Benzoate ion determination in beverages by using a potentiometric sensor immobilized in a graphite matrix

Analytica Chimica Acta 433 (2001) 281–288

Benzoate ion determination in beverages by using a potentiometric sensor immobilized in a graphite matrix夽 L. Pezza∗ , A.O. Santini, H.R. Pezza, C.B. Melios, V.J.F. Ferreira, A.L.M. Nasser Instituto de Qu mica, UNESP, Caixa Postal 355, CEP 14801-970, Araraquara, SP, Brazil Received 6 September 2000; received in revised form 19 December 2000; accepted 3 January 2001

Abstract The characteristics, performance, and application of an electrode, namely, Pt | Hg | Hg2 (Bzt)2 | graphite, where Bzt stands for benzoate ion, are described. This electrode responds to Bzt with sensitivity of 57.7 ± 1.0 mV/decade over the range 5 × 10−4 –1 × 10−1 mol l−1 at pH 6.0–8.0 with a detection limit of 1.6 × 10−4 mol l−1 . The electrode shows easy construction, fast response time (between 10–30 s), low-cost, and excellent response stability (lifetime > 6 months, in continuous use). The proposed sensor displayed good selectivity for benzoate in the presence of several carboxylate and inorganic anions. It was used to determine benzoate in various beverages by means of the standard additions method. The results obtained by using this electrode compared very favorably with those given by the official AOAC spectrophotometric method and by a HPLC procedure as well. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Benzoate sensitive electrode; Potentiometry; Beverages' analysis.

1. Introduction Benzoic acid (BA) presents low human toxicity and since the early-1920s it has been used to a growing extent throughout the world to protect a variety of foods and beverages against deterioration by microorganisms [1]. Addition of BA as a preservative for beverages within the 100–1000 ppm (parts per million) range has been a fairly common trade practice [1], lower concentrations are often of little inhibitory effect on the growth of molds, yeasts and bacteria [1] whereas higher ones promote quite significant changes concerning smell and taste of most beverages, especially when fruit juices are involved [2]. Thus, an 夽 This paper is dedicated to the memory of Professors Waldemar Saffioti (1922–1999) and Manuel Molina Ortega (1931–1999). ∗ Corresponding author. Fax: +55-16-2227932. E-mail address: [email protected] (L. Pezza).

effective quality assurance and surveillance program for beverages calls for the examination of a large number of samples. Consequently, it requires a fast, simple, low-cost, selective and reliable method that could be used routinely for BA determination. Most of the methods used for the estimation of BA in fruit juices, soft drinks and synthetic syrups include titrimetry, UV-spectrophotometry, gas–liquid chromatography (GLC), thin-layer chromatography (TLC) and high-performance liquid chromatography (HPLC), while the last two procedures are often coupled with UV-detection and quantification devices. Almost all these methods present significant limitations and drawbacks. For example, the general titrimetric method [3,4] has long been used to determine sodium benzoate in aqueous media through its successive extraction, as BA, into several portions of chloroform, these extracts contain many other acidic compounds that interfere [3–5]. Since this method

0003-2670/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 3 - 2 6 7 0 ( 0 1 ) 0 0 7 9 4 - 2

