JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 159 Analyte Volatilization Procedure for the Determination of Low Concentrations of Iodine by Inductively Coupled Plasma Atomic Emission Spectrometry* Invited Lecture Taketoshi Nakahara and Toshio Mori Department of Applied Chemistry University of Osaka Prefecture Sakai Osaka 593 Japan A simple method is described for the generation of a continuous flow of volatile iodine by the oxidation of aqueous iodide for the determination of low concentrations of iodine by inductively coupled plasma (ICP) atomic emission spectrometry in the normal ultraviolet and vacuum ultraviolet (VUV) regions of the spectrum. For measuring spectral lines in the VUV region the monochromator and the enclosed external optical path between the ICP source and the entrance slit of the monochromator were both purged with nitrogen to minimize light absorption by atmospheric oxygen.The iodine atom emission lines at 178.28 183.04 and 206.1 6 nm were selected as the analytical lines of interest. Of the various oxidation reactions investigated an oxidizing solution of 5.0 mmol I-’ of sodium nitrite in 8.0 mol I-’ sulfuric acid was found to be the most appropriate for the generation of elemental iodine. The gaseous iodine is separated from the solution in a simple gas-liquid separator and swept into the argon stream of an ICP for analysis. The best attainable detection limits (30 criterion) for iodine at 178.28 183.04 and 206.16 nm were found to be 0.39 0.55 and 2.1 ng ml -’ respectively. Typical calibration graphs obtained under the optimized experimental conditions are rectilinear over approximately four orders of magnitude of concentration.The present method has successfully been applied to the determination of total iodine (i.e. iodide + iodate) in several brine samples. Keywords Inductively coupled plasma atomic emission spectrometry; iodine determination; analyte volatiliz- ation; gas-phase sample introduction; brine sample In natural environments iodine is known to be widely distrib- uted. However the amount of iodine contained in natural environmental materials such as natural water rocks and vegetables is often at trace concentration levels,’-4 and sufficient sensitivity is required for the determination of iodine at such levels. Inductively coupled plasma atomic emission spectrometry (ICP-AES) is the current method of choice for the determi- nation of over 70 elements.However the direct determination of non-metals such as iodine has to date been quite limited.5 A major reason for the difficulties encountered is that the sensitive resonance lines for most non-metals lie in the vacuum ultraviolet (VUV) and/or near-infrared spectral regions making their determination by AES difficult. Also the high excitation energies of non-metals cause excited-state popu- lations to be low. Recently for example Nakahara and Wasa6 reported the direct determination of iodine in aqueous solu- tions with an argon ICP where typical detection limits for iodine were 0.092 0.088 and 0.559 pg ml-’ at 178.28 183.04 and 206.16 nm respectively.In addition the same workers described a prior-oxidation procedure6 which improved the detection limits by a factor of approximately 40 at both 183.04 and 206.16 nm. One of the most important features in any sample introduc- tion system is transport efficiency. For normal pneumatic nebulizers this efficiency is l-2%.7 If the analyte is present in solution in a volatile form analyte transport efficiency could approach 100% and hence the detection limit would be improved by a corresponding factor. However most non-metal microwave induced plasma (MIP) AES has been commonly performed with a gas-phase sample introduction mode for a relatively low power (<200 W) MIP. This gas-phase sample introduction mode includes indirect determination of iodine based on the cold-vapour method for mercury,8 electrothermal vaporization ( ETV)9 and analyte volatilization On the other hand little work has been reported on the * Presented at the XXVIII Colloquium Spectroscopicurn Internationale (CSI) York UK June 29-July 4 1993.application of this analyte volatilization