CHAPTER 4 METAL-LIGAND COMPLEXES

Transcription

1 CHAPTER 4 METAL-LIGAND COMPLEXES

2 SECTION-1 DETERMINATION OF BINARY STABILITY CONSTANTS Introduction Stability constants are well known tools for solution chemists, biochemists and chemists in general to help determine the properties of metal-ligand reactions in water and biological systems. They have important medicinal implication to measure the metal ligand selectivity in terms of relative strength of metal-ligand bonds 1. Metal coordination complexes have been extensively used in clinical applications as enzyme inhibitors 2, anti-bacterial 3,4, antiviral 5-7 and as anti-cancer 8-1 drugs. Transition metal ion chelate complexes are also exploited by industry in the large-scale purification of amino acids and a wide range of drug and drug precursors containing an amino carboxylic acid moiety 11,12. The formation of metal complexes can be represented by Equilibrium 4.1. pm + ql + rh MpLqHr The stability of complexes can be quantified by using equilibrium constant as expressed in Eq The equilibrium constant is also called as stability constant or formation constant. βpqr = [MpLqHr]/[M] p [L] q [H] r

3 The concentration of metal ion depends on the stability constant of the complex and free concentration of the ligand, which is dependent upon corresponding pk and values. Very low stability constant values mean that the metal-ligand complex is not only soluble in water but readily dissociates into the metal ion and the ligand at physiological. Consequently these metal ions are available for absorption from the digestive tract and allow life to be sustained in the case of metals that are nutrients, and harm life or terminate life if the metal is toxin or promote tissue injury in the case of biologically absorbable complexes. Even though metal-ligand systems can be quite complex, the stability constant of equal molar metal-ligand complexes decides the availability of the ionic metal in aqueous solution, particularly in the acidic to neutral range applicable to biological species. Investigations of acido-basic equilibria of proline and valine and their interaction with metal ions in media of varying ionic strength, temperature and dielectric constant throw light on the mechanism of enzyme catalyzed reactions. Although it is known that the polarity is lower in some biochemical micro environments, such as active sites of enzymes and side chains of proteins than that of the bulk, a direct measurement of the dielectric constant is not possible. Comparison of formation constants of metal complexes with those at biological centers offers a way to estimate the effective dielectric constant or equivalent solution dielectric constant for the active site cavity. This 115

4 brought a renaissance in the study of complex equilibria in aquaorganic mixtures. Extinctometric and electrometric techniques have been extensively employed for the evaluation of formation constants of mononuclear metal ion complexes formed in solution by the interaction of aqua-metal cation and the ligand. In extinctometric (spectrophotometric) method, the absorbance of the colored complex is monitored at a specified wavelength where the metal ion and ligands have little extinction coefficients. But the interpretation of data is rather difficult for systems containing two or more species with overlapping spectral profiles. Since most of the ligands used are conjugate bases of weak acids, liberation of protons is involved in the complex reaction resulting in change in. Hence an appropriate constant ionic strength and buffer are employed in the spectrophotometric investigation and the constants thus obtained are only the conditional (apparent) stability constants. However, in electrometric methods this change in the proton activity or one of the free reacting components is used as probe in monitoring the complexation using an appropriate ion selective electrode. Bjerrum showed that glass electrode is especially useful for the measurement of hydrogen ion concentration. Though it is difficult to apply metric technique with high accuracy for strong acids 16,17 (pka<2) or strong bases (pkb>12) the method gained wide popularity. Bjerrum evaluated the stepwise equilibrium constants by measuring the of a series of solutions containing a definite amount of metal 116

5 salt and ligand but different concentrations of strong alkali. Calvin and Wilson 18 simplified the procedure by titrating a mixture of strong mineral acid, metal salts and ligand with a standard alkali. Later Calvin and Melchior 19 investigated the systems in the absence of initial strong acid. L-Proline (2-pyrrolidine carboxylic acid) is an imino acid containing pyrrole type nitrogen rather than the amino nitrogen of the amino acids 2. It is a bidentate ligand with a high affinity to metals like Co(II), Ni(II) and Cu(II). Speciation in aqueous solutions containing proline and transition metal ions has been the subject of intense study because of their biological importance Several literature reports have been devoted to the study of stability constants of metal-proline complexes. Durrani 24 carried out metric studies on the complex formation of proline with Mn(II), Fe(II) and Ni(II). Morzyk and Zelichowicz 25 studied Co(II), Ni(II) and Cu(II) complexes potentiometrically. Ni(II) complexes were studied potentiometrically by Ammar et al. 26. Stability constants of Ni(II) and Cu(II) were studied by Tanaka and Tabata 27. Aliyu and Naaliya 28 during the course of the potentiometric studies reported the formation of proline complexes of Cr(II), Mn(II), Fe(II), Co(II), Ni(II), Cu(II) and Zn(II). El-Gaber et al. 29 studied binary and ternary complexes of M(II) potentiometrically. A thermodynamic study of Ce(III) and Y(III) complexes was carried out potentiometrically by Sekhon and Chopra 3. Sanaie et al. 31 studied Cu(II) complexes potentiometrically 117

6 and evaluated stability constants using the computer program CHEMEQ. Paper electrophoretic technique was developed by Tewari 32 for the study of Be(II) and Co(II) complexes. metric study of Co(II), Ni(II), Cu(II), Cd(II) and Zn(II) complexes was carried out and thermodynamic parameters were established by Park et al. 33. Batch equilibrium method with cation exchange resin was employed to determine the stability constants of Co(II) by Hirano and Koyanagi 34. Cu(II) and Pb(II) complexes were studied potentiometrically by Shoukry et al. 35 and evaluated with MINIQUAD75 by El-Sherif et al. 36. Hallman et al. 37 used SCOGS computer program to calculate the stability constants of Cu(II) and Zn(II) complexes. Sanaie et al. 39 calculated stability constants of Cu(II) using CHEMEQ in aqueous and methanol-water systems. They also studied the effect of temperature and solvent on the stability of complexes. The equilibrium studies involving various metals and proline were not exhaustive and many of these studies were performed in aqueous medium. Further, there are some differences in the nature of the reported species. There are several instances where different authors reported different species for the same system under comparable conditions. Probably this may be attributed to the nature of probes, the errors associated with the probing techniques or the inadequacies associated with the modeling strategies. Most of the authors reported + species, some reported 2 species and very few 118

