Spectroscopic, potentiometric and theoretical studies on the binding properties of a novel tripodal polycatechol-imine ligand towards iron(III)

Spectrochimica Acta Part A 71 (2008) 1452–1460

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Spectroscopic, potentiometric and theoretical studies on the binding properties of a novel tripodal polycatechol-imine ligand towards iron(III) B.K. Kanungo a,∗ , Suban K. Sahoo a , Minati Baral b a b

Department of Chemistry, Sant Longowal Institute of Engineering & Technology, Deemed to-be University, Longowal 148106, India Department of Chemistry, National Institute of Technology, Kurukshetra 136119, India

a r t i c l e

i n f o

Article history: Received 20 July 2007 Received in revised form 28 April 2008 Accepted 6 May 2008 Keywords: Tripodal polycatechol-imine ligand Iron(III) Stability constants Molecular mechanics and semi-empirical method

a b s t r a c t A novel multidentate tripodal ligand, cis,cis-1,3,5-tris[(2,3-dihydroxybenzylidene)aminomethyl]cyclohexane (TDBAC, L) containing one catechol unit in each arms of a tripodal amine, cis,cis-1,3,5tris(aminomethyl)cyclohexane was investigated as a chelator for iron(III) through potentiometric and spectrophotometric methods in an aqueous medium of 0.1N ionic strength and 25 ± 1 ◦ C as well as in ethanol by continuous variation method. From pH metric in water, three protonation constants characterized for the three-hydroxyl groups of the catechol units at ortho were used as input data to evaluate the stability constants of the complexes. Formation of monomeric complexes FeLH3 , FeLH2 , FeLH and FeL were depicted. In ethanol, formation of complexes FeL, Fe2 L and Fe3 L were characterized. Structures of the complexes were explained by using the experimental evidences and predicted through molecular modeling calculations. The ligand showed potential to coordinate iron(III) through three imine nitrogens and three catecholic oxygens at ortho to form a tris(iminocatecholate) type complex. © 2008 Published by Elsevier B.V.

1. Introduction Research on design and synthesis of biomimetic simple synthetic chelators received considerable interest in the recent years due to their similar selective and strong binding efficiency like the natural chelators performed for a specific metal ion in presence of the other metal ions in the living systems. Such synthetic molecules are also known to explain the essential structural features that are responsible for the performance of natural compounds and can be implemented as metal ions sequestration. One of the most studied biomolecules for designing biomimic synthetic chelators for iron(III) is enterobactin (Fig. 1), which produced and excreted by bacteria in iron-deficient media in order to bind and assimilate extracellular iron [1,2]. It contains three catechol groups (the binding units, domain III) appended to a tripodal cyclic l-serine (domain I) through amide linkage (domain II). Enterobactin is found to be the best iron-chelating agent with highest formation constant (log K = 52) [3,4] and its efficiency as Fe(III) ion scavenger and carrier has stimulated the synthesis of many analogues containing three catechol units in tripod with respect to their medicinal use [5,6] and for elucidation of biological process [7,8].

∗ Corresponding author. Tel.: +91 1672 284840; fax: +91 1672 284840. E-mail address: [email protected] (B.K. Kanungo). 1386-1425/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.saa.2008.05.014

Since, catechol is the binding units in enterobactin, the two parts: domains I and II can be modified to design new biomimic molecules. However, most of the synthetic analogous were designed previously by replacing the cyclic l-serine unit with a suitable tripodal amine [3,4,9–14], but very few attempts were made to see the effect due to change in amide linkage [15]. It is well known that amide group is a very weak acid [16,17] and shows very nil coordination behaviour towards transition metal ions [18,19]. So, altering the amide linkage with some other functional having potential to coordinate transition metal ions may provide a system of different electronic properties with compartments for encapsulating the metal ions. In search of such ligands, a novel tripodal ligand, cis,cis-1,3,5-tris[(2,3dihydroxy-benzylidene)aminomethyl]cyclohexane (TDBAC) was introduced (Fig. 1), where both domains I and II of enterobactin has been replaced with a tripodal amine, cis,cis1,3,5-tris(aminomethyl)cyclohexane and imine (–N C–) linkage, respectively. The imine nitrogens are more acidic than the catecholic oxygens and also known to formed complexes with a number of tripositive transition metal ions [20,21]. Therefore, it may be expected that ligand TDBAC is resulting two N3 O3 -donors and O3 O3 -donors compartments for the encapsulation of metal ion. Keeping the above facts in view, present investigation was carried out to see the effect due to introduction of imine group in place of the amide linkage in enterobactin analogues and to explore the possible coordination modes. The interaction between ligand TDBAC with iron(III) was studied potentiometrically and

