A green analytical procedure for sensitive and selective determination of iron in water samples by flow-injection solid-phase spectrophotometry

Talanta 71 (2007) 1507–1511

A green analytical procedure for sensitive and selective determination of iron in water samples by flow-injection solid-phase spectrophotometry Leonardo S.G. Teixeira a , F´abio R.P. Rocha b,∗ a

Departamento de Engenharia e Arquitetura, Universidade Salvador, Av. Cardeal da Silva 132, 40220-141, Salvador, BA, Brazil b Instituto de Qu´ımica, Universidade de S˜ ao Paulo, PO Box 26077, 05513-970 S˜ao Paulo, SP, Brazil Received 21 March 2006; received in revised form 12 July 2006; accepted 13 July 2006 Available online 14 August 2006

Abstract A greener analytical procedure based on flow-injection solid-phase spectrophotometry is proposed for iron determination. Iron(II) is reversibly retained on 1-(2-thiazolylazo)-2-naphthol immobilized on C18-bonded silica, yielding a brown complex. The metal ion is eluted as iron(II) with a small volume of a diluted acid solution without removing the immobilized reagent, which can be used for at least 100 determinations. Other chemicals (buffer and reducing agent) were carefully selected taking into account the analytical performance and toxicity. The developed procedure is 10-fold more sensitive in comparison to the analogous procedure based on measurements in solution, being suitable for the determination of iron in water samples with good accuracy and precision. The detection limit (99.7% confidence level), sampling rate and coefficient of variation (n = 10) were estimated as 15 ␮g L−1 , 25 measurements per hour and 4.0%, respectively. The proposed procedure involves a reduced effluent generation (3.6 mL per determination) and consumes micro amounts of reagents. © 2006 Elsevier B.V. All rights reserved. Keywords: Green chemistry; Flow injection analysis; Solid-phase spectrophotometry; Iron; Water

1. Introduction Green chemistry, defined as the use of chemistry for pollution prevention [1], has been applied mainly in the context of organic and inorganic synthesis, resulting in new synthetic routes, replacement of toxic solvents and minimization of side products [2]. However, several analytical methods currently in use generate large amounts of toxic residues, causing environmental impact. In this sense, the development of greener analytical procedures is widely desirable and this aspect should be taken into account by the analytical chemists [1]. Waste amount and toxicity are parameters as important as any other analytical feature. Several strategies are available to achieve this goal, such as reagent replacement, recycling and waste treatment. However, minimization of both reagent consumption and waste generation is the more general approach. Analytical methodologies can be revisited to drastically minimize waste generation, exploiting automation and miniaturization, for example.

∗

Corresponding author. Fax: +55 11 3815 5579. E-mail address: [email protected] (F.R.P. Rocha).

0039-9140/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2006.07.025

Amounts of iron exist widely in river, tap, pond, well, and underground water and this metallic ion is essential for biological systems [3,4]. There are many spectrophotometric methods for iron determination, but some problems arise when these methods are applied to samples having complex matrices or containing traces of iron, owing to their poor sensitivity and selectivity [5]. Thus, it is important to develop procedures that can be directly applied to these samples and that incorporate aspects of green chemistry. Conventional spectrophotometry, based on measurement of light transmission through solutions of absorbing species, is a widespread instrumental technique whose scope can be expanded by measurements in solid-phase. This approach, named solid-phase spectrophotometry (SPS), is based on absorbance measurements directly on a solid support on which the analyte is retained, thus avoiding tedious preconcentration and separation analytical steps [6,7]. Measurements can be carried out with conventional spectrophotometers, resulting in simple and low cost procedures. SPS also offers the advantage of in situ concentration of the analyte due to its accumulation in a small volume of the solid support, thus resulting in better sensitivity and lower detection limits in comparison to measurements

