Molybdenum ( VI ) Binded to Humic and Nitrohumic Acid Models in Aqueous Solutions . Salicylic , 3-Nitrosalicylic , 5-Nitrosalicylic and 3 , 5 Dinitrosalicylic Acids , Part 2

Apresentam-se neste trabalho, estudos eletroquímicos e espectroscópicos na região do Ultravioleta-Visível de modelos derivados do ácido salicílico de ácidos húmicos e nitrohúmicos, esse último um artefato laboratorial, na presença de molibdênio para determinar a afinidade desses modelos com esse íon metálico. Molibdênio tem um papel importante na química de solo e conjuntamente com as substâncias húmicas, promove fertilidade ao solo e água e é um elemento chave na enzima nitrogenase. Os resultados obtidos mostraram que pelo menos uma espécie complexada está presente na faixa de pH 6,3 a 8,0, mesmo para o modelo menos básico escolhido, dos ácidos nitrossalicílicos. Estudo anterior mostrou que modelos derivados do ácido ftálico e nitroftálico, modelos também para os ácidos húmicos e nitrohúmicos, apresentaram espécies complexadas apenas até pH 6,5. As constantes de formação calculadas mostraram que a substituição do grupamento nitro na posição orto desfavorece a complexação, quando comparada à substitução para, devido provavelmente ao impedimento estérico no primeiro caso, uma vez que esse impedimento se manifesta indubitavelmente no composto duplamente substituído com o grupamento nitro. Os espectros obtidos por voltametria cíclica e por espectroscopia no Ultravioleta-Visível mostraram que a química do molibdênio em solução aquosa é muito complexa com a variação do pH, e que o ânion molibdato deixa de ser um ânion em valores de pH próximos a 4, quando então passa a se comportar como cátion, MoO 2 2+ (M).


Introduction
Chemistry of soil molybdenum (Mo) plays an important role in reactions where the nitrogenase enzyme is involved.
Thus, each Mo acts as a separate catalytic center and is a significant redox enzyme essential for microorganisms, plants and animals. At the centre of the nitrogenase enzyme is the FeMo cofactor, where the actual reduction of dinitrogen takes place. The Mo-Fe-S is biologically important in nitrogen fixation catalysed by nitrogenases.
One of its component is the Mo-Fe protein which contains a unique Fe-Mo-S cluster, which is believed to be the active site for the dinitrogen reduction. Figure 1 presents a structure of this nitrogenase FeMo cofactor where after core rearrangement there is the formation of two isomers (Mo 2 Fe 6 S 9 ), having Mo or Fe as the metal ion. 1 The [MoO 3 (OH)] -+ H + + 2 H 2 O Mo(OH) 6 or MoO 2 (OH) 2 (H 2 O) 2 (7) At pH values higher than 4 the most important species is MoO 2 2+ , when molybdenum is chemically acting as a cation according to previous equations 3 and 7. [3][4][5] Soil organic matter (OM) promotes soil aggregation and its chemical function is recognized by its ability to interact with metals, oxides, metal hydroxides and mineral clays to form organometallic complexes. Soil OM improves N, P, and S bioavailability, therefore it has the potential use as a slow release fertilizer. The OM is very heterogeneous and it can be functionally grouped into nonhumic (carbohydrate, proteins, fats, waxes, etc) and humic structures. 6 It is widely stated in the literature that soil humic acid (HA) composition is dependent on soil composition and organic precursors, resulting in differences in HA molecular mass and chemical structures. Consequently metal binding will vary with HA sources. [7][8][9][10][11][12][13] Soil fertility plays a role by improving metal ion mobility and transport in the environment. [11][12][14][15][16] Cation binding to humic substances include the extreme binding heterogeneity of these natural materials, the variable stoichiometry of binding, the competition between specifically-bound ions, especially protons and metal ions, and electrostatic effects which give rise to ionic strength effects and the non specific binding of couterions. 17 All the mathematical models presented in the literature so far [18][19][20][21][22][23][24][25] cannot deal with all those parameters, sometimes employing empirical equations and bacterial and HA competition. Also the complexation of metals by heterogeneous ligands is dependent on metal loading (the rate of bound metal to the binding site concentration). To overcome some of these difficulties this study deals with some molecular models of HAs according to the sorption sites following others in the literature 5,26-28 presenting data on those HAs main sorption sites that can reliably be used in those mathematical models as well as the speciation along a wide range of pH values.
