Determination of traces of iron by thin-layer spectroelectrochemistry

411

Analytica Chimica Acta, 276 (1993) 411-417 Elsevier Science Publishers B.V.. Amsterdam

Determination of traces of iron by thin-layer spectrodectrochemistry Qingii Xie, Weidong Kuang, Lihua Nie and Shouzhuo Yao Department of Chemistry and Chemical Engineering, Hunan University, Chungsha 410082 (China) (Received 8th February 1992; revised manuscript received 16th September 1992)

Abstract The complex of iron with Chrome Azurol S (CAS) was studied using a long path-length thin-layer spectroelectrochemical cell with dual working electrodes. A method for the determination of traces of iron is proposed, based on the variation in the absorbance between the oxidized and reduced state of the complex (AA). AA was proportional to iron concentration over the range O-3 pg ml-r. Compared with the conventional speetrophotometric determination of iron using CAS, the selectivity was improved because the analytical signal here depended on both the spectral and the electrochemical behaviour of the tested species. Iron was determined in water samples by this method. A concept characterizing the sensitivity of the spectroelectrochemieal signals is also presented. Keywords: W-Visible

spectrophotometry;

Voltammetry; Iron; Spectroelectrochemistry;

Spectroelectrochemistry (SEC) has been widely used in the investigation of the kinetics and mechanisms of electrode reactions and the measurement of electrochemical parameters [1,2]. However, there are fewer reports on quantitative analysis, probably because the traditional optically transparent electrode (OTR) has low optical sensitivity. Tyson and West [3-51 developed a long path-length spectroelectrochemical cell with an optical sensitivity equivalent to that in conventional spectrophotometry. They also initiated quantitative analysis by spectroelectrochemistry based on the W absorption of the complexes formed between the investigated metal ions in solution and the OH- ions on the surface of a platinum electrode. However, this method showed poor selectivity because of the strong overlapping of the spectra of these complexes and was not used for analyses of practical samples. Correspondence to: Shouzhuo Yao, Department of Chemistry and Chemical Engineering, Hunan University, Changsha 410082 (China). 0003-2670/93/$06.00

Waters

In this paper, a spectroelectroanalytical method is proposed and its selectivity is discussed in detail. Traces of iron were determined based on this method using Chrome Azurol S (CAS). A concept demonstrating the sensitivity of the spectroelectrochemical signals is also presented.

THEORY

For an electrode reaction 0 + ne + R, a variation in absorbance (AA) will be recorded for a long path-length thin-layer spectroelectrochemical cell immediately with a single potential step from the potential at which the oxidized state is the only species present to that at which the reduced state is the only species present. According to the Lambert-Beer law, A A = A, -A,

= E&Z

- E~C,Z = ( lR -

lo)Cl (1)

where e. and ln are the molar absorptivity of the oxidized and reduced species, respectively,

0 1993 - Elsevier Science Publishers B.V. All rights reserved

Q. Xie et al. /And.

412

Co and C, the concentration of the electroactive species which only exist in the oxidized and the reduced form, respectively, after a potential-step perturbation, C is the total concentration investigated and I is the light path length. For an electrode reaction R + 0 + rre, we can similarly obtain AA = A, -A,

= l,C,l

- E&I

= ( lo - c&Cl (2)

It is obvious that AA varies linearly to C according to these equations. If AA is taken as the analytical signal, an analytical method can be established. It can also be found that the analytical sensitivity can be enhanced only when the absorptivity of the product, reduced (as in Rqn. 1) or oxidized (as in Eqn. 2) species by electrochemical methods, is greater than that of the reactant being determined, otherwise the analytical sensitivity will be decreased. Hence it is easy to adjust the analytical sensitivity for the spectroelectroanalytical method presented here. The selectivity of traditional polarography can be improved by adjusting the potential so as to avoid side electrode reactions. If two ions have polarographic waves that differ by 100 mV, they are considered not to interfere with each other. Also, spectrophotometry presents a certain selectivity because the tested species usually has its particular spectrum and selective chromogenic reagents have been widely used. However, the analytical selectivity may be further improved by combining the two methods, i.e., in a spectroelectroanalytical method, because only those species that not only have the same spectra but also show the same electrochemical behaviour as the tested species will interfere, otherwise interference signals do not exist or are not recorded. In our view, this method can be applied for quantitative analysis for metal ions, organic and biological species etc. This method is also useful for the determination of the oxidized and reduced species in samples. The analytical signal here can be expressed as follows:

AA = l ,Col + eRCR1 - E,CI = (e. -

l,)Col

for an electrode reaction 0 + ne + R and

(4) for an electrode reaction R * 0 + ne, where Co and CR are the concentration of the oxidized and reduced species existing in the sample respectively. &rowing lo, En, 1 and C, Co and CR can easily be determined. In this work, based on the color development reaction between Fe3+ and CAS [6] and the electrode reaction of this complex, traces of iron were determined. As is well known, the spectroelectrochemical signal from a long path-length cell is more sensitive than that from an OTE. However, it seems that there is no quantitative demonstration of this sensitivity. Hence an attempt is made to propose a concept to characterize the, sensitivity of the spectroelectrochemical signal (S,,,), i.e., the molar absorbance/potential ratio. It is defined as follows:

AA AI,-A,., SsEc= CUE = C(E,,_E,,,) A.U.lmol-‘mV_’

(5) where AI,,, E,,and Ao.I, E,., arethe absorbance and the potential when [O]/[R] = 10 and 0.1, respectively, C is the total concentration of the tested species, A A = A,,-A,, and AE = E,, E 0.1. According to the following equation [ll:

2.303RT E=E"+ =EO'+

nF 2.303RT nF

where A, and AR are the absorbance when the oxidized or reduced state is the only species present, respectively, and A is the absorbance when the oxidized and the reduced state co-exist. Hence we have

E,,=E"' + (3)

Chh. Acta 276 (1993) 411-417

2.303RT 2.303RT log 10 = E"'+ nF nF

(7)

Q. Xie et aL/Anul. Chim. Acta 276 (1993) 411-417

2.303RT

E 0_1= E” +

nF

log 0.1 = E”’ -

413

2.303 RT nF

(8) 40 -A, = 10 A0 -40 Ao.1-AC A0 --A,.1

=

(9)

0.1

Then we have SsEc= =

Fig. 1. Vertical section of long path-length thin-layer spectroelectrochemical cell with dual working electrodes. 1~ PTFE plates; 2 = thin polyethylene spacers; 3 - hole for the counter electrode; 4 = hole for the reference electrode; 5 - dual work* ing electrodes; 6 = thin-layer solution; 7 = solution inlet; 8 = light windows.

1.776 X 10-4nFel RT

1.776 ;;-4nF

( Ao;A”)

EXPERIMENTAL

xA.U. 1 mol-’ mV_’

(11) Hence SsEc is related to the constant of the spectroelectrochemical cell and the optical absorption behaviour of the tested species. However, it does not depend on the tested concentration. Molar absorbance/potential ratios obtained from the present experiments and the literature are given in Table 1. They varied with different species and different cells. However, if Fe(CN)i-/ Fe(CN)i- is taken as a reference system, the enhancement of the spectroelectrochemical sensitivity of the long path-length thin-layer ceh is demonstrated by this concept, as their S, values are much larger than that of an OTE cell.

Fabrication of the spectroelectrochemicacalcell The spectroelectrochemical cell was fabricated

as described [7,8]with some improvements, and is shown in Fig. 1. The cell body was made from PTFE. The dimensions were 1.1 X 1.1 X 3 cm, i.e., nearly the same as for a l-cm cell in conventional spectrophotometry. The light path Iength was ca. 0.7 cm and the width of the thin-layer solution was ca. 250 pm measured by an optical microscope. The tested solution was injected into the thin-layer space confined by the dual working electrodes and two light windows. The cell was positioned and fixed in the optical path of the spectrophotometer and the light beam was passed

TABLE 1 Molar absorbance/potential ratios Celltype *

Optical path length (mm)

A

9.0 9.0 9.0 16.2 0.24 0.145

B

Species tested

Wavelength monitored (nm)

Ssnc (AU. I mol-’ mV-t)

Ref.

Methylene Blue Cu(NH,)f-/Cu(NH,): FeKN$ /Fe(CN# Fe(Cw,-/FdcN):o-Tolidine

420 666 600 420 420 438

6.34 f 0.01 594 lt4 0.416 f 0.001 10.8 0.400 9.03

7 7 7 8 9 10

8 A.- long path-length thin-layer cells; B = normal OTE cells.

Q. Xzk et aL /Anal. Chim. Acta 276 (1993) 411-417

414

Factory) or Hitachi 557 double-wavelength/ double-beam spectrophotometer and a CSH-2 double-electrode potentiostat (Sanming Electronic Factory). The cyclic voltammetric (CV) experiment was carried out with an XJP-821 (B) polarograph (Jiangshu Electroanalytical Instrument Factory). All potentials were measured with respect to an Ag/AgCl, saturated KCl reference electrode and read out from a PHS3C ion analyser (Shanghai Leici Analytical Instrument Factory) which was precalibrated using a standard electrical cell. A platinum wire was used as the counter electrode and glassy carbon was used as the dual working electrode. All chemicals were of at least analytical-reagent grade. All solutions were prepared with doubly distilled water. The experiments were done at room temperature.