282

L. Pezza et al. / Analytica Chimica Acta 433 (2001) 281–288

has the main disadvantages of being very poorly selective and time-consuming, it is not particularly recommended for BA analyses in beverages [5]. Low selectivity is also noted for methods based on UV-absorption [5–11] as all unsaturated compounds display one or more bands (often broad ones) in that region of the spectrum. The main problem for the analysis of beverages is the interference resulting from sorbic acid (sometimes used in combination with benzoic acid, increasing the antimicrobial activity spectrum) [6–10]. Interference from many other UV-absorbing substances such as vanillin [4,6,7,12], 3,4-dehydrocoumarin [12], phenolic compounds [7], fumaric acid [6] and some ill-defined materials arising from biological sources [7] have been reported. The official AOAC–UV-spectrophotometric method for determining BA in beverages [5] requires 15 consecutive solvent extractions, it is therefore, not suitable for processing a large number of samples. Moreover, consistent results are obtained only in the absence of sorbic acid and vanillin. Thin-layer chromatography has been applied both for detection and quantification of BA in mixtures with other preservatives [13–16]. The reported procedures are generally lengthy and subject to losses due mainly to steam distillation of BA (and other preserving acids), subsequent solvent extractions and clean up steps. Additionally, BA and sorbic acid are poorly separated by applying the available methods [13–16] and, thus, these are best regarded as semi-quantitative tools [15]. Steam distillation and extractions with several solvents are also required before BA derivatization and its subsequent determination by GLC [5,12,17–21]. Several difficulties have been faced, especially in the analysis of fruit juices, extraction of BA from these samples and its determination by GLC, has only been attempted crudely to date. In no case (except for apple juice [5]) has the determination been duplicated by a recognized, reliable method. On the other hand, HPLC, including reversed-phase, ion-exclusion and ion-exchange chromatography are used for the analysis of BA. In these cases, aqueous samples of beverages can often be directly injected in the column without prior derivatization or partitioning into an organic solvent [22–31]. The proposed procedures can also resolve BA from sorbic acid, should both be present in the same sample. Recoveries of added BA to samples ranged from 93 to 101.2%, with overall

average of 98.4% [22–31]. A recent interlaboratory study [32] involving 10 laboratories was conducted for testing the performance of one of the methods [31] concerning the estimation of BA content in a selected sample of orange juice. For 1–10 ppm of BA added to the sample, the recoveries were within 96.1–101% [32]. Although some of the proposed HPLC methods present drawbacks such as troublesome, consecutive extractions [33], co-elution of BA with sorbic acid [35], low BA recoveries concerning the analyses of certain products [33,34,36] and requirement for a previous sample clean up step [32], the reports mentioned above show that HPLC is a powerful technique in the area of beverages' preservative analysis. The major disadvantages are the high cost of both the basic equipment and its subsequent maintenance services. These limitations also apply to GLC, particularly when attached to mass spectrometry, as well as to capillary electrophoresis [37]. So far, there are a few reports on the use of ion-selective potentiometric sensors for the determination of benzoate [38,39]. However, it has been not reported on the determination of benzoate in beverages using potentiometric sensors. For many such electrodes the problems, mainly those concerning interference, reproducibility, stability of the measurements, linear response range and lifetime, were poorly solved. Previous work from this laboratory dealt with the development of mercury(I) carboxylate electrodes and their application to solution equilibrium [40–44] and pharmaceutical analysis [45,46] involving carboxylate bearing compounds. In this work, the preparation of an electrode, namely, Pt | Hg | Hg2 (Bzt)2 | graphite, where Bzt stands for benzoate ion, is described. The investigation of the experimental variables that contribute to the electrode response led to the development of a simple, selective and reliable method for benzoate determination. Studies on beverages, particularly fruit juices and carbonated soft drinks, were carried out to illustrate the feasibility of the proposed method. Furthermore, as both the electrode and the standard potentiometric equipment are low-cost, the developed procedure also allows small laboratories with limited resources to run benzoate analyses for the aforementioned samples.

