite the sample in the Trace-O-Mat combustion chamber were cut from new Whatman No. 542 ashless filters (20 mm × 4 mm pieces) in a class-100 clean room using acid-cleaned tools. Reagent grade sodium hydroxide pellets (Fisher Scientific, Pittsburgh, PA) or a mixture of water-soluble tertiary amines (CFA-C amines, Spectrasol, Warwick, NY) adjusted to pH 8 with subboiled nitric acid were used to prepare the matrix solution for samples and standards. The iodine stock solution was prepared by dissolving an appropriate amount of KIO3 (Fisher Scientific) in deionized distilled water. The working standard solutions were obtained daily by diluting the stock solution in 0.05 mol L-1 NaOH or in 5% (v/v) CFA-C (pH 8). External calibration and standard additions with iodine standards between 2 and 20 ng/mL were used to determine the iodine concentration in samples. The linearity of the calibration curves extended beyond 100 ng mL-1 on both instruments. Standard reference materials (SRM) were obtained from the National Institute of Standards and Technology (NIST). Sample Preparation. Samples (200-500 mg) were pressed into a pellet, placed on the precleaned quartz sample holder with a match cord, and introduced into the combustion chamber of the Trace-O-Mat. The condenser was then filled with liquid nitrogen, and the match cord was ignited using the two infrared lamps in their upper position. Following the combustion of the sample, the quartz test tube containing the sample holder was disconnected from the combustion chamber and filled with 2 mL of 0.05 mol L-1 NaOH or 2 mL of 5% CFA-C amine solution. The sample holder was turned upside down inside the test tube with a pair of plastic tweezers, and the test tube was then reconnected to the combustion chamber. After thawing of the condensate, the infrared lamps were relocated into their lower position to boil the NaOH or CFA-C solution under reflux for 30 min. The volatile fraction of iodine released during the combustion was leached from the walls of the condenser and collected in the solution. The reflux solution in the test tube was allowed to cool for 5 min and quantitatively transferred into a 25-mL volumetric flask. The walls of the cooling jacket, coldfinger, and combustion chamber were rinsed twice with two additional aliquots of the NaOH or CFA-C solution. The rinsing solution was added to the flask that was then completed to exactly 25.0 mL with the NaOH or CFA-C solution. The resulting solution was analyzed without further dilution. Collecting the sample in only 10.0 mL resulted in a nonquantitative recovery particularly for material with a relatively high iodine content (>1 mg kg-1) because of the incomplete leaching of the adsorbed iodine on the condenser. Between each sample run, the entire device was thoroughly rinsed with 5% HNO3 followed by deionized distilled water to remove any residual iodine. The sample holder designed for liquid samples35 was slightly modified to obtain a cylindrical cavity (37) Krushevska, A.; La´sztity, A.; Kotrebai, M.; Barnes, R. M. J. Anal. At. Spectrom. 1996, 11, 343-52. (38) Users Manual, Perkin-Elmer Model 5000 ICP.-MS System, Perkin-Elmer, Norwalk, CT, Version 1, May 1992, Rev B. (39) Amarasiriwardena, D.; Durrant, J. F.; La´sztity, A.; Krushevska, A.; Argentine, M. D.; Barnes, R. M. Mikrochem. J. 1997, 56, 352-72.
Table 2. ICP-MS Limits of Detection (LOD) and Quantification (LOQ), and Blank Values (ng L-1) Elan 5000a NaOH LOD (3σ) LOQ (10σ) blank valuesa Trace-O-Mat memory effectb
18.3 60.8
amines
Elan 250 NaOH
14.8 37.9 49.2 126.2 110 ( 50 (n ) 21) 140 ( 80 (n ) 13)
amines 22.8 76.0
a Blank values for the match cord and CFA-C solution boiled under reflux. b Values obtained for the CFA-C solution boiled under reflux following a sample.
