Electrometric Investigation of the Nature and Stability of Mixed Ligand Complexes of L-Ornithine and 1,10-Phenanthroline With Some Essential Metal Ions in Aqua-TBAB or PEG-400 Surfactant

Complexes of some essential metal (M) ions (Co2+, Ni2+, Cu2+) of L-ornithine as primary (L) and 1,10-phenanthroline as secondary (X) ligands in various concentrations (0.0–2.5% v/v) of aqueous tetrabutylammonium bromide (TBAB) or polyethylene glycol-400 (PEG-400) surfactant were synthesized pH-metrically at 298 K and 0.16 mol L-1 ionic strength. The relative amounts of L:M:X were 2.5:1.0:2.5; 5.0:1.0:2.5; 2.5:1.0:5.0. The data acquisition of acid-base equilibria and determination of stability constants were performed using MINIQUAD75 algorithm. The distribution patterns of the complexes with varying pH and compositions of surfactants were investigated from the plots of SIM run data. The best fit chemical models were found to be MLXH, MLX2H in the lower pH, and MLX, ML2X in the higher pH ranges for all the metals. For all these species, the logged values of stability constants decreased linearly with increasing concentrations of surfactants, indicating the dominance of electrostatic factors. The log of the disproportionation constant and the change in log values of the mixed ligand constant indicated additional stability of the mixed ligand complexes, compared to the parent binary complexes due to interactions outside the coordination sphere. This makes the mixed ligand complexes more amenable to metal ion storage and transport, while less easily bio-available. Significant change on the magnitudes of the stability constants, high values of standard deviation and rejection of some of the proposed chemical models were observed due to pessimistic error, indicating the sufficiency of the models to represent the data and accuracy of the method employed.

Mononuclear binary complexes involving interaction of single metal with one ligand and mixed ligand complexes that contain more than one ligand have been studied for long time (Ramanaiah et al. 2013;Ghazy et al. 2010). The stabilities of the synthesized complexes are a direct function of many factors such as: dielectric constants (Sakurai et al. 1976;Watters and Dawitt 1996;Esteves et al. 2015;Irving and Rossotti 1954;Mounir et al. 2007), stacking interactions between ligands, and metal-ligands (Mounir et al. 2008;Guerra et al. 2015), and electrostatic and non-electrostatic interactions (Prenesti et al. 2004;Ramanaiah et al. 2013;Sakurai et al. 1976;Watters and Dawitt 1996;Esteves et al. 2015;Irving and Rossotti 1954;Mounir et al. 2007. Literature reports indicated that most of these factors were mainly dependent on the composition of organic solvents, which could alter the dielectric constant of the medium and solvating power of the solvents (Watters and Dawitt 1996; Irving and Rossotti 1954).
Despite chemical speciation have been exhaustively studied in aqua-organic medium, the synthesis of metal complexes of ligands in aqua-micellar medium has sustained due attention owning to their untouchable applications and versatile altering capacity of the medium in solution chemistry.
1,10-Phenanthroline (Phen) is bidentate hetrocyclic ligand having two electron donor N-atoms to the vacant shell of transition metal ions (Nandihalli and Rebeiz 1991;Chetana et al. 2012). Studies have also revealed that Phen forms both binary and mixed ligand complexes with a number of metal ions in wide pH ranges (Bendi et al. 2012;Nandihalli and Rebeiz 1991;Chetana et al. 2012;Padmaja and Rao 2012;Atnafu et al. 2018). It is capable of forming stable metal complexes of even lower oxidation states (Ghazy et al. 2010;Cotton and Wilkinson 1972;McWhinnic and Miller 1969) in mixed ligand system. Several reports have indicated that Phen metal complexes have varieties of applications. For example, copper complexes of Phen inhibit the proteasome activity housed into tumor cells (Zhang et al. 2012). Phen complexes of metals were not only synthesized in aqueous solvent, but also in aqua-organic solvents and used as potent photodynamic herbicide modulators (Bendi et al. 2012;Nandihalli and Rebeiz 1991; in chemotherapeutic applications such as building muscles, reducing body fats, removing toxic ammonia from liver, etc (Rose et al. 1998). Recently, groups of authors have published the capability of L-ornithine to form stable mononuclear binary complexes with some essential metal ions in surfactant media (Atnafu et al. 2019). The current work of mixed ligand complex of metals is continuation of the protonation and binary complexes of the same ligand and metal ions in the same media.
Phen as bidentate and Orn as tridentate ligands play prominent roles in maintaining the stability of complexes in various solvents (Bendi et al. 2012;Padmaja and Rao 2012). Despite the availability of information on the effect of aqua-surfactants on the stability of binary complexes (Nandihalli and Rebeiz 1991;Atnafu et al. 2018;Atnafu et al. 2019), data related to mixed ligand complexes of Phen and Orn with essential metals in aqua-surfactant medium is scarce. The aim of this study was, therefore, to determine the stabilities of the mixed ligand complexes of some essential metal ions (Co 2+ , Ni 2+ , Cu 2+ ) with Phen and Orn in aqua-TBAB and PEG-400 surfactant media with an eye on the prediction of the bioavailability the metal complexes.

