Transition Met Chem (2008) 33:39–53 DOI 10.1007/s11243-007-9012-4
A new octahedral cobalt(III) complex of (1,8)-bis (2-hydroxybenzamido)-3,6-diazaoctane incorporating phenolate-amide-amine coordination: synthesis, X-ray crystal structure, ligand substitution and redox activity with sulfur(IV) and L-ascorbic acid Suprava Nayak Æ Anadi C. Dash Æ Gautam K. Lahiri Received: 14 July 2007 / Accepted: 9 August 2007 / Published online: 7 December 2007 Ó Springer Science+Business Media B.V. 2007
Abstract The octahedral complex, [CoIII(HL)]9H2O (H4L = (1,8)-bis(2-hydroxybenzamido)-3,6-diazaoctane) incorporating bis carboxamido-N-, bis sec-NH, phenolate, and phenol coordination has been synthesized and characterized by analytical, NMR (1H, 13C), e.s.i.-Mass, UV– vis, i.r., and Raman spectroscopy. The formation of the complex has also been confirmed by its single crystal X-ray structure. The cyclic voltammetry of the sample in DMF ([TEAP] = 0.1 mol dm-3, TEAP = tetraethylammonium perchlorate) displayed irreversible redox processes, [CoIII(HL)] ? [CoIV(HL)]+ and [CoIII(HL)] ? [CoII(HL)]- at 0.41 and -1.09 V (versus SCE), respectively. A slow and H+ mediated isomerisation was observed for the protonated complex, [CoIII(H2L)]+ (pK = 3.5, 25 °C, I = 0.5 mol dm-3). H2Asc was an efficient reductant for the complex and the reaction involved outer sphere mechanism; the propensity of different species for intra molecular reduction followed the sequence: [{[CoIII(HL)],(H2Asc)}–H]- \\\ {[CoIII(H2L)],(H2Asc)}+ \ {[CoIII(HL)],(H2Asc)}. A low value (ca. 3.7 9 10-10 dm3 mol-1 s-1, 25 °C, I = 0.5 mol dm-3) for the self exchange rate constant of the couple [CoIII(HL)]/[CoII(HL)]- indicated that the ligand HL3- with amido (N-) donor offers substantial III stability to the CoIII state. HSO3 and [Co (HL)] formed an outer sphere complex {[CoIII(HL)],(HSO3 )}, which was Electronic supplementary material The online version of this article (doi:10.1007/s11243-007-9012-4) contains supplementary material, which is available to authorized users. S. Nayak A. C. Dash (&) Department of Chemistry, Utkal University, Bhubaneswar 751 004, India e-mail:
[email protected] G. K. Lahiri Department of Chemistry, Indian Institute of Technology, Powai, Mumbai 400 076, India
slowly transformed to an inner sphere S-bonded sulfito complex, [CoIII(H2L)(HSO3)] and the latter was inert to reduction by external sulfite but underwent intramolecular SIV ? CoIII electron transfer very slowly.
Introduction Multidentate ligands endowed with carboxamido functions have proved to be excellent complexing agents for several metal ions such as FeIII/IV [1], MnIII/IV [2], CrIII [3], CoIII/IV [4], VIVO [5], OsIV [6], and CuII [7]. The amide function is ambidentate in nature as it can coordinate to a metal centre via N, O or both [1, 8]. Of particular interest is the stabilization of higher oxidation state metal ions via N-coordination of the deprotonated carboxamides. Collins and coworkers [4] have reported some carboxamido complexes of CoIII and CoIV with octahedral and square planar geometry derived from polyanionic carboxamido ligands, which can be considered as novel inorganic oxidizing agents. There has been a great upsurge of interest in recent years on the study of the carboxamido complexes of metal ions due to the prevalence of the carboxamido function as ligational site in several metalloenzymes such as nitrile hydratase (NHase) [9, 10], superoxide dismutase (Ni-SOD) [11, 12], antitumor drug bleomycins [13], and the iron centres in the P cluster of nitrogenase [14]. The N-ligation of the deprotonated carboxamide interestingly stabilizes the higher oxidation state of the metal ion, more so when more than one such moieties are included in the framework of a polydentate ligand. In consequence, the resulting complexes become less prone to reduction but exhibit Lewis acid characterstics in promoting hydrolytic
123
40
reactions. As a matter of fact, FeIII in FeIII-NHase is redoxresistant but catalyses the hydrolysis of organic nitriles. We have been interested in the study of the binding mode of the carboxamide in a polydentate ligand frame and how it modulates the stability and reactivity of the corresponding metal complexes with regard to redox, ligand substitution and catalytic efficiency in hydrolysis reactions. In continuation of our earlier work [1f, 15, 16], we now report the synthesis, characterization, substitution, and redox reactions of a CoIII complex of (1,8)-bis(2-hydroxybenzamido)–3,6-diazaoctane (H4L) using SIV and L-ascorbic acid (H2Asc) as potential secondary ligands. To the best of our knowledge, the synthesis, characterization, and reactivity studies of this complex have not been reported earlier.
