ANALYST NOVEMBER 1984 VOL. 109 1375 Instability of Analytical Ligands in Solution Part 1. Hydrolysis Reactions and Interchange Reactions of C=N Groups A Review Maria Dolores Luque de Castro Manuel Silva and Miguel Valcarcel Department of Analytical Chemistry Faculty of Sciences University of Cordoba Cordoba Spain Summary of Contents Introduction Hydro I ysi s reactions Partial hydrolysis Hydrolysis followed by recondensation Complete hydrolysis Erroneous determinations of acid - base constants Determination of metal ions acting as catalysts in hydrolysis reactions Improper use of amines as prior reducers Diacetyl monoxime azomethinic derivatives - hydroxylamine systems 6-Methylpicolinaldehyde azine - amine systems in the presence of metal ions 2-Hydroxybenzaldehyde azine - amine systems Interchange reactions of C=N groups Keywords Review; instability of analytical ligands; hydrolysis reactions; interchange reactions of G N groups Introduction When new analytical methods are established the possibility of the ligand experiencing irreversible changes (total or partial chemical transformations which hinder its action as such) under the working conditions is frequently ignored-we have not found any systematic studies concerning the lability of analytical ligands in solution.In recent monographs devoted to the use of organic compounds in analytical ~hemistryl-~ there is no clear reference made to the instability of ligands in solution even though the behaviour of the ligand with respect to its stability in solution can be of great importance in determining its suitability as an analytical reagent.It has frequently been observed that when the reaction conditions are empirically optimised to favour the formation and/or the stability of a complex serving as the basis of a determinative method conditions favouring the stability of the ligand (as far as possible) and avoiding side-reactions are actually being created. Thus there are many methods in the literature involving the use of a large excess of reagent (up to 100-200-fold over the stoicheiometric amount) which cannot be explained purely on the basis of a low stability constant. Firstly we should like to make it clear that we are not referring to the well known reversible changes in the behaviour of the ligand such as acid - base reactions and prototropic or keto - enolic equilibria but those side-reactions of the ligand in which it may undergo transformations that prevent it from acting as such.This behaviour can be regarded as a negative factor in the general behaviour of unstable ligands although in some instances this instability is a positive factor for their use. The following reactions are the most important arising from the lability of analytical ligands: 1. Hydrolysis reactions. 2. Interchange reactions of C=N groups. 3. Redox reactions. 4. Molecular aggregate formation reactions. 5. Photochemical reactions. This paper discusses the first two types whereas the other three will be discussed in the following paper (p. 1383). In neither of the papers have we attempted to present a complete compilation of all the work presented in the literature but in showing the most significant examples of each type of reaction we have aimed at bringing this common but at the same time so little studied subject to the attention of analytical chemists.Hydrolysis Reactions This type of reaction is especially important for those Iigands resulting from the condensation of carbonyl compounds with amines such as Schiff bases and oximes. Hydrolysis an inverse reaction to condensation is catalysed by H+ and OH-ions: H’ (OH-) RR’CNR’’ + HOH * RR’CO + HZNR” Ligand Carbonyl compound Amine The greater or lesser tendency of the C-N bonds to hydrolyse depends on the nature of the R R’ and R” groups. In spite of being a reaction that takes place in a solution of the majority of Schiff bases (at a pH below 1-2 and above 10-11), on reviewing the literature of these compounds,S no specific references to their lability have been found.Some examples in which hydrolysis conditions favour or hinder the development of analytical methods are described below. Partial Hydrolysis The Fe(I1) - methyl 2-pyridyl ketone azine system is an example undergoing this type of hydrolysis. When the cation and ligand solutions are mixed a soluble and very unstable red ferroin-type (1 3 metal to ligand ratio) chelate is formed. The terminal C=N bond which does not take part in the chelation 1376 ANALYST NOVEMBER 1984 VOL. 109 is quickly hydrolysed giving rise to the stable orange ferroin-type (1 3 metal to ligand ratio) Fe(I1) - hydrazone chelate which is suitable for photometric purposes.6 A molecule of the corresponding ketone results from this reaction as shown below: 2+ I $C-CH3 6 A,,, = 520 nm -2-t + QL0 I CH3 It should be noted that (i) the C=N bonds in the chelate rings are not hydrolysed as this bond type is stabilised by the formation of these rings.