7 reported 3 -. But none of them reported protonated species for any of the metal. These literature reports are recorded in Table 4.1. L-Valine is a non-polar, essential and branched chain amino acid. It acts as a bidentate ligand and has wide applications in the field of pharmaceutical and food industry 4. Even though proline has the greatest avidity for metal ions, almost all the amino acids exert the same order of preference for various metal ions. Hence speciation studies of valine with transition metal ions have been an active field of research 41,42. Valine complexes of Mn(II) were studied by Kumar et al. 43 polarographically and they evaluated electrode kinetic parameters also. Ammar et al. 26 used MINIQUAD75 to calculate stability constants of Ni(II) complexes. Aliyu and Naaliya 28 used ORIGIN5 for Fe(II), Ni(II), Co(II), Zn(II), Cu(II), Cr(II) and Mn(II) complexes. Ce(III) and Y(III) complexes were studied by Sekhon and Chopra 3 potentiometrically and they determined thermodynamic parameters also. Sanaie et al. 31 studied Cu(II) complexes potentiometrically and evaluated stability constants using CHEMEQ. Co(II) complexes were studied by Hirano and Koyanagi 34 using batch equilibrium method. Cu(II) and Pb(II) complexes were studied potentiometrically by Shoukry et al. 35 and El-sherif et al. 36 evaluated with MINIQUAD75. Rengaraj et al. 44 studied binary and ternary complexes of Co(II), Ni(II) and Zn(II) complexes metrically. Co(II) and Ni(II) complexes were studied potentiometrically and evaluated with SUPERQUAD by Khatoon and Din 45. The literature survey indicates that valine 119

8 complexes were not extensively studied. Most of the studies were performed in aqueous medium. + and 2 Species were common for most of the systems. Some of these data are recorded in Table 4.2. The earlier literature reports and the distribution patterns of different forms of proline and valine in PG- and AN-water mixtures reported in Chapter 3 indicate a high probability for the formation of various protonated metal complexes of different stoichiometries. Hence, the author has investigated the complex equilibria of proline and valine with Co(II), Ni(II) and Cu(II) in PG- and AN-water mixtures. This type of study throws light on 1) the effect of electrostatic and non-electrostatic interactions (denaturation) on complex formation in vitro, 2) the nature of active site cavities of enzymes, 3) detection of specific solute-solvent interactions, and 4) understanding the selectivity and sensitivity of various protonated complexes and relative abundance of free components, 5) the type of complex formed by the metal ion, and 6) the bonding behavior of the protein residues with the metal ion. The presence of PG or AN in aqueous solution considerably decreases the dielectric constant of the medium and these solutions are expected to mimic the physiological conditions where the concept of the equivalent solution dielectric constant 46 for protein cavities is applicable. The species refined and their relative concentrations under the present experimental conditions represent the possible forms of metal ions in the biological fluids. 12

9 Metal ion Table 4.1: Stability constants of complexes of proline with metal ions reported in literature. log β mlh Ionic strength mol dm NaClO 4 Medium 121 Instru mental technique Calculation method T C Ref. Mn(II) 3.81 Aqueous metry Irving Rossotti technique Fe(II) 4.92,,,,,,,,,, 24 Ni(II) 3.66,,,,,,,,,, 24 Mn(II) 5.5,, 5%,,,, Ethanol,, 24 Fe(II) 8.61,,,,,,,,,, 24 Ni(II) 7.22,,,,,,,,,, 24 Mn(II) 3.1,, 1%,,,, Ethanol,, 24 Fe(II) 2.98,,,,,,,,,, 24 Ni(II) 3.8,,,,,,,,,, 24 Co(II) Aqueous Potentio Bjerrum KNO 3 metry method Ni(II) 5.6,,,,,,,,,, 25 Cu(II) 7.6,,,,,,,,,, 25 Ni(II) NaNO 3,,,, MINIQUAD Ni(II),, --- Mechanistic Consideration,, 27 Cu(II) ,,,, ---,,,, 27 Mn(II) ,, Potentio ORIGIN5 metry Fe(II) ,,,,,, Ni(II) ,,,,,, Co(II) ,,,,,, Zn(II) ,,,,,, Cu(II) ,,,,,, Cr(II) ,,,,,, Co(II),,,, Irving Rossotti KCl technique Ni(II) 6.25,,,,,,,,,, 29 Cu(II) 9.,,,,,,,,,, 29 Ce(III) 6..1 KCl,,,,,,,, 3 Ce(III) 6.8,,,,,,,, 35 3 Y(III) ,,,,,,,, 25 3 Y(III) ,,,,,,,, 35 3 Cu(II) KNO 3,,,, CHEMEQ Cu(II) ,,,,,,,, Cu(II) ,,,,,,,, Cu(II) ,,,,,,,, Cu(II) ,,,,,,,, 6 31 Be(II),, PET Graphical method Co(II) ,,,,,,,, 32 Cd(II).1,, metry Least squares 3.52 NaClO 4 analysis Co(II) 3.49,,,,,,,,,, 33 Ni(II) 3.62,,,,,,,,,, 33 Cu(II) 3.51,,,,,,,,,, 33 Zn(II) 3.3,,,,,,,,,, 33 Co(II) ,, BEM Cu(II) NaNO 3,, Potentio metry MINIQUAD