B.K. Kanungo et al. / Spectrochimica Acta Part A 71 (2008) 1452–1460

1453

Fig. 1. Structural diagram of enterobactin with one arm of the ligand emphasized (domains of significance for receptor recognition are: (I) ligand backbone; (II) amide linkage; (III) catecholate binding units) and structure of ligand TDBAC (L).

spectrophotometrically. The stability constants of the complexes depicted in solution as well as the coordination modes of ligand were explained.

The synthesis and characterization of ligand, cis,cis-1,3,5tris[(2,3-dihydroxy benzylidene)aminomethyl]cyclohexane (TDBAC, L) were reported in our previous communication [22]. All solvents and chemicals required for the experimental work were obtained from Merck and used directly. All solutions were prepared prior to the experiments in double distilled deoxygenated water. KOH solution of 0.1 M was prepared and standardized against potassium hydrogen phthalate. HCl solution (0.1 M) was prepared and standardized against standard KOH. Stock solutions of 0.01 M ligand, standardized [23] 0.01 M metal ion (anhydrous FeCl3 ) and 1 M KCl were also prepared in deoxygenated water.

spectrophotometric titrations are similar to the potentiometric titration, except a dilute solution of ligand (5 × 10−5 M) with HCl (5 × 10−4 M) and metal ion (5 × 10−5 M) was titrated with 0.1 M KOH at an ionic strength of 0.1 M KCl and 25 ± 1 ◦ C in aqueous medium. After each adjustment of pH, an aliquot was removed and electronic spectra were recorded on an Agilent-8453 diode array UV–vis spectrophotometer. The equilibrium constants from the spectral data were calculated using a non-linear least-square fitting program, pHAb [28]. The metal–ligand complex formation by continuous variation method was investigated in ethanolic medium. All the solutions taken for experiments were prepared by dilution from the stock solution of ligand (0.001 M). For determination of formation constants of TDBAC with Fe(III), the spectra were recorded by keeping the ligand concentration fixed (1 × 10−5 M) with increasing the metal ions concentration from 0 to 3 × 10−5 M. The ratio between ligand to metal was maintained from 1:0 to 1:3. The formation constants were calculated by global fitting of the whole spectral data using a non-linear least-square fitting program, SPECFIT [29].

2.2. Titration procedures

2.3. Computational methods

In the potentiometric titrations, the observed pH was measured as −log [H+ ] using a Thermo Orion 720A+ pH meter equipped with a combined glass electrode calibrated with standard buffer solutions [24]. The potentiometric titrations were carried out at 25 ± 1 ◦ C maintained from a double wall glass jacketed titration cell connected to a constant temperature circulatory bath. The ionic strength was maintained to 0.1 M KCl with the addition of appropriate amount of 1 M KCl solution. The final concentration of 1 × 10−3 M ligand and 1 × 10−3 M metal ion solution was maintained for both titrations the with metal–ligand ratio 1:0::L:M and 1:1×::L:M. The titrations were carried out at the pH range of 2–11 and sufficient time was given for the attainment of equilibrium to give a stable pH reading. The equilibrium constants from potentiometry were calculated using computer program Hyperquad 2000 [25]. Spectrophotometric titrations were carried out through two methods: (i) by varying the pH in aqueous medium and (ii) by applying continuous variation method [26,27] in ethanolic medium. In case pH metric, the apparatus and method used for

All calculations were carried out on a Pentium IV 3.0 GHz machine on windows 2000 environment using the computer program CAChe Version 6.1.1 [30]. The initial geometry of TDBAC leading to minimum strain energy was achieved through molecular mechanics using MM3 force field followed by semi-empirical PM3 self-consistent fields (SCF) method, at the restricted HartreeFock (RHF) level. The geometry optimizations were obtained by the application of the steepest descent method followed by the Polak–Ribiere method with convergence limit of 0.0001 kcal/mol and RMS gradient of 0.001 kcal/mol. The electronic spectra of the various protonated and deprotonated species of TDBAC were calculated through semi-empirical methods applying ZINDO Hamiltonian. The various possible initial 3D structure related to the geometry of each metal complex was drawn through constrain geometry and model builder option of the computer program, and related minimum strain energy was calculated. The geometry of the complex was energetically optimized through molecular mechanics by applying MM3 force field and then the least strain structure was re-optimized through semi-empirical PM3 method.