(ii) The hydrolysis of the Fe(I1) -hydrazone chelate is very slow and negligible. When this chelate boils for a long time at a pH below 1.5 it is destroyed, forming hydrazine and ketone as the orange colour disappears from the solution. This behaviour is also observed without heating in a concentrated hydrochloric acid medium. (iii) If the original azine ligand employed is dissolved in nitrobenzene and the complex is formed by extraction from an aqueous solution of Fe(II) the red organic solutions of the Fe(I1) -azine are perfectly stable (Amax = 525 nm) because their hydrolytic degradation does not take place in this organic solution.Perchlorate is used as a forming agent for the ion pair making the extraction of the charged chelate possible. One example that differs from the above owing to the complexity of the reactions resulting from hydrolysis is the Pd(I1) - picolinaldehyde azine system. The reason for this different behaviour could be the greater ease with which the Pd(I1) ion forms complexes. Picolinaldehyde azine (PAA) is very easily hydrolysed between pH 0 and 5 yielding the corresponding hydrazone. In the presence of Pd(I1) and at pH 1.7-2.5 the formation of the following complexes takes place: NH2 A,,, = 660 and 730 nm (El A binuclear 2 1 (metal ligand) complex (Amax = 420 nm) (A) is attributable to the azine ligand which is very unstable and is transformed in less than 1 h into a very stable 1 1 complex (Amax = 400 nm) (B) attributable to the hydrazone a 2 2 complex (kmax = 545 nm) (C) attributable to the azine a 2 1 2 Pd - azine - hydrazone mixed ligand complex that corresponds to the 2 3 Pd(I1) to azine ratio initially added (Amax = 520 nm) (D) and shows lower chromaticity than the previous one because of its lower degree of conjugation and a 1 3 Pd(I1) to hydrazone complex (E) which has an intense blue colour with two characteristic maxima at 660 and 730 nm attributable to the hydrazone.The occurrence of this wide variety of complexes has been shown by different spectroscopic and chromatographic tech-niques.7 Other azines such as those of 2-benzoylpyridine8 and that of di-2-pyridyl ketone9 show similar behaviour. This type of hydrolysis is fairly common in compounds bearing the azine group. 10 Hydrolysis Followed by Recondensation When diluted solutions of Fe(1I) and picolinaldehyde azine (PAA) are mixed a red ferroin-type (1 3 metal to ligand ratio) chelate (Amax. = 475-480 nm) is formed [as shown in equation (l)] that without heating is rapidly transformed into the ferroin-type 1 3 Fe(I1) - hydrazone chelate (Amax = 435 nm). This behaviour is similar to that of the first example described under Partial Hydrolysis.The peculiarity of this system is that if an excess of picolinaldehyde is added to the solution of the hydrazone chelate the azine is again formed (by condensation with the terminal amine group) but in this instance a chelate with an intense blue colour and 1 2 metal to ligand stoicheiometry is formed the azine acting as a terdentate ligand bound to two pyridinic nitrogens and to one azinic nitrogen. This behaviour can be accounted for on the basis of the existence of three isomers of the azine according to the position of the pyridinic rings with respect to the two C=N bonds. The trans - trans form (which can have two rotational isomers) is stable both in the solid state and in solution and responsible for the formation of the initial red complex.When 2+ A,,, = 420 nm pH 1-3 Azine + Pd2+ A,,,. = 400 nm (€3) 14+ A,,, = 545 nm (C) -1 A,,, = 520 nm (D ANALYST NOVEMBER 1984 VOL. 109 0- Am i ne p he n o I a te NH2 /DH-\\ // c-c I I 1377 1 .N \c \ I H 2+ r Fe/3 I L L h,,,. = 475-480 nm Amax. 435 nrn . + O C H O the azine is formed again by recondensation the Fe(I1) bound to the hydrazone facilitates the azine formation as a terdentate ligand yielding the blue chelate which is much more stable than the red one. This blue chelate has excellent analytical photometric