method in ICP- AES.13,14 Thus in this study the on-line generation of volatile iodine by the oxidation of aqueous iodide was examined for argon ICP-AES. Iodine emission lines were observed in the UV and VUV spectral regions. A nitrogen-purged optical system was used for the observation of the VUV spectra.” The optimiz- ation and analytical performance of the method and the application of the proposed technique to the determination of total iodine (ie. iodide +iodate) in several brine samples are described below.Experimental Instrumentation and Apparatus A Nippon Jarrell-Ash Model ICAP-SOSM inductively coupled plasma emission spectrometer was used in combination with a gas-phase sample introduction system. Specifications of the instrumentation used in this work are given in Table 1 and the complete analytical system is schematically represented in Fig. 1. A detailed description of the gas-liquid separator made of borosilicate glass is given in Fig. 2. The internal mono- chromator path and the enclosed external optical path between the ICP source and the monochromator were both purged with nitrogen to reduce light absorption by atmospheric oxygen. The detailed schematic diagram of this simple nitrogen- purged optical system has been described previously by Nakahara.” With the use of this system the VUV spectrum between 170 and 190 nm could be observed. Reagents All the reagents used in this study were of analytical-reagent grade.High-purity water (Milli-Q water) was obtained by passing distilled water through a Milli-Q ion-exchange and membrane-filtering system (Millipore). A standard stock solution of iodine (1000pgml-’) was prepared daily by dissolving potassium iodide in Milli-Q water.160 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Table 1 Experimental instrumentation Component Description Generator (r,f.) Crystal-controlled type; 27.12 MHz; automatic power control; automatic impedance matching network; maximum output power 2.0 kW All quartz; Fassel type Argon for coolant plasma and carrier gases Ebert 0.5 m mounting; 1180 lines mm-' grating blazed at 240 nm; reciprocal linear dispersion (first order) 1.6 nm mm-' Plasma source focussed as 1 1 image onto an entrance-slit with a 60 mm focal length Suprasil lens D.c.amplification/lO s integration; digital voltmeter Rikadenki Model R-21 (1.0 V full scale) Plasma torch Gas for ICP Nebulizer Pneumatic cross-flow type Spectrometer Optics Photomultiplier Hamamatsu Photonics R-106UH Signal measurement Chart recorder Peristaltic pump Tokyo Rikakikai MP-3 Y Y Y F C A B Fig. 1 Schematic diagram of the ICP-AES system with gas-phase sample introduction for the continuous-flow determination of iodine A sample solution; B oxidizing solution; C peristaltic pump; D gas-liquid separator; E liquid waste; F carrier gas; G elemental iodine +carrier gas; H coolant gas; I plasma gas; J gas-flow controller; K argon tank; L plasma torch; M ICP; N nebulizer chamber; 0 monochromator; P photomultiplier; Q high voltage supply; R amplifier; S digital integrator; T chart recorder; U tuning and coupling; V r.f.power generator; W lens; X light cell; and Y nitrogen purge gas Sample Oxidizing solution solution ___._I) ICP Waste 5 cm Fig. 2 Gas-liquid separator for continuous-flow generation of elemental iodine From this standard solution working standard solutions were prepared immediately before use by serial dilution. Oxidizing solutions sodium nitrite potassium persulfate hydrogen peroxide potassium permanganate potassium dichromate and potassium perbromate were prepared acidified with sulfuric acid and used for the continuous-flow generation of elemental iodine from iodide.Argon (99.99% pure) served in a dual role i.e. as a plasma- sustaining gas for the ICP and as a carrier gas for introduction into the ICP of the iodine vapour that was generated. Nitrogen (99.99% pure) was used as a purge gas for measuring the VUV lines of iodine. Procedure A previously acidified oxidizing solution and a sample solution containing iodine as iodide were conveyed continuously into the iodine generator (Fig. 2). The chemically evolved iodine vapour was separated