In an earlier report' on complexes of iron(III) with 2,3_dihydroxypyridine in various media which were all nominally 1 M in hydrogen ion, phosphoric acid was found to produce anomalous effects. Further investigations of such solutions are reported here.
In phosphoric acid media, iron(III) and 2,3_dihydroxypyridine reacted to give a blue-violet solution, unlike the clear blue produced in other mineral acids; further acidification caused fading of the colour without any significant change toward blue. With the other mineral acids, it was practical to prepare all solutions in the desired final acid concentration as solvent, but the low solubility of iron(II1) and the reagent in phosphoric acid made it necessary to use perchloric acid also. Stock solutions of iron(II1) perchlorate and of 2,3-dihydroxypyridine were prepared so that the fmal solutions for measurements were 0.2 M in phosphoric acid and 0.4 M in perchloric acid. The preparation of samples containing the two reactants at total molarities of 9.824 -lo-3 and 5.044 -10B3, the preparation of similar Series lacking one or the other reactant, the measurements of spectra with the .Cary 14 spectrophotometer, and the treatment of the digitized data by the method of Varga and Veatch*, wereall carried out as before'. The computer piogram RANK indicated the presence of one. absorbing complex (estimated absorbance error, 0.010) and the program CONTOUR displayeda pattern supporting this result. However, the contour map indicated a mole ratio of Fe: 2,3-dihydroxypyridine = 2 :3 in the complex (Fig. ). Replotting the mapped data as conventional Job curves at four selected wavelengths showed no peculiarities, and the agreement in shape of Job curves at the two working concentrations suggested that the apparent 2:3 single species was unlikely to be a mixture.
A conventional mole-ratio experiment, with up to 200 moles of reagent per mole of iron(II1) being added, gave a very rounded curve (Fig. ). Since it was impractical to increase the reagent:metal ratio further, in order to establish the maximal absorbance obtainable from a given amount of iron(III), the data already obtained were replotted to estimate the maximal sensitivity of the system; the value found, by extrapolation to an infinite excess of reagent, was 0.0259 pg of iron. From this value, by calculation from the iron(II1) concentration in the mole-ratio