properties which have allowed the development of a selective method for the determination of trace amounts of Fe(I1) in the 1-8 p.p.m. range." Complete Hydrolysis As an example the well known Ca(I1) - glyoxal bis(2-hydroxyanil) system is used and the hydrolysis mechanism and absorbance profile are shown in Fig.1. This is widely accepted for the photometric determination of trace amounts of calcium in many samples. The difficulty involved in this photometry due to the instability of the coloured solutions is well known; Linstrom and Milligan12 have demonstrated that the loss of absorbance with time can be explained 2s follows. (i) The 1 1 I I H H (GBHA)Ca It H H GBHA G lycolate Ca(C2H303h 1 HO-CH2 -COO Calcium glycolate BSHB(A = 370 nm) 1 2 3 4 Tirne/h Fig. 1. Complete hydrolysis mechanism for the Ca(I1) - glyoxal bis(2-hydroxyanil) system and the corresponding absorption spectra 2+ Excess ____, +o H u A,,, = 660 nrn .chelate has a low stability constant and it dissociates even if an excess of ligand is used. (ii) The C=N bonds hydrolyse very easily yielding 2-aminophenol and glyoxal. This reaction is catalysed by OH- ions and it is necessary to take into account that the reaction only occurs in an alkaline medium. (iii) Glyoxal undergoes an internal regrouping mechanism (one aldehyde group is oxidised and the other is reduced) in a basic medium the glycolate anion being formed as a result. The glycolate anion reacts with the calcium liberated in (i) to form calcium glycolate Ca(C2H303)2 whose solubility depends on the percentage of organic solvent present in the medium. Therefore the hydrolysis of the ligand (not only that resulting from the chelate dissociation but also that in excess) is favoured by the conversion of glyoxal into another ligand, which is bonded to the calcium from the complex.From a comparative point of view it is interesting to mention the different behaviour of another imine ligand, diphenylglyoxal bis(2-hydroxybenzoyl hydrazone) (BSHB) ,I3 which we have suggested as a photometric reagent for the determination of calcium with which it forms an intense yellow chelate of 2 3 metal to ligand stoicheiometry. Both the BSHB (Amax = 370 nm) and the calcium chelate = 435 nm) are perfectly stable for long periods of time. There are two probable reasons for this higher stability in general the complexes of BSHB14-16 are much more stable than those of glyoxal bis(2-hydroxyanil) (GBHA) and the C=N bonds in the hydrazone tend to be hydrolysed to a lesser extent.Erroneous Determinations of Acid - Base Constants Incorrect data on acid - base constants (pK,) concerning organic ligands17-19 are sometimes found in the literature because the hydrolytic degradation of the compound under study at the end of the pH scale has not been taken into account. In these situations one must ensure that the measurements of the analytical signal are free from this disturbing effect. Therefore the stability of the values of the analytical signal must be tested with time for the moment when the samples are prepared if these samples show extreme pH values; if these samples are unstable the value of the analytical signal must be tested with time from the moment started ( t = 0) must be calculated and it is only the ligand that generates this signal that has to be taken into account (this calculation can be carried out by extrapolation of the analytical signal versus time graphs).Fig. 2 shows the absorbance versus pH graph for 6-methyl-picolinaldehyde azine. 19 This compound apparently shows three pK values but two of them are calculated for pHs at which the absorbance is unstable. Therefore each point must be calculated in these zones for t = 0 prior to its measurement. This study is shown in Fig. 2(b) and (c) in which an exponential decrease in absorbance with time corresponding to the extreme zones of pH can be observed. This decrease is more important in the acidic zone owing to the higher rate at which the hydrolytic degradation of this azine occurs.Once these corrections have been made the absorbance versu 1378 ANALYST NOVEMBER 1984 VOL. 109 0.60 b’ 8 p 0.58 -0.56 -n Q 0.54 0.52 m 0, --I I I I 0.096 8 0.094 e 0.092 n C m Z J a 0.090 \ pH = 12.2 0.088 I I I I 1 2 2 4 6 8 Timeimi n Fig. 2. (a) Absorbance - pH curve for 6-methylpicolinaldehyde azine and ( h ) and (c) influence of the hydrolytic cleavage of the C=N group on the determination of acid - base constants of 6-methylpico-linaldehyde