Although higher valence cations such as molybdenum ions can undergo either hydrolysis or ligand binding, potentiometric titrations can provide the conditional stability constants as a way to try to understand how biological systems binds so selectively to ligands or metal ions. The study of model compounds has to be in such a way that they represent properly the known physical and chemical properties of the basic sites of the complex structure of the substance under investigation. Such is the case for humic and nitrohumic acids, geopolymers mainly composed of phenolic hydroxyl groups, hydroxycarboxylic and carboxylic aromatic groups as basic sites. [7][8][9] Therefore, in this work the MoO 4 2anion was chosen to be studied with the humic and nitrohumic acid modelssalicylic and nitro-salicylic acids. The nitrohumic acids are a laboratory artifact. Although the high oxidation species of molybdenum (such as in molybdate) is not the only one involved in the biological reactions, it is water soluble and undergoes many interconvertible reactions (refer to equations 1 to 7). It is possible to determine how the pH values changes can act on the equilibrium of all species. In order to determine all equilibrium possibilities, it was performed potentiometric, cyclic voltammetric (CV) and ultraviolet-visible (UV-Vis) titrations. A previous work reports a study with this same metal ion with models of phthalic, nitro-phthalic, catechol and nitro-catechol humic acids-like sorption or basic sites. 5
A proper mass of NH 4 Mo 7 O 24 .4H 2 O (Carlo Erba -Italy) was dissolved in water and its content was measured by back titration with disodium edetate (Carlo Erba -Italy) following the literature. 29 Potentiometric equilibrium measurements Following the procedures described in the literature 30 the errors arising in calculating unknown equilibrium constants by least-square techniques were minimized. Three titrations were performed with the metal ion and each ligand for all metal to ligand ratios analysed. The potentiometric studies were carried out in a Micronal (model B-375 -Brazil) pH meter fitted with a glass and a calomel reference electrode calibrated with standard HCl, 10 -2 mol L -1 (I= 0.100 mol L -1 (KCl)) and KOH solutions to read -log of the concentration of H + (pH) directly, under a KOH -aqueous saturated stream of N 2 (White-Martins -Brazil) and recording the pH values only after stabilization. The pH reproducibility was <0.005 pH at acidic pH region and <0.015 pH at the basic pH region. The temperature was maintained at 25.0±0.1 o C (Microquímica -MQBCT -99-20, Brazil) and the ionic strength adjusted to 0.100 mol L -1 with KCl.

Calculations
The protonation constants of the ligands employed were reported previously 26,27 according to equations 8 and 9 with charges omitted for simplicity.
The hydrolysis constants of molybdenum (VI) were taken from the literature. 31,32 The species considered in the calculations were metal to ligand ratios (metal being MoO 2
The calculations were made with Best7 program and species distributions were calculated with SPE program. 5,30 Aware of the region where the uptaking of protons by molybdate is seen, only when MoO 2 2+ cations were present the calculations began. This initial portion of the pH profiles produced no complexed species rather than polyanions, under the experimental conditions employed. The mathematical model for the calculations took into account the possible rates of bound metal to the binding site concentration (the metal loading parameter) 5 in order to simulate an heterogeneous ligand such as humic and nitrohumic substances, the hydrolysis constants of molybdenum ions 31,32 and the standardized concentrations of KOH -the titrant employed -and the concentration of the mineral acid (HNO 3 ) added when necessary.

Cyclic voltammetry
The cyclic voltammetric measurements were obtained with a cyclic voltammeter (Bioanalytical Systems Inc. -USA, model CV-27). A three electrode system was utilized: vitreous carbon as the working electrode, saturated Ag/AgCl-KCl as the reference and a platinum wire, the auxiliary electrode. The voltammograms were obtained in aqueous solutions, at room temperature of 25 o C, in ionic strength of 0.10 mol L -1 (NaNO 3 ), and bubbled with ultra pure argon at pH range of 2.5 to 10.0. The initial solutions were acidified with HNO 3 ( 0.1 mol L -1 ) and titrated with KOH 0.1 mol L -1 delivered by a microburet (Gilmont -USA) until the desired pH was reached. The cyclic voltammograms were obtained as a function of each pH value measured with a 5 mm diameter combined glass and Ag/AgCl reference electrode (Analyser -Brazil) and a Corning pH meter (UK), at an accuracy of 0.01 pH units. The final concentration in the reaction cell for all solutions was 10 -4 mol L -1 and the ligand to metal ratios were 1:1, 2:1 and 3:1. The experimental conditions of the cyclic voltammograph were as follows. The optimum range of swept potential was from 0.0 V to -1.1 V. All scans were made from -1.1 to 1.0 V; the scan-rate = 10 mV s -1 ; and initial potential applied = 0.0 V. 485 Molybdenum (VI) Binded to Humic and Nitrohumic Acid Models Vol. 17, No. 3, 2006 Ultraviolet-Visible spectra Separated spectrophotometric titrations were carried out with the ligands alone and in the presence of the molybdate starting solution. By following the pH values where there is the presence of complexed species, aliquots were taken and analysed by UV-Vis spectroscopy and it was possible to monitor the formation and decomposition of species in the equilibria studied.