1508

L.S.G. Teixeira, F.R.P. Rocha / Talanta 71 (2007) 1507–1511

in solution. Selectivity is also improved due to different retention abilities of the analyte and the interfering substances on the solid support [8,9]. Coupling flow injection with SPS (FI-SPS) makes feasible such measurements without attaining (physical or chemical) equilibrium conditions, thus providing faster and more selective procedures in comparison to those implemented in batch. In addition, small amounts of samples and reagents are employed and, in some situations, reagents immobilized on the solid support can be used for the reversible retention of the analyte, resulting on inherently greener procedures [7,8]. In this work, a greener analytical procedure is proposed for sensitive and selective determination of iron in water samples exploring its retention on 1-(2-thiazolylazo)-2-naphthol (TAN)/C18-bonded silica support in a flow injection system. Elution, buffering and reduction steps are performed with nontoxic chemicals in order to drastically minimize waste toxicity. 2. Experimental 2.1. Apparatus The flow system was constructed with a lab-made sliding-bar injector, similar to one previously described [10,11], polyethylene tubes (0.7 mm i.d.) and Perspex joint points. Alternatively, the sliding-bar injector can be replaced by two commercially available six-way injection valves coupled in series. An Ismatec IPC-4 peristaltic pump equipped with Tygon tubes was used for fluid propulsion. An UV–vis spectrophotometer (Varian, Cary 1E) equipped with a lab-made 1-mm optical path cell for FISPS similar to one previously described [7] was employed for signal measurements. This flow cell consisted of two Perspex blocs separated by a rubber strip (1 mm thickness) and attached with four screws. A 10-mm diameter circular hole was made in the rubber strip to support the C18-bonded silica. In previous experiments, it was observed that higher optical paths caused an excessive attenuation of the radiation beam by absorption and scattering. 2.2. Reagents and solutions All solutions were prepared with analytical grade chemicals and freshly distilled-deionized water. Iron reference solutions were prepared in the range 50.0–1000 ␮g L−1 by dilution of a 1000 mg L−1 stock solution prepared after dissolving metal iron in 5 mL concentrated HCl under heating. Sodium acetate/acetic acid, hexamine/hydrochloric acid and tris(hydroxymethyl)aminomethane/maleic acid were evaluated as buffer solutions. Ascorbic acid and hydroxylamine solutions were evaluated as reducing agents. A 0.1 mol L−1 HCl solution was employed as eluent. TAN solution was prepared by dissolving 1.0 mg 1-(2-thiazolylazo)2-naphthol (Merck) in 1.0 mL ethanol and bringing the volume to 100 mL with a 5% (m/v) Triton X-100 solution. Water was used as the carrier stream. Freshwater samples were collected in polyethylene vessels and filtered through 0.45-␮m cellulose acetate membranes (Millipore) before analysis. Sample solutions were prepared contain-

Fig. 1. Flow diagram of the system for iron determination by FI-SPS: (a) sampling position and (b) injection position. I: sliding-bar injector; S: sample; C: carrier stream (H2 O, 1.5 mL min−1 ), E: eluent (0.1 mol L−1 HCl), L1 : 125-cm sample loop (625 ␮L), L2 : 20-cm eluent loop (100 ␮L), FC: lab-made flow cell placed at the spectrophotometer optical path, W: waste vessel.