L. Pezza et al. / Analytica Chimica Acta 433 (2001) 281–288

283

2. Experimental 2.1. Reagents High purity deionized water (resistivity 18.2 M cm) obtained by using a Milli-Q Plus system (Millipore Corporation, Bedford, MA, USA) was used throughout. All solutions were prepared with analytical reagent grade chemicals, obtained from E. Merck, Darmstadt, Germany. Standardizations of carbonate-free sodium hydroxide, perchloric acid, nitric acid and sodium perchlorate solutions were performed as described elsewhere [40,46]. Metallic mercury was purified according to a previously reported procedure [40]. The sodium benzoate stock solution was standardized potentiometrically with standard perchloric acid solution. Mercury(I) benzoate was prepared by mixing, in aqueous solution, the corresponding nitrate with an excess of sodium benzoate. The resulting precipitate was filtered through a sintered glass funnel, washed with deionized water until nitrate free, and then dried in a desiccator, over calcium chloride under reduced pressure, at room temperature, to constant mass. A white powder was obtained as the final product. 2.2. Electrode preparation and conditioning The mercury(I) benzoate indicator electrode was prepared as follows: mercury(I) benzoate (1.4 g) and metallic mercury (ca. 0.2 g) were mixed in an agate mortar and the material was crushed until a homogeneous solid was obtained. Pure powdered graphite (0.7 g) was then added and the crushing process was continued until perfect homogenization was attained. Part of the resulting solid was transferred to a press mold, being compressed at 9 tonnes for about 5 min. The black pellet (1.5 mm thick, 12 mm o.d., and 0.6 g mass) was fixed at one end of a glass tube (12 mm o.d., 80 mm long) with a silicone-rubber glue (``Rhodiastic'', Rhône-Poulenc, France) and allowed to dry for 48 h. Sufficient metallic mercury (ca. 0.6 g) was then introduced into the tube to produce a small pool on the inner pellet surface, electric contact was established through a platinum wire plunged into the mercury pool and a subsequent conductor cable. The resulting electrode is diagrammed in Fig. 1, showing that it is sealed. This feature, coupled with the small

Fig. 1. Mercury(I) benzoate electrode: (A) conductor cable, (B) banana plug, (C) metallic mercury, (D) Pt wire, (E) silicone glue, (F) sensor pellet (graphite | Hg2 (Bzt)2 | Hg).

amount of metallic mercury placed inside the electrode (ca. 0.6 g), stresses that the considered sensor does not offer significant risk to the operator's health and can, thus, be recognized as safe. When not in use, the electrode's pellet was kept immersed in a small volume of 0.100 mol l−1 sodium benzoate solution whose ionic strength (µ) was adjusted to 0.500 mol l−1 with a sodium perchlorate solution. Before carrying out each experiment, the external surface of the aforementioned pellet was polished with an alumina paper (polishing strip 30144-001, Orion Instruments Inc., Cambridge, MA, USA),

284

L. Pezza et al. / Analytica Chimica Acta 433 (2001) 281–288

washed with deionized water and dried with absorbent paper. 2.3. Instruments The electromotive force (emf) values were read to the nearest 0.1 mV with a Metrohm model 692 pH|ion meter (Metrohm Ltd., Herisau, Switzerland). The reference electrode was a Metrohm Ag|AgCl double junction, model 6.0726.100. The pH of aqueous solutions was adjusted and monitored with the aid of a Metrohm pH electrode, model 6.0234.100. A thermostated titration cell (25.0 ± 0.1◦ C) was employed. Volume measurements (±0.001 ml) were performed with a Metrohm model 665 automatic burette. A Cary model 1E spectrophotometer (Varian, Pty. Ltd., Canberra, Australia) with 1 cm matched quartz cells was used for all absorbance measurements. The HPLC system consisted of a Shimadzu model SPD-10A liquid chromatograph (Shimadzu Seisakusho, Kyoto, Japan), equipped with a LC-10 AS pump (Shimadzu), variable UV–VIS detector (model SR-10A, Shimadzu) set at 230 nm, gradient control (Waters, model 680; Waters Chromatography Division, Milford, MA, USA) and a ``Rheodyne'' 20 ␮l injector (Rheodyne Inc., Berkeley, CA, USA). A stainless steel ``Supelcosil LC-18'' analytical column (150 mm × 4.6 mm i.d., Supelco, Bellefonte, PA, USA) with 5 ␮m particle size packing material was used. Before injection the samples were filtered through a Millex unit (Millex-HV, 0.45 ␮m, Millipore). Chromatograms were recorded and peak heights measured with integrator (Waters, model 746 recording integrator). All experiments were performed in a thermostated room, maintained at 25 ± 1◦ C. 2.4. Potentiometric Cell The following cell was used:

where Bzt stands for benzoate ion and x was in the range 10−1 –10−5 mol l−1 . The ionic strength of the cell compartments was kept constant at 0.500 mol l−1 . No flow of chloride ions from the reference electrode into the test solution could be detected during the measurements. The performance of the mercury(I) benzoate electrode was assessed by measuring the emf of the aforementioned cell for 10−1 –10−5 mol l−1 sodium benzoate solutions. These solutions were freshly prepared by serial dilution of a 0.100 mol l−1 stock standard solution with deionized water, at constant pH (7.0±0.1). The emf readings were obtained for solutions under stirring and recorded when they became stable. The data were plotted as observed emf versus benzoate ion concentration. 2.5. Determination of benzoate ion in beverages The analyzed products were purchased from food stores and all were tested prior to the listed expiration date. All samples are first homogenized and degassed ultrasonically by using a Thorton T14 apparatus (Inpec Eletrônica Ltd., Vinhedo, SP, Brazil). Next, oxygen is bubbled into the juices' samples for 20 min (rate: 200 cm3 min−1 ) to avoid the possible interference of added sulphite and/or naturally occurring ascorbic acid. A 25 ml aliquot of a previously treated sample is transferred to a 125 ml separatory funnel, add 1 ml of 4 mol l−1 HNO3 , 25 ml of deionized water and mix well. The resulting solution is extracted twice with 25 ml portions of chloroform (shaking time: at least 1 min for each extraction). The aqueous phase is discarded and the combined chloroform extract is evaporated in a rotary evaporator under reduced pressure, at 40◦ C, just to dryness. The resulting residue is dissolved in 5 ml of 2.50 mol l−1 NaOH. An amount of 5 ml of 2.45 mol l−1 HClO4 are added and the solution pH is adjusted within 6.0–8.0 (preferably to 7.0 ± 0.1) with 10−2 mol l−1 HClO4 . Transfer the resulting

L. Pezza et al. / Analytica Chimica Acta 433 (2001) 281–288

285

Table 1 Potentiometric response characteristics of the mercury(I) benzoate electrodea Slope (mV/decade)b

Intercept, E0 (mV)b

Linear range (mol l−1 )

Detection limit (mol l−1 )

57.7 ± 1

96.9 ± 3

5 × 10−4 –10−1

1.6 × 10−4

a

T = 25.0 ± 0.1◦ C; pH = 7.0 ± 0.1; µ = 0.500 mol l−1 (NaClO4 ). of 20 determinations over a period of 6 months. Number of data points: 20–22. Mean linear correlation coefficient: 0.995±0.004.

b Average

solution quantitatively to a 25 ml volumetric flask using deionized water (pH = 7.0 ± 0.1) for rinsing and volume completion. An aliquot of 20 ml is employed for analysis with the benzoate-sensitive electrode.

3. Results and discussion 3.1. Electrode response Experiments carried out as described in Section 2.4 led to the following linear relationship between the measured emf (E, in mV) and benzoate ion concentration E = E 0 + S p[Bzt] where E0 is the formal cell potential and S represents the Nernst coefficient (59.16 mV/decade, at 25◦ C, for monovalent ions). Potentiometric parameters and other features associated with the mercury(I) benzoate electrode are given in Table 1. The above calibration equation and the slope value (Table 1) show that the electrode provides a near-Nernstian response to the benzoate ion. The sensor response displayed good stability and reproducibility over the tests, the last mentioned feature is illustrated by the standard deviation values shown in Table 1. 3.2. Response time and lifetime of the electrode The response time of the electrode was tested by measuring the time required to achieve a steady state potential (within ±0.5 mV min−1 ), for 10−1 –5 × 10−4 mol l−1 sodium benzoate solutions at pH 7.0 [47]. The electrode yielded steady potentials within 10–15 s at high concentrations (≥5 × 10−3 mol l−1 ) and about 30 s at concentrations near the detection limit. The useful lifetime of the electrode for the