Chemicals and Solutions
All solutions were prepared in triply distilled water free of dissolved gases by nitrogen gas purged in it.
Tetrabutylammonium bromide, TBAB (Avra, India) and polyethylene glycol 400, PEG-400 (Merck, India), murexide and fast sulphon black F (both in ALPHA CHEMIKA, India), concentrated ammonia (Mysore Ammonia, India), hexamine, hydroxylamine hydrochloride (Prochem, Inc, Medley, USA), Buffer Solutions (pH 4, pH 7 and pH 9; Merck, India) were used as received while NaOH was washed 3-4 times with double distilled water and discarded prior to its standardization against potassium hydrogen phthalate and Gran-titrated regularly (Gran 1952) to check the absence of carbonates. Then, NaOH was used to determine concentration of HCl (0.05 mol dm -3 ). HCl solution was added to the titration mixture to ensure the solubility of the analytes, and suppress hydrolysis (Irving and Rossotti 1954;Mounir et al. 2007;Mounir et al. 2008;Guerra et al. 2015;Asagba et al. 2007;Veeraswami et al. 2014) of the synthesized complexes. A 2 mol dm -3 solution of sodium chloride (Merck, India) was prepared to maintain the desired ionic strength (0.16 mol dm -3 ) of the titration mixture. Chemicals used in this investigation were of analytical reagent grade.

Instruments and Methods of Titrations
A combined glass electrode of aqueous range conjugated with Metrohm 877 Titrino plus Auto-titrator (Switzerland, readability 0.001) was used to carry out electrometric titrations. Data acquisition was carried out at an ionic strength of 0.16 mol dm -3 and temperature of 298 K. The steady flow of purified nitrogen gas was applied through the titration mixture (Padmaja and Rao 2012;Atnafu et al. 2019;Kiptoo and Ngila 2005). Periodic calibration of the electrode was done to ensure reliable measurement under stable and reproducible response. The electrode had been checked for its stable response through careful equilibration in well stirred TBAB or PEG-400 solutions. Then, free strong acid titration (HCl) with strong base (NaOH) was carried out from which correction factor could be calculated using based on their contribution to the designed total ionic strength used in this study (0.16 mol dm -3 ) with modified experimental design and titration assembly given elsewhere (Irving and Rossotti 1954).
The effects of dielectric constant, variations in liquid junction potential, asymmetry potential, dissolved carbon dioxide, sodium ion error, and activity coefficient on the response of the electrode were explained by correction factor estimated by SCPHD program (Rao 1989;Sylva and Davidson 1979). For a reliable determination of stability constants corresponding to mixed ligand model complexes, pooled data refinement of all the experimental trial using MINIQUAD75 computer program was chosen. The results of the free acid correction factor, values of Kw, protonation constants of the ligands [22] and binary M-L complex constants (Atnafu et al. 2019) were fixed and used as initial imputes in the analysis of the mixed ligand complexes using MINIQUAD75 program (Sylva and Davidson 1979;Shoukry and Ezzat 2015;Esteves et al. 2015) with inbuilt statistical parameters of chemical model refinement tools.
Typical mixed ligand metal complexes were published by different researchers in the form of Orn-M(II)ligand and phen-M(II)-ligand with brief account of the systems (Sakurai et al. 1978;Hadgu et al. 2015;Moussa et al. 1987;Shehata et al.2004) and the present study is presented for comparison in Table 1.

