Anal. Chem. 1997, 69, 3628-3632
Determination of Iodide in Seawater by Ion Chromatography Kazuaki Ito
Department of Environmental Science, Faculty of Engineering, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima 739, Japan
A simple and highly sensitive ion chromatographic method with UV detection was developed for iodide (I-) in seawater. A high-capacity anion-exchange resin with polystyrene-divinylbenzene matrix was used for both preconcentration and separation of iodide. Iodide in artificial seawater (salinity, 35‰) was trapped quantitatively (98.8 ( 0.6%) without peak broadening on a preconcentrator column and was separated with 0.35 M NaClO4 + 0.01 M phosphate buffer (pH 6.1). On the other hand, the major anions in seawater, chloride and sulfate ions, were partially trapped (5-20%) and did not interfere in the determination of I-. The detection limit for I- was 0.2 µg/L for 6 mL of artificial seawater. The present method was applied to determination of I- (ND - 18.3 µg/L) and total inorganic iodine (I- + IO3- - I, 50.0-52.7 µg/L) in seawater samples taken near Japan. Iodine species in seawater exist as iodide (I-) and iodate (IO3-). Iodide, which is thermodynamically unstable in oxygenated water, is usually a minor species in seawater compared to iodate, and total inorganic iodine concentration is approximately in the range of 50-60 µg/L.1-14 However, iodide concentration ranges from 30 µg/L (near the shore and in ocean surface and bottom waters) to below 1 µg/L (deep ocean water).1-3,6-8,11,13,14 Iodine is an essential micronutrient for many organisms. Iodide in seawater is produced by biologically mediated reduction of iodate15 and is also produced under reducing conditions. Thus, the distribution of iodide and iodate gives clues for understanding the marine environment.4,5,12-14 Iodide has been determined mainly by difference between total inorganic iodine (I- and IO3-) and IO3-:iodate has been determined by differential pulse (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Tsunogai, S. Deep-Sea Res. 1971, 18, 913-919. Tsunogai, S.; Henmi, T. J. Oceanogr. Soc. Jpn. 1971, 27, 67-72. Wong, G. T. F.; Brewer, P. G. Anal. Chim. Acta 1976, 81, 81-90. Smith, J. D.; Butler, E. C. V.; Airey, D.; Sandars, G. Mar. Chem. 1990, 28, 353-364. Wong, G. T. F.; Zhang, L. Cont. Shelf Res. 1992, 12, 717-733. Nakayama, E.; Kimoto, T.; Okazaki, S. Anal. Chem. 1985, 57, 1157-1160. Nakayama, E.; Kimoto, T.; Isshiki, K.; Sohrin, Y.; Okazaki, S. Mar. Chem. 1989, 27, 105-116. Herring, J. R.; Liss, P. S. Deep-Sea Res. 1974, 21, 777-783. Takayanagi, K.; Wong, G. T. F. Talanta 1986, 33, 451-454. Truesdale, V. W. Mar. Chem. 1978, 6, 253-273. Luther, G. W., III; Swartz, C. B.; Ullman, W. J. Anal. Chem. 1988, 60, 1721-1724. Jickells, T. D.; Boyd, S. S.; Knap, A. H. Mar. Chem. 1988, 24, 61-82. Wong, G. T. F.; Takayanagi, K.; Todd, J. F. Mar. Chem. 1985, 17, 177183. McTaggart, A. R.; Butler, E. C. V.; Haddad, P. R.; Middleton, J. H. Mar. Chem. 1994, 47, 159-172. Tsunogai, S.; Sase, T. Deep-Sea Res. 1969, 16, 489-496.