10 Pd(II) 11.16,,,,,,,,,, 36 Cu(II) KNO 3,, metry SCOGS Zn(II) ,,,,,,,,,, 37 Cd(II),, Potentio n method metry 2 38 Co(II) 9.3,,,,,,,,,, 38 Ni(II) 11.3,,,,,,,,,, 38 Cu(II) 16.8,,,,,,,,,, 38 Zn(II) 1.2,,,,,,,,,, 38 Fe(II) 8.3,,,,,,,,,, 38 Mn(II) 5.5,,,,,,,,,, 38 Cu(II) KNO 3,,,, CHEMEQ Cu(II) ,,,,,,,, Cu(II) ,,,,,,,, Cu(II) ,,,,,,,, 6 39 Cu(II) ,, 2%,,,, Cu(II) ,, Methanol 4% Methanol,,,, PET = Paper electrophoretic technique; BEM = Batch equilibrium method,,

11 Table 4.2: Stability constants of complexes of valine with metal ions in aqueous medium reported in literature. Metal ion log β mlh Ionic strength mol dm -3 BEM = Batch equilibrium method Instrumental technique Calculation method T C Ref. Ni(II) NaNO 3 Potentiometry MINIQUAD Fe(II) ,, ORIGIN Ni(II) ,,,, Co(II) ,,,, Zn(II) ,,,, Cu(II) ,,,, Cr(II) ,,,, Mn(II) ,,,, Ce(III),, Irving KCl Rossotti 25 3 technique Ce(III) 5.22,,,,,, 35 3 Y(III) ,,,,,, 25 3 Y(III) ,,,,,, 35 3 Cu(II) KNO 3,, CHEMEQ Cu(II) ,,,,,, Cu(II) ,,,,,, Cu(II) ,,,,,, Cu(II) ,,,,,, 6 31 Co(II) BEM Cu(II) NaNO 3 Potentiometry MINIQUAD Pd(II) 1.33,,,,,,,, 36 Cu(II) KNO 3 metry SCOGS Zn(II) ,,,,,,,, 37 Co(II) Potentiometry n method 2 38 Cu(II) 15.1,,,,,, 2 38 Fe(II) 6.8,,,,,,,, 38 Mn(II) KCl Polarography Ni(II) KNO 3 metry Bjerrum method 3 44 Cu(II) ,,,,,,,, 44 Zn(II) ,,,,,,,, 44 Co(II) KNO 3 Potentiometry SUPERQUAD Ni(II) ,,,,,,,,

12 FORMATION OF SPECIES It was observed from the data in Tables 4.1 and 4.2 that, there were several instances where different species were reported for the same chemical system by different researchers. The probes utilized and the computational procedures adopted by them were sometimes the same and sometimes different. In order to rationalize the contradictory models reported by different authors, greater details are required regarding the 1) ingredient concentration, 2) method of pruning primary data for the refinement, 3) range and the number of points in each sub range and 4) proportion of the points corresponding to each species in the model. Further, different species proposed for the same metal-ligand system are many a time judged on the best fit criteria. It is very difficult to say a final word about it, as it is unequivocally proved that different algorithms (or same algorithm with different weighting schemes) produce different species. In many chemical systems it was observed that, the predicted species based on the range and the form of the ligand were in fact not detected. This unexpected behavior could be attributed to the i) hard and soft acid base theory of the metal ion and ligand, ii) coordination number of the metal ion, and iii) stabilities of the complexes formed. If the ligand contains one or more donor atoms than those participating in complexation, there is possibility for its participation in acido-basic equilibria, thus resulting in the formation of protonated 124

13 species, lhh in addition to unprotonated l type of complexes. With progressively increasing, the protonated species lose protons to form unprotonated complexes and if the complex is not broken, it may undergo hydrolysis resulting in l(oh)k. Finally, at high the precipitation of the metal ion as hydroxide will takes place. Typical alkalimetric titrations curves are given in Figs. 4.1 and 4.2 from which the stability constants of binary metal-ligand complexes were determined. Since the ratio of TL to TM is greater than unity in all the experiments, as clear from the data in Table 2.2 of Chapter 2, formation of polynuclear complexes is not considered. The formation of polynuclear species is possible if TL : TM 2. Also dimeric and trimeric species are ruled out since the experimental conditions are not favorable for their formation. Hydroxylated species are not considered in the present study because the titrations are discontinued if precipitation occurs. 125

14 12 (A) 12 (D) 9 a b c 9 a b c Vol. of Alkali (cm 3 ) Vol. of Alkali (cm 3 ) 12 (B) 12 (E) 9 a b c 9 a b c Vol. of Alakli (cm 3 ) Vol. of Alkali (cm 3 ) 12 (C) 12 (F) 9 a b c 9 a b c Vol. of Alkali (cm 3 ) Vol. of Alkali (cm 3 ) Fig. 4.1: Alkalimetric titration curves for proline complexes of (A) Co(II), (B) Ni(II) and (C) Cu(II) in 2% v/v PG-water mixture and valine complexes of (D) Co(II), (E) Ni(II) and (F) Cu(II) in 2% v/v PG-water mixture; (a).25, (b).38 and (c).5 mmol. 126

15 12 (A) 12 (D) a b c a b c Vol. of Alkali (cm 3 ) Vol. of Alkali (cm 3 ) 12 (B) 12 (E) 9 a b c 9 a b c Vol. of Alkali (cm 3 ) Vol. of Alkali (cm 3 ) 12 (C) 12 (F) 9 a b c 9 a b c Vol. of Alkali (cm 3 ) Vol. of Alkali (cm 3 ) Fig. 4.2: Alkalimetric titration curves for proline complexes of (A) Co(II), (B) Ni(II) and (C) Cu(II) in 2% v/v AN-water mixture and valine complexes of (D) Co(II), (E) Ni(II) and (F) Cu(II) in 2% v/v AN-water mixture; (a).25, (b).38 and (c).5 mmol. 127