2. Experimental 2.1. Materials

1454

B.K. Kanungo et al. / Spectrochimica Acta Part A 71 (2008) 1452–1460

Fig. 2. Potentiometric titration curves of TDBAC (L) in absence and presence of Fe(III) in 1:1 ligand–metal molar ratio, where ‘a’ is moles of base added per mole of ligand present.

3. Results and discussion 3.1. Ligand protonation constants The ligand L (TDBAC), was insoluble in water; appropriate amount of standard 0.1 M HCl was added to make it water soluble by converting into its hydrochloride salt. In the potentiometric method, the acidified ligand solution was titrated against standard KOH at an ionic strength of 0.1 M KCl and 25 ± 1 ◦ C in aqueous medium. The titration curve of the ligand is shown in Fig. 2. The solid lines with symbols represent equilibrium points collected when no solid phase was present in solution, while dotted lines represent points collected when turbidity or precipitation appeared. The equilibrium points collected before a = 0, where ‘a’ is moles of base added per mole of ligand present, is due to the neutralization of excess acid present in the solution. After a = 3, appearance of dotted line indicates the release of three moles of protons from the ligand prior to its precipitation. Analysis of the potentiometric titration curve using the program Hyperquad 2000 gave best fit for three protonation constants (Table 1) for the ligand as proposed by Eq. (1) (charges are omitted for clarity): LHn−1 + H  LHn ,

K1n =

[LHn ] [LHn−1 ][H]

(1)

The ligand TDBAC is expected to undergo deprotonation from its fully protonated form upon addition of base possibly from the three tertiary imine nitrogen and six catecholic oxygen atoms to give total nine protonation constants. However, only three constants were calculated in the present investigation. It has been reported that, the imine nitrogens of Schiff bases with aromatic diols and triols show high acidity with very low log protonation constants

Fig. 3. Experimental electronic spectra of TDBAC (L) as a function of pH (4.35–8.01) during a spectrophotometric titration: [L] = 0.05 mM; [KCl] = 0.1 M and T = 25 ± 1 ◦ C.

lie between 1.53 and 2.84 [31] because substituted phenyl groups attached to the imine nitrogen inductively reduce the electron density of imine nitrogen. Again, due to very low acidity of catechol (pKa1 = 9.13, pKa2 = 13.00) [32] subunits, only one of the two acidic functions of catechol at ortho can be deprotonated within in the present experimental conditions adopted here (pH < 11). Thus, the three-protonation constants are due to the release of protons from the catecholic units. In order to supplement the result of the deprotonation process, a pH versus absorbance measurements was carried out. The experimental electronic spectra of the ligand within the pH range 4.35–8.01 are shown in Fig. 3. No appreciable change in the electronic spectra was observed below the experimental pH < 4.35 and the spectra overlapped with each other, but after pH > ∼8.01, solution became turbid. The state of equilibrium between protonated and deprotonated ligand was examined from the spectral changes, shifting of peaks towards longer wavelength upon increase in pH, and formation of isosbestic points. Inclusion of whole range of spectral data for calculation using non-linear least square fitting program pHAb gave best fit for three species whose protonation constants agreed well with the potentiometric results (Table 1). The predicted electronic spectra of the various species whose protonation constants have been determined and the protonated ligands are shown in Fig. 4. The speciation curve of the ligand is shown in Fig. 5. It is clear from the curve that the first deprotonation started from pH ∼4.5 with the formation of LH2 from the fully protonated form LH3 . Subsequently, second and third protons were released from LH2 and LH in step from pH ∼6 to ∼7, respectively. At low pH (≤ ∼4.35), the solution was colourless and exhibited three peaks with maxima at 214, 265 and 345 nm, respectively. The first and the second peaks were attributed to ␲ → ␲* transitions

Table 1 Protonation (log K) and formation constants (log ˇ) of TDBAC at 25 ± 1 ◦ C and  = 0.1 M KCl (A = potentiometry and B = spectrophotometry) in aqueous medium Equilibrium

L + H  LH L + 2H  LH2 L + 3H  LH3 a

log K

Equilibrium

A

B

8.11 7.43 6.93

8.13 7.45 6.85

Fe + L  FeL Fe + L + H  FeLH Fe + L + 2H  FeLH2 Fe + L + 3H  FeLH3

Values obtained through continuous variation method in ethanolic medium.

log ˇ

Equilibriuma

A

B

20.39 23.83 26.61 29.15

20.47 23.89 26.65 29.17

log ˇa B

Fe + L  FeL 2Fe + L  Fe2 L 3Fe + L  Fe3 L

16.89 6.91 7.11

B.K. Kanungo et al. / Spectrochimica Acta Part A 71 (2008) 1452–1460

1455

Fig. 5. pH-dependent species distribution curves for TDBAC.

