Extraction-Spectrophotometric Studies on the Complex Formation of Iron(III) with 4-(2-Thiazolylazo)Resorcinol and Tetrazolium Salts

Four liquid-liquid extraction-chromogenic systems containing Fe(III), 4-(2-thiazolylazo)resorcinol (TAR), tetrazolium salt (TZS), water and chloroform were studied. 2,3,5-Triphenyl-2H-tetrazolium chloride (TTC), 3-(4,5-dimethyl-2-thiazol)-2,5diphenyl-2H-tetrazolium bromide (MTT), 3-(2-naphtyl)-2,5-diphenyl-2H-tetrazolium chloride (TV), and 2-(4-iodophenyl)-3(4-nitrophenyl)-5-phenyl-2H-tetrazolium chloride (INT) were the examined TZSs. Optimization experiments for iron extraction were performed and the following parameters were found for each system: pH(opt), CTAR(opt), CTZS(opt), shaking time (opt), and (opt). Under the optimum conditions, the molar ratio of the reacting Fe(III), TAR and TZS is 1:2:2 and the general formula of the extracted species is (TZ + )2[Fe II (TAR 2– )2]. Some equilibrium constants (constants of association, constants of distribution, and constants of extraction) and analytical characteristics (molar absorptivities, Sandell’s sensitivities, Beer’s law limits, etc.) were calculated. Linear relationships involving the molecular mass of TZ +


Procedure for establishing the optimum conditions
Aliquots of Fe(III) solution, TAR solution (up to 2.0 cm 3 ), TS solution (up to 2.0 cm 3 ) and buffer solution (3 cm 3 ; pH ranging from 3.9 to 9.2) were introduced into 100-cm 3 separatory funnels. The resulting solutions were diluted with distilled water to a total volume of 10 cm 3 . Then 10 cm 3 of chloroform were added and the funnels were shaken for a fixed time (up to 4.0 min). A portion of each organic extract was transferred through a paper filter into a cell and the absorbance was read against a blank.

Procedure for determining the constants of distribution
The distribution constants KD were found from the ratio KD = A1/(Ax-A1) where A1 is the absorbance (measured against a blank) after a single extraction under the optimum operating conditions ( Table 2) and Ax is the absorbance obtained after a triple extraction (x=3; TZS=MTT, TV and INT) or double extraction (x=2; TZS=TTC) under the same conditions. The single extraction and the first stage of the double or triple extraction were performed with 10 cm 3 of chloroform. The organic layers were transferred into 25 cm 3 calibrated flasks and the flask for the single extraction was brought to volume with chloroform. The second stage was performed by adding 10 cm 3 (x=2) or 7 mL (x=3) of chloroform to the aqueous phase that remained after the first stage. The shaking time was 2 min. The third stage (x=3) was performed in the same manner (with 7 cm 3 of chloroform). The organic layers (x=2 or 3) were transferred to the flask containing the extract obtained during the first stage. The volume was brought to the mark with chloroform and shaken for homogenization.

The choice of the optimum wavelength
Spectra of the extracted ternary complexes are shown in Figure 1. Since the absorbance of the blank (curves 1', 2', 3' and 4') is high at <550 nm, the maxima of the Fe-TAR-TTC (curve 1) and Fe-TAR-INT (curve 4) complexes at about 495 nm were inappropriate for absorbance measurements. The maxima at about 545-550 nm of the same two complexes were also unsuitable; they are sharp, pH-dependent, and TAR-concentration dependent. The nature of the mentioned maxima can be understood if we take into account the competitive extraction of coloured ion-pairs (TZ + )(TAR -) [40] in the Fe-TAR-TSZ systems. As a result, minima appear in the recorded spectral curves of the complexes (Ai = AFe-TAR-TZS (i) -ATAR-TZS (i)) when i is close to max of (TZ + )(TAR -). In the previous paper [40] we showed that the constants of association   [41,42], V(IV) [43], Ga(III) [44], and Co J u l y 3 1 , 2 0 1 4 [45]; V(IV,V)-TAR-TZS=550 nm [41][42][43], Ga-TAR-TTC=530 nm [44], and Co-TAR-TTC=525 nm [45]} are rather far from this region. These specifics gave us a reason to perform all further absorbance measurements at 615-620 nm (max=615 nm when TZS=TTC; max=620 nm when TZS=MTT, TV or INT).

Effect of pH
The effect of pH on the absorbance of the extracted ternary complexes is shown in Figure 2 (curves 1-4). The widest optimum pH-interval was recorded for the TTC-complex (pH 5.5-6.6; curve 1) and the narrowest optimum pH-interval was recorded for the INT-complex (pH 6.3-6.5; curve 4). Buffer solutions, prepared from acetic acid and ammonium hydroxide, were applied to control pH throughout the work.

Effect of Shaking Time
Extraction equilibrium is reached for about 35-40 seconds. However, in order to avoid accidental errors caused by a combination of short shaking times and different shaking rates, we extracted for 2 min in our experiments J u l y 3 1 , 2 0 1 4

Effect of Reagents Concentration
The effect of TAR and TZS concentrations on the absorbance of the extracted species is illustrated in Figure 3 and Figure  4. The saturation is reached most easily in the extraction system containing TV (Figure 3, curve 3 and Figure 4, curve 3). The optimum reagents concentrations deduced from the Figures 3 and 4 are shown in Table 2.