evaluated concentration range is at least 6 months, in continuous use. 3.3. pH effect The influence of pH on the electrode response was tested over the pH range 4.0–9.0 for 1×10−1 , 1×10−2 and 1×10−3 mol l−1 benzoate ion concentrations. The resulting solutions' pH(s) were adjusted with diluted HClO4 or NaOH solutions. For pH values below 6 significant fractions of benzoate ion (pK a = 4.20) changes to the corresponding protonated form which is not detected by the electrode. For pH > 8.0, the hydroxide ion interferes with the electrode's response. The emf values are independent of pH in the range 6–8, this can be taken as the working pH range of the electrode. 3.4. Electrode selectivity The most important characteristic of any ion sensitive sensor is its response to the primary ion in the presence of other ions present in solution, which is expressed in terms of the potentiometric selectivity coefficient. The potentiometric selectivity coefficients for the mercury(I) benzoate electrode (KBzt, M ) were determined, for a number of anions (M), by the matched potential method (MPM) [48–50]. In this method, the selectivity coefficient is defined by the ratio of the activity of the primary ion relative to an interfering ion, when they generate identical potentials in the same reference solution. In the MPM method, both monovalent and divalent ions are treated in the same manner and the valence of the ions does not influence the selectivity coefficient. Furthermore, the MPM can be used with no regard to the electrode slopes being Nernstian or even linear [51]. Mainly for these reasons, it has increased in popularity in the last few years [52]. The determined KBzt, M values are presented in Table 2.

286

L. Pezza et al. / Analytica Chimica Acta 433 (2001) 281–288

Table 2 Selectivity coefficients (KBzt, M ) for various anionsa KBzt, M Formate Acetate Propionate Citrate Lactate Sorbate Vanillin Salicylate Phthalate Dehydroascorbate Chloride Sulfate Perchlorate Nitrate a

1.3 1.8 2.5 4.6 3.8 9.1 2.0 1.3 8.4 No 3.9 1.8 No No

× 10−5 × 10−3 × 10−3 × 10−3 × 10−3 × 10−3 × 10−2 × 10−1 × 10−3 interference × 10−6 interference interference

T = 25 ± 0.1◦ C; pH = 7 ± 0.1.

The results comprised in the aforementioned Table show that the selectivity of the mercury(I) benzoate electrode towards all tested organic acid anions is good. Alkanemonocarboxylic acids (e.g. acetic and propionic acids) are used as food preservatives [13,14], lactic and citric acids are found in many food products [53,54], moreover, the latter is used as an additive in almost all carbonated soft drinks. From the selectivity coefficient values of sorbate (often used in combination with BA) and vanillin (added to synthetic syrups as a flavoring agent), it can be seen that the benzoate electrode is about 110 and 50 times more responsive to benzoate than to sorbate and vanillin, respectively. As previously pointed out, the aforementioned substances seriously interfere with BA analysis by titrimetry [3,4], spectrophotometry [4,6–10] and, in the case of sorbate, by TLC [13–16] as well. Little interference is also presented by phthalate (Table 2). For salicylate, which is less often used as a food preservative [13,14], the associated selectivity coefficient (Table 2) shows that the benzoate electrode can tolerate ca. 7 times higher concentration of salicylate than benzoate without suffering interference. No interference was noted for glucose, sucrose, dehydroascorbate, glycerol, menthol, o-methoxyphenol, methyl-, ethyl- and n-propyl-p-hydroxybenzoate. The mentioned esters of p-hydroxybenzoic acid are also extensively used as food preservatives [11,13,14]. Sulfate has a very low selectivity coefficient