with closed walls to avoid melting and leaking of fatty material during the combustion step. Blank solutions were prepared in the same way with a match cord but without sample. At least one blank was processed every day to check for contamination. RESULTS AND DISCUSSION Detection Limits. The limit of detection (LOD, 3σ, where σ is the standard deviation of the blank) and the limit of quantification (LOQ, 10σ) routinely obtained with both instruments for the 0.05 mol L-1 NaOH and the 5% CFA-C amine solutions are listed in Table 2. The values for the Elan 5000a were averaged over the complete analysis period (∼2 months), using the intensities obtained for the blank solution of each standard addition curve. They thus represent the true practical limits that can be expected with routine analysis of nutritional and biological samples. Lower LOD and LOQ were obtained for both instruments with the CFA-C amine solution mainly because of the organic content of the matrix solution (Table 2). Organic carbon is known to significantly enhance the sensitivity (ions s-1 per concentration unit) of the elements with a high ionization potential such as iodine, resulting in a higher signal-to-noise ratio.40 Values found for the ICP-MS Elan 250 were surprisingly close to those of the Elan 5000a despite a ∼10-fold lower sensitivity. This is explained by the lower relative standard deviation (RSD) of the mean blank intensity measured on the former instrument and by the fact that the LOD and LOQ of the Elan 250 were determined on a single day with carefully optimized measurement parameters and cleanliness conditions of the introduction system/interface components. Determination Blank Levels. Throughout this work, fairly constant blank values of ∼0.1 ng mL-1 were found (Table 2), corresponding to a determination limit of ∼10 µg kg-1 dry material in a 250-mg sample diluted to 25.0 mL. This values compares well with the range of 2-50 found in the literature.13,20,24,28,30 This is despite a thorough cleaning of the Trace-O-Mat quartz parts with nitric acid between each combustion. This contamination stemmed mainly from the filters used as match cords and from the sample manipulations that must be carried out in open laboratory air during the combustion and collection steps. Contamination was minimized by using new filters and preparing the match cords on a class 100 laminar flow bench in a clean room. Iodine contamination could probably be further reduced by installing the Trace-O-Mat instrument on a class-100 or -10 clean (40) Krushevska, A.; Kotrebai, M.; La´sztity, A.; Barnes, R. M.; Amarasiriwardena, D. Fresenius J. Anal. Chem. 1996, 335, 793-800.
Analytical Chemistry, Vol. 70, No. 5, March 1, 1998
1023
Table 3. Determination of Iodine in Standard Reference Materials by the Trace-O-Mat and the Scho 1 niger Methods, and Comparison of the Semiquantitative Method Using External Standardization and the Quantitative Method Using Standard Additions
material oyster tissue (SRM 1566a) non-fat milk powder (SRM 1549a) whole milk powder (SRM 8435) whole egg powder (SRM 8415) tomato leaves (SRM 1573a) peach leaves (SRM 1547) a
certified value (mg kg-1)
Trace-O-Mat quantitative
found valuesa Trace-O-Mat semiquantitative
Scho ¨niger quantitative
4.46 ( 0.42
4.24 ( 0.17
4.26 ( 0.21
4.19 ( 0.34
3.38 ( 0.02
3.26 ( 0.10
3.24 ( 0.13
3.27 ( 0.18
2.3 ( 0.4
2.17 ( 0.12
ndc
nd
1.97 ( 0.45
1.83 ( 0.16
nd
nd
(0.85)b
0.84 ( 0.04
0.89 ( 0.07
0.79 ( 0.13
(0.3)b
0.32 ( 0.02
0.36 ( 0.03
0.38 ( 0.07
Mean and 95% confidence interval for five independent determinations. b Noncertified value. c nd, not measured.
surface (this was not done for practical reasons), since Schramel and Hasse reported that the potentially high iodine concentration in the laboratory air was the main source of iodine contamination when an oxygen combustion device was used.24 Choice of Collecting Solution. The Trace-O-Mat device was specially designed to minimize the loss of volatile elements when solid biological samples were mineralized. However, when iodine is being mineralized, an alkaline solution must be used to avoid the loss of a fraction of the volatile iodide, which precludes the simultaneous determination of most of the cationic trace metals.25 The use of an alkaline solution such as the tertiary amine solution to stabilize cations in the liquid phase also can permit multielemental determinations. This eliminates the need for a wet acid digestion of the sample when the concentration of other trace metals is also sought. The tertiary amine solution also improves the quality of the results obtained for iodine because of the following factors: (a) The organic content of the CFA-C amine solution results in a significant enhancement of the sensitivity (ions s-1 per concentration unit) of elements with a high first ionization potential such as As, Se, Hg, I, Zn, and Be40 (i.e., only 25-29% of iodine is ionized in an argon plasma41). This leads, in turn, to lowered detection and quantification limits (Table 2), as well as to lowered relative standard deviations owing to the higher ions counts generated. (b) The action of the amine solution stabilizes the dissolved iodine in the liquid phase.36 The relative standard deviations obtained for similar iodine intensities were generally lower with the amine solution compared to the NaOH solution, and the correlation coefficients of the calibration curves were generally closer to unity. The combination of a simple oxygen combustion method with the advantages of using the tertiary amine solution and ICP mass spectrometry makes this approach very powerful for multielemental determinations in biological and nutritional samples. It eliminates the problems usually encountered when iodine is solubilized from these materials (i.e., volatilization in acid solution (41) Horlick, G.; Tan, S. H.; Vaughan, M. A.; Shao, Y. Inductively Coupled Plasmas in Analytical Atomic Spectrometry; Montaser, A., Golightly, D. W., Eds.; VCH: New York, 1987; pp 361-86.