16 The formation function ( n the number of moles of ligand bound per mole of metal ion) is a good parameter for the detection of protonated species. If there are no protonated species the plots of n versus pl at all the ligand concentrations, with fixed metal ion concentration should overlap and any deviation indicates the presence of the protonated species. Some typical plots given in Figs. 4.3 and 4.4 show spread in some regions, indicating the presence of protonated metal complexes. Although the stability constants calculated from the formation function data are vitiated in presence of protonated, hydroxylated or polymeric species, still a wealth of information is hidden in n versus or pl plots which can have many applications. For example, PSEUDOPLOT 47 technique has become very popular in detecting species other than those invoked in the model and even in rejecting a species refined by a computer program. Detection of systematic errors in the concentrations of mineral acid and ligand and non-linear response of glass electrode in extreme ranges etc. is another interesting application of this auxiliary function. Figs. 4.5 and 4.6 give the relation between the number of moles of alkali consumed per mole of metal ion (a) and the. The number of protons released in the equilibrium can be guessed depending on the moles of alkali consumed. 128

17 2 (A) 2 (D) n 1 n 1 12 pl 12 pl (B) 2 (E) n 1. n pl 12 pl 2 (C) 2 (F) n 1 n 1 12 pl 12 Fig. 4.3: Formation curves of proline complexes of (A) Co(II), (B) Ni(II) and (C) Cu(II) and valine complexes of (D) Co(II), (E) Ni(II) and (F) Cu(II) in 3% v/v PG-water mixture; ( ).25, ( ).38 and ( ).5 mmol. pl 129

18 2 (A) 2 (D) n 1 n 1 12 pl 12 pl 2 (B) 2 (E) n n pl 12 pl 2 (C) 2 (F) n 1 n 1 12 pl 12 pl Fig. 4.4: Formation curves of proline complexes of (A) Co(II), (B) Ni(II) and (C) Cu(II) and valine complexes of (D) Co(II) (E) Ni(II) (F) Cu(II) in 2% v/v AN-water mixture; ( ).25, ( ).38, and ( ).5 mmol. 13

19 (A) 8 4 (D) a -4 a (B) 8 (E) 4 4 a -4 a (C) 8 (F) 4 4 a a Fig. 4.5: Number of moles of alkali versus curves of proline complexes of (A) Co(II), (B) Ni(II) and (C) Cu(II) and valine complexes of (D) Co(II), (E) Ni(II) and (F) Cu(II) in 4% v/v PG-water mixture; ( ).25, ( ).38, and ( ).5 mmol. 131

20 8 (A) 8 (D) 4 4 a a (B) 8 (E) 4 4 a a (C) 8 (F) 4 4 a a Fig. 4.6: Number of moles of alkali versus curves of proline complexes of (A) Co(II), (B) Ni(II) and (C) Cu(II) and valine complexes of (D) Co(II), (E) Ni(II) and (F) Cu(II) in 2% v/v AN-water mixture; ( ).25, ( ).38, and ( ).5 mmol. 132

21 Now-a-days knowledge based or expert systems are gaining importance in chemical research, as they mimic the human expert in typical tasks. Some of the thumb rules embedded in computer assisted modeling studies are documented in the form of IF-THEN- ELSE rules in Table 4.3. Selection of species Based on the earlier reports the maximum number of ligands bound to metal ion is restricted to three. In all the cases the maximum n values observed were less than three, with no ascending slope 48. The n values greater than three or ascending slope indicates the presence of a complex species with higher ligand number. All the possible expected species are represented in Table 4.4 for consideration in developing different models that are to be tested. 133

22 Table 4.3: Heuristics in the detection of metal complexes by metry Protonated/hydroxylated species IF A single experiment with any ratio of TL and TM is performed, and n exp n > TL for more than 5 consecutive cal points, THEN IF Species other than n may be present. Experiments with more than one TL to TM ratio are performed and TL/TM>1, and n versus (or pl) curves do not coincide at least in some regions and formation curves are not equidistant(parallel) THEN Protonated or hydroxylated species may be present. IF TL/TM=1, and n exp n deviate near n =, cal THEN Hydroxylated species may be formed. Polynuclear species IF THEN IF THEN TL/TM=1 in more than one concentrations and formation curves are equidistant Dimers and trimers may be formed. TL/TM 2, and formation curves are not overlapping for different concentrations, Polynuclear complexes may be present. 134

23 Table 4.4: Some of the possible binary metal complex species in M(II)- proline (L) and M(II)-valine(X) systems. Constraints: 1) Maximum number of metal ions =1 2) Maximum ligand number = 3 Ligand form Ligand number 1(a) 2(b) 3(c) H2 2+, 2H + 3H3 2+, 3H2 +, 3H XH2 + MXH 2+ MX2H2 2+, MX2H + MX3H3 2+, MX3H2 +, MX3H XH MX + MX2 MX3 - Ligand No. of Ligands Number of species (n) Exhaustive No. of models n n C r r = 2 proline 2 5(a+b ) (a+b+c ) 52 valine 2 5(a+b ) (a+b+c ) 52 In view of the large number, it is not pragmatic to consider all models for refinement process. Hence, several thumb rules based on chemical principles and well established practices in multiple linear regression analysis are resorted to, for this purpose. Hence the species considered to develop different models are 2+, +, 2H2 2+, 2H +, 2, 3H and 3 -. The preliminary screening of a number of species based on several thumb rules 49 resulted in a short list of 2+, +, 2H2 2+, 2H + and