Chemical models of mixed ligand complexes
Reliable and best representative models for the experimental data were obtained by increasing the number of species to be refined, satisfying statistical refinement parameters. This indicates that the final model appropriately fits with the experimental data. The plausible chemical models for Orn-M(II)-Phen mixed ligand systems in aqueous TBAB or PEG-400 medium for Co(II), Ni(II) and Cu(II) along with the magnitude of the corresponding stability constants are given (Tables 2 and 3).

System
Log β mlxh (T 0 C/ μ) The species successfully converged and finally refined in the active pH ranges were MLX, MLXH, MLX 2 H and ML 2 X for all Co(II), Ni(II) and Cu(II) in PEG-400-and TBAB-water mixture. The results of refinement under the parameters of the best-fit models are presented in Tables 2 and 3. The best fit models were found to be chemically consistent with better statistical fits to the electrometric titration data without showing systematic trends in the overall magnitude of the residuals. The standard deviations in log β values were also very low and U corr were very small, indicating the precision of the parameters; the good agreement of the models converged to the titration data and consistency of the model, respectively (Rao and Murthy 2004). The slight left or right distribution of errors has been clearly shown by the magnitude of the skewness close to zero and hence a least squares method may be applied to the generated data. The values of kurtosis were virtually observed to be greater than 3 for which the residuals form part of leptokurtic pattern of distribution errors (R.S. Rao and G.N. Rao 2005;Rao and Sudarson 2006).
The representativeness of the chemical models and its sufficiency has been indicated by the very low values of crystallographic R factor given in Table 2

Validation of the chemical models and interpretation of systematic errors
The sufficiency and quality of the best fit chemical model have been evaluated through the introduction of pessimistic errors in the concentrations (2 % and 5 %) of acid, alkali, ligand molecules and metal ions of interest. The validation was made in order to test the reliability and accuracy of data acquisition under varied experimental conditions. The results of refinements for the data subjected to pessimistic errors were found to show high standard deviation in log β values, significant deviation in the magnitude of the where m = number of species; NP = Number of experimental points stability constants and rejection of some of the proposed models by MINIQUAD75 algorithm. The effects of errors were more profoundly observed in alkali and acids than ligands and the metal ions as their concentration changed systematically. This shows the sufficiency of the models and accuracy of the method. The results of typical data in 1.5 % PEG-400 and TBAB are given in Table 4. Table 4: Effect of errors in concentrations of influential parameters on the stability constants of Orn-Ni(II)-Phen complexes in 1.5 % (v/v) surfactant-water mixtures.

Effect of surfactant
The additions of PEG-400 and TBAB lowered the dielectric constant (Bendi et al. 2012;Raju et al. 2012;Singh and Kalamdhad 2013) (35) system were found to be linearly decreasing with increasing percentage of PEG-400 or TBAB (Fig. 1), which causes destabilization of the complexes with net predominant effect of electrostatic factors.