1024 Analytical Chemistry, Vol. 70, No. 5, March 1, 1998
and memory effect during measurement) without hampering the determination of most of the other major and trace elements.36,37 Precision and Accuracy. The results obtained using the standard addition method for the analysis of certified reference and other nutritional and biological materials containing iodine at concentrations from 0.3 to 4.46 mg kg-1 are listed in Table 3. The values obtained for the standard reference materials (SRM) agree well with the certified concentrations at the 95% confidence. They were also in agreement (95% confidence interval) with those obtained using the Scho¨niger combustion method. The reproducibility (RSD) of separate ICP-MS determinations of iodine in a single sample is less than 3%, while that of the complete combustion and determination procedure is below 10%. In an earlier stage of this study, the recoveries obtained for samples with a high fat content using the Trace-O-Mat conventional holder for solid samples were 84-95%. This was due to the incomplete combustion/dissolution of the carbon residue remaining on the sampler holder following the incineration step and to the melting and leaking of a fraction of the sample’s fat content through the wall openings of the upper part of the sample holder. By using a modified sample holder with a closed-wall cylindrical top cavity, the incineration rate could be slowed by a factor of 2 without loss resulting from the melting and leaking of the fatty material. Improved destruction of the carbon matrix and better recoveries of iodine and other trace metals resulted. As shown by the results of Table 4, the quantitative recovery of trace metals of biological of toxicological interest (e.g., Mn, Co, Ni, Cu, Zn, Rb, Cd, and Pb) demonstrated the multielemental potential of this technique. Although only three determinations were made, most of the results agreed well with the certified values for a wide range of materials and concentrations. Moreover, iodine and other trace elements are stable for more than one month in the mixed amine solution. Evaluation of the Semiquantitative TotalQuant II Program. Since cationic trace metals are not stable in most alkaline solutions, the evaluation of the TotalQuantII multielemental analysis subroutine is reasonable only when the 5% tertiary amine solution is used.36 The semiquantitative results obtained for four standard reference materials are listed in Table 3. A standard solution containing iodine and the external calibration subroutine
Table 4. Trace Elements Determination in Standard Reference Materials Using the Trace-O-Mat Instrument (Mean and 95% Confidence Interval, n ) 3) concn (mg kg-1) material
55Mn
59Co
62Ni
63Cu
66Zn
85Rb
114Cd
208Pb
oyster tissue (SRM 1566a) certified whole milk powder (SRM 8435) certified whole egg powder (SRM 8415) certified tomato leaves (SRM 1573a) certified
11.5 ( 0.7 12.3 ( 1.5 0.15 ( 0.02 0.17 ( 0.05 1.56 ( 0.25 1.78 ( 0.37 239 ( 12 246 ( 7
0.5 ( 0.05 0.57 ( 0.11 nda (0.003)b 0.019 ( 0.007 0.012 ( 0.005 0.55 ( 0.03 0.57 ( 0.02
2.14 ( 0.32 2.25 ( 0.44 nd (0.01)
60.9 ( 3.6 66.3 ( 4.3 0.41 ( 0.1 0.46 ( 0.08 2.4 ( 0.3 2.7 ( 0.4 4.4 ( 0.3 4.7 ( 0.2
788 ( 34 830 ( 57 25 ( 4 28 ( 3 62.4 ( 4.1 67.5 ( 7.4 29.4 ( 0.5 30.9 ( 0.6
2.9 ( 0.2 (3) 15.3 ( 0.5 (16)
3.92 (0.22 4.15 ( 0.38 nd (0.0002) nd (0.005) 1.48 ( 0.04 1.52 ( 0.05
0.343 ( 0.012 0.371 ( 0.014 0.1 ( 0.02 0.11 ( 0.05 0.049 ( 0.021 0.061 ( 0.012
a
1.51 ( 0.04 1.59 ( 0.06
14.8 ( 0.3 14.9 ( 0.3
Below limit of determination. b Noncertified values.
were used to optimize and update the response factor according to the instrumental operating conditions.39 For samples with an iodine concentration of >2 mg kg-1, the accuracy and precision were the same as those obtained with the standard addition measurement. The measurement of the 127I isotope is not hampered by any major isobaric or spectroscopic interferences. Only for low concentrations (