(Table 2), no interference at all is caused by nitrate or perchlorate and they can, therefore, be used as background electrolytes or ionic strength adjusters for benzoate solutions before performing potentiometric measurements. Chloride ion interferes seriously as shown in Table 2. It should be noted that the analytical procedure adopted in this work is based on chloroform extraction of BA from aqueous matrices (pH 1.1) followed by its reversion to the aqueous phase (pH 7 ± 0.1) as benzoate. The chloride content found in the last mentioned aqueous phase (which originates from the analyses of fruit juices, syrups and soft drinks) was always <1 ␮g l−1 , as analyzed by the mercury thiocyanate method [55]. Therefore, the working procedure completely removes chloride interference. Strong reducing agents, such as sulfite and ascorbic acid, convert mercury(I) to elemental mercury at the electrode's surface and seriously affect its response. Previous oxidation of these species, coupled with the subsequent chloroform extraction of BA from the aqueous samples, as described in the analytical procedure (Section 2.5) completely eliminates their interference. Moreover, the oxidation products, i.e. sulfate and dehydroascorbate, do not interfere (see Table 2). 3.5. Analytical application A standard additions method [56,57] was employed for potentiometric BA estimation in some fruit juices, soft drinks and gooseberry syrup by using the presently proposed benzoate-sensitive electrode. The results, along with those obtained by applying the official AOAC–UV method [5] and a HPLC procedure to the same samples, are given in Table 3. The adopted HPLC method was that recommended by Trifiro et al. [27], except for the replacement of the original ``Lichrosorb RP-18'' column [27] by a ``Supelcosil LC-18''. Comparison of the results presented in Table 3 shows very good agreement among the applied methods thereby reflecting the utility of the proposed electrode. In no instance was sorbic acid found in the analyzed samples, as proven by HPLC. The times required for the analyses by spectrophotometry [5], HPLC [27] and potentiometry were about 2 h, 20 min and 35–55 min per sample, respectively. Concern-

L. Pezza et al. / Analytica Chimica Acta 433 (2001) 281–288

287

Table 3 Benzoate ion determination in commercial beverages Sample (manufacturer)

Benzoatea,b found (ppm) Electrode methodc

Grape juice (Maguari) Acerolad juice (Maguari) Guava juice (Maguari) Passionfruit juice (Maguari) Gooseberry syrup (Primor) Guarana soft drinke (Boituva) Guarana soft drinke (Brahma) Guarana soft drinke (Ciomino) Lemon soda/``sprite'' (coca-cola)

688.4 401.3 408.3 412.8 723.8 283.7 311.9 390.4 245.3

± ± ± ± ± ± ± ± ±

AOAC [5]

7.0 4.5 4.0 6.1 8.0 4.0 6.3 5.2 4.2

692.3 405.8 415.0 420.2 715.0 280.1 316.3 395.1 239.8

± ± ± ± ± ± ± ± ±

4.8 3.4 2.9 5.2 4.6 2.7 3.7 4.5 2.8

HPLC [27] 680.1 391.0 411.0 423.2 718.0 286.9 309.1 405.2 237.4

± ± ± ± ± ± ± ± ±

1.0 2.8 3.3 5.2 5.0 2.7 2.8 2.1 3.1

Average ± S.D. of six determinations. Expressed as sodium benzoate. Qualitative label declaration only. c The reproducibility of the measured emf values in samples, analyses is very close to that found for standard benzoate solutions (Table 1). d A typical Brazilian fruit, rich in ascorbic acid. e For the features of these products see [58]. a

b

ing analyses' costs the UV-spectrophotometric, TLC, GLC and HPLC techniques are more expensive than the electrode method, if expenses regarding reagents, solvents and initial investment on good quality standard equipment associated with each of the named techniques are considered. In conclusion, the proposed electrode exhibits long lifetime, good stability, sensitivity, precision and selectivity. It is low-cost, easy to prepare and to use. Its usefulness for benzoate determination in real samples, particularly for some commercial fruit juices and carbonated soft drinks, was demonstrated suggesting its use as a reliable and advantageous alternative to the official [5] as well as to most other previously reported methods in the routine control of benzoate concentration in these samples. The most important characteristics of the electrode developed in this laboratory are superior (especially concerning lifetime, tolerance to perchlorate and nitrate ions, simplicity and Nernstian behavior) as compared with other benzoate electrodes described in [38,39]. Acknowledgements Financial support from FAPESP through grant (Process 97/01953-7) and fellowship (Process 97/12267-7, to A.O.S.), FUNDUNESP (Process 080/98-DFP) and CNPq Foundations (Brazil) is gratefully acknowledged.