ing 0.05 mol L−1 acetate buffer pH 5.5 and 0.1% (m/v) ascorbic acid. The blank solution was prepared in freshly distilleddeionized water, containing the same concentrations of buffer and the reducing solutions. 2.3. Flow diagram and procedure The flow-cell was filled with ca. 35 mg C18-bonded silica (60–100 ␮m) obtained from Sep-Pak cartridges (Waters). TAN reagent was immobilized on the solid support by pumping the solution through the flow cell at 0.5 mL min−1 for 10 min. The cell was then sequentially washed with 0.1 mol L−1 HCl and water. In the flow diagram shown in Fig. 1, sample and eluent aliquots are alternately introduced into the carrier stream. In the sampling position (Fig. 1a), the loop L1 (625 ␮L) is filled with buffered sample or reference solutions (S), while the acid eluent aliquot contained in L2 is transported towards the flow cell by the water carrier stream. Metallic ions retained at the solid support are then eluted and the solid-phase is regenerated previously to the sample injection. By sliding the central bar of the injector to the other resting position (Fig. 1b), the sample aliquot (loop L1 ) is inserted into the analytical path and transported by the carrier towards the flow cell. The iron(II)–TAN complex is formed at the solid support, yielding an analytical signal proportional to the metal amount, measured at 780 nm. Simultaneously, the loop L2 is filled with the eluent solution (0.1 mol L−1 HCl). After attaining the signal maximum, the central bar is moved back to the sampling position to start another measurement cycle. The reference method was based on complex formation between iron(II) and 1,10-phenantroline in ammonium acetate buffer after reduction of iron(III) with hydroxylamine [12]. For achieving the required sensitivity, after sample processing according to the reference method, absorbance measurements were carried out in a 100-cm optical path flow cell constructed with a liquid core waveguide (LCW) based on Teflon AF2400® [13]. A similar strategy was adopted by Zhang et al. that employed an LCW with 200 cm optical path for iron determination in a gas segmented flow system [14].

L.S.G. Teixeira, F.R.P. Rocha / Talanta 71 (2007) 1507–1511

3. Results and discussion 3.1. Chemical variables TAN reacts quickly with iron(II) in aqueous solution to form a brown complex with absorption maxima at 575 and 787 nm [15]. The reaction also occurs instantaneously when the reagent is adsorbed on C18 bonded silica and the formed complex yields slightly different absorption maxima −580 and 780 nm. Measurements at 780 nm can be exploited for selectivity improvement, because other metallic ions (e.g., copper, zinc, nickel and cobalt) react with TAN forming complexes with absorption maxima between 550 and 600 nm [8,9]. Iron retention on C18-TAN is reversible – the complex is decomposed in acid media and iron(II) is eluted without removing the chromogenic reagent from the solid support. This characteristic allows the development of a greener analytical assay also incorporating the advantages of using SPS such as improvements in sensitivity and selectivity. Eluent, buffer and reducing solutions were carefully selected in order to minimize waste amount and toxicity. It was observed that both ascorbic acid and hydroxylamine solutions could be used for iron(III) reduction with the same efficiency. Ascorbic acid was selected in view of its very low toxicity. The retention of iron(II) on C18-TAN was studied at various pH values for different buffers (Fig. 2). Acetate buffer was selected taking into account the sensitivity and its low toxicity when compared with other solutions evaluated. As higher signals were obtained at pH 6.3, ammonium acetate buffer can be used in view of the buffer capacity. Different mineral acids can be used for iron elution from the solid support. Hydrochloric acid was selected because it allowed quantitative iron elution without removing the immobilized chromogenic reagent. The effects caused by eluent concentration and volume were evaluated and it was observed that 100 ␮L of 0.1 mol L−1 HCl was sufficient for complete iron elution. Thus,

Fig. 2. Effect of the pH on iron(II) retention. (a) Tris(hydroxymethyl)aminomethane/maleic acid; (b) acetate/acetic acid and (c) hexamine/hydrochloric acid buffer. A: absorbance measured at solid-phase.