Fig. 4. Electronic spectra of the three species as a function of molar absorptivity and wavelength (a) predicted from pHAb using experimental data and (b) calculated by applying semi-empirical PM3/ZINDO methods for TDBAC (L).

associated with the phenyl ring and the azomethine chromophore, respectively, whereas the third band in the spectrum of the ligand was assigned to n → ␲* transition of the imine nitrogen atom in conjugation with the catechol group. On increase in pH progressively, the peak due to n → ␲* transition showed bathochromic shift,

whereas other two peaks shifted hypsochromically. The absorption bands at 345 nm shifted to 387 nm with concomitant rise in absorbance and the colour of the solution changed to yellow. The shifting of ligand peaks towards lower energy on deprotonation of the chelating unit can be explained by two ways: (i) the interaction of enolimine with a hydrogen-bond forming solvent presumably reduces the O–H bond strength and facilitates proton transfer to the nitrogen centre favouring the formation of the ketoamine, which gives characteristic n → ␲* transition at longer wavelength than the corresponding enolimine; (ii) on deprotonation, formation of catecholate ion from catechol stabilizes the ␲* excited state due to charge delocalization and brings the lowest excited state closer to the highest ground state and thus permits a lower energy (longer wavelength) for transition. The predicted spectra (Fig. 4a) for the different species LH3 , LH2 , LH and L of TDBAC obtained in the program pHAb using the log K1 , log K3 and log K3 values were also calculated theoretically through semi-empirical PM3/ZINDO method (Fig. 4b). The predicted spectra of TDBAC gave major variations for the n → ␲* transition; the peak of neutral (LH3 ) at 345 nm shifted bathochromically with formation of anionic (LH2 , LH and L) forms of TDBAC. Similar variations were also observed in the calculated electronic spectra. The calculated n → ␲* transition for LH3 obtained at 300 nm whose HOMO and LUMO diagram are shown in Fig. 6, shifted to 336, 350 and 352 nm for the species LH2 , LH and L, respectively.

Fig. 6. (a) HOMO and (b) LUMO diagram for TDBAC at 300 nm obtained through semi-empirical PM3/ZINDO method.

1456

B.K. Kanungo et al. / Spectrochimica Acta Part A 71 (2008) 1452–1460

3.2. Metal complex formation

resented by Eqs. (6)–(8).

Potentiometric titration curves of ligand TDBAC (L) alone and in presence of Fe(III) in 1:1 metal–ligand molar ratio at  = 0.1 M KCl and 25 ± 1 ◦ C in aqueous medium are shown in Fig. 2. The deviation in the metal–ligand titration curve from the free ligand curve implies the formation of metal complexes. Also, the shape of titration curve qualitatively indicates that the ligand has considerable affinity for the metal ion. The first break in the titration curve obtained at a = 3 (a = moles of base added/moles of ligand), indicates the release of three moles of protons from the ligand upon complexation with the metal ion. When the pH is raised further, turbidity appeared which may be due to insolubility and/or hydrolysis of the complex formed. Keeping in view these preliminary observations, many sets of possible models were tested in the minimization programme and the best-fit model was obtained when formation of species MLH3 , MLH2 , MLH and ML were considered. The overall formation constants (log ˇ) of the species were calculated using Hyperquad 2000, are summarized in Table 1. The equilibrium reactions for the overall formation of the complexes in terms of M, L and H (M = Fe(III) in the present case are given by the following equations (charges are omitted for clarity): M + L  ML,

ˇ110 =

M + L + H  MLH,

[ML] [M][L]

ˇ111 =

(2)

[MLH] [M][L][H]

M + L + 2H  MLH2 ,

ˇ112 =

M + L + 3H  MLH3 ,

ˇ113 =

[MLH2 ] [M][L][H]2 [MLH3 ] [M][L][H]3

(3) (4)

(5)