Composition of the Complexes
The saturation curves presented in Figure 3 and Figure 4 allowed determination of the TAR-to-Fe and TZS-to-Fe molar ratios in the ternary complexes. The mobile equilibrium method [46], which is appropriate to distinguish mononuclear species from polynuclear species, and the straight-line method of Asmus [47] (Table 3) were used. The results led to the conclusion that mononuclear species are formed in all systems; the molar Fe:TAR:TZS ratio in these species is 1:2:2 if CTAR>(6.0-8.0)×10 -5 mol dm -3 and CTZS>(3.0-8.0)×10 -5 mol dm -3 (see Table 3,*). J u l y 3 1 , 2 0 1 4   ; *-At reagent concentrations lower than the specified (references a, b, c, d, e, f, g, and h) the determined squared correlation coefficient values for n=1 were closest to 1.

Suggested Reaction Scheme and General Formula of the Ternary Complexes
The obtained Fe:TAR:TZS molar ratios (1:2:2 at the optimum extraction conditions) and the ability of TZS to form cations (TZ + ) in aqueous medium suppose a complex formation according to one of the following two schemes: iii) "The colour of PAR chelates of … Fe(III) faded gradually. But, the color of PAR chelates of Cr(III), Fe(II), Co(III) … did not change for two hours … The inert nature of d 3 (Cr(III)), diamagnetic d 6 (Fe(II), Co(III)), and d 8 (Ni(II)) is due to the large crystal field activation energies for the substitution reaction and the diamagnetic d 6 system has the largest activation energy among these systems." [5] J u l y 3 1 , 2 0 1 4 iv) "The IR data support that the ligand is coordinated to the metal ions [Fe II , Cu II , Zn II and Cd II ] in a terdentate manner and complexation with 4-(2-thiazolylazo)resorcinol through the resorcinol OH, azo N and thiazole N." [25] v) "Co-ordination of two molecules of a terdentate ligand would lead to an octahedral complex, lacking a free position for attaching hydroxyl groups." [6] It should be added, that our previous investigations on cobalt-containing extraction systems, namely Co(II)-TAR-TTCwater-chloroform [45], Co(II)-PAR-TTC-water-chloroform [48], Co(II)-PAR-INT-water-chloroform [48], and Co(II)-PARnitron-water-chloroform [49] indicated fast and trouble-free oxidation of Co(II) (d 7 ) to Co(III) (d 6 ). Hence, the suggested Fe(III)Fe(II) reduction, which ensures the same diamagnetic d 6 electron configuration, can hardly even be regarded as a surprise.

Equilibrium constants and recovery
The constants of association (), characterising the processes of association, were calculated by two different methods: the mobile equilibrium method [46] and the Holme-Langmihr method [50]. The constants of distribution (KD) characterizing the processes of distribution of the ternary complexes between the aqueous and chloroform phases were determined as described above. It should be mentioned that the triple extraction cycle for the Fe-TAR-TTC system was found inappropriate for calculating KD {KD=A1/(A3-A1); A3A1}. The reason for this anomaly (A3A1) is probably a timedependent transformation of the ternary complex during its dilution with chloroform [42,51]. In fact, the complex deriving from TTC has the lowest association constant (Table 4); hence, it should be the most susceptible to alteration.
The recovery factors (R%) and the constants of extraction (Kex) were determined by the formulae R%=100×KD/(KD+1) and Kex=KD× [52,53], respectively. The results are presented in Table 4. All experiments were performed at room temperature of 22°C and the calculations were carried out at a probability of 95 %.

Beer's Law and Analytical Characteristics
The adherence to Beer's law for each Fe(III)-TAR-TZS system was examined under the optimum extractionspectrophotometric conditions ( Table 2). Calculated molar absorptivities () are listed in Table 5, along with some important characteristics concerning the application of the ternary complexes for extraction-spectrophotometric determination of iron. The couples TZS-TAR ensure high sensitivity of determination. In this criterion, they are better than many reagents used in similar systems:  [55]}. J u l y 3 1 , 2 0 1 4 Table 5. Characteristics concerning the application of the ion-association complexes for extractivespectrophotometric determination of iron(III)

Extraction system
Analytical characteristics

Relationships Involving the Molecular Mass of the Tetrazolium Cation
Continuous investigations on tetrazolium ion-association complexes have revealed that at least two factors noticeably influence the values of : molecular mass (M) [58] (factor I) and the presence of nitro group(s) in the tetrazolium cation [59] (factor II). According to [58], a linear relationship exists between the M and : the higher molecular mass, the higher the association constant. However, when TZ + contains nitro group(s), the values of  are considerably lower [38,[59][60][61] than those expected by the authors of Ref. [58].
The obtained in the present paper results confirm the significance of both factors (Fig. 5, points 1,1' and 2,2'). The full markers (points 1 and points 2) represent the data for the ion-associates of TT + , MTT + and TV + (i.e. TZs + which do not contain -NO2 group) and the hollow markers (point 1' and point 2') represent the data for the ion-associate of INT + . Straight lines can be drawn through the points 1 or 2 with squared correlation coefficients of 1.000 (line 2) or close to 1 (line 1; R 2 =0.9401) (confirmation of factor I) and the points 1' and 2' are situated below these straight lines (confirmation of factor II).
The rest two experimental plots in Fig. 5   3. The key equilibrium constants and analytical characteristics were determined: constants of association , constants of distribution KD, constants of extraction Kex, recovery factors R%, molar absorptivities  (opt) , Sandell's sensitivities, Beer's law upper limits, limits of detection, and limits of quantification. Linear relationships appear to exist between the logarithm of the molecular mass of the tetrazolium cation (Log MTZ+) and some of the constants (Log , Log Kex, and  (opt) ) for the complexes with TZ + which do not contain NO2 groups (i.e. TZ + =TT + , MTT + and TV + ): the higher Log MTZ + , the higher the constant. The poorer extraction-spectrophotometric performance of (INT + )2[Fe(TAR)2] can be explained with a negative influence of the nitro group.