azine. ( b ) Acidic hydrolysis; and (c) alkaline hydrolysis pH graph shows a single pK due to the single acid - base equilibrium: H L e L - + H+ It proves to be more advisable to carry out the calculation of the pK of unstable ligands in solution by the flow injection analysis (FIA) technique which is very much faster.20 FIA permits measurements to be taken immediately after reagent -acid (or base) mixing in such a manner that in most situations the hydrolysis process is not developed to any appreciable extent.Determination of Metals Ions Acting as Catalysts in Hydrolysis Reactions This is an example in which the hydrolytic degradation of the ligand proves useful as it is the basis for the determination of metal ions acting as catalysts in the reaction. It is a special catalytic effect because the catalyst does not change its oxidation state during the catalytic cycle,21 but its action involves the weakening of the hydrolysable bonds in the molecule by the formation of complexes or chelates. There is some literature available concerning this type of reaction the most studied examples being those concerning the hydrolysis reaction of amino acids or esters of phosphoric acid deriva-tives.22-24 Amongst the metal ions that exhibit a catalytic effect on these compounds Cu(II) Zn(I1).Mn(I1) and Cd(I1) are the most important. Other species involved in these hydrolysis degradation reactions catalysed by metals are for example thioesters and Schiff bases.ls For instance the hydrolysis of 2-hydroxybenz-aldehyde azine is discussed. This reaction takes place slowly at pH 12 (basic hydrolysis) in which the corresponding hydraz-one and 2-hydroxybenzaldehyde are formed and the fluores-cent properties of the solution change from A, = 420 nm and A, = 520 nm to A, = 355 nm and A,, = 465 nm with a remarkable increase in the fluorescence intensity.If trace amounts of copper(I1) (i.e. [copper] [azine] 2 1 1000) are added to alkaline solutions of 2-hydroxybenz-aldehyde azine the hydrolysis rate is greatly increased because this metal ion forms a chelate with the ligand. As copper(I1) is a charge inductor the nitrogen of the chelate ring bears a residual positive charge which is displaced to the carbon adjoining the C=N group. Because of this effect the hydrolysis reaction cannot be carried out in an acidic medium as the positive polarisation of the C=N bond prevents the electrophilic attack of the hydrogen ion. This polarising effect of the catalyst produces an electrophilic site in the molecule, which is easily attacked by the hydroxide ion.Therefore the hydrolytic reaction rate is increased with respect to the reaction in the absence of the metal ion which permits the development of catalytic - kinetic method for the determina-tion of trace amounts of copper.26 The reaction mechanism is shown in equation (2) on the opposite page. This reaction proceeds to the extent that the concentration of 2-hydroxybenzaldehyde hydrazone formed in the catalysed reaction is higher than the remaining concentration of 2-hydroxybenzaldehyde azine. In this example the chelate between copper(I1) and hydrazone prevents the reaction of this metal ion with the azine and therefore the catalytic cycle is prevented. The catalytic action of copper in this hydrolytic reaction is in disagreement with the study carried out by Bozhevol’nov et al.,*7 who pointed out that the reaction between copper(I1) and 2-hydroxybenzaldehyde azine corresponds to the forma-tion of a fluorescent chelate rather than to a catalytic reaction. From our point of view the fluorescence measurements at 490 nm that have been reported by Bozhevol’nov et al. correspond to a reaction stage in which 2- hydroxybenzaldehyde azine is being partially hydrolysed to 2-hydroxybenzaldehyde hydra-zone. It is interesting to note that when the pH was adjusted with amine bases the reaction rate sharply increased and the catalytic effect of copper(I1) was not detected. The enhance-ment of the reaction rate in the presence of amine bases could be assumed to be due to interchange reactions of C=N groups, according to the following reaction: The analytical scope of these reactions for the kinetic determination of amine bases and for those of the metal ions, due to their effects on these reactions is expanded in the following section.Interchange Reactions of C=N Groups ‘There are some types of reactions that have already been described28 and applied to the