from the reaction solution and swept into the plasma by a continuous flow of argon carrier gas while the resulting solution was drained to waste. The iodine emission lines in the UV and VUV spectral regions were monitored.All emission intensities when stabilized were inte- grated at least in triplicate over 10 s. For off-line background correction at the same wavelength as the analyte the back- ground emission (which included any emission from the reagent blank) integrated over the same period of time (10s) was subtracted from all the emission data. Under the standard operating conditions the iodine emission signal begins to rise from the baseline ,within 20s of the sample (or standard) solution being pumped in and it reaches a plateau after about 30 s. Furthermore the tailing edge of the signal returns to the baseline well within 20 s of the sample (or standard) solutions being exchanged for Milli-Q water. Six methods for chemical vapour generation of iodine were investigated; the oxidation reactions along with the standard oxidation-reduction potentia1sl6 are shown in Table 2. The influence of the six oxidizing agents shown in Table 2 at 0.005-1000 mmol l-.' and acidities of 1.0-8.0 moll-' in sulfuric acid on the iodine emission intensity at 178.28 183.04 and 206.16 nm was briefly examined.These parameters in addition to other experimental conditions were optimized by a univari- ate search technique with the use of 0.2 1.0 and 5.0 pg ml-' iodide solutions at 178.28 183.04 and 206.16 nm respectively unless otherwise stated. Results and Discussion Optimization of Experimental Parameters In a preliminary study the VUV emission spectrum in the 175-186nm range was obtained from an iodine solution of 0.2 pg ml-' witlh the use of the nitrogen-purged optical system developed previou~ly'~ (see Fig.3). For comparison the normal UV spectrum at 205-207 nm was measured under the same conditions and is depicted in Fig. 3 also. In an attempt to obtain a maximum line-to-background intensity ratio In/Ib (where I is the background-corrected net analyte intensity and Ib is the background emission intensity) Table 2 Iodine generation reactions with various oxidizing agents ~~~~~ Oxidizing agent Iodine generation reaction K2SzOs (+ 2.123)* H,O (+ 1.776) KMnO (+ 1.507) KBrO (+ 1.423) K,Cr20 (+ 1.232) S20s2- + 21- + 2Hf = I 2 + 2HSO,- H,O + 21- + 2Hf = I + 2H20 2Mn0,- + 101- + 16H+ = 2Br03- + 101- + 12H' = Cr,'07'- +6I- + 14H' = 51,+ 2Md' + 8H20 51 + Br + 6H20 31 + 2Cr3+ + 7H,O NaNO (+ 0.938) 2N02- + 21- +4Ht =I,+ 2N0 + 2H,O *Standard oxidation-reduction potentials ( V versus a normal hydro- gen electrode) for each oxidizing agent. Standard oxidation-reduction potential of iodine ( I + 2e- = 21-) is + 0.536 V.All potential values were taken from ref. 16JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 161 C Baseline bA 207 205 185 180 175 Wavelengthhm Fig.3 ICP emission spectral regions of 175-185 and 205-207 nm obtained for a 0.2 pg ml- ' iodide solution by continuous-flow iodine generation. A S I 180.73; B S I 182.04; C S I 182.62; D I I 206.16; E I I 184.44; F I I 183.04; G I I 179.91; and H I I 178.28 nm for iodine various operating parameters were examined and optimized individually while the other parameters were kept at their optimum values.The parameters investigated were r.f. power analytical wavelength observation position in the ICP source argon flow rates of coolant plasma and sample gases and nitrogen purge gas flow rates. The selection of the most sensitive lines for iodine determi- nations was achieved by the comparison of major iodine lines shown in Fig. 3. The I IJIb and background equivalent concentration (BEC) obtained for major lines of iodine after gas-phase sample introduction or conventional solution nebul- ization are given in Table 3. The BEC is defined as the concentration of the analyte expressed in pgml-' of iodine that yields a net intensity signal equal to the intensity of the background. As a consequence three major iodine lines I I 178.28 I I 183.04 and I I 206.16 nm were the analytical lines selected and the optimized operating conditions at these three wavelengths