If MLH3 is assumed to be the first species formed by the reaction of M and LH3 , then the other species MLH2 , MLH and ML are expected to be formed due to dissociation of protons in steps from MLH3 . The stepwise dissociation of the complex MLH3 can be rep-

[MLH3 ]+3  [MLH2 ]+2 + H+ , [MLH2 ]2+

K

3+

[MLH3 ]

=

[MLH2 ]2+ [H+ ] [MLH3 ]3+

= −2.54,

(p K1 = 2.54)

(6)

[MLH2 ]+2  [MLH]+ + H+ , K

[MLH]+ [MLH2 ]

2+

=

[MLH]+ [H+ ] [MLH2 ]2+

[MLH]+  [ML] + H+ ,

= −2.78,

[MLH]+

K [ML]

=

(p K2 = 2.78) [ML][H+ ] [MLH]+

(p K3 = 3.44)

(7)

= −3.44 (8)

It is obvious that the successive log K decrease in stepwise formation of ML from MLH3 . Although water is the most suited solvent for potentiometric titrations, one of the most important drawbacks while investigating the thermodynamic parameters of Schiff bases in aqueous medium is the pH-dependent hydrolysis of imine linkage. Hydrolysis of imine linkage led to the formation of starting organic fragments. In the present study, various models were tried to fit the experimental metal–ligand potentiometric curve by considering the possible hydrolysis products of TDBAC during refinement process; but no such models gave a suitable fit. Further, the precipitate obtained from the metal–ligand potentiometric titration was analyzed (Fig. 7b). The infrared spectrum of precipitate was found to be more or less similar to that of TDBAC (Fig. 7a). The stretching vibrational bands of the precipitate and TDBAC in the region 2700–3700 cm−1 appeared nearly in the same position. All characteristics bands in the region 1700–800 cm−1 of TDBAC were also observed in the precipitate but their position shifted slightly. Such variations in the infrared spectra are usually observed when a ligand is coordinated to a metal ion. No remarkable band indicating the presence of hydrolysis products of

Fig. 7. Infrared spectra (KBr pellet) of (a) TDBAC and (b) precipitate from Fe–TDBAC potentiometric titration.

B.K. Kanungo et al. / Spectrochimica Acta Part A 71 (2008) 1452–1460

TDBAC, i.e., 2,3-dihydroxyaldehyde and TMACH was obtained from the spectrum of the precipitate. Complexation of various tripodal hexadentate N3 O3 -donors Schiff bases with Fe(III) have been studied by Evans and Jakubovic in aqueous medium [33] in which it has been reported that the complexes formed were quite stable in the acidic and slightly basic medium. Keeping an analogy with the above observations, it may be inferred that the formation of precipitate during the titration above pH ∼7.5 is due to the insolubility of the complex formed. Further, the appearance of band in the region 505–516 cm−1 (Fig. 7b) assigned to M–N [34] and the shifting of C N peak of TDBAC from 1632 to 1653 cm−1 in the precipitate indicates the coordination of imine nitrogens. Again, the bands obtained between 1300 and 1200 cm−1 in ligand due to ıC–O shifted towards lower wavenumber in the precipitate indicates the coordination of catecholic oxygens. Also, the presence of ıO–H band in the region 1400–1300 cm−1 in the precipitate indicates the presence of free hydroxyl group. Thus, it may be suggested that the precipitate formed during titration above pH ∼7.5 is a Fe–TDBAC complex and the ligand may be coordinated to Fe(III) through the imine nitrogen and one of the catecholic oxygens of each pendant unit. Spectrophotometric titration was also carried out to study the interaction between Fe(III) and TDBAC (L) and to supplement the potentiometric result. The titration was carried out for a solution containing 1:1 metal:ligand with the ligand concentration [L] = 5 × 10−5 M and the metal ion concentrations [M(III)] = 5 × 10−5 M at a pH range ∼2.5–7.5. After pH ∼7.5, the solution became turbid. Major spectral changes were observed between pH ∼2.5 and 7.5, which may be due to the formation of metal complexes. The experimental electronic spectra of TDBAC–Fe(III) system was shown in Fig. 8, which showed no additional band specific to the metal chelate(s) at higher wavelength due to the low concentration of metal–ligand systems. However, a change in the colour of the solution from violet-to-red (low intensity) was observed as the pH raised from 2.67 to 7.46. Global fitting of whole spectral data using computer program pHAb gave best fit for the model contains MLH3 , MLH2 , MLH and ML, whose calculated formation constants are given alongwith the potentiometric results in Table 1. The result obtained from the spectrometric study agrees well with that obtained from potentiometric method. With increase in pH the ligand peaks of solution obtained at pH 2.67 was shifted bathochromically with the formation of isosbestic

Fig. 8. Experimental electronic spectra of 1:1 Fe(III) to ligand ratio with increasing pH.