synthesis of metal chelates,29 but which we have shown to be a cause of ligand instability, i.e. the transformations that some organic reagents ligands or chelates can undergo in the presence of several different amines. In the interchange reactions of C=N groups an azomethinic ligand [ligand (l)]. with one or several groups of thi5 type (Schiff bases for example) undergoes a change in the presence of an excess of amine which is added to the solution to transform it into another azomethinic ligand [ligand (2)], but with a net change of the group R3 by another group R4 as follows”: RlR2C=NRj + RjNHz -+ RIRzC=NR4 + RiNH2 Ligand 1 Amine 2 Ligand 2 Amine ANALYST NOVEMBER 1984 VOL.109 1379 Lex = 420 nm kern = 520 nm r OH- [Azinel << [Hydrazone] Aex. = 355 nrn kern. = 465 nm I Table I Transformations due to interchange reactions of C=N groups with four different amine 2 compounds (ammonia hydroxylamine, hydrazine and thiosemicarbazide) Ligand 2 Ligand 1 Ammonia Hydroxylamine Hydrazine Thiosemicarbazide Azine . . . . . . Hydrazone Oxime Hydrazone Thiosemicarbazone Hydrazone . . . . Hydrazone Oxime Other hydrazone Thiosemicarbazone Oxime .. . . . Hydrazone - Hydrazone Thiosemicarbazone Thiosemicarbazone . . Hydrazone Oxime Hydrazone -These reactions take place at a moderately high rate in an aqueous medium at a suitable pH provided an excess of amine 2 is added. The transformations that we have verified are given in Table 1. These transformations are very interesting from an ana-lytical point of view in the presence of transition metal ions, which form chelates that have different properties with ligands 1 and 2. The following behaviour must be taken into account when considering these reactions. (i) The metal ion also reacts with amines 1 and 2. Undoubtedly this can complicate the system as these amines can also form complexes (soluble or not) with the cation competing with the main ligands or form mixed-ligand complexes or reduce the cation to an oxidation state which may or not be suitable for chelate formation.(ii) R3 does not take part in the formation of the chelate with the initial azomethinic ligand. Its replacement by R4 can give rise, for exampIe to a change in the chelate solubility a change in the absorption spectra of the chelate-batho- or hypso-chromic shifts depending on whether the conjugation of the bond system increases or decreases and a change in the selectivity as the steric effect may undergo a change either making possible or hindering the tris-chelate formation on transposing R3 by R4. (iii) R3 has some active “sites” in the formation of the first chelate so that its replacement by R4 may involve a cleavage of the chelate ring and therefore a fundamental change.R4 can have or not have acidic or basic groups with differences in activity or position which can either favour or hinder the transformation. Some of our experiments in this area are discussed below. Improper Use of Amines as Prior Reducers Hydroxylamine and hydrazine are frequently used as prior reducers in the photometric determination of some metal ions such as iron(I1) and copper(I) which has allowed us to describe the interchange reactions of C=N groups in the presence of metal ions. Suggi et al.32 reported the photometric determination of iron(I1) with PAA using hydroxylamine as a prior reductant. However we have demonstrated” that they did not work with the original azine ligand which forms an unstable reddish (Amax = 485 nm) ferroin-type [Fe(PAA)3] chelate with iron(II) but that the real ligand was picolinaldehyde oxime, which also produces a yellow 1 3 metal to ligand chelate (Amax.= 420 nm). This is the suggested wavelength for the photometric method. Suggi et al.32 indicated that to achieve the fast stabilisation of the absorbance values it is necessary to heat for 20 min at 40-50 OC with addition of azine in a very large excess above the stoicheiometric amount. All of these experimental data indicate the occurrence of the following interchange reactions of C=N groups: Reducer 3L + Fe2+ - Fe3+ - FeL32+ Red chelate Hvdroxvlamine i 3L + Fez+ - Fe3+ - 3L’ + Fez+ -+ FeL’32+ (or hydrazine) Schiff base 1 Schiff Red base 2 chelate Diacetyl Monoxime Azomethinic Derivatives - Hydroxylamine Systems Diacetyl monoxime hydrazone (DMH) semicarbazone (DMS) and thiosemicarbazone (DMT) undergo respective interchange reactions of their C=N groups upon treatment with an excess of hydroxylamine in an acidic