0.6

700 Xfnm) Fig. 2. Thin-layer spectra at different potentials for 1 pg ml-’ Fe3+ +OS mM CAS in 0.02 M acetic acid-O.1 M sodium acetate buffer. Potential (mV): (a) 350, 450; (b) 180, (c) 130; (d) 105; (e) 90; (0 65; (g) 40; (h) - 150, -200. 500

550

650

600

Procedures The tested solutions were injected into the spectroelectrochemical cells carefully in order to avoid gas bubbles. The cell was positioned in the optical path. Each of the electrodes was connected to the corresponding pole of the potentiostat. The wavelength and the potential step amplitude were set and optical signals were recorded in situ.

over the dual working electrode surfaces at graxing incidence. Inm-umentation and reagents Spectroelectrochemical experiments were carried out with a Model 751G UV-visible spectrophotometer (Shanghai Analytical Instruments

, 1.0

0.8

.

.

0.6

0.4

1

0.2

0

-0.2

-0.4

-0.6

-0.8

-1.0

VOLTS Fig. 3. Thin-layer cyclic voltammetric trace for 0.2 mM Fe3++ 1 mM CAS in 0.02 M acetic acid-O.1 M sodium acetate buffer. Scan rate, 100 mV s-l.

Q. Xii et al. /Ad.

415

Chim. Acta 276 (1993) 411-417

The results from the thin-layer CV experiment are shown in Fig. 3. In spite of the distortion due to the iR drop of the response from the thin-layer solution, two pairs of potential peaks can easily be found that are located at 0.1, -0.12 V and -0.27, -0.43 V. Considering the conclusions drawn from spectroelectrochemistry, it is readily concluded that the former pair of peaks is due to the oxidation and reduction of the complex of iron with CAS and the latter corresponds to the oxidation and reduction of CAS co-existing in the solution.

RESULTS AND DISCUSSION

Investigation by CV and SEC

Thin-layer absorption spectra at different potentials are shown in Fig. 2. The absorption maximum of the complex of Fe3+ with CAS is 575 nm, while the complex of Fe2+ with CAS does not absorb in the visible region, which agrees with the reported data [ll]. Moreover, the absorption maximum of CAS alone was measured to be ca. 452 nm, which agrees well with the published value [6]. According to Eqn. 6, we can obtain E”’ and n from the intercept and slope of the E vs. lo$(A -A,)/(A, -A)] plot after measuring E, A,, A, and A at 575 nm. Results of three determinations were E”’ = 91, 92 and 91 mV and n = 0.88, 0.93 and 0.89. The results of thin-layer derivative cyclic voltabsorptiometry, as shown in Fig. 4, show that two pairs of peaks were observed at 575 mn but only one pair of peaks at 480 nm. As is known, the complex of Fe3+ with CAS hardly absorbs at 480 nm but CAS absorbs more strongly at 480 nm than at 575 nm. Hence the peaks at 140 and 40 mV are due to the oxidation and reduction of the complex of iron with CAS, respectively, and these two peaks disappear at 480 nm; the peaks at - 350 and - 380 mV are due to the oxidation and reduction of CAS, respectively, and they become stronger at 480 nm than at 575 nm.

Potential step amplitude in the process of analySk

Based the spectroelectrochemical properties of the complex of iron with CAS, AA between 300 and -200 mV was taken as the analytical signal here. The time to reach a steady absorbance after a single potential step is about 5 min when the width of thin layer is about 300 pm. Because of the dual working electrodes and therefore a large electrode surface area/solution volume ratio, it is certain that faster exhaustive electrolysis will be achieved with the present cell than that with only one working electrode, and this was verified experimentally. Effects of the pH and the amount of CAS

pH is an important factor affecting the colour development reaction. In this work, in order to

A b

a

500

400

300

200

loo

0

-100

-200

-300

-400

-500

M illi volts Fig. 4. Thin-layer derivative cyclic voltabsorptiometric trace for 3.3 pg ml-’ Fe3++ 0.5 mM CAS in 0.02 M acetic acid-O.1 M sodium acetate buffer. Observation wavelength: (a) 575 and (b) 480 nm. Potential scan rate, 1.5 mV s-l.

Q. Xii et d /Anal. Chim. Acta 276 (1993) 411-417

416 4

AA t aA 0.6.

0.6 9

Q4.

0.2

(--* * 3

i

5

6

7pl-

Fig. 5. Effect of pH for 1 pg ml-’ Fe3++0.5 acetic acid-sodium acetate buffer.