1457

Fig. 9. Speciation diagram (% of formation vs. pH) calculated for 1:1 Fe(III)–TDBAC system.

points. Appreciable change was observed for the peak at 345 nm, assigned for n → ␲* transition from the imine nitrogen atom in conjugation with the catechol group, which shifted to 386 nm with concomitant rise in the absorbance. Again, the variations observed in Fig. 8 was more or less similar to that obtained for free ligand (Fig. 3). This provides an indication for the participation of more acidic catecholic hydroxy group at ortho for complexation as catecholate anion whose protonation constants were calculated from the acid–base spectrophotometric titration. From the pH-dependent speciation diagram (Fig. 9), it is seen that the complexation starts below pH 2 with the formation of MLH3 and followed by MLH2 and MLH which are expected to be formed by successive deprotonation of MLH3 . After pH ∼5, only ML species is predominant. Appearance of violet colouration at initial experimental pH indicates the complexation of ligand with iron(III), and possibly due to the coordination of three most acidic imine nitrogen atoms whose protonation constants were not evaluated due to their low acidity within the experimental pH adopted. It has been well established that FeCl3 exist as [Fe(H2 O)6 ]3+ in aqueous solution [35]. Hence, in MLH3 , the three remaining coordination sites of iron may be occupied by water molecules to complete the octahedron. As the pH increases, coordinated water molecules are replaced by the deprotonated catecholic hydroxyl groups of L at ortho in three steps to form MLH2 , MLH and ML. The coordination of catecholate oxygen anion can also be verified from the bathochromic spectral shifting of ligand peaks (Fig. 8). In order to supplement the results obtained in the aqueous medium, the formation of complex in non-aqueous medium was also studied. Complexation of TDBAC was carried out spectrophotometrically in ethanol medium by continuous variation method. The TDBAC concentration was kept constant (1 × 10−5 M) and metal ion concentration was varied from 0 to 3 × 10−5 M. The experimental electronic spectra of TDBAC with Fe(III) is shown in Fig. 10. Analyses of spectra using global fitting of the whole spectral data with SPECFIT gave the best fit for the model which includes ML, M2 L and M3 L. The calculated formation constants are given in Table 1. The possibility of formulation of other type was ruled out, because inclusions of additional species worsen the fit. It was found that the yellow colour of the free ligand solution changed into red upon addition of metal ion. Shifting and disappearance of the ligand peaks and appearance of new peaks upon addition of metal ion indicate the formation of metal complexes.

1458

B.K. Kanungo et al. / Spectrochimica Acta Part A 71 (2008) 1452–1460

Fig. 10. Experimental electronic spectra of TDBAC with (a) [TDBAC] = 1 × 10−5 M in ethanol with increasing [Fe(III)] = 0–3 × 10−5 M.

The ligand showed four peaks at 220, 264, 301 and 427 nm. The first and the second peaks were attributed to ␲ → ␲* transitions associated with the phenyl ring and the azomethine chromophore, respectively. The third and fourth bands in the spectrum of the ligand were assigned to n → ␲* transition form the nitrogen atom of azomethine in conjugation with catechol group. Appearance of an additional peak in the spectra of TDBAC in ethanol as compared to water is an indication of existence of keto and enol tautomers. The peaks at 301 and 427 nm, respectively can be assigned to enolimine and ketoamine tautomeric form of TDBAC, respectively, which was not observed in the aqueous medium. The interaction of enolimine with a hydrogen bond forming solvent (like ethanol) would presumably reduce the O–H bond strength and facilitate proton transfer to the nitrogen centre, which led to the formation of ketoamine tautomer [31]. Semi-empirical PM3/ZINDO method was applied to get the electronic spectra of the two tautomers, enolimine and ketoamine form of TDBAC. The calculated spectra exhibited n → ␲* transition at 300 and 375 nm for enolimine and ketoamine form of TDBAC, respectively, which corroborate well with the assignment made here. Upon successive addition of metal ion into ligand solution and increasing the ligand to metal ratio from 1:0 to 1:3, three different set of changes were identified as shown in Fig. 10b and c. Addition of metal ion to the solution of the free ligand (L:M = 1:0)