Experimental Materials and methods This hexadentate ligand (H4L) was received from our earlier work [1f]. H2O was doubly distilled; the second distillation was made from alkaline KMnO4 using an all-glass distillation apparatus. Analar grade chemicals (E. Merck) were used for the kinetic study. Sodium metabisulfite (Na2S2O5) was the source of SIV. NaClO4 used to adjust ionic strength, was prepared by neutralizing HClO4 with NaOH, both of which were previously standardized. The pH of the stock NaClO4 solution (1 mol dm-3) was adjusted to 6. The resin Dowex 50 W X 8 (Na+) used for ion exchange experiments was pre-treated with alkaline H2O2. The cation exchange resin was converted into H+ form with HClO4. Cobalt was estimated iodometrically after converting a known amount of the sample to CoII sulfate by pyrosulfate fusion as described earlier [17].
Preparation of [CoIII(HL)], 9H2O MeOH + H2O (50% v/v) solution of CoCl2, 6H2O (2.38 g, 10 mmol), the (H4L) (3.86 g, 10 mmol) and LiOHH2O (1.26 g, 30 mmol) was stirred and aerated for 24 h at ambient temperature when a uniform deep purple solution developed. This solution was concentrated at 50 °C to a small volume and set aside at room temperature for a few days to yield a deep purple solid, which was filtered on a glass sintered G-2 funnel. It was re-crystallized from water and stored over silica gel being protected from light [[Co(C20N4H23O4)], 9H2O (F. Wt. 603.9) calcd.: C, 39.7; H, 6.7; N, 9.3; Co, 9.7%. Found: C, 39.6; H, 6.2; N, 9.2;
123
Transition Met Chem (2008) 33:39–53
Co, 9.5%]. The complex is highly soluble in water and the pH of its 2 9 10-3 mol dm-3 aqueous solution was 5.6.
Instrumentation The rapid scan u.v.-vis spectra were run on a HI-Tech (UK) SF 61 MX stopped flow spectrophotometer using diode array (MG 6000) attachment. The sampling unit was mounted inside a thermostatted bath compartment maintained at 25.0 ± 0.2 °C by circulating water from a water bath at the same temperature. The wavelength range was 299–699 nm. The concentration of the complex was fixed at 1.0 9 10-4 mol dm-3. The cell path length was 10 mm and no delay time was introduced at the start of the scan while the dead time of the instrument was 3 ms. In all 96 scans (for *6 s) were recorded with Dt = 60 ms between two successive scans for any composition. A conventional u.v.-vis spectrophotometer (Cecil CE 7200, UK) with peltier controlled thermostatted cell block housing a pair of 10 mm quartz cells was used for all other spectral measurements and kinetic study. The i. r. spectra were recorded on a Nicolet spectrophotometer with samples prepared as KBr pellets. Raman spectra were recorded in dispersive mode (10 B m (cm-1) B 3200) on a Renishaw Raman spectrometer (model in VIA Reflex). The light source was Argon-ion laser (514 nm excitation) and data analysis was done by the WIRE 2 spectral acquisition wizard. The mass spectral measurements (100 B m/z B 1000) were done at IIT Madras on a Micromass Q-TOF mass spectrometer adopting electron spray ionization mode (esi); sample solution was prepared in water. 1H and 13C NMR spectra were recorded in D2O (22 °C) with a Jeol AL 400 FT NMR spectrometer operating at 399.65 MHz; TMS was used as the external reference. Some 1H NMR measurements were also made using a Bruker AVANCE 300 NMR spectrometer, D2O was the solvent ([complex]T = 3.07 9 10-3 mol dm-3). The cyclic voltammetric measurements were carried out in DMF using a PAR model 273A electrochemistry system. Platinum wire working and auxiliary electrodes and an aqueous saturated calomel reference electrode (SCE) were used in a threeelectrode configuration. The supporting electrolyte was [NEt4] ClO4 (TEAP, 0.1mol dm-3) and the complex concentration was *10-3 mol dm-3; a N2 atmosphere was maintained and no corrections were applied for junction potentials. The elemental analysis (C, H, N) was carried out with a Perkin Elmer 240C element analyzer. The pH measurements were made with a Systronics (India) pH meter model 335 with a combined glass-Ag/AgCl, Cl(3 mol dm-3 NaCl) electrode CL 51. Standard buffers of pH 4.01 and 9.2 were used to calibrate the instrument. The
Transition Met Chem (2008) 33:39–53
41
pH data (i.e., meter readings) at pH \ 7 were converted to -log [H+] by a calibration curve as described earlier [18].