24 SELECTION OF BEST FIT MODEL The models containing different number of species were tried from the primary alkalimetric titration data. Only a few species were refined while other species were rejected by MINIQUAD75 5. The models with combinations of different species at a time were also tested. After arriving at a valid model, the species rejected in the primary scrutiny were again tried. Some species were removed from the model if their percentage is less than 1. Existence of species was determined by performing exhaustive modeling 51 and the result of one such typical system was given in Table 4.5. The models were evaluated assuming the simultaneous existence of different combinations of species. Models containing various number and combinations of species were generated using an expert system package CEES 52 and these models were refined using MINIQUAD75. As the number of species was increased, the models gave better statistics denoting better fit. This indicates that the final model appropriately fits the experimental data. Such exhaustive modeling was performed for all the systems and the final models are given in Tables for Co(II), Ni(II) and Cu(II) complexes of proline and valine in PG- and AN-water mixtures. The tables contain the stoichiometric coefficients and stability constants of the complex species, standard deviations in the stability constants and residual statistics of the best fit models. 136

25 proline complexes in 3% v/v AN-water mixture ( range = ; Number of points = 61). Model num ber log βmlh(sd) Ucorr *1 8 Ucorr= U/(NP-m); NP = Number of points; m = Number of formation constants; SD = Standard deviation. χ 2 Skew -ness Table 4.5: Results of some exhaustive modeling studies for Cu(II)- Kurtosis R- factor (8) (14) (37) (7) 16.84(18) (13) 13.36(13) (7) (1) (3) 17.5(7) 13.2(4)

26 Table 4.6: Parameters of best fit chemical models of Co(II), Ni(II) and Cu(II)-proline complexes in PG-water mixtures. PG % v/v log βmlh(sd) Range NP Ucorr *1 8 χ 2 Skewness Kurtosis R- factor Co(II). 5.3(8) 9.12(7) 11.88(14) (9) 9.51(8) 12.18(11) (8) 9.6(7) 12.27(7) (19) 9.42(19) 12.73(12) (9) 9.86(9) 12.57(6) (12) 1.49(12) 13.15(6) (9) 1.89(1) 13.4(4) Ni(II). 6.51(12) 11.51(13) 12.73(8) (8) 11.43(9) 13.11(6) (6) 11.72(6) 13.3(3) (5) 11.89(6) 13.5(3) (7) 12.22(8) 13.19(4) (8) 12.71(8) 13.52(4) (8) 12.74(9) 13.84(4) Cu(II). 8.92(8) 16.23(9) 11.9(55) (9) 16.33(9) 12.27(33) (4) 16.4(8) 12.48(5) (11) 16.54(16) 12.81(2) (3) 17.14(5) 12.85(5) (3) 17.37(6) 12.75(7) (5) 17.17(11) 13.18(9)

27 Table 4.7: Parameters of best fit chemical models of Co(II), Ni(II) and Cu(II)-proline complexes in AN-water mixtures. R- AN log βmlh(sd) - NP Ucorr χ 2 Skewnesosis Kurt- % v/v Range *1 8 factor Co(II). 5.3(8) 9.12(7) 11.88(14) (22) 9.15(21) 13.7(15) (17) 9.65(17) 13.4(1) (15) 1.35(16) 13.71(1) (19) 1.65(21) 13.86(15) (15) 11.3(17) 14.14(11) (22) 11.4(27) 14.19(13) Ni(II). 6.51(12) 11.51(13) 12.73(8) (13) 11.6(16) 13.7(9) (13) 11.75(14) 13.65(8) (12) 11.88(15) 13.99(7) (19) 12.31(23) 14.2(12) (22) 12.38(3) 14.49(13) (22) 13.23(23) 14.2(1) Cu(II). 8.92(8) 16.23(9) 11.9(55) (1) 15.92(17) 12.72(16) (6) 16.31(1) 12.63(1) (3) 17.5(7) 13.2(4) (7) 17.36(17) 13.2(7) (6) 17.59(2) 13.52(8) (6) 18.36(17) 13.88(8)

28 Table 4.8: Parameters of best fit chemical models of Co(II), Ni(II) and Cu(II)-valine complexes in PG-water mixtures. PG log βmlh(sd) - NP Ucorr χ 2 Skewnesosis Kurt- R-factor % v/v Range *1 8 Co(II). 4.44(1) 8.7(9) 1.97(17) (4) 7.98(4) 11.25(4) (7) 8.3(5) 1.96(16) (21) 8.13(19) (8) 8.35(8) 11.16(12) (11) 8.91(1) 11.9(2) (8) 8.98(7) 11.51(8) Ni(II). 5.31(9) 9.52(9) 11.36(9) (6) 9.78(6) 11.84(4) (7) 1.46(7) 11.97(4) (62) 11.62(6) 12.1(92) (13) 11.5(13) 12.69(8) (18) 12.27(18) 12.89(11) (2) 12.52(21) 13.61(15) Cu(II). 8.35(5) 14.8(8) 12.4(4) (3) 15.15(6) 11.73(4) (4) 15.41(9) 11.94(5) (74) 15.59(74) 12.48(91) (5) 15.97(9) 12.29(5) (6) 16.44(13) 12.59(7) (8) 16.76(15) 12.66(8)

29 Table 4.9: Parameters of best fit chemical models of Co(II), Ni(II) and Cu(II)-valine complexes in AN-water mixtures. AN log βmlh(sd) NP Ucorr χ 2 Skewnesosis Kurt- R-factor % v/v *1 8 Co(II) ( range 2.-9.). 1.97(17) 4.44(1) 8.7(9) (17) 4.24(2) 7.79(16) (16) 4.39(16) 7.99(14) (13) 4.63(11) 8.28(11) (19) 4.83(15) 8.62(15) (22) 4.84(14) 8.7(15) (26) 5.4(21) 9.5(22) Ni(II) ( range 2.-9.) (9) 5.31(9) 9.52(9) (6) 5.45(6) 9.77(6) (11) 5.47(13) 1.22(11) (5) 5.73(6) 1.17(6) (15) 5.88(2) 1.64(21) (14) 6.1(17) 1.66(22) (13) 6.35(15) 11.16(19) Cu(II) ( range ). 12.4(4) 8.35(5) 14.8(8) (5) 8.31(4) 14.86(8) (6) 8.63(5) 15.52(11) (6) 9.19(4) 16.2(12) (6) 9.29(3) 16.54(9) (9) 9.67(8) 17.6(19) (12) 1.2(8) 17.97(16)