resulted appearance of new peak at 525 nm with the disappearance of peak at 427 nm. The variation continued till the ratio L:M::1:1 was attained (Fig. 10b). The appearance of new peak at 525 nm can be attributed to coordination of three most acidic catecholic oxygens, and the disappearance of ligand peak at 427 nm is due to the coordination of three-imine nitrogen. Thus, ligand TDBAC presumably form an encapsulated complex as shown in Fig. 11a. Similar result has also been observed by Hiratani and co-workers in which it has been reported that tripodal catecholate ligand, when titrated with Fe(III) in acetonitrile forms a neutral complex which gives characteristic peak at 528 nm, where three of the most acidic hydroxyl groups in the ligand are deprotonated and the others are protonated [15]. On increase of metal ligand ratio from 1:1 to 1:2, the intensity of the peak at 525 nm decreases with simultaneous reappearance of the ligand peak at 427 nm (Fig. 10c). The decrease in intensity of peak at 525 nm and reappearance of the ligand peak at 427 nm suggests that some of the –OH groups of the ligand which was coordinated to the metal ion in a 1:1 (ML) complex, became free on addition of excess metal ion providing an electronic environment similar to that of the free ligand. The formation of M2 L type of complex then be suggested as (b) in Fig. 11, in which two arms of TDBAC bonded to two moles of metal ions and one arm remains free that gives the characteristic peak of the ligand (Fig. 10a). Very little variation was observed, when concentration of metal ion was

B.K. Kanungo et al. / Spectrochimica Acta Part A 71 (2008) 1452–1460

1459

Fig. 11. Probable structure of the complexes formed in ethanolic medium.

raised thrice compared to the ligand (Fig. 10d): intensity of peak at 427 nm was decreased while that at 525 nm was increased. The structure for the M3 L is suggested in Fig. 11c. From the pH potentiometric titrations, infrared spectral study and spectrophotometric measurements made both in aqueous and ethanolic medium, it may be inferred that coordination of ligand TDBAC occurs through three imine nitrogens and three-catecholic oxygen at ortho to give a uncharged hexa-coordinated complex. Molecular modeling calculations were carried out for ML to obtain its possible structure. Ligand TDBAC, which was isolated in most stable equatorial form, may encapsulate the metal ion without any change in its conformation, by changing its conformation into

axial or mixed equatorial–axial. All theoretically possible geometries were optimized through molecular mechanics using MM3 force field. The least steric energy was obtained for the structure, where the ligand preferred to flip it conformation from equatorial to axial in order to encapsulate the metal ion in form of a distorted octahedron. The least strain structure was re-optimized through semi-empirical method by applying PM3 Hamiltonian (Fig. 12). Some important calculated bond length for Fe (TDBAC) are: Fe–O, ˚ whereas bond angles (in degree) are: 2.034 A˚ and Fe–N, 2.156 A, N–Fe–N, 93.81; O–Fe–O, 89.38 and N–Fe–O, 174.47. 4. Conclusion The ligand TDBAC, a biomimetic analog of enterobactin, formed monomeric complexes FeLH3 , FeLH2 , FeLH and FeL in aqueous solution, whereas in ethanol, it formed complexes of the type FeL, Fe2 L and Fe3 L. Unlike a tris(catecholate) complex of enterobactin, the ligand coordinated to iron(III) through three imine nitrogens and three catecholic oxygens at ortho to give a uncharged tris(iminocatecholate) complex. The ligand preferred to flip its conformation from equatorial to axial to encapsulate iron(III) to form a distorted octahedal Fe(III)–TDBAC, thereby reducing its stability as compared to Fe(III)–enterobactin complex. Nevertheless, the potential of TDBAC to form monomeric complex and its resistance towards the hydrolysis of imine linkages will be useful to implement TDBAC as a chelator for iron(III). References

Fig. 12. Optimized structure of Fe (TDBAC) obtained through semi-empirical calculation using PM3 Hamiltonian.

[1] J.R. Telford, K.N. Raymond, Comprehensive Supramolecular Chemistry, Elsevier Science Ltd., Oxford, 1996. [2] G. Winkelmann, CRC Handbook of Microbial Iron Chelates, CRC Press, Boca Raton, 1991. [3] W.R. Harris, C.J. Carrano, K.N. Raymond, J. Am. Chem. Soc. 101 (1979) 2213. [4] W.R. Harris, C.J. Carrano, S.R. Cooper, S.R. Sofen, A.E. Avdeef, J.V. McArdle, K.N. Raymond, J. Am. Chem. Soc. 101 (1979) 6097. [5] J. Neilands, J. Biol. Chem. 270 (1997) 26723.