are summarized in Table 4.The ICP discharge was stable even at lower r.f. power (<l.OkW) probably because the perturbation of the plasma by excessive water vapour aerosol could be avoided in the present gas-phase sample introduction method. Unknown emission lines marked with A B and C in Fig. 3 were found to be S I 180.73 S I 182.04 and S I 182.62 nm Table 3 Emission characteristics of major iodine (I I) lines Table 4 iodine Optimized operating conditions for the determination of Parameter I 178.28 I 183.04 1206.14 R.f. forward power/W 0.9 0.9 1.2 Carrier gas flow rate/l min-' 1.1 1.2 1.2 Coolant gas flow rate/l min-' 17.0 18.0 18.0 Plasma gas flow rate/l min - ' 0.3 0.3 0.8 Observation height/nm 17.0 17.0 18.0 Slit-height/mm 5.0 5.0 2.0 Purge gas flow rates/l min-' Light cell 6.0 6.0 0 Monochromator 6.0 6.0 0 Sample solution flow rate/ml min-' 18.0 18.0 18.0 Oxidizing solution flow rate/ml min-' 9.0 9.0 9.0 (above load coil) respectively.These emission lines arose as a result of a small amount of sulfuric acid vapour being conveyed together with the generated elemental iodine. This was confirmed by the fact that no appreciable emission lines of sulfur were observed when conventional solution nebulization of a standard solution of much higher iodine concentration (ie. 25.0 pg ml-' of iodide) was carried out. Analyte Volatilization Reactions For sodium nitrite the concentration dependencies of the oxidizing agent and sulfuric acid on background-corrected emission intensity and line-to-background intensity ratio at the I 1 183.04nm line are shown in Fig.4.The optimum concentration and acidity of each of the six oxidizing solutions tested (see Table 5) were selected taking into consideration the emission characteristics of the iodine lines as well as the lifetime of the tube conveying the acidic solution. It can be seen from Table 5 that except for sodium nitrite the optimized concen- trations for the oxidants are to some extent different depending upon the iodine 178.28 183.04 and 206.16 nm lines; the reason for these differences is not known at present. Furthermore the use of these oxidizing solutions did not appear to produce any significant cumulative memory effects.Analytical Calibration Graphs and Detection Limits Under the optimized experimental conditions listed in Tables 4 and 5 double logarithmic analytical calibration graphs were obtained for iodine at I I 178.28 183.04 and 206.16 nm by the six analyte volatilization reactions. All analytical working curves showed relatively large dynamic ranges whereas those achieved with potassium dichroma te potassium permanganate and potassium perbromate gave somewhat shorter dynamic ranges. Typical calibration curves obtained with the other three oxidants H202 NaNO and K2S,0 are shown in Fig. 5. Each point on the plots was determined from at least five replicate measurements. The lowest point of each graph gener- Iodine generation with 0.2 pg ml-' of iodine Wavelength/nm In* zn/lbf BEC/pg ml-' 178.28 1 .oo 1 .oo 0.01 179.91 0.26 0.16 0.07 183.04 0.88 0.27 0.04 184.44 0.12 0.03 0.34 206.16 0.58 0.04 0.29 Solution nebulization with 25.0 pg ml-' of iodine In$ In/Ib§ BEC/pg ml ~ 0.74 1 .oo 2.4 0.26 0.21 11.4 1 .oo 0.32 7.6 0.17 0.05 53.2 0.67 0.05 45.5 *In is relative to the intensity of the 178.28 nm line. tlJlb is relative to the intensity ratio for the 178.28 nm line.$1 is relative to the intensity of the 183.04 nm line. $ I J I is relative to the intensity ratio for the 178.28 nm line.162 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 20 -= 10 0 ( C - B A I I I I I I I 20 ,= -= 10 -. 