medium, dimethyl glyoxime (DMG) being formed in situ as a result.If palladium(I1) is present in the reaction medium these reac-tions serve as a basis for the homogeneous precipitation of palladium dimethylglyoximate as well as for the subsequent gravimetric determination of micro-amounts of this metal DMT is an exception and although it undergoes inter-change reactions of the C=N groups with an excess of hydroxylamine in acidic medium the formation of very stable soluble complexes between thiosemicarbazide DMT and palladium(I1) hinders the formation of Pd(DMG)*.The kinetic study of these systems has permitted the establishment of reactivity sequences related to the ease of formation of dimethyl glyoxime in solution from the i0n.3 1380 ANALYST NOVEMBER 1984 VOL. 109 Table 2. Influence of some metal ions on the 6-MePAA - amine systems Effect of metal ion Type of transformation* Metal ion pH E,/kJ mol- on reaction rate 6-MePAA into 6-MePAO . . Cu(1) 4.5 33.56 Slight enhancement 6-MePAA into 6-MePAH . . Cu(1) 4.5 52.50 Moderate delay 8.6 31.60 Slight delay 6-MePAA into 6-MePAT . . Cu(1) 4.5 45.39 High delay Ni(I1) 4.5 36.70 Moderate delay Co(I1) 4.5 12.58 Moderate enhancement * 6-MePAA = 6-methylpicolinaldehyde azine; 6-MePAO = 6-methylpicolinaldehyde oxime; 6-MePAH = 6-methylpicolinaldehyde hydrazone; and 6-MePAT = 6methylpicolinaldehyde thiosernicarbazone.aforementioned azomethinic compounds. In addition this study allows a definite mechanism for these reactions to be hypothesised which is similar to that of hydrolysis reactions or Schiff base formation.31 6-Methylpicolinaldehyde Azine - Amine Systems in the Presence of Metal Ions 6-Methylpicolinaldehyde azine (6-MePAA) undergoes inter-change reactions of its C=N groups in the presence of amines such as hydrazine hydroxylamine and thiosemicarbazide , yielding the corresponding azomethinic derivative. The influ-ence of several metal ions on these systems is shown in Table 2. From these observations some interesting conclusions can be drawn the maximum reaction rate is obtained for moderately acidic pHs (about pH 4.5); copper(1) is the most important metal ion because of the presence of the cuproin group in the azomethines involved; the azine - thiosemicarba-zone transformation is special as the azomethine obtained, although bearing the cuproin group does not show selectivity towards copper(1).In the mechanism by which the transformation takes place three steps can be established (1) protonation of the azomethinic compound; (2) nucleophilic attack of the amine molecule on the protonated azomethinic compound followed by an internal rearrangement of a proton; and (3) formation of the second amine followed by deprotonation. The role of copper(1) (the only cation forming complexes with the azine molecule) in this system is specifically to facilitate the formation of the initial carbocation which most readily reacts with the C=N group that participates in the coordination.As can be observed this effect is similar to that indicated for the hydrolytic degradation of the 2-hydroxy-benzaldehyde azine catalysed by copper(I1). In contrast, nickel(I1) and cobalt(I1) ions do not form complexes with the azine but they do form them with the thiosemicarbazone, which accounts for the great rapidity of the azine - thiosemi-carbazone interchange in the presence of these metal ions a behaviour that is more marked in the presence of cobalt(I1) as the pH at which the interchange reaction occurs is the optimum for the formation of the Co(I1) - 6-MePAT com-plex.3435 The different reaction rates with which these chelates are formed by their respective interchange reactions of C=N groups permits the application of differential kinetic analysis methods to the resolution of mixtures of these cations. The interference of EDTA in these systems permits their determi-nation as well as the indirect determination of zinc(I1) and bismuth(II1) on the basis of the different stability of the NiYZ- ZnY2- and BiY - complexes and on the displacement effected by zinc(I1) and bismuth(II1) cations on nickel(I1) complexed with EDTA. 2-Hydroxybenzaldehyde Azine - Amine Systems The use of interchange reactions of C=N groups in the 2-hydroxybenzaldehyde azine (2-OHBAA) - amine systems Table 3. Features of the kinetic - fluorimetric determination of amines based on interchange reactions of C=N groups EJkJ System pH mol-1 Determination rangeh 2-OHBAA - NH3 .. 