The UV-Vis spectra were recorded on a Hewlett-Packard model 8450-A diode array spectrophotometer (USA) in the range of 190-600 nm. Aliquots of about 3.0 mL of a separate titration in the ratios of 1:1, 1:2: 1:3 metal to ligand were performed in a quartz cell (1.000 cm path length) appropriately coupled to a 10 mL vessel where the pH values were adjusted by adding small volumes of KOH (0.1 mol L -1 ) and measured accordingly to cyclic voltammetry section. The main striking feature of these profiles in the presence of metal is that at the beginning of the titrations, specially when the proportion favours the higher concentration of metal ion, the molybdate, an anion, first takes protons from the system, and then at pH values around 4.5 or greater, starts to act like a cation, by then complexing with the ligands. 32 Although the ML 3 species is highly thermodinamically unfavoured, potentiometric titration ratio M:L of 1:3 was conducted, but it was not possible to detect this species in percentages above 10% as required for the mathematical and chemical models employed in this study. 30 Formation constants for the complexed species were derived from the potentiometric data and are shown in Table 1.

Results and Discussion
Looking at the values in Table 1 the non nitro ligand (SALA) has greater binding affinity among the four studied ligands because the electrons of the potentially binding sites (carboxylic and phenolic groups) are not involved in the resonance effect of the aromatic ring. Such is not the case of the nitro ligands (NSA's) since the electron withdrawing effect of the -NO 2 substituent pulls the electron from the binding groups towards the aromatic ring and consequently depleting the basicity effect of these groups. In addition when the -NO 2 group is in C-3 (3-NSA and 3,5-DNSA) there is the stereochemical influence in the formation of ML and ML 2 , making the complexation a little less favoured compared to when the electron pulling group is in para position (5-NSA). Both ML and ML 2 logarithms of the formation constants showed this trend.    In the species distribution diagrams (Figures not  shown) there was at least one complexed species at pH near 6.5 and until pH 8, for each ligand tested.
The cyclic voltammetric titrations added only limited information due to the irreversibility of the obtained voltammograms. In Figure 5 it is shown the voltammograms of all ligands, 10 -4 mol L -1 . SALA is not electroactive and the salt (NH 4 ) 2 MO 2 is pseudoreversible at acidic pH values. All ligands caused a pH dependent cathodic peak (E pc ), but the definition became lower with the increased pH. At very high pH values the voltammograms lost their original shape and the E pc wave was no longer shown. All voltammograms showed a second wave at ~0.95 V that was assigned to amine group formation and does not undergo an oxidation process at the cathodic sweep (positive portion of the voltammogram not shown). The irreversibility may be due to the reduction of the nitro group, according to what occurs to nitrobenzene, which exhibits a single, pH-dependent and irreversible 4electron polarographic wave in aqueous media below pH 5. This 4-electron wave is accompanied by a second 2-electron wave at more negative potentials giving way to p-aminophenol, formed by rearrangement of phenylhydroxylamine. 34 The cathodic peak potential of all irreversible obtained voltammograms (Figures 5 to 9) showed a similar potential at acidic pH values (from 2.4 to 3.7) with E pc varying from ~ -480 mV to ~ -1000 mV depending on the ligand, when the metal to ligand ratio increased from 1:1 to 1:3 metal to ligand ratios, indicating more difficulty in reducing ML 2 than ML.
In Figure 6 the SALA complexes voltammograms are shown with varying pH in 1:1 metal to ligand ratio. In all metal to ligand ratios studied, the voltammograms showed a pseudo-reversible curves at low pH values (from 2.88 to 3.32) and above pH = 5 it was observed undefined regions impossible to visualize neither the cathodic nor the anodic peaks.   Figure 9 presents the cyclic voltammograms of molybdenum and 3,5-DNSA in the ratio 1:1. A second and a third peaks were observed in this voltammogram, E pc = -0.75 V and at -0.975 V, the latter being assigned to the second nitro group being reduced in 3,5-DNSA ligand.
Although the metal to ligand ratios of 1:3 have shown different values for E pc than any other metal to ligand ratio studied (Figure not shown), it was not possible to identify new species formed in this proportion neither by the cyclic voltammetry nor by potentiometric titrations. C items in Figures 5 to 9 are due to the uptaking of protons by the molybdate anion as well as the complexation at some extent by the system of any metal to ligand ratio.