30 Residuals analysis The residuals-differences between what is actually measured or observed (EMF,, volume, concentration of ingredients etc.) and that predicted by a model-measure the unexplained variation in the dependent variable by the regression equation. The goal of rigorous residual analysis is to conclude that the assumptions in regression analysis and of chemical laws (mass-balance equations, electrode response, equilibrium expression etc.) are valid for a data set on hand. The residuals should follow normal distribution for the best fit model. The use of plots of residuals versus calculated volumes of dependent variables (EMF or volume) and independent variables (TM, TL) is popular. Gans et al. 53 applied sample standard deviation (SD) in weighted least squares analysis for the calculation of β s and suggested that any value less than three is satisfactory. Of course, the ideal value of 1 was observed only in a single titration curve (unlike pooled data) inferring that the data have been correctly weighted. The SD and confidence intervals in β are meaningful only when unweighted residuals follow χ 2 distribution, which measures the possibility of residuals forming a part of standard normal distribution with zero mean and unit standard deviation. Higher χ 2 values than expected are due to 1) the inadequacy of the model although the experimental data are of high quality, 2) use of 142

31 poor data even though the model is appropriate and 3) invoking optimistic estimates of errors in primary data. A perusal of Tables indicates that the χ 2 values range between for PG-water and for AN-water mixtures. The values of kurtosis and skewness range from and for PG-water mixtures and and for AN-water mixtures, respectively. Deviation of the values of kurtosis and skewness from three and zero, respectively, show the tendency of these residuals to concentrate more to the left or right of the mean and broadening of the peak. However, the values of Ucorr in all the three mass-balance equations, are very low confirming the adequacy of the chemical model to represent the experimental data. Retrieval of protonation constants Protonation constants were retrieved from the metal-ligand titrations and compared with those obtained from proton-ligand titration data. The proximity of the two values confirms the existence of only reported metal-ligand species and accuracy of the titration data. Such comparisons for some typical systems are given in Table 4.1. Then simultaneous refinement of all the constants revealed that when the approximate constants are very close to the true values, either fixing some of the species or ingredient concentrations do not have any ill- effects on modeling studies. 143

32 Table 4.1: Comparison of protonation constants determined from System proton-ligand and metal-ligand titration data. From proton-ligand titration data From metal-ligand titration data log β1 log β2 log β1 log β2 Ni(II)-proline in 1% v/v PG Co(II)-proline in 2% v/v PG Ni(II)-proline in 3% v/v PG Cu(II)-proline in 3% v/v PG Cu(II)-proline in 4% v/v PG Cu(II)-proline in 5% v/v PG Cu(II)-proline in 1% v/v AN Ni(II)-proline in 3% v/v AN Co(II)-valine in 1% v/v PG Cu(II)-valine in 1% v/v PG Ni(II)-valine in 3% v/v PG Co(II)-valine in 4% v/v PG Co(II)-valine in 5% v/v PG Cu(II)-valine in 6% v/v PG Ni(II)-valine in 1% v/v AN Cu(II)-valine in 1% v/v AN Co(II)-valine in 2% v/v AN Co(II)-valine in 3% v/v AN Ni(II)-valine in 3% v/v AN Co(II)-valine in 4% v/v AN Co(II)-valine in 5% v/v AN Cu(II)-valine in 5% v/v AN Co(II)-valine in 6% v/v AN Perturbation in Stability Constants due to Systematic Errors The computer program refines the stability constants by minimizing the random errors in the data. But in the presence of 144

33 considerable systemic errors, not only the β s are in errors; even some species may be rejected. MINIQUAD75 has no provision to vary the influential parameters. Hence, some representative systems were studied in order to have a cognizance of the effect of errors in concentrations of ingredients on the stability constants of binary metal complexes. These results are given in Tables 4.11 and The data shows that the magnitudes of stability constants are more affected by acid and alkali than ligand and metal. The increased standard deviation in stability constants and even rejection of some species on the introduction of errors confirms the correctness of the proposed models. This type of investigation is significant as the data acquisition was done under varied experimental conditions with different accuracies. 145

34 Table 4.11: Effect of errors in influential parameters on the stability constants of Co(II)-proline binary complexes in 4% v/v PG-water mixture. Ingredient % log βmlh(sd) Error (9) 9.86(9) 12.57(6) Alkali Acid Ligand Metal Log F (36) 7.3(62) Rejected (15) 8.61(17) Rejected (8) 11.3(8) 13.21(4) (8) 15.99(17) 13.61(1) -5 Rejected Rejected Rejected (8) 11.61(9) 13.65(4) (17) 8.7(18) Rejected (47) 7.34(62) Rejected (8) 9.7(8) 11.13(75) (8) 9.79(8) 12.31(8) (9) 9.93(1) 12.78(5) (9) 1.5(1) 13.3(5) (1) 1.1(9) 12.61(6) (9) 9.92(9) 12.59(6) (8) 9.8(8) 12.56(6) (8) 9.71(9) 12.54(6) (1) 1.5(1) 12.92(5) (9) 9.94(1) 12.73(6) (8) 9.79(8) 12.41(7) (8) 9.7(8) 12.8(11) 146