1509

the reversible adsorption of the analyte was assured and no significant change on iron retention efficiency was observed even after 100 measurements. After this, the C18 solid support was replaced and a new aliquot of TAN was pumped through the cell for immobilization of the reagent. By considering the TAN amount on the solid support (ca. 9 × 10−7 mol), the estimated reagent consumption is lower than 1 ␮g per determination. A systematic evaluation of the effect of potentially interfering species on iron determination in natural water samples was carried out. This study was performed by adding known amounts of foreign species usually found in water samples to a 500 ␮g L−1 iron solution. The threshold was established ±5% variation in the analytical signal. When the study was carried out at 580 nm, interferences were not observed for the following species up to 50 mg L−1 : Al(III), Ba(II), Be(II), Ca(II), Cr(III), Hg(II), Li(II), Mg(II), Mn(II), Mo(VI), Pb(II), Sb(IV), Sn(IV), Ti(IV), V(IV), W(IV), chloride, bromide, iodide, sulfate, nitrate and carbonate. Positive interferences were observed in the presence of Cd(II), Cu(II), Co(II), Ni(II) and Zn(II) even in the same concentration of the analyte. However, with measurements at 780 nm, selectivity was improved and these metallic ions were tolerated up to 2 mg L−1 . For higher concentrations (unusual in water samples), negative interferences were observed in view of the competition of the cations by the immobilized reagent. 3.2. Hydrodynamic variables The flow system was designed to provide the removal of the analyte from the solid support after each sample measurement to avoid saturation of the adsorbent sites. This was carried out by introducing the sample and the eluent aliquots alternately into an inert carrier stream (see Fig. 1). As the total flow rate can affect both the retention efficiency and sampling rate in FI-SPS, the effect of this parameter was studied over the range 0.4–2.1 mL min−1 (Fig. 3). Higher sensitivities were obtained when lower flow rates were employed. Flow rates >2.1 mL min−1 caused fluid leakage in the joints due to the increase in back-pressure. A flow-rate of 1.5 mL min−1 was

Fig. 3. Effect of the carrier flow rate on the analytical signal. A: absorbance measured at solid-phase.

1510

L.S.G. Teixeira, F.R.P. Rocha / Talanta 71 (2007) 1507–1511

selected as a compromise between sample throughput and sensitivity. An important advantage of solid-phase spectrophotometry is the potential to improve sensitivity by increasing the sample volume from which the analyte is concentrated in the solid support [7,16]. This alternative can be exploited to adjust the sensitivity of the SPS procedure as function of the iron concentration in the samples and also to obtain analytical curves from a single standard solution. In this experiment, a linear correlation between the absorbance signal and the sample volume was observed in the range 200–800 ␮L. On the other hand, by increasing the sample volume, an inherent decrease in the sampling rate was also observed. Thus, a 625-␮L sample volume was chosen as a compromise between sensitivity and sample throughput. 3.3. Analytical features and application In the optimized conditions, linear response was observed between 50 and 1000 ␮g L−1 for a 625-␮L sample volume. The calibration curve can be described by the equation: A = 0.341C + 0.0278, r = 0.997, in which A represents the absorbance measured as peak height and C the iron concentration in mg L−1 . The flow cell filled with TAN-C18 and equilibrated with water (carrier stream) was employed to zero-setting the spectrophotometer in the single beam mode to establish the baseline. Under this condition, low blank values were measured, as described in the calibration curve equation. Apparent molar absorptivity was estimated as 1.9 × 105 L mol−1 cm−1 , which is ca. 17-fold higher than that achieved in the Fe(II)-1,10-phenantroline reference method (1.1 × 104 L mol−1 cm−1 ) [12] and ca. 10-fold higher than that obtained in the analogous batch procedure with measurements in solution [15], in which higher sample and reagents volumes were employed. The assay presented in this paper uses only small volumes of sample and reagents, showing low sample and reagents consumption (see Table 1) with a waste generation of 90 mL h−1 . The detection limit at the 99.7% confidence level was estimated as 15 ␮g L−1 , according to the recommendations of IUPAC [17]. Sampling rate and coefficient of variation (n = 10) Table 1 Reagent consumption per determination in procedures using TAN as chromogenic reagent for iron determination Reagent

TAN (␮g) Ascorbic acid (mg) Hydroxylammonium chloride (mg) Sodium acetate (mg) Glacial acetic acid (␮L) Concentrated hydrochloric acid (37.5%) (␮L)

Consumption Proposed procedure

Conventional spectrophotometry [15]

<1.0 0.6 –

1000 – 100

3.6 0.4

340 147

0.8

–

Table 2 Mean values and standard deviations (n = 3) for total iron determination by the proposed and reference methods [12] Sample