1460

B.K. Kanungo et al. / Spectrochimica Acta Part A 71 (2008) 1452–1460

[6] C. Caris, P. Baret, C. Beguin, G. Serratrice, J.-L. Pierre, J.-P. Laulhere, J. Biochem. 312 (1995) 879. [7] A. Shanzer, J. Libman, in: A. Sigel, H. Sigel, M. Dekker, Iron Transport and Storage in Microorganisms, Plants and Animals, Marcel Dekker, New York, 1998, p. 239 (ch. 8). [8] K.N. Raymond, Coord. Chem. Rev. 105 (1990) 135. [9] A. Shanzer, J. Libman, S. Lifson, C.E. Felder, J. Am. Chem. Soc. 108 (1986) 7609. [10] N. Cheraiti, M.E. Brik, G. Kunesch, A. Gaudemer, J. Organomet. Chem. 575 (1999) 149. [11] N. Cheraiti, M.E. Brik, A. Gaudemer, G. Kunesh, Bioorg. Med. Chem. Lett. 9 (1999) 781. [12] N. Cheraiti, M.E. Brik, L. Bricard, B. Keita, L. Nadjo, A. Gaudemer, Bioorg. Med. Chem. Lett. 9 (1999) 2309. [13] S.J. Rodgers, C.W. Lee, C.Y. Ng, K.N. Raymond, Inorg. Chem. 26 (1987) 1622. [14] E.J. Corey, S.D. Hurt, Tetrahedron Lett. 45 (1977) 3923. [15] M. Hayashi, K. Hiratani, S.I. Kina, M. Ishii, K. Saigo, Tetrahedron Lett. 39 (1998) 6211. [16] H. Sigel, R.B. Martin, Chem. Rev. 82 (1982) 285. [17] R.B. Martin, J. Chem. Soc., Chem. Commun. (1972) 793. [18] L. Fabbrizzi, A. Perotti, A. Poggi, Inorg. Chem. 22 (1983) 1411. [19] E. Kimura, T. Koike, R. Machida, R. Nagai, M. Kodama, Inorg. Chem. 23 (1984) 4181.

[20] N.W. Alcock, D.F. Cook, E.D. McKenzie, J.M. Worthington, Inorg. Chim. Acta 38 (1980) 107. [21] D.F. Cook, D. Cummins, E.D. McKenzie, J. Chem. Soc., Dalton Trans. (1976) 1369. [22] S.K. Sahoo, M. Baral, B.K. Kanungo, Polyhedron 25 (2006) 722. [23] D.A. Scoog, D.M. West, Fundamentals of Analytical Chemistry, 2nd ed., Holt, Rinehart and Winston, New York, 1969, p. 690. [24] A.E. Martell, R.J. Motekaitis, The Determination and Use of Stability Constants, VCH Publishers, New York, 1992. [25] P. Gans, A. Sabatini, A. Vacca, Talanta 43 (1996) 1739. [26] Y. Shibata, B. Inoue, Y. Nakatuka, Nihon Kagakukaishi 42 (1921) 983. [27] R. Tsuchida, Bull. Chem. Soc. Japan 10 (1935) 27. [28] P. Gans, A. Sabatini, A. Vacca, Ann. Chim. 89 (1999) 45. [29] H. Gampp, M. Maeder, C.J. Meyer, A.D. Zuberbuhler, Talanta 32 (1985) 95. [30] User Guide Manual for CAChe Version 6.1.1, Fujitsu Limited, 2003. [31] A. Golcu, M. Tumer, H. Demirelli, R.A. Wheatley, Inorg. Chim. Acta 358 (2005) 1785. [32] A.E. Martell, R.M. Smith, Critical Stability Constants, vols. 1–6, Plenum, New York, 1974–1989. [33] D.F. Evans, D.A. Jakubovic, J. Chem. Soc., Dalton Trans. (1988) 2927. [34] M. Dolaz, M. Tumer, M. Digrak, Trans. Met. Chem. 29 (2004) 528. [35] F.A. Cotton, G. Wilkinson, C.A. Murillo, M. Bochmann, Advanced Inorganic Chemistry, 6th ed., John Wiley & Sons, Inc., Singapore, 1999.