35 Table 4.12: Effect of errors in influential parameters on the stability constants of Ni(II)-valine binary complexes in 3% v/v ANwater mixture. Ingredient % log βmlh(sd) Error (5) 5.73(6) 1.17(6) Alkali Acid Ligand Metal Log F -5 Rejected 4.34(26) 6.45(127) -2 Rejected 5.6(1) 8.77(7) (5) 6.55(1) 11.48(1) (11) 9.62(9) 15.13(19) -5 Rejected Rejected Rejected (5) 6.77(1) 11.65(1) +2 Rejected 5.6(12) 8.9(16) +5 Rejected 4.43(33) 7.5(79) (5) 5.75(6) 1.34(6) (5) 5.74(6) 1.24(6) (5) 5.73(6) 11.11(6) (5) 5.72(5) 1.1(7) (22) 5.51(7) 1.3(7) (6) 5.64(6) 1.11(6) (4) 5.83(6) 1.25(7) (3) 5.99(6) 1.37(7) (4) 5.83(6) 1.28(7) (4) 5.78(6) 1.22(7) (5) 5.69(6) 1.13(6) (6) 5.64(6) 1.8(6) 147

36 SECTION-2 EFFECT OF SOLVENT ON METAL-LIGAND EQUILIBRIA Cosolvent influences the equilibria in solution due to change in the dielectric constant (D) of the medium that varies the relative contribution of electrostatic and non-electrostatic interactions which in turn vary the magnitude of stability constants. The variation of overall stability constant values or change in free energy with cosolvent content depends upon electrostatic and non electrostatic factors. Born s classical treatment 54 holds good in accounting for the electrostatic contribution to the free energy change. According to this treatment, the energy of electrostatic interaction is related to dielectric constant. Hence, the log β value should vary linearly as a function of reciprocal of the dielectric constant (1/D) of the medium. PG-water mixtures PG is an amphiprotic and coordinating solvent. It is a structure former and enhances the water structure in PG-water mixtures. Hence it removes water from the coordination sphere of metal ions, making them more reactive towards the ligands. As a result, the stability of the complex is expected to increase. It is also a coordinating solvent and competes with the ligands for coordinating the metals. This decreases the stability of the complexes. Hence, the stability of the complex is expected to either increase or decrease linearly. 148

37 The linear trend observed in the present study (Fig. 4.7) indicates that electrostatic forces are dominating the equilibrium process under the present experimental conditions. The linear increase in the stabilities of the complexes with 1/D confirms the dominance of structure forming nature of PG over its coordinating nature. AN-water mixtures AN is a protophobic, dipolar aprotic and coordinating solvent. It is a structure breaker of water and disrupts the water structure to form AN-water complex 55 of the formula AN.H2O. When small amount of AN is added to water, the water structure breaks down resulting in more basic monomeric water molecules. Hence water molecules compete with the ligands for coordination with metal ions, decreasing the stability of the complexes. But the formation of solvent-water complex decreases the coordinating power of water thereby increases the stability of the complex. The linear increase in the stabilities of the complexes (Fig. 4.8) with 1/D confirms the dominance of complexing ability of AN with water over its coordinating nature. Since complex formation can be viewed as a competition between the pure and solvated forms of ligand and the metal ion, the solute-solvent interactions, relative thermodynamic stabilities and kinetic labilities are also expected to play an important role. Different types of electrostatic forces dominate in different ranges of the composition of PG- and AN-water mixtures. With the increase in the percentage of PG and AN from.-6.% v/v, the dielectric constant 149

38 of the medium decreases from 78.5 to 51.2 and 5.8, respectively. Thus the variation of stability constants was studied over a range of the dielectric constant from 78.5 to 5.8. It is concluded from these studies that there is considerable increase in the stabilities of proline metal complexes in AN than in PG. But the increase in the case of valine metal complexes is negligible. This is attributed to the dominance of complex forming ability of AN with water than the structure forming nature of PG. 15

39 15 (A) 12 (D) 12 1 log β 9 log β /D /D 15 (B) 14 (E) log β 9 log β /D /D 18 (C) 18 (F) log β 12 log β /D /D Fig. 4.7: Variation of stability constants of proline complexes (A) Co(II), (B) Ni(II) and (C) Cu(II) and valine complexes of (D) Co(II), (E) Ni(II) and (F) Cu(II) with reciprocal of dielectric constant (1/D) of PG-water mixtures: ( ) log β, ( ) log β2 and ( ) log β. 151

40 15 (A) 12 (D) 12 9 log β 9 log β /D /D 15 (B) 12 (E) 12 1 log β 9 log β /D /D 2 (C) 18 (F) log β 12 log β /D /D Fig. 4.8: Variation of stability constants of proline complexes (A) Co(II), (B) Ni(II) and (C) Cu(II) and valine complexes (D) Co(II), (E) Ni(II) and (F) Cu(II) with reciprocal of dielectric constant (1/D) of AN-water mixtures: ( ) log β, ( ) log β2 and ( ) log β. 152

41 SECTION- 3 DISTRIBUTION DIAGRAMS The percentage of metal ion in the form of various complex species is given by Eq l(j) h(j) m(j) β m(j)l(j)h(j ) FL FH PS = i i N l(j) h(j) β m(j)l(j)h(j ) FL FH j= i i 1 (4.3) Using the formation constant in the best fit model, distribution diagrams were obtained with the computer programs DISPLOT 56 and SCPHD 57. These programs can be used to output distribution contours in any required range and to study the effect of errors in stability constants or variation in the concentration of ingredients. This information is highly useful in choosing ingredient concentrations to increase or decrease the concentrations of selected species in PG- and AN-water mixtures. The variation of species concentration with is shown in Figs for metal-proline and metal-valine complexes in PG- and AN-water mixtures. The patterns of the distribution of species with show that the concentrations of species are affected by solvent. Proline exists as 2 +, and L - in the ranges , and , and valine in the ranges , and , respectively 58. Therefore, the stability constants of metalligand complexes of proline and valine will depend on the range of study. Under the present experimental conditions, the predominant 153