0 I I 1 1 1 1 0.1 0.5 1 5 10 50 ~NaNO,l/mmol I-' Fig.4 Dependence of (a) I and (b) ZJIb on the concentration of oxidizing agent (NaNO,) and sulfuric acid using the I 1 183.04nm line.Concentration of H,S04 A 3.0; B 5.0; and C 8.0 moll-'. Table 5 Optimized concentration and acidity of oxidizing solutions Oxidizing agent acidified with H,S04 H,O,/mmol 1-' H,SO,/mol 1 -' NaNO,/mmoll-' HzS04/mol 1- ' K,SzO,/mmol 1 - H,S04/mol 1- ' KBrO,/mmoll-' H,SO,/moll- ' K,Cr,O,/mmoll- ' H,SO,/mol 1-' KMnO,/mmoll-' H,SO,/mol I-' Wavelength/nm 178.28 10 8 5 8 5 8 10 1 1 8 0.01 3 183.04 100 8 5 8 10 8 50 1 10 8 0.1 8 206.1 6 50 8 5 8 50 8 5 3 5 8 0.1 8 ally corresponds to the determinable concentration limit of the analyte. The lowest concentration range was limited mostly by spectral background noise and by little or no significant reagent blank. Moreover it was considered that the upper concentration limit would be set by the chemistry of iodine generation and transport efficiency rather than the ICP detec- tion system.The difference in linear dynamic range between the iodine 178.28 183.04 and 206.16 nm lines can be attributed to the differences in the nature of these three lines. Relatively poor linearity at I I 178.28 nm has also been pointed out in the study with MIP-AES in this laboratory.'' The calibration graphs obtained at these three wavelengths show good linearity with correlation coefficients of better than 0.999 as illustrated in Fig. 5. Detection limits for iodine were extrapolated from the linear calibration graphs (Fig. 5 ) and are defined as the con- centration of analyte that would produce a net signal (Le. background-corrected line intensity) equal to three times the standard deviation of the background emission intensity in accordance with IUP'AC recommendation^.'^ The detection limits for the iodine lines studied using the three iodine generation methods are summarized in Table 6 where the results achieved in the present work are compared with those \obtained by the conventional solution nebulization method.6 'The detection limits were improved by two or three orders of magnitude. As a result in this study the best attainable limit of detection and the smallest BEC for iodine were achieved by using the iodine 178.28 nm line in combination with iodine generation by use of hydrogen peroxide as an oxiaant. In addition the best attainable limits of detection achieved here are approximately an order of magnitude in concentration better than that already reported in the 1iterat~re.l~ The instrumental precisions expressed as relative standard deviations (RSDs) were obtained at the three wavelengths and by the three methods for iodine generation with the use of different concentrations of iodine as iodide and are shown in Table 7.In conclusion taking the optimized oxidant concentrations (Table 5 ) and some analytical figures of merit into consider- ation an oxidizing solution of 5.0 mmol 1-' of sodium nitrite in 8.0 moll-' of sulfuric acid was found to be the most suitable of the different oxidation reactions for the generation of elemental iodine and was used subsequently. In addition to an oxidizing solution the use of the I 1 178.28 nm line was discarded for the remainder of this work because of its smaller linear dynamic range (Fig.5 ) . Interference Study Using the iodine lines at 183.04 and 206.16 nm with continu- ous-flow gas-phase sample introduction a study of various concomitant ions at concentrations 1 10 100 and 1000-fold greater than iodine (0.5 pg ml-' of iodide) was carried out. Several foreign species (i.e. cations and anions) at different concentrations were selected taking the constituents of brines into account. The results of the effect of some foreign ions interfering with the emission intensity of iodide at 183.04 and 206.16 nm by the proposed procedure are shown in Tables 8 and 9. Interference is considered to have occurred when the change in emission intensity is greater than f 3 % from that for iodine alone.The following ions did not interfere SO4,- NO,- C1- F- NH4+ and C204*-. And the following elements did not interfere Al As B Ba Be Bi Ca Cd Cs Cu Ge In K La Li Mg Mn Mo Na Ni P Pb Rb Sb Se Sr Th T1 W Y and Zr. Relative intensity is defined in this context as the ratio of the iodine emission intensity obtained in the presence of the foreign ion to that obtained when no foreign ion was present in the analyte solution. The presence of C032- and HC03- in particular caused an increased background due to copious amounts of simul- taneously evolved carbon dioxide which in the ICP source produced CO molecular band emission'' observed at 206.77 and 184.15 nm in the vicinity of the 206.16 and 183.04 nm lines