References [1] J.R. Chipley, in: A.L. Branen, P.M. Davidson (Eds.), Sodium Benzoate and Benzoic Acid in Antimicrobials in Foods, Marcel Dekker, New York, 1983. [2] D.F. Chichester, F.W. Tanner, in: T. Furia (Ed.), Antimicrobial Food Additives in Handbook of Food Additives, The Chemical Rubber Co., OH, USA, 1968. [3] C.M. Sowbhagya, K. Mayura, K.G. Nair, L.V.L. Sastry, G.S. Siddappa, J. Assoc. Off. Anal. Chem. 46 (1963) 767. [4] E.M. Hoshall, J. Assoc. Off. Anal. Chem. 47 (1964) 68. [5] Official Methods of Analysis of AOAC International, 16th Edition, Vol. II, Arlington, VA, USA, 1995, Chapter 47, pp. 7–10. [6] W.M. Gantenbein, A.B. Karasz, J. Assoc. Off. Anal. Chem. 52 (1969) 738. [7] H. Zonneveld, J. Sci. Food Agric. 26 (1975) 879. [8] T. Gutfinger, R. Ashkenazy, A. Letan, Analyst 101 (1976) 49. [9] J. Lopez, J. Simal, Ann. Bromatol. 34 (1982) 113. [10] L. Almela, J.M. Lopez-Roca, Sci. Aliment. 4 (1984) 37. [11] C. Cantoni, G.M. Longoni, Ind. Aliment. 17 (1978) 25. [12] M.E. Neale, J. Ridlington, J. Assoc. Publ. Anal. 16 (1978) 135. [13] S.J. Pinella, A.D. Falco, G. Schwartzman, J. Assoc. Off. Anal. Chem. 49 (1966) 829. [14] G.H. Tjan, T. Konter, J. Assoc. Off. Anal. Chem. 55 (1972) 1223. [15] N.K. Sabbagh, R.G. Coelho, Y. Yokomizu, Colet. Inst. Tecnol. Aliment. 8 (1977) 19. [16] B. Mandrou, F. Bressole, J. Assoc. Off. Anal. Chem. 63 (1980) 675. [17] E. Fogden, M. Fryer, S. Uroy, J. Assoc. Publ. Anal. 12 (1974) 93. [18] K. Isshiki, S. Tsumura, T. Watanabe, Agric. Biol. Chem. 44 (1980) 1601.