Kinetics The reactions of the complex were monitored under pseudo-first order conditions as follows. The ionic strength adjusted solution of the reaction mixture was thermally equilibrated in a 25 cm3 measuring flask and then a known volume (1 cm3) of the stock complex solution was added and volume was made up. This reaction mixture was quickly (ca. B 40s) transferred to the cell placed in the thermostated cell block of the spectrophotometer. The progress of the reaction was monitored at a pre-selected wavelength (see Tables 4–6) by recording absorbance with time against solvent blank. The [complex]T was varied as (1–3) 9 10-4 mol dm-3. The wavelength dependence was also checked for some runs by rate measurements at 540 nm ([complex]T = 6 9 10-4 mol dm-3). It may be noted that the redox reaction by SIV was extremely slow which did not interfere with the relatively faster initially observed nonredox reaction. The redox reaction with ascorbic acid ({H2Asc]T/[Complex]T C 50) was also mono phasic as for the reactions of the complex in the presence of acid and SIV (for the early part in the latter case); all absorbance (At)–time(t, s) data fitted well to the pseudofirst order kinetic equation: At = C1 exp(-kobst) + A0 where A0 denotes At for long time intervals to get the best fit. The data fitting involved 45–100 data points in most
cases (ca. 4t1/2). The reaction of the complex in acid medium (i.e., in absence of [SIV] and L-ascorbic acid) approached equilibrium rapidly at low [H+]T showing a decreasing trend 0 0 of (A0-A ) with the decrease of [H+] (i.e., A0-A = 0.079 and + -3 0.220 for [H ]T = 0.002 and 0.1 mol dm , respectively for [complex]T = 3 9 10-4 mol dm-3). Representative plots are shown in Fig. S1(a–c) (see supplementary material); r(kobs) from individual runs was \ ± 2% of kobs. All measurements were made at constant ionic strength, I = 0.5 mol dm-3. All calculations were made on a PC using linear and nonlinear least squares programs.
X-ray crystallography Single crystals of the complex were grown by slow evaporation of its dilute aqueous solution at room temperature in the dark. A violet colored block shaped crystal 0.25 9 0.20 9 0.15 mm in size was used for X-ray data collection at 293(2) K by OXFORD XCALIBUR-S CCD single crystal X-ray diffractometer equipped with MoKa radiation. A total of 13,410 reflections (2.9587 B H B 31.8987°) were collected. Crystal data and data collection parameters are collected in Table 1. The structure was solved and refined by full-matrix least squares on F2 using SHELX-97 (SHELXTL) [19]. The absorption corrections were done by multi scan. All the data were corrected for Lorentz, polarization effect and the nonhydrogen atoms were refined anisotropically. All hydrogen atoms except the NH’s were included in the refinement process as per the
Table 1 Crystallographic data for the complex Empirical formula
CoC20N4H30O9.5
Abs coeff (l)
0.780 mm-1
fw T
F(000) H range
1124 2.96–25.00
k
537.4 293 (2) K ´˚ 0.71073 A
No. of reflns/restraints/parameters
4023/0/352
Crystal system
Monoclinic
Unique reflns
3251 (R(int) = 0.0725)
Space group
P21/c
Limiting indices
-9 B h B 8
Unit cell dimensions
´˚ a = 7.6237(13) A ˚´ b = 18.957 (3) A ´ ˚ c = 16.745 (4) A
-21 B k B 22, -19 B l B 19
a = c = 90° V
b = 100.673(17)° ´˚ 3 2378.3 (8) A
Goodness of fit on F2
1.369
Z
4
R indices [I [ 2r(I)] (3251 data)
Ra1 = 0.1442 Rbw1 = 0.2065
qcalcd
1.481 g cm
-3
R indices (all data)
Ra2 = 0.1818 Rbw2 = 0.2190
a b
P P R ¼ h kF o jFc k= jFc j i 1=2 Rw ¼ wðjFo Fc jÞ2 =wjFo j2
w-1 = [r2(F2o) + (0.0000P)2 + 30.27P]; P = (F2o + 2F2c )/3
123
42
riding model. The four hydrogen atoms of two water molecules and one hydrogen atom associated with each of the three water molecules could be detected.
Results and discussion Characterization of the complex Crystal data and data collection parameters are collected in Table 1. The X-ray structure of the complex is shown in Fig. 1. Although the analytical data of the bulk sample were consistent with 9H2O as the water of crystallization, only 5.5H2O could be detected by X-ray. This discrepancy might be due to the poor quality of the crystal. However, in spite of our best efforts, we failed to generate suitable