of iodine respectively. As shown in Table 8 a slight enhance- ment in emission intensity of iodine lines in the presence of iodate was produced by oxidation of iodide with iodate as follows; 1 0 3 - +51-+6H+=312+3H20 whereas the presence of bromate and cyanide ions caused a depressing interference according to the following reactions respectively 2Bt-0,- + I = Br + 210,- 2CN- +12=21CN+2e- In addition the great decrease in iodine emission intensity in the presence of Ag Au Hg Pd and Pt ions as shown in Table 9 can be attributed to the formation of their stable and insoluble iodides.Furthermore there was little or no significant difference in the extent of interferences between the 183.04 and 206.16 nm iodine lines.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 163 (a) 108 - ( b ) ( C) A .- c A 10' 103 10 l o I 10 102 lo3 10" 1 10 lo2 lo3 l o A lo5 lo6 10 lo2 lo3 10" 105 106 Iodine concentrationhg ml-' Fig.5 Calibration graphs for iodine with various oxidizing agents A hydrogen peroxide; B sodium nitrite; and C potassium persulfate. (a) I I 178.28; (b) I I 183.04; and (c) I I 206.06 nm Table 6 Detection limits for iodine Wavelength/nm Oxidizing agent I 178.28 NaN02 H202 None* I 183.04 NaN02 H202 None* I 206.16 NaNO K2S208 K2S208 H202 K2S,O* None* Detection limit/ng ml- ' 0.392 0.34 1 0.820 0.553 0.676 1.18 2.12 3.10 11.1 92 88 559 BEC/pg ml-' 0.0 18 0.008 0.015 4.34 0.034 0.028 0.033 7.82 0.337 0.274 0.582 37.9 *Direct solution nebulization (ref. 6 ) . Table 7 Reproducibility* of iodine emission intensity measurements Iodine concentration/pg ml-l I 1 178.28 nm I I 183.04 nm 11 206.16 nm Oxidizing agent 0.1 1 .o 10 0.1 1 .o 10 0.1 1 .o 10 NaNO 0.9 3 0.92 0.51 1.74 1.21 1.12 2.82 1.94 1.82 H202 2.37 1.63 1.61 2.03 1.36 1.56 2.29 1.45 1.23 K23208 1.35 0.94 0.78 1.41 1.22 1.01 2.65 1.07 1.03 *Expressed as relative standard deviation (%); n = 10 Table 8 Effect of foreign ions on iodine emission intensity Species IO - Br- BrO,- ClO,- co,2- HC03- S2 - Compound added Na2S0 KIO KBr KBrO NaC10 Na2C0 NaHC0 Na2S Amount added/ pg ml-' 50 0.5 50 0.5 5 50 50 5 Relative intensity* I1 183.04 nm 103 107 100 99 99 97 102 101 I 1 206.16 nm 103 105 100 97 99 101 97 99 *Relative to 100 for the emission intensity of iodine (0.5 pg ml-l) alone.Applications iodine (iodide and iodate) in brine iodate must be reduced to - - Pre-reduction of iodate to iodide iodide prior to use of the present gas-phase sample introduction method.After optimization of the concentration of ascorbic It is well known that iodine is present in natural waters both as iodide and iodate." Therefore for the determination of total acid as a pre-redu~tant,'~ 0.003% m/v ascorbic acid was added to the sample solution to reduce iodate to iodide.164 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Table 9 Effect of foreign elements on iodine emission intensity Element Au Ce c o Cr Fe Ga Pd Pt Sn Te Ti V Zn Ag Hg Compound added 4 0 0 3 HAuC1 * 4H20 Ce2(S04)3' 8H20 Metal in HCI Feel * 6H20 Metal in HCI PdCI H2PtCl Metal in HC1 Na,Te03 Metal in HCI Metal in HCI KzCrzO7 NH4VOj Amount added/ pg ml-' 0.5 0.5 50 50 50 5 50 0.5 0.5 0.5 50 50 50 5 50 Relative intensity* I I 183.04 nm 0 31 98 99 101 95 99 0 0 31 100 102 97 98 101 I I 206.16 nm 1 47 98 100 100 99 100 0 0 35 100 100 98 99 101 *Relative to 100 for the emission intensity of iodine (0.5 pg m1-I) alone.Table 10 Determination of iodine (pg m1-l) in brines This work* Standard additions method Calibration graph method Sample 183.04 nm 206.16 nm 183.04 nm 206.16 nm Reference method A 90.8 -t 2.0 87.6 f 2.2 90.9 1.3 91.0k 1.9 90.5t 92.4$ B 70.1 k 2.6 68.7 & 3.5 71.4 & 2.2 70.1 f3.4 70.6t 69.61 C 105.9 f 2.4 104.7 f 3.2 106.0 k 2.7 107.3 2.9 106.5t 107.31 *Mean value f standard deviation (n = 5). ?Determined by ICP-AES with prior-oxidation (ref. 6). $Determined by atmospheric pressure helium MIP-AES with continuous-flow gas-phase sample introduction (ref.10). Determination of total iodine in brines The accuracy and precision of the present continuous-flow gas-phase sample introduction method for ICP-AES were established by analysing several samples of natural water. To this end the