42 forms of the ligands are 2 + and which limits the probable metalligand species to be +, 2, 2+, 2H2 2+ and 2H +. The species that are refined are +, 2 and 2+ for Co(II), Ni(II) and Cu(II) with proline and valine in PG- and AN-water mixtures. The stability constants of these species are found to follow the trend Co(II) < Ni(II) << Cu(II). This is in conformity with the Irving-Williams order 59. The additional high stability of Cu(II) complex may be due to Jahn- Teller distortion. + and 2 species were reported in the literature for Co(II), Ni(II) and Cu(II) with proline and valine. But the present study reports 2+ species also for both the ligands. The formation of various binary complex species is represented in the following general Equilibria ( ) for both the ligands. M(II) H + (4.4) H + (4.5) M(II) H + (minor process) (4.6) M(II) H + (4.7) M(II) H + (4.8) H + (4.9) H + (4.1) 154

43 2+, + and 2 species are formed in the range (Figs ). For all metal-ligand systems in PG- and AN-water mixtures, 2+ species is formed by the interaction of metal ion with 2 + (Equilibrium 4.4), because the percentages of metal ion and 2 + are decreasing with increasing percentage of species can be formed by the deprotonation of 2+ (Equilibrium 4.5), the interaction of metal ion with 2 + (Equilibrium 4.6) and the interaction of metal ion with (Equilibrium 4.7). Equilibria 4.5 and 4.7 are more predominant than Equilibrium 4.6 because the concentrations of metal ion, 2+ and are decreasing with increasing concentration of +. Equilibrium 4.6 is responsible for the initial formation of +. The simultaneous formation of + and 2 suggests the existence of Equilibria 4.7 and is formed by the interaction of metal ion with (Equilibrium 4.8), 2+ with (Equilibrium 4.9) and + with (Equilibrium 4.1). The Equilibria 4.8 and 4.9 are more appropriate because the concentrations of metal ion, 2+ and are decreasing where 2 species is increasing. An observation made from the distribution diagrams (Figs ) is that the free metal ion concentration is more in the case of valine than proline for Co(II) and Ni(II) in PG- and AN-water mixtures. This infers the stronger complexing ability of proline than valine, even though the distinctive cyclic structure of proline's side chain gives proline an exceptional conformational rigidity compared to other amino acids. 155

44 1 (A) 1 (D) (B) (E) (C) 2 8 (F) Fig. 4.9: Distribution diagrams of M(II)-proline/valine in aqueous media (A) Co(II)-proline, (B) Ni(II)-proline, (C) Cu(II)-proline, (D) Co(II)-valine, (E) Ni(II)-valine and (F) Cu(II)-valine. 156

45 (A) (B) 2 1 (C) 1 (D) (E) 2 9 (F) Fig. 4.1: Distribution diagrams of Co-proline in PG-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

46 9 (A) 8 (B) (C) 2 9 (D) (E) 2 9 (F) Fig. 4.11: Distribution diagrams of Ni-proline in PG-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

47 9 6 3 (A) (B) (C) (D) (E) 9 (F) Fig. 4.12: Distribution diagrams of Cu-proline in PG-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

48 9 (A) 1 (B) FL FL 1 (C) 1 (D) FL 5 25 FL 1 (E) 2 1 (F) FL 5 25 FL Fig. 4.13: Distribution diagrams of Co-valine in PG-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

49 9 (A) 2 9 (B) FL FL (C) 2 FL 9 6 (D) 2 3 FL (E) 2 FL Fig. 4.14: Distribution diagrams of Ni-valine in PG-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F) (F) 2 161

50 8 6 (A) 8 6 (B) (C) (D) (E) (F) Fig. 4.15: Distribution diagrams of Cu-valine in PG-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

51 9 (A) 2 8 (B) (C) 2 9 (D) (E) 2 9 (F) Fig. 4.16: Distribution diagrams of Co-proline in AN-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

52 9 6 3 (A) (B) 2 9 (C) 2 9 (D) (E) 2 9 (F) Fig. 4.17: Distribution diagrams of Ni-proline in AN-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

53 9 6 3 (A) (B) (C) 2 9 (D) (E) 2 9 (F) Fig. 4.18: Distribution diagrams of Cu-proline in AN-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

54 9 (A) 9 (B) (C) 9 (D) (E) 2 9 (F) Fig. 4.19: Distribution diagrams of Co-valine in AN-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

55 9 (A) 2 9 (B) (C) 2 9 (D) (E) 2 9 (F) Fig. 4.2: Distribution diagrams of Ni-valine in AN-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

56 9 6 3 (A) (B) (C) 2 8 (D) (E) 2 9 (F) Fig. 4.21: Distribution diagrams of Cu-valine in AN-water mixtures. % v/v: (A) 1., (B) 2., (C) 3., (D) 4., (E) 5. and (F)

57 Structures of binary complexes Although it is not possible to elucidate or confirm the structures of complex species metrically, they can be proposed based on the literature reports and chemical knowledge. In aqueous solutions metal ions are coordinated by six water molecules. Amino acids replace water molecules and form metal-amino acid complexes Depending upon the nature of the ligands and metal ions and based on the basic chemical knowledge tentative structures of the complexes are proposed as shown in Figs and 4.23 for proline and valine complexes. Carboxyl oxygen and amino nitrogen of ligands are bonded to the metal ions. Amino nitrogen can associate with hydrogen ions in physiological ranges and it results in the formation of protonated species. Hence protonated complex species are detected in the present study. The Cu(II) ion forms distorted octahedral or square planar complexes due to Jahn-Teller effect. 169

58 S S M S NH 2 S S M NH S O S 2+ O S O S + O O O S M NH NH S O O 2 Fig. 4.22: Proposed structures of L-proline complexes, where S is either solvent or water molecule. O O NH 3 S O M S S O M H 2 N S S S S S S 2+ O + S O M H 2 N N H 2 O S O 2 Fig. 4.23: Proposed structures of L-valine complexes, where S is either solvent or water molecule. 17