10.5-10.9 46.98 5 X 10p3-3 X 2-OHBAA - NHZNHZ 10.5-10.8 30.01 2.5 X 10-5-6 X 10-3 2-OHBAA-NHZOH . . 10.6-11.0 27.75 1 X 10-3-1 X 10 - 2 2-OHBAA -NH2NHCSNH2 . . 10.8-11.1 23.41 5 X 10-‘-1 X lo-’ has permitted the direct kinetic - fluorimetric determination of amines.35 As we have indicated above these reactions can be used for the determination of copper(I1) by its catalytic effect on the hydrolysis of 2-hydroxybenzaldehyde azine.26 The action of various amines such as ammonia hydrazine, hydroxylamine and thiosemicarbazide has been studied. Table 3 summarises the most important features of these determi-nations. The different kinetic behaviour of each of these C=N interchanges has permitted the simultaneous determination of hydrazine - hydroxylamine ammonia - hydroxylamine and ammonia - hydrazine mixtures.In this section it is worth mentioning the behaviour of 2-hydroxybenzaldehyde hydraz-one in solution a common product resulting from the interchange reactions of C=N groups in acidic medium. This behaviour can be used for the determination of the acidity constants of this compound as below pH 4.5 fluctuations in the measurement made (fluorescence intensity or absorb-ance36) are observed. The process corresponds to an auto-transformation of the hydrazone resulting from an inter-change reaction of the C=N groups into the corresponding azine as is shown in the following scheme: pH < 4.5 I The mechanism by which this reaction and other similar reactions take place has been reported by Szmaut and McGinnis.37 References Flaschka H.A. and Barnard A. J . “Chelates in Analytical Chemistry,” Volume 2 Marcel Dekker New York 1969. Holzbecher Z . Divis. L . Kral M. Sucha L. and Vlacil F., “Organic Reagents in Inorganic Analysis,” Wiley New York, 1976. Shilt A. A. “Analytical Applications of 1 ,lo-Phenanthroline and Related Compounds,” Pergamon Press Oxford 1969 ANALYST NOVEMBER 1984 VOL. 109 1381 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15, 16. 17. 18. 19. 20. 21. 22. Burger K. “Organic Reagents in Metal Analysis,” Pergamon Press Oxford 1973. Jungreis E . and Thabet S . “Chelates in Analytical Chern-istry.” Volume 2 Marcel Dekker New York 1976 chapter IV.Luque de Castro M. D. and Valcarcel M. Afinidad 1977, 348. 405. Garcia-Vargas M. and Valcarcel M. An. Quim. 1980 76, 471. Luque de Castro M. D. Valcarcel M. and Pino F. An. Quim. 1976 72 382. Martinez M. P. Valcarcel M. and Pino F. Analyst 1975, 100 33. Luque de Castro M. D. Garcia-Vargas M. and Valcarcel, M. Quim. Anal. 1982 1 1. Luque de Castro M. D. and Valcarcel M. Anal. Lett. 1978, 1 1. Linstrom F. and Milligan C. W. Anal. Chem. 1967,39,132. Silva M. and Valcarcel M. Analyst 1980 105 193. Silva M. and Valcarcel M. An. Quim. 1980 76 129. Silva M. and Valcarcel M. Microchem. J. 1980 25 289. Silva M. and Valcarcel M. Microchem. J. 1980 25 117. Green R. W. Hallman P. S . and Lions F.Znorg. Chem., 1964,3 376. Green R. W. Hallman P. S . and Lions F. Znorg. Chem., 1964 3 1541. Valcarcel M. Perez-Bendito D. and Pino F. inf. Quim. Anal. 1971 25 1. RGiiCka J and Hansen E. H. ”Flow Injection Analysis,” Wiley New York 1981. Bontchev P. R. Talanta 1970 17 449. Kroll H. J. Am. Chem. Soc. 1952 74 2036. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Tikhonova L. P. and Yatsimirskii K. B. Zh. Neorg. Khim., 1966 11 2259. Yatsimirskii K. B. and Tikhonova L. P. Zh. Neorg. Khim., 1965 10 2070. Fernando Q. Ad. Inorg. Chem. Radiochem. 1965,7,234. Alarcon R. Silva M. and ValcBrcel M. Anal. Lett. 1982, 15 891. Bozhevol’nov E. A. Kreongol’d S. U. and Sovenkova L. I., Tr. Vses. Nauchno Zssled. Znst. Khim. Reakt. Osobo Chist. Khim. Veshch. 1967 30 176;Anal. Abstr. 1968 15 3806. Zymalkowski F. Editor “Methodicum Chimicum. C-N Compounds,” Volume 6 Academic Press New York 1975. Martin D. F. “Metal Complexes of Ketimine and Aldimine Compounds,” in Jolly W. L. Editor “Preparative Inorganic Reactions,” Volume 1 Wiley-Interscience New York 1964. Valcarcel M. and Pino F. Talanta 1973 20 224. Patai S. “The Chemistry of the Carbon-Nitrogen Double Bond,” Interscience London 1970 p. 81. Suggi A. Dan M. Inone Y. and Nakamura H. Bunseki Kagaku 1965 16 1133. Rios A. and Valcarcel M. Analyst 1982 107 737. Rios A. PhD Thesis University of Cordoba 1983. Rios A. and Valcarcel M. Quim. Anal. 1983 1 227. Alarcon R. Thesis University of Cordoba 1982. Szmaut H. H. and McGinnis C. J. Am. Chem. SOC. 1950, 72,2890. Paper A41100 Received March 15th 1984 Accepted April 16th 198