An evolution 300 UV-Vis spectrophotometer with 1.0 cm matched cells fitted with vision pro software of Thermo Electronic Corporation (Cambridge, U.K.) was used for electronic spectral measurements. pH measurements were made with Jenway 3040 ion analyzer-pH meter, equipped with Jenway 924,005 combined glass electrode.

All chemicals were of analytical reagent grade. Deionized water was used to prepare all solutions. The tartrazine dye (E102) was supplied from Alfa Aesar (Germany). Stock standard solution containing 2.672 g・L⁻¹ of the dye was prepared in deionized water. The working solutions were prepared daily by appropriate dilution. A 5 × 10⁻³ mol・L⁻¹ of copper nitrate was prepared by dissolving the required amount in deionized water and standardized complexmetrically with EDTA [25] . The ionic strength of the solutions was maintained at a constant value of I = 0.1 mol・L⁻¹ (NaClO₄). Acetic acid (0.2 mol・L⁻¹) sodium acetate (0.2 mol・L⁻¹) buffer (pH = 5.9) was prepared in 100 ml volumetric flask. Potassium permanganate (Merck, Germany) 1 × 10⁻² mol・L⁻¹ was prepared in deionized water.

Powdered gelatin samples (Lemon, pineapple, green apple, apricot, banana and peach) and powdered drinks (apple, orange and tamarind) were bought from local supermarket in Assiut city (Egypt). These samples contained sugars, fumaric and citric acids, beef gelatin, artificial flavour, colour (E102), tricalcium phosphate and aspartame. Each gelatin or drink samples were weighed exactly (about 1 g) and dissolved in 50 ml of deionized water. Then each solution was centrifuged (10 min, 3500 rpm) to remove the insoluble particles. The filtrate was collected in 100 ml volumetric flask and diluted to the mark with deionized water. For determination of Tz in the above real samples, 0.2 ml of sample solution was used and analyzed according to method I. Finally, the Tz content in food samples were determined using the calibration equation and standard addition calibration curves procedure.

Samples were prepared by mixing 0.5344 mg of Tz (method I) or 0.267 mg (in case of method II) with various amounts of common matrix cations, anions, sugar, gelatin, aspartame and dyes such as sunset yellow, and allura red. The procedure was continued as described under general procedures.

The decomposition of stability constant into its enthalpy change (DH), Gibbs free energy change (DG) and entropic (DS) contributions is of fundamental importance to understand the various factors that may influence coordination [26] [27] . Enthalpy changes due to coordination reactions can always be obtained from the determination of stability constants at different temperatures according to the Van’t Hoff [28] relationship (1, 2):

where R is ideal gas constant (8.314 JK⁻¹・mol⁻¹), K_i is the stability constant of the complex and T is the absolute temperature in Kelvin (K).

The absorption spectra of tartrazine salt (HL³⁻) solution (5 × 10⁻⁵ mo・L⁻¹) in aqueous medium at I = 0.1 mol・L⁻¹ (NaClO₄), 25˚C were recorded at various pH values (Figure 1). The food additive tartrazine salt (HL³⁻) yields four acid―base forms in solution of pH 0.86 - 12.4, H₅L⁺, H₂L²⁻, HL³⁻ and L⁴⁻ exhibiting the absorption maxima at 440, 430, 420 and 400 nm. The protonated acid form H₅L⁺ at azo-nitrogen (l_max = 440 nm) predominates in strongly acidic medium (pH < 3). The solution spectra of food colorant display a symmetrical and highly intense band (l_max = 430 nm) at pH 3.2 - 4.3 corresponding to the protonated form of the dye salt (H₂L²⁻). On decreasing, the acidity of the medium (4.2 - 6.5), a blue shift of the latter band at 420 nm is obtained resulting from dissociation of protonated nitrogen azo group (HL³⁻). The electronic spectra of Tz in aqueous medium within pH 7.2 - 10.4 exhibit a broad band at 400 nm with a shoulder at 425 nm corresponding to tautomeric equilibria of monoanionic form of tartrazine salt (L⁴⁻), Scheme 1. The absorbance versus pH graph were interpreted [29] assuming that a particular equilibrium established under selected conditions. Under our experimental condition pKa₁ [H₂L⁻²/HL⁻³ (-HN=N-)] = 5.15 ± 0.10, pKa₂ [HL⁻³/L⁻⁴ (OH)] = 9.25 ± 0.16.