288

L. Pezza et al. / Analytica Chimica Acta 433 (2001) 281–288

[19] A. Bertrand, C. Sarre, Ann. Falsif. Expert. Chem. 71 (1978) 35. [20] J. Lopez, J. Simal, Ann. Bromatol. 34 (1983) 123. [21] R.G. Coelho, D.L. Nelson, L. Jokl, Ciênc. Tecnol. Aliment. 3 (1983) 1. [22] J.J. Nelson, J. Chromatogr. 11 (1973) 28. [23] M.C. Bennet, D. Petrus, J. Food Sci. 42 (1977) 1220. [24] A. Collinge, A. Noirfalise, Anal. Chim. Acta 121 (1980) 337. [25] A. Collinge, A. Noirfalise, Anal. Chim. Acta 132 (1981) 201. [26] J. Carnevale, Food Technol. Aust. 32 (1980) 302. [27] A. Trifiro, D. Bigliardi, R. Bazzarini, S. Gherardi, Ind. Conserve 56 (1981) 22. [28] D.H. Fröhlich, J. High Resol. Chromatogr. 5 (1982) 158. [29] H. Terada, K. Isada, Y. Maruyama, Y. Sakake, J. Hyg. Chem. 29 (1983) 297. [30] S. Donhauser, K. Glas, B. Gruber, Monatsz. Brauwiss 37 (1984) 252. [31] H.S. Lee, L.R. Rouself, J.F. Fischer, J. Food Sci. 51 (1986) 568. [32] H.S. Lee, J. Assoc. Off. Anal. Chem. Int. 78 (1995) 80. [33] M.L. Puttemans, L. Dryon, D.L. Massart, J. Assoc. Off. Anal. Chem. 67 (1984) 880. [34] M.L. Puttemans, L. Dryon, D.L. Massart, J. Assoc. Off. Anal. Chem. 68 (1985) 80. [35] J.F. Fischer, J. Agric. Food Chem. 31 (1983) 66. [36] H.M. Pylypiw Jr., M.J. Grether, J. Chromatogr. A 883 (2000) 299. [37] J.C. Walker, S.E. Zaugg, E.B. Walker, J. Chromatogr. A 781 (1997) 481. [38] J.L.F.C. Lima, M.C.B.S.M. Montenegro, M.G.F. Sales, Mikrochim. Acta 124 (1996) 35, and references therein. [39] X.M. Lin, K. Umezawa, K. Tohda, H. Furata, J.L. Sessler, Y. Umezawa, Anal. Sci. 14 (1998) 99, and references therein.

[40] L. Pezza, M. Molina, M. Moraes, C.B. Melios, J.O. Tognolli, Talanta 43 (1996) 1689. [41] L. Pezza, M. Molina, C.B. Melios, M. Moraes, J.O. Tognolli, Talanta 43 (1996) 1697. [42] M. Moraes, L. Pezza, C.B. Melios, M. Molina, H.R. Pezza, A.C. Villafranca, J.C. Filho, Ecl. Qu m. 21 (1996) 133. [43] L. Pezza, M. Molina, C.B. Melios, M. Moraes, J.O. Tognolli, H.M. Gomes, Int. J. Environ. Anal. Chem. 68 (1997) 295. [44] A.C.V. Cavalheiro, M. Moraes, L. Pezza, Ecl. Qu m. 25 (2000) 123. [45] A.M. Peres, M. Moraes, L. Pezza, H.R. Pezza, C.B. Melios, Microchem. J. 60 (1998) 184. [46] V.J.F. Ferreira, A.C.V. Cavalheiro, E. Fagnani, M. Moraes, L. Pezza, H.R. Pezza, C.B. Melios, Anal. Sci. 15 (1999) 249. [47] R.P. Buck, E. Lindner, Pure Appl. Chem. 66 (1994) 2527. [48] V.P. Gadzekpo, G.D. Christian, Anal. Chim. Acta 164 (1984) 279. [49] G.D. Christian, Analyst 119 (1994) 2309. [50] Y. Umezawa, K. Umezawa, H. Sato, Pure Appl. Chem. 67 (1995) 507. [51] E. Bakker, P. Bühlmann, E. Pretsch, Chem. Rev. 97 (1997) 3083. [52] M. Mazloum, M.K. Amini, I.M. Baltork, Sens. Actuators B63 (2000) 80. [53] L. Restaino, K.K. Komatsu, M.J. Syracuse, J. Food Sci. 47 (1982) 134. [54] R.B. Smittle, R.S. Flowers, J. Food Prot. 45 (1982) 977. [55] W.J. Williams, Handbook of Anion Determination, Butterworths, London, 1979, p. 307. [56] G.D. Christian, J.E. O'Reilly, Instrumental Analysis, Allyn and Bacon, New York, 1986, p. 42. [57] O.M. Bader, J. Chem. Educ. 57 (1980) 703. [58] M. Carlson, R.D. Thompson, J. Assoc. Off. Anal. Chem. Int. 81 (1998) 691.