Elsevier

Talanta

Volume 72, Issue 2, 30 April 2007, Pages 626-633
Talanta

Pervaporation-flow injection with chemiluminescence detection for determination of iodide in multivitamin tablets

https://doi.org/10.1016/j.talanta.2006.11.033Get rights and content

Abstract

This paper describes the use of a pervaporation (PV) technique in a flow injection (FI) system for selective improvement in iodide analysis. Iodide in the sample zone is oxidized to iodine, which permeates through a hydrophobic membrane. Detection of the diffused iodine is achieved using the chemiluminescent (CL) emission at 425 nm that results from the reaction between iodine and luminol. The method was applied for the analysis of some pharmaceutical products, such as nuclear emergency tablets and multivitamin tablets. Ascorbic acid present in multivitamin samples interfered seriously with the analysis, and off-line sample treatment using anion exchange resin was employed to successfully remove ascorbic acid before the analysis. Ascorbic acid was flushed from the column using 0.4 M sodium nitrate followed by elution of iodide with 2 M sodium nitrate. The detection limit (3S.D.) of the system was 0.5 mg l−1, with reproducibility of 5.2% R.S.D. at 5 mg l−1. Sample throughput was determined as 30 injections h−1. There was good agreement between iodide concentrations from extracted samples determined using four different methods, i.e., PV-FI, gas diffusion-flow injection, potentiometry and ICP-MS. A comparison of the analytical features of the developed pervaporation system with these of the previously reported chemiluminescence gas diffusion-flow injection previously reported is also described.

Introduction

Iodine is an essential micronutrient as a part of thyroid hormone, which is necessary for normal brain development. Inadequate iodine during prenatal and early development periods, can lead to several diseases, including spontaneous abortion, increased infant mortality, hypothyroidism and cretinism [1]. Iodine supplementation in food and the use of multivitamins containing potassium iodide is used to avoid these symptoms. However, excessive iodine intake can reduce thyroid function because large amounts of iodine block the thyroid's ability to produce the hormones thyroxin and triiodothyronine [2]. Hence, there has been an increase in interest in the analytical control of iodide in food and pharmaceutical products.

In order to determine low level of iodide, many methods based on different principles have been proposed. These include ion [3] and ion-pair reversed-phase [4], [5], [6] high performance liquid chromatography with either post column reaction [3] or electrochemical detection [4], [5], [6]. Neutron activation analysis [7], [8] and inductively coupled plasma-mass spectrometry [9], [10], [11] have also been described for determination of iodide. These last two techniques offer high sensitivity and selectivity for iodide measurement but require a high level of specialization and the facilities are expensive to establish and operate. Direct determination of iodide can be performed using ion selective electrode (ISE). Nevertheless, for the complicated sample, such as in urine [12] and milk [13], the matrix interference was suspected to contribute some electrode response.

A number of colorimetric methods for the quantitative determination of iodide in aqueous samples have been proposed [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24]. The most frequently used methods are based on the catalytic effect of iodide on a reaction between Ce(IV) and As(III) [14], [15], [16] or the decomposition of the Fe(III)–thiocyanate complex in nitric acid solution [17], [18]. In these approaches, determinations were very sensitive but selectivity of the methods was not satisfied. Direct colorimetric methods for determination of iodide have also been presented. Agrawa et al. [19] reported that leucocrystal violet was selectively oxidized with iodine (I2) to form crystal violet dye which was then extracted in chloroform before measuring of absorbance at 588 nm. A procedure for the determination of iodide in charcoal impregnated with potassium iodide was developed by Taylor et al. [20]. An aqueous extract of iodide in the charcoal was converted to iodine with iodate in acidic media followed by spectrophotometric measurement at 460 nm. Similarly, tri-iodide detection by flow injection was reported by Kamson [21] and subsequently by Ensafi and Dehaghi [22]. To improve the sensitivity of the tri-iodide detection, the measurements of tri-iodide starch complex have been proposed [23], [24].

Chemiluminescence (CL) detection is attractive in terms of the relatively low cost and the simplicity of the equipment involved. Moreover, the previously reported chemiluminescent iodine–luminol reaction permits detection of iodine to as low as 1 × 10−7 M to be achieved [25]. Iodide does not react with luminol, and it is therefore necessary to first oxidize iodide to elementary iodine, in order to initiate chemiluminescence. A flow injection system with CL detection for the determination of iodide was reported by Burguera et al. [26]. The iodine was generated in a closed headspace device and was carried out in a nitrogen gas flow to a vial, where it was trapped in a KI solution. The trapped iodine was then mixed with Co(II) and luminol solutions in a FI system to produce CL light. This was applied for urinary iodine determination and gave a detection limit of 10 μg l−1. Use of on-line oxidation and solvent extraction coupled with reversed micellar mediated CL detection has also been described for the determination of iodine and iodide in commercial gargle products [27]. A detection limit of 0.02 ng ml−1 was achieved.