0

mM CAS in

0.15

030

0.45

0.60

CCAS(mM)

Fig. 6. Effect of the amount of CAS for 1 pg ml-r Fe3+ plus different amounts of CAS in 0.02 M acetic acid-O.1 M sodium acetate buffer.

investigate the acidity effect and find the optimum concentration of acetic acid and sodium acetate, several supporting electrolytes of various pH were prepared from these components. Then AA values when the tested solutions contained 1 pg ml-l iron, 0.5 mM CAS and one of these supporting electrolytes were measured. The results showed that the AA vs. pH plot (Fig. 5) had two parts. In the first, at pH < 4.8, AA increased abruptly with increasing PH. In the second, AA had a steady value with increasing pH from 4.8 to 7.0, but experiments proved that a precipitate formed in the solution at pH > 6.0. These results agree with those in [HI. A buffer solution of 0.02 M acetic acid-O.1 M sodium acetate (pH 5.3) was chosen in this work, as in [ll]. The amount of chromogenic reagent is also an important factor in spectrophotometry. Investigation of the effect of the concentration of CAS (Fig. 6) showed that AA increased rapidly with increasing concentration of CAS and eventually reached a steady value at CAS concentrations

> 0.15 m&I. The concentration of CAS was selected as 0.5 mM in subsequent work. Calibration, reproducibility and inte~emces The linear calibration graph of A A versus iron concentration, CFe (pg ml-‘), is described by the equation AA = 0.013 + 0.51OC,( r = 0.998, n = 8)

(12)

where AA was measured in the test solution under the chosen conditions. The relative standard deviation for ten complete repetitive determinations of 1 pg ml-l of iron was 1.0%. The effects of other ions on the determination of iron were investigated. No significant interferences in the determination of 1 pg ml-’ of iron were caused by the following anions and cations (in pg ml-‘): Zn*+ @OO), Ca*+ (5001, Ni2+ WXIO), Cd*+ (lOO), Pb*+ (600), Mg*+ (lOOO), Be*+ (l), A13+ (l), Cu*+ (20), NH: (lOOO),S,O;-

TABLE 2 Results of determination of iron in water samples (I.cg ml-l) * Sample Tap water Spring water from Yuelu mountain Synthetic sample

Iron concentration

Found by expt.

Added

Found

Recovery (%I

0.266 + 0.018 b

0.264,0.252,0.271

1.00

0.972

97.2

0.338 f 0.013 b

0.336,0.347,0.343

1.00

0.963

96.3

1.20

1.23, 1.17, 1.18

1.00

0.990

99.0

’ Average of three determinations. b Obtained by 1,l~phenanthrohne spectrophotometry, mean f standard deviation (n = 3).

417

Q. Xie et al. /Anal. Chim. Acta 276 (1993) 411-417

(lOO), F- (80) and PO:- (10). As in conventional spectrophotometry, A13+ (1 pg ml-‘), Be2+ (1 pg ml-‘) and Cu2+ (20 lg ml-r) interfere strongly to the determination of 1 pg ml-’ of iron using CAS, so the present technique is more selective, as discussed earlier. Detemination of iron in practical samples A test solution was prepared by taking different water samples and then adding a supporting electrolyte and CAS. AA was then measured and the concentration of iron in samples was calculated using Eqn. 2. Iron was also determined by the l,lO-phenanthroline spectrophotometric method [12] with 4-cm cells. The results of the two methods were in good agreement (Table 2). It can be concluded that the proposed method can be applied for the determination of iron in natural waters and tap water. The authors record their thanks for financial assistance from the National Education Connnission and the Natural Science Funds of the People’s Republic of China in support of this work.

REFERENCES 1 T. Kuwana and N. Winograd, in A.J. Bard (Ed.), Electroanalytical Chemistry, Vol. 7, Dekker, New York, 1974, p. 1. 2 W.R. Heineman, F.M. Hawkridge and H.N. Blount in A.J. Bard (Ed.), Electroanalytical Chemistry, Vol. 13, Dekker, New York, 1983, p. 1. 3 J.F. Tyson and T.S. West, Talanta, 26 (1979) 117. 4 J.F. Tyson and T.S. West, Talanta, 27 (1980) 335. 5 J.F. Tyson, Talanta, 33 (1986) 51. 6 F.J. Langmyhr and KS. Klausen, Anal. Chii. Acta, 29 (1963) 149. 7 Q.-J. Xie, Thesis for Master’s Degree, Hunan University, Changsha, 1990. 8 J. Zak, M.D. Porter and T. Kuwana, Anal. Chem., 55 (1983) 2219. 9 S.-H. Song, G.-J. Chong and S.-J. Dong, Fengxi Huaxue, 15 (1987) 461. 10 D.A. Scherson, S. Sarangapani and F.L. Urbach, Anal. Chem., 57 (1985) 1501. 11 Education and Research Institute of Analytical Chemistry, Department of Chemistry, Huangzhou University, Handbook of Analytical Chemisty, Vol. 3, Publishing House of Chemical Industry, Beijing, 1983, p. 480. 12 Z. Marczenko, Spectrophotometric Determination of Elements, Horwood, Chichester, 1976, Chap. 27.