3628 Analytical Chemistry, Vol. 69, No. 17, September 1, 1997
polarography4,5,8,9,13,14 and the spectrophotometric method after conversion of IO3- to I3- in acidic solution.1,2,10,12 Total inorganic iodine was determined after converting iodide to iodate by using chemical or photochemical means.1,2,4,5,8-10,12,13 However, as the detection limit of iodide by the difference technique is not very good (for example, 1.3-2.5 µg/L for differential pulse polarography5,8), it is desirable to detect iodide directly. To date, a few methods have been proposed for direct determination of trace iodide in seawater. The first involved the use of neutron activation analysis,3 where iodide in seawater was concentrated by a strongly basic anion-exchange column, eluted by sodium nitrate, and precipitated as palladium iodide. The second involved the use of automated electrochemical procedures;6,7 iodide was electrochemically oxidized to iodine and was concentrated on a carbon wool electrode. After removal of interference ions, the iodine was eluted with ascorbic acid and was determined by a polished Ag3SI electrode. The third method involved the use of cathodic stripping square wave voltammetry.11 Iodide reacts with mercury in a one-electron process, and the sensitivity is increased remarkably by the addition of Triton X. The three methods have detection limits of 0.7 (250 mL of seawater sample), 0.1 (50 mL), and 0.02 µg/L (10 mL), respectively, and could be applied to almost all the samples. However, neutron activation analysis is not generally employed. The second method uses an automated system but is a special apparatus just for determination of iodide. The first and third methods are timeconsuming. Recently, attention has been paid to ion chromatography (IC) as a simple method for determining trace iodide in seawater.14,16-19 Two factors are essential in order to attain good sensitivity: (1) separation of iodide from an excess of anions in seawater and (2) highly sensitive detection of iodide. Iodide was determined by an iodide-selective electrode (Ag2S/AgI) after other anions were separated by NaNO3 eluent.16 However, the electrode that was stabilized by 0.5 µM iodide responded to chloride ion in seawater, and the detection limit of iodide was ∼2 µg/L. We showed that 0.1 M sodium chloride eluent was effective for the separation of iodide in seawater, due to faster elution of an excess of anions in seawater and slow elution of iodide with hydrophobicity on conventional low-capacity anion-exchange columns (polymethacry(16) Butler, E. C. V.; Gershey, R. M. Anal. Chim. Acta, 1984, 164, 153-161. (17) (a) Ito, K.; Sunahara, H. Bunseki Kagaku, 1988, 37, 292-295. (b) Ito, K.; Sunahara, H. J. Chromatogr. 1990, 502, 121-129. (18) Ito, K.; Shoto, E.; Sunahara, H. J. Chromatogr. 1991, 549, 265-272. (19) Branda˜o, A. C. M.; Buchberger, W. W.; Butler, E. C. V.; Fagan, P. A.; Haddad, P. R. J. Chromatogr., A 1995, 706, 271-275. S0003-2700(97)00078-4 CCC: $14.00
© 1997 American Chemical Society
late and styrene-divinylbenzene copolymer)17 and an ODS-column coated with cetyltrimethylammonium.18 However, the detection limits were ∼5 µg/L (100 µL of sample) for UV detection and amperometry using a glassy carbon electrode. For a 500 µL sample volume, a detection limit of 1.3 µg/L was obtained.14 A lower detection limit (0.8 µg/L for 150 µL of sample) was obtained for postcolumn reaction of iodide with 4,4′-bis(dimethylamino)diphenylmethane and 0.1 M NaCl eluent system.19 In this study, sample injection of high volume was examined for a simple and highly sensitive determination of iodide in seawater by conventional LC with UV detection. However, a large amount of seawater sample itself contains an excess of Cl- and SO42- and might exert as eluent.17-19 Thus, it is necessary to optimize the IC system to depress both the broadening of the iodide peak and the elution of iodide by seawater. Styrenedivinylbenzene copolymer with high anion-exchange capacity and NaClO4 eluent with strong eluting power were selected for this purpose. The resin has strong affinity for iodide with hydrophobicity17 and also is useful to avoid overloading of seawater samples due to its high anion-exchange capacity. Two sample injection methods, concentrator column and sample loop systems, and the effect of salinity in seawater were examined for determination of iodide, and the optimized IC system was applied to seawater samples.