Prediction of extra stability of the mixed ligand complexes
The relative stabilities of mixed ligand complexes, compared to their binary counterparts, can be predicted from disproportionation equilibria ( Irving and Rossotti 1954;Shoukry and Ezzat 2015;Pyreu et al. 2016 The formation of a mixed ligand complex of two different ligands, MLX, is statistically more favored than the formation of the parent ML 2 and MX 2 binary complexes when equal concentrations of M, L and X are available in solutions (Watters and Dawitt 1996). Additionally, ML 2 (25%) and MX 2 (25%) complexes can readily form mixed ligand complexes, 2MLX (50%), and the value of log X was reported to be 0.6 (Mounir et al. 2008;Guerra et al. 2015;Ramanaiah et al. 2013;Sigel et al. 1974;Kumar et al. 2012).
The value of Δlog K should be negative when either of the ligands coordinates with the free metal ion in comparison to the simple complexes in binary system, where the log value of first stepwise stability constant K 1 is always greater than the log value of second stepwise stability constant K 2 . The change in log values of the mixed ligand constant (Δlog K) is usually positive. Thus, the Δlog K could be taken as better measure of the tendency towards the formation of mixed ligand complexes than severely criticized log values of overall stability constants (Pyreu et al. 2016;).
Based on the electrostatic theory of binary complex formation and statistical arguments, the additional site of coordination of given multivalent hydrated central metal ion could easily be available for the first ligand than for the second, probably due to steric hindrance. Hence, the usual order of stability >  (Pyreu et al. 2016; for O-donor (malonic acid, pyrocatechol, etc.), negative for N-donors (ethylenediamine) and intermediate or negative ( Rao and Sudarson 2006) (Sakurai et al. 1976;Sakurai et al. 1977).
The values of Δlog K and log X are calculated from the magnitude of stability constants of binary and mixed ligand models using the selected equations given in Table 5. When the charges of the two different ligands are not equal, electrostatic factor also contributes for the formulation of log X. The log X values for all Orn-M(II)-Phen system were observed to be higher than the known table value of the disproportion constant (usually 0.6) which reveals the extra stability of the mixed ligand complex (Table 6). It also indicates that statistical and electrostatic factors are applied favourably for the formation of mixed ligand complexes.   and ML 2 X for all the metals. MLXH and MLX 2 H model species exist at lower pH while MLX and ML 2 X were found at higher pH (cf. Figs. 2 and 3).
As shown in Equilibrium (4), MLXH is formed by direct combination of the free metal ion, M(II), with LH and XH 2 forms of primary and secondary ligands in the pH ranges of 1.5-11.0; 1.5-7.0 and 1.5-11.0 with max extent of formation, resulting in the formation of complexes with different amounts, e.g. 60-80% for Co(II); 35-80% for Ni(II) and 55-80% for Cu(II) system, respectively.    Deprotonation of MLXH at higher pH gives MLX chemical model, Equilibrium (5). MLX could also be formed by the interaction of free metal ion M(II), primary ligand (LH) and secondary ligand (XH 2 ) (cf. Equilibrium (6) and ML with XH 2 , Equilibrium (7), in the pH ranges of 7.0-11.5; 1.5-8.0 and 2.0-8.5 with maximum extent of formation varying as low as 15 to max of 90% for Co(II), 30-80% for Ni(II) and 40-90% for Cu(II) system, respectively.
The other species which exists predominantly at higher pH region is ML 2 X. In the present study, ML 2 X might be formed by the interaction of ML simultaneously with LH and XH 2 (Equilibrium (9)) and MLX with LH (Equilibrium (10)

Structures of mixed ligand complexes
The stability of a complex is directly dependent on the nature of the central metal ion (

Conclusion
In this study, the chemical models refined were only MLXH, MLX 2 H, MLX and ML 2 X for Co(II), Ni (II) and Cu(II) since the active Orn exists as LH 3 2+ , LH 2 + and LH active forms in the pH ranges of 2.0-3.5, 2.0-9.0, 8.0-11.0  and Phen as XH 2 2+ and XH + active forms in the pH ranges of 1.6-7.0 and 1.48-7.0 (Atnafu et al. 2018), respectively. The extra stability of mixed ligand complexes compared to the parent binary complexes was determined based on disproportination constant (log X) and change in the stability constant (Δlog K) of the complexes (Atnafu et al. 2019). This extra stability may be attributed to the interactions outside the coordination sphere, such as the formation of hydrogen bonds between the coordinated ligands, chelate effect, charge neutralisation, the electrostatic interaction between non-coordinated charge groups of the ligands and stacking interactions. Thus, this study threw light on the bioavailability, transport and storage of the metal ions in the bio-fluids. The less stable binary complexes could easily be bioavailable while the mixed ligand complexes could be stored and transported in the biofluids. The sufficiency of the best fit chemical model has been validated by the introduction of pessimistic errors. Pessimistic errors due to change in the concentration of alkali and acids affected the values of the stability constants, standard deviations, and caused rejection of some of the chemical species more significantly than the ligands and metal ions. The linear decrease in the values of stability constants with concentration of surfactants may be ascribed to the synergistic effects of lowered dielectric constant values and destabilising influences of the neutral and cationic surfactants on the positively charged complexes.

Declarations
Author contribution statement All authors have contributed to this work in accordance with their order put in this article. Moreover, the first and last authors have contributed in the conception of the research ideas.

Funding statement
There has not been any grant received for this work except material support from Andhra University, India.

Data availability statement
Any data included in the article are acknowledged.

Additional information
No additional information is needed for this paper.