According to previous reports, detection of iodine by luminol CL may be susceptible to a number of interferences including metal ions, and for this reason, a separation technique is usually carried out in the analysis of real samples [26], [27]. Manual separations are always time consuming, laborious and difficult to perform in micro scale, but these can be automated by incorporation of hydrophobic membrane-based gas diffusion (GD) and pervaporation (PV) techniques into a flow injection (FI) system, and used to separate volatile analyte form. Improvement in selectivity is thus achieved because fewer species are converted to the gaseous form at room temperature [28].

To date, only a few methods involving GD-FI have been proposed for determination of iodide [29], [30], [31]. All of these detections of iodine were based on a simple photometric measurement of either I3 or the I3–starch complex. We have also reported use of the GD-FI method for determination of iodide in some pharmaceutical samples but in this instance, CL detection of diffused iodine was employed for detection. The method was successfully applied for nuclear emergency tablets (potassium iodide tablet) and Thai liquid medicine formulations, but could not be used for multivitamin sample determination because of interference by ascorbic acid [34].

Pervaporation coupled with FI has been described for the direct measurement of volatile and semi-volatile analytes in samples that may cause deterioration of the gas permeable membrane if used in GD-FI [32], [33]. In PV-FI, the presence of an air gap between donor solution (sample) and the membrane ensures that contamination or deterioration of the latter is avoided. This paper presents PV-FI with CL detection for determination of iodide in pharmaceutical products including multivitamin tablets containing ascorbic acid that interfered in the GD-FI method described earlier [34]. Off-line sample treatment with an anion exchange column was used to separate ascorbic acid from iodide before analysis. Using the PV unit brings an advantage that prolongs the life-time of the PTFE membrane by avoiding the direct contact of high salt content solution of sodium nitrate, which was used in the sample treatment.

Section snippets

The PV-FI manifold

The FI system with a pervaporation unit is depicted in Fig. 1. An AS-90 series autosampler (Perkin-Elmer, USA) was used for automatically loading of standard or sample solutions into the FI injection valve. A FIAS-300 module (Perkin-Elmer, USA) was employed for pumping the reagents. A home-made pervaporation unit consisted of two circular Perspex blocks (61 mm diameter, 25 mm deep) held together by stainless steel ring clamps and four stainless steel bolts. Both the acceptor chamber (0.3 mm deep)

Manifold design and optimization

Our previous work using GD-FI system [34] clearly demonstrated that molecular iodine was readily adsorbed onto the surface of the manifold tubing and the pores of the hydrophobic membrane. Without a conditioning stream of iodide solution (depicted as dashed line in the FI manifold shown in Fig. 1), signals within this range were not reproducible, and showed a gradual increase to reach a plateau as subsequent injections were made. Moreover, the adsorption of iodine resulted in a non-linear

Conclusions

In this work, a PV-FI method was developed for the determination of iodide based on the chemiluminescent reaction of I2 with luminol. The method is directly applicable for nuclear emergency tablets, which are basically potassium iodide with pharmaceutical binders. For multivitamins, an off-line clean-up using AER was employed for the separation of ascorbic acid from iodide. Separate elution between iodide and the Vitamin C interference was optimized using a so-called dual-detection FI manifold.

Acknowledgements

The authors would like to thank the Royal Golden Jubilee Ph.D. scholarships for the grant given to N. Ratanawimarnwong. We also would like to thank the Thailand Research Fund (TRF) and the Postgraduate Education and Research Program in Chemistry (PERCH) for their financial supports of this project.

References (37)

  • W. Hu et al.

    J. Chromatogr. A

    (2002)
  • J. Odink et al.

    J. Chromatogr.

    (1988)
  • F. Moussa et al.

    J. Chromatogr. B

    (1995)
  • G.T.F. Wong et al.

    Anal. Chim. Acta

    (1976)
  • J. Melichercik et al.

    J. Dairy Sci.

    (2006)
  • N. Choengchan et al.

    Talanta

    (2002)
  • W.M. Hardy et al.

    Talanta

    (1977)
  • J.L. Burguera et al.

    Talanta

    (1996)
  • S. Motomizu et al.

    Anal. Chim. Acta

    (1992)
  • M.D. Luque de Castro et al.

    Trends Anal. Chem.

    (1998)
  • L. Wang et al.

    Anal. Chim. Acta

    (2000)
  • N. Ratanawimarnwong et al.

    Talanta

    (2005)
  • K. Grudpan et al.

    Talanta

    (1999)
  • J.T. Dunn, IDD Newsletter, November 1,...
  • M.G. Wohl et al.

    Modern Nutrition in Health and Desease

    (1983)
  • J. Rendl et al.

    Clin. Chem.

    (1994)
  • X. Hou et al.

    Anal. Chem.

    (1999)
  • V. Poluzzi et al.

    J. Anal. Atom. Spectrom.

    (1996)
  • Cited by (0)

    View full text