EXPERIMENTAL SECTION Apparatus. The ion chromatographic system consisted of (a) a Tosoh Model CCPM pump (Tokyo, Japan), (b) a Rheodyne 7125 injector with a concentrator column (10 mm × 4 mm i.d.) or sample loop of 2 or 5 mL, (c) a Hitachi Model L-4200 UV-visible detector (Tokyo, Japan), and (d) a Tosoh Model SC-8010 chromatoprocessor. The resin with high anion-exchange capacity was TSKgel SAX (Tosoh; styrene-divinylbenzene copolymer with ∼3.7 mequiv/g anion-exchange capacity; functional group, benzyltrimethylammonium type; particle size, 5 µm). The slurry of resin in 0.5% sodium acetate solution was packed in a concentrator column (10 mm × 4 mm i.d.) and in a separation column (150 mm × 4.6 mm i.d.) at flow rates of 4 and 2 mL/min, respectively, by the slurry packing technique. Reagent and Mobile Phase. All inorganic salts for standard solutions and mobile phase preparation were sodium salts of analytical reagent grade. Stock solutions of each anion (10 g/L) were prepared by dissolving the salts in deionized water and were mixed and diluted to the desired concentrations. The mobile phase was 0.35 M NaClO4 + 0.01 M sodium phosphate buffer (pH 6.1). Stock solution of 0.5 M phosphate buffer was prepared from 0.15 M Na2HPO4 + 0.35 M NaH2PO4. Artificial seawaters were prepared according to the Lyman-Fleming formula.20 Analysis of Iodide and Total Iodine. For iodide, seawater samples (2 or 6 mL), filtered through a 0.45 µm membrane filter, were passed through a concentrator column (10 mm × 4 mm i.d.), and trapped iodide was eluted with 0.35 M NaClO4. For total inorganic iodine, ascorbic acid and acetic acid (final concentrations, 4 × 10-4 and 0.06 M, respectively) were added to seawater samples for reduction of iodate to iodide. After that, the treated (20) Lyman, J.; Fleming, R. H. J. Mar. Res. 1940, 3, 134-146.
Figure 1. Ion chromatograms of inorganic anions. Conditions: concentrator and separation columns, TSK gel SAX resin, 10 mm × 4 mm i.d. and 150 mm × 4.6 mm i.d., respectively; mobile phase, 0.35 M NaClO4 + 0.01 M sodium phosphate buffer (pH 6.1); detection, UV absorbance at 226 nm; flow rate, 1 mL/min.
samples were left standing for 1 h to complete of the reaction. Ascorbic acid (0.2 M) was made just before addition to the samples. The filtered samples (1 mL) and subsequently H2O (2 mL) were passed through a concentrator column, and trapped iodide was eluted again with 0.35 M NaClO4. Seawater samples (salinity, 34.32-34.74‰; depth, 0-400 m) taken near Japan were collected with a Van Dorn sampler during a cruise (May 31-June 7, 1996) of the R&T/V “Toyoshiomaru” (Hiroshima University). Samples were stored refrigerated or frozen just after sampling and were analyzed after 1-2 weeks in the laboratory. Sampling stations (St1 and St2) were 31°17′70′′ N, 131°53′00′′ E (June 1, 1996) and 30°18′74′′ N, 130°14′10′′ E (June 2, 1996), respectively.
RESULTS AND DISCUSSION Separation of Iodide. Figure 1 shows the separation of UVabsorbing anions with 0.35 M NaClO4 eluent at 226 nm. The dependence of the retention volume of each anion on the eluent concentration was examined in the range of 0.1-0.7 M NaClO4. Iodide was completely separated from the anions IO3-, NO2-, NO3-, and Br- with hydrophilicity and SCN- with hydrophobicity in the range of eluent concentration examined, although the retention volume of each anion decreased with increase in NaClO4. However, tentative experiments showed that separation of trace iodide peak from major anions in 35‰ artifitial seawater was not very good with higher NaClO4 eluent. Thus, 0.35 M NaClO4 was selected for the separation of iodide with shorter retention time and without interference by anions in seawater. The elution order of anions, IO3- < NO2- < NO3- < Br- < I- < SCN-, was the same as that with 0.1 M NaCl eluent and low-capacity anionexchange columns (polymethacrylate and styrene-divinylbenzene Analytical Chemistry, Vol. 69, No. 17, September 1, 1997
3629
Figure 2. Ion chromatograms of iodide in artificial seawater (salinity, 35‰). (a) Concentraor column system: solid line, 1 µg/L (sample volume, 6 mL); dashed line, 0 µg/L (6 mL). (b) Sample loop injection system: solid line, 1 µg/L (5 mL); dashed line, 0 µg/L (5 mL). Other conditions were the same as in Figure 1.