procedure described below was followed for the determination of total (iodide and iodate) concentrations in brines taken at Mobara City Chiba Prefecture Japan. The presence of some matrix components would be expected to interfere in the determination of iodine by the proposed method depending upon whether or not the gas-phase sample introduction procedure and the off-peak background correc- tion techniques were used. Recovery studies therefore were made by adding known amounts of iodine as iodide or iodate to diluted samples and applying the recommended procedure with 5 mmoll-' of sodium nitrite in 8.0 moll-' sulfuric acid.Mean recoveries of added iodine were in the range 97.1-104.6% depending on the brine sample being measured. Therefore both the calibration graph method and the method of standard additions were employed for the determination of iodine. A sample was filtered through a No. 2 filter-paper (Toyo Roshi) and the filtrate was analysed for total iodine after spiking with iodine in the range 1-10 pg m1-l. A dilution factor of 50 was used for the analysis of brine samples. The results obtained by the present method for several samples of brines are presented in Table 10. At the same time the total iodine in the samples was determined by ICP-AES with the prior-oxidation procedure6 and quantified by atmospheric- pressure helium MIP-AES with continuous-flow gas-phase sample introduction." It should be noted that there is little or no appreciable difference in the iodine concentration between the 183.04 and 206.16 nm iodine lines and that the results obtained by the present method are in good agreement with those obtained by other techniques.Conclusions A simple analyte volatilization procedure for ICP-AES has been developed and the results presented which allows the continuous-flow determination of iodine at low concentrations. As well as can be established this on-line system with a simple gas-liquid separator for the generation of elemental iodine by oxidation of aqueous iodide has been proposed for the first time and the best attainable limits of detection obtained in this work were approximately an order of magnitude better than those reported in the 1iterat~re.l~ In conclusion the proposed method has been successfully applied to the determination of total iodine (Le.iodide plus iodate) in samples of brines giving RSDs of 1.4-5.1%. The authors thank DP. I. Makino of Asahi Glass for supplying the brine samples investigated in this work. References 1 Niazi S. B. and Mozammi M. Anal. Chim. Acta 1991 252 115. 2 Koh T. Ono M. and Makino I. Analyst 1988 113 945. 3 Ebihara M. Saito N. Akaiwa H. and Tomura K. Anal. Sci. 1992 8 183. 4 Cox R. J. Pickford C. J. and Thompson M. J. Anal. At. Spectrom. 1992 7 635. 5 McGregor D. A. Cull K. B. Gehlhausen J. M. Viscomi A. S. Wu M. Zhang L. and Carnahan J . W. Anal. Chem. 1988 60 1089A. 6 Nakahara T. and Wasa T. Appl. Spectrosc. 1987 41 1238. 7 Browner R. F. and Boorn A. W. Anal. Chem. 1984 56 786A. 8 Nakahara T. and Wasa T. Microchem. J. 1990 41 148. 9 Barnett N. W. and Kirkbright G. F. J. Anal. At. Spectrom. 1986 1 337.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 165 10 Nakahara T. Yamada S. and Wasa T. Appl. Spectrosc. 1990 44 1673. 11 Quintero Ortega M. C. Cotrino Bautista J. Saez M. Menendez Garcia A. Sanchez Uria J. E. and Sanz Medel A Spectrochim. Acta Part B 1992 47 79. 12 Dolores Calzada M. Carmen Quintero M. Gamero A. and Gallego M. Anal. Chem. 1992 64 1374. 13 Dolan S. P. Sinex S. A. Capar S. G. Montaser A. and Clifford R. H. Anal. Chem. 1991 63 2539. 14 Cave M. R. and Green K. A. J. Anal. At. Spectrom. 1989,4 223. 15 Nakahara T. Spectrochim. Acta Part B 1985 40 293. 16 CRC Handbook of Chemistry and Physics ed. Weast R. C. CRC Press Boca Raton 66th edn. 1985 p. D-151. 17 IUPAC Nomenclature Symbols Units and Their Usage in Spectrochemical Analysis Revision 1975 Part 11 Spectrochim. Acta Part B 1978 33 248. 18 Gaydon A. G. The Spectroscopy of Flames 2nd edn. Chapman and Hall London 1974 p. 351. 19 Nakayama E. Kimoto T. and Okazaki S. Anal. Chem. 1985 57 1157. Paper 3/048140 Received August 1 I 1993 Accepted August 31 1993