Tartrazine likes Sudan dyes possess two reducible groups, a nitrogen-nitrogen double bond and a phenol group [30] . A reduction process, which corresponds to a spontaneous auto-reduction of the Cu²⁺ into Cu⁺, was observed in cases of complexation of Cu²⁺ ions with sudan(I) [31] and sudan(II) [32] . Because of these features of sudan dyes, tartrazine is considered as chromogenic chelating and redox sensitive agent reacting with copper(II) to form Cu⁺-Tz complex under optimum conditions. On the basis of redox sensitivity in the Cu(II)-tartrazine

Scheme 1. Protonation process of tartrazine.

interaction in aqueous medium at I = 0.1 mol・L⁻¹ (NaClO₄), pH = 5.9, 30˚C, solution spectrum shows an absorption band with a maximum at 332 nm corresponding to Cu(I)-Tz complex (MLCT) (Figure 2(a)).

Figure 2(b) shows the absorption spectrum obtained for alkaline KMnO₄ solution (l_max = 530 nm). With addition of tartrazine solution, an absorption spectrum with two absorption bands at 390, 410 (double headed) and 610 nm was obtained. The latter band is attributed to the formation of manganate ion as a result of the oxidation of tartrazine with KMnO₄ in alkaline medium [33] [34] and the highly intense double headed band is related to Tz oxidized form.

The complexation equilibria of Cu²⁺ with Tz were investigated at I = 0.1 mol・L⁻¹ (NaClO₄) in aqueous medium over the pH range 3.5 - 10 at 293, 303, 313 and 323 K. The solution spectra were recorded in presence of an excess of the metal ion and in equimolar solutions. The absorption spectra for both metal/Tz ratios at various pH are analogous at the same temperature and exhibit an absorption band at 332 nm.

The absorbance versus pH graphs for the above solutions at different temperatures show the range of complex

formation (3.0 - 6.5) and existence of one complex equilibria within the pH range studied (Figure 3(a)). All graphs exhibit a similar descending branch above pH 6.5 which is due to the hydrolysis of the complexed Tz. The absorbance versus pH graphs for copper-Tz system at different temperatures were interpreted using relations derived earlier by Sommer et al. [35] , Idriss et al. [36] [37] and Saleh et al. [34] . The following Equations (3, 4) were valid for equimolar and solutions with an excess of metal ion respectively:

All symbols have their usual meanings. The logarithmic transformation are linear with a slope of q = 2, indicating the release of two protons during complexation and an intercept including K^*. By considering the values of the dissociation constants of Tz under our experimental conditions, one can assume the following complexation equilibria with Cu(II) in the pH range 3.0 - 6.5.

The mean equilibrium constant K^* was determined at different temperatures by considering equilibrium A. the stability constant of Cu(I) complex is related to the equilibrium constant K^* by the expression. The calculated values of log K₁ for the complex at 293, 303, 313 and 323 K are given in Table 1. The proposed binding model of Tz-copper complex is shown below.

The proposed binding model of Tz with copper(I).

According to Van’t Hoff equation the values of overall thermodynamic parameters DG, DH and DS accompanying complex formation have been determined (Figure 3(b)). The DH and DS values can then be considered as a sum of three contributions, Cu(II)-Tz redox reaction, release of H₂O molecules from octahedral copper(II) ions and metal-ligand bond formation. From Table 1, the -ve value of DG for the complexation process Cu(I)-Tz suggests a spontaneous nature of such process [38] . The +ve value of DH means that this reaction is endothermic favourable at high temperature. The +ve value of DS for the complexation process confirm that the Cu(I)-Tz complex formation is entropically favourable [39] .

The complexation method was successfully applied to the determination of Tz in two different commercial food products (powered drink and powdered gelatin samples). To investigate the applicability of the proposed method, recovery experiments were performed using multiple points standard addition method. For this purpose, a known amount of Tz was spiked to the formulated preparations and the total amount of the dye was estimated. The results are summarized in Table 5 and Table 6. The t- and F-tests indicated no significant differences between the calculated and theoretical values of both the proposed and the reported methods [24] [44] at 95% confidence level, which indicate good precision and accuracy in the analysis of investigated Tz in commercial food products.