copolymer and an ODS-column coated with cetyltrimethylammonium), except for NO3- > Br-.17,18 However, for the present column, NaCl eluents with even higher concentration (0.5-2.0 M) were not successful due to strong affinity of iodide to the resin, thus suggesting the usefulness of NaClO4 eluent with strong eluting power. The iodide peak obtained with a new column packed with the slurry packing method was relatively low at first. Thus, 0.1 mg/L iodide (2 mL) was injected successively (about five times) to attain constant peak height and area. This was only one procedure after packing the resin and was not necessary any more. Sample Injection Method. Figure 2 shows the ion chromatograms of trace iodide (1 and 0 µg/L) in artificial seawater (salinity, 35‰) by two sample injection methods, a concentrator column and sample loop systems. With a concentrator column (10 mm × 4 mm i.d.), a good iodide peak was obtained without interference by an excess of anions:20 19 300 mg/L for Cl-, 2710 mg/L for SO42-, 67 mg/L for Br-, and 142 mg/L for HCO3- (6 mL, Figure 2a). Iodide was quantitatively trapped by the column: when 100 µg/L iodide (4 mL) in 35‰ artificial seawater was passed through the small column, the trapping efficiency of iodide was 98.8 ( 0.6%, n ) 4. This was in contrast to the lower trapping efficiency (4.4 ( 0.2%, 4 mL of artificial seawater) of a concentrator column (10 mm × 4 mm i.d.) of same matrix with low anion-exchange capacity (Yokogawa AX-1, 0.02 mequiv/g). This result indicates that high anion-exchange capacity is essential for concentration of iodide in seawater of high volume. On the other hand, the trapping efficiency of Cl- and SO42- was low, even for the resin with high exchange capacity, and the trapping efficiency decreased with the increase in sample volume: 17.7 ( 0.8% and 11.7 ( 1.0% (n ) 4 each) for 2 mL of sample and 6.6 ( 1.3% and 4.6 ( 2.5% (n ) 3 each) for 6 mL of sample, respectively. Thus, a concentrator column in this study has two advantages: 3630
Analytical Chemistry, Vol. 69, No. 17, September 1, 1997
(1) quantitative concentration of iodide in seawater and (2) removal of coexisting anions. For the sample loop system (5 mL of sample), the iodide peak height was smaller than that for a concentrator column system and overlapped with salinity (Figure 2b). However, the iodide peak for 2 mL sample did not overlap with salinity. This indicated that all salinity was introduced and spread in a separation column for the sample loop injection system, and thus the complete separation of iodide from salinity was not enough for the large volume of sample, especially for 5 mL of sample. This was in contrast to a concentrator column system: there, salinity was partially trapped in a concentrator column and was easily separated from iodide on a separation column (Figure 2a). Effects of Sample Volume. For a concentrator column system, the dependence of both number of theoretical plates per column and peak area for iodide on injection volume was examined using 35‰ artifitial seawater with the same amount of iodide (0.1 µg of I-, from 100 (sample volume, 1 mL) to 10 µg/L (10 mL)). The number of theoretical plates increased a little with the volume (y ) 1573 + 16.7x, R2 ) 0.782; y, theoretical plates/ column; x, sample volume in mL), and the retention time also increased from 4.98 (1 mL of sample) to 5.12 min (10 mL). On the other hand, iodide peak area decreased with an increase in the volume (y ) 478.7-10.7x, R2 ) 0.973; y, peak area; x, sample volume in mL). The same pattern was also obtained for lower iodide (0.02 µg of I-), 20 (1 mL)-2 µg/L (10 mL). These results can be explained as follows: although 35‰ artifitial seawater has an excess of anions, Cl- ) 0.54 M () 17 300 mg/L), SO42- ) 0.03 M () 2710 mg/L), it hardly exerted for elution of iodide in spite of a larger sample injection, as is clear from the smaller change in retention volume. However, the introduction of salinity to a concentrator column brought the ClO4- form of the column
Figure 3. Ion chromatograms of seawater samples. (a) St1 (I- ) 0.82 µg/L; sample volume, 6 mL); (b) St2 (I- ) 5.1 µg/L; 2 mL); (c) St2 (I+ IO3- - I, 52.7 µg/L, 1 mL). IO3- was reduced to I- by ascorbic acid. Other conditions were the same as in Figure 1.