The species distribution diagrams of all ligands in the presence of the metal ion were obtained (Figures not   shown) using the input of potentiometric data calculations in SPE program. 30 The maximum peak absorbances, shoulders and isosbestic points of the UV-Vis spectra of all ligands and the metal ion are collected in Tables 3 to 6. Aliquots of pH varying from 2 to 12 were taken and analysed by UV-Vis spectroscopy of all ligands (Supplementary Figure S1, and each ligand in the presence of the metal ion, SALA, 3-NSA and 3,5-DNSA, supplementary Figures S2, S3 and S4, respectively) where it can be seen changes in the spectra due to complexed species in agreement with the species distribution diagrams. At pH values far below and far above those shown, no significant presence of complexed species were detected in the equilibria studied as the spectra in these values showed a sharp decrease in the charge transfer bands characteristic of the Molybdenum complexes with the nitro ligands studied (300-370 nm) and SALA (270-300 nm). 35 The spectra of HL and L forms of SALA are at pH values of 4.67 and 12.19, where there are the presence of one deprotonated and the fully deprotonated forms, respectively (refer to Table 3). The spectra of MoO 4 2-(MO 2 ) alone has shown at pH = 5.45 the cation MoO 2 2+ (M) form which is the responsible for the complexed species found in all studied systems. The pH values of 4.07 and 5.52 represent the spectra of the complexed forms MHL and ML, respectively, in the 1:1 metal:SALA ratio. In 1:2 metal to ligand ratio it is seen the formation of ML 2 (pH=6.86). In the 1:3 metal:SALA ratio it is seen both at pH 4.16 and 10.19 that other species rather than ML, MHL or ML 2 are present, but it was not possible to clearly identify them by potentiometric titrations under the experimental conditions employed, as the equilibrium in those pH values might be a mixture of all species involving polyanions in the former pH value and hydrolysis products in the basic pH value.
The UV-Vis spectra of HL and L were taken at pH 4.66 and 10.63 (refer to Table 4) of 3-NSA. The potentiometric profile and the complete UV-Vis spectra of 3-NSA are published elsewhere. 26 In pH values of 4.11, 4.77 in 1:1 metal:3-NSA ratio shows the complexed species MHL and ML. In 1:2 metal to ligand ratio the UV-Vis spectra taken at pH = 5.01 depict the ML species and in 1:3 ratio at acidic and basic pH values there is a mixture of species including hydrolytic ones in the equilibrium (refer to supplementary Figure S3).
The UV-Vis spectra of HL and L were taken at pH 4.74 and 10.81 of 5-NSA, respectively (refer to Table 5).
The UV-Vis spectra of HL and L forms of 3,5-DNSA was presented at pH = 4.42 and 7.32, respectively. The pH values of 4.71 and 7.01 refer to the complexed forms MHL, ML being coincidental the two obtained curves due to an almost equimolar presence of MHL and ML from pH 4 to 7. In metal to 3,5-DNSA 1:2 ratio, ML 2 is present at pH around 7.6 (refer to Table 1).
The great number of isosbestic points in all taken UV-Vis spectra indicating interconvertion of complexed species, coincided with the pH and species presented by the distribution diagrams. It was seen that 3,5-DNSA has a sterical difficulty in forming ML 2 species even presenting a stability constant similar to 3-NSA. This can be explained by the small percentage of formation in the equilibrium of this species. The change in the CV and UV-Vis spectra caused by complex formation was used to support potentiometric data and was not used to calculate formation constants. Other complexed species were suggested by UV-Vis and CV analyses but it was not attempted to characterize them due to the large number of species in the equilibria as well as in the spectra. Some proposed complexes had a M:L ratio of 2:2, 2:1 and 4:2, as well as their protonated counterparts. The variation in the stoichiometry can be a direct consequence of the behaviour of molybdate and this metal ion towards α-hydroxycarboxylate ligands as previously reported. 35,36 Conclusions The use of this study showed the existence of at least one complexed model species of HA and NHA with molybdenum (VI) at a pH value commonly found in soil. Use of a salicylic acid model may contribute to the understanding of other complexed species involving this metal ion in aqueous solutions with varying pH values.
Making a comparison among salicylic HA and nitrosalicylic model compounds and ligands derived from phthalic and nitrophthalic based model compounds, the presence of the complexed species is seen in pH values until 8 only for the salicylic derived models and 4-   nitrocatechol, since the complexed species for phthalic derivatives are decomposed after pH 6.5. 5 The nitro compounds are weaker Lewis bases than the non nitro model compound -SALA, due to the electron withdrawing effect of -NO 2 in the aromatic ring. 5 Nonetheless, the nitrohumic substances, laboratory artifacts, presented as an alternative fertilizer by providing more N content when the mineralization of the soil occur, can also bind differently to metal ions when compared to non nitro organic matter in a wider range of pH values.