Iron concentration (␮g L−1 ) FI-SPS

Lake water River water Weir water 1 Weir water 2

51 80 21 80

± ± ± ±

2 1 1 1

Reference method 53 92 21 101

± ± ± ±

6 1 3 6

were estimated as 25 measurements per hour and 4.0%, respectively. The FI-SPS procedure was applied to total iron determination in freshwaters from lake, river or weirs. The results shown in Table 2 agreed with those achieved by the modified reference method [12] at the 95% confidence level. 4. Conclusions The developed procedure provides a highly sensitive, selective and simple method for iron determination at ␮g L−1 level. The proposed FI-SPS procedure was suitable for the determination of iron in water samples, yielding good accuracy and precision. The reversible retention of iron(II) on the C18-TAN solid support was explored to develop a greener analytical procedure, with reduced sample and reagent consumption and minimized waste generation. As TAN immobilized on C18 does not react with iron(III), this species can be determined by difference from measurements carried out with and without the reducing solution. This strategy can be then adopted for redox speciation of iron that is often required in environmental studies. Acknowledgements The authors acknowledge the fellowships and financial support from the Brazilian agencies Conselho Nacional de Desenvolvimento Cient´ıfico e Tecnol´ogico (CNPq), Financiadora de Estudos e Projetos (FINEP) and Fundac¸a˜ o de Amparo a` Pesquisa do Estado de S˜ao Paulo (FAPESP). References [1] P.T. Anastas, Crit. Rev. Anal. Chem. 29 (1999) 167. [2] P. Tundo, T.C. Williamson, Green Chemistry: Challenging and Perspectives, Oxford University Press, 1998. [3] S. Ohno, M. Tanaka, N. Teshima, T. Sakai, Anal. Sci. 20 (2004) 171. [4] S. Kawakubo, A. Naito, A. Fujihara, M. Iwatsuki, Anal. Sci. 20 (2004) 1159. [5] P.K. Tarafder, R. Thakur, Microchem. J. 80 (2005) 39. [6] F.R.P. Rocha, L.S.G. Teixeira, Quim. Nova. 27 (2004) 807. [7] L.S.G. Teixeira, F.R.P. Rocha, M. Korn, B.F. Reis, S.L.C. Ferreira, A.C.S. Costa, Anal. Chim. Acta 383 (1999) 309. [8] L.S.G. Teixeira, F.R.P. Rocha, M. Korn, B.F. Reis, S.L.C. Ferreira, A.C.S. Costa, Talanta 51 (2000) 1027. [9] L.S.G. Teixeira, A.C.S. Costa, S. Garrigues, M. de la Guardia, J. Braz. Chem. Soc. 13 (2002) 54. [10] P.B. Martelli, B.F. Reis, M. Korn, I.A. Rufini, J. Braz. Chem. Soc. 8 (1997) 479. [11] F.R.P. Rocha, J.A. N´obrega, Chem. Educator 4 (1999) 179.

L.S.G. Teixeira, F.R.P. Rocha / Talanta 71 (2007) 1507–1511 [12] L.S. Clesceri, A.E. Greenberg, A.D. Eaton, Standard Methods for the Examination of Water and Wastewater, 20th ed., American Public Health Association, Washington, 1998. [13] R.H. Byrne, W. Yao, E. Kaltenbacher, R.D. Waterbury, Talanta 50 (2000) 1307. [14] J.Z. Zhang, C. Kelble, F.J. Millero, Anal. Chim. Acta 438 (2001) 49.

1511

[15] S.L.C. Ferreira, R.M.W. Nano, Talanta 41 (1994) 1937. [16] L.S.G. Teixeira, E.S. Le˜ao, A.F. Dantas, H.L.C. Pinheiro, A.C.S. Costa, J.B. de Andrade, Talanta 64 (2004) 711. [17] M. Thompson, H.M. Bee, R.V. Cheeseman, W.H. Evans, D.W. Lord, B.D. Ripley, R. Wood, J.J. Wilson, Analyst 112 (1987) 199.