partially to Cl- and SO42- forms. As the selectivity of I- to Cland SO42- for ion-exchange reaction is higher than that to ClO4-, iodide was concentrated around the inlet of a concentrator column and was back-flushed with NaClO4 eluent, resulting in a sharp iodide peak. This result was different from the data obtained with the low anion-exchange capacity column using sodium chloride eluent (0.1 M):17-19 salinity in sample exerted as eluent, and the iodide peak became broad with an increase in salinity. For a salinity of 0.7 M NaCl, for example, the theoretical plates of iodide decreased to one-fourth with an increase in sample volume from 100 to 200 µL.19 On the other hand, the decrease in iodide peak area with injection volume is probably attributed to the decrease in trapping efficiency due to an increase in salinity, viz, an increase in the ratio of salinity to iodide. Calibration Curve, Repeatability, and Detection Limit. Peak area of iodide decreased with salinity: y ) 211.7-0.520x, R2 ) 0.940; y, peak area (20 µg/L I-, 2 mL); x, salinity (x ) 0-45‰). On the other hand, retention volume increased from 4.56 (0‰) to 4.82 min (45‰), and also the theoretical plates per column increased from 823 to 960. This also indicates that salinity in seawater hardly exerts as eluent and assists in concentrating iodide effectively. However, as the decrease in peak area with salinity is small, it is possible to use external standard for determination. The calibration graph of iodide by peak area was y ) 10.03x-2.93 (x ) 1-100 µg/L), R2 ) 1.000, by a concentrator column for 2 mL of sample (35‰). The reproducibility of replicate injections of 5 (2 mL of sample) and 1 µg/L (6 mL of sample) was good: the relative standard deviations (RSDs) were 1.9% (n ) 10) and 1.6% (n ) 5), respectively. The detection limit for a concentrator column system was 0.2 µg/L (signal-to-noise ratio of 2) for 6 mL of artificial seawater. Although this value is higher than the cathodic stripping square wave voltammetry11 (detection limit of 0.02 µg/L (50 mL)), it is good enough for application of real samples. In addition, the present method is simple to operate, and no special techniques are necessary.
Figure 4. Vertical profiles of iodide and total inorganic iodine at St1 and St2.
Application to Seawater Samples. Figure 3 shows the separation of iodide in seawater samples taken near Japan. IO3was reduced to I- by ascorbic acid and acetic acid.6,7,18 Good chromatograms were obtained without any interferences. The determination was done using calibration curves in relatively narrow ranges (0-2 µg/L, 2-20 µg/L for iodide and 40-70 µg/L for total inorganic iodine), because the resin showed a small background iodide peak (shown in Figure 2a), and the range of iodide concentration in seawater samples was relatively limited. The corresponding standard solutions were injected after every two or three sample injections.19 Figure 4 shows the analytical results of seawater samples taken near Japan. Although the concentration of iodide in seawater varied significantly with location and depth, the patterns were similar with previous data:1-3,6-8,11,13,14 (1) iodide concentration decreased steeply with an increase in depth and (2) the total inorganic iodine concentration was conservative with salinity. The accuracy of data obtained was also supported by a recovery test: for St2 (0 and 200 m deep), Analytical Chemistry, Vol. 69, No. 17, September 1, 1997
3631
99.5% and 99.0% for addition of 20 µg/L I- (iodide determination) and 99.2% and 102.4% for the addition 50 µg/L IO3--I (total iodine determination), respectively. In conclusion, the present method for determination of iodide in seawater is simple and easy to operate, compared to existing methods. Iodide in seawater was quantitatively concentrated without broadening of the iodide peak on styrene-divinylbenzene resin with high anion-exchange capacity. Thus, this method will become a more acceptable technique for determination of iodide in seawater.
3632
Analytical Chemistry, Vol. 69, No. 17, September 1, 1997
ACKNOWLEDGMENT I am grateful to Dr. S. Ohtsuka of Fisheries Laboratory, Faculty of Applied Biological Science, Hiroshima University, for offering the seawater samples and his useful discussion. Received for review January 22, 1997. Accepted June 6, 1997.X AC9700787 X
Abstract published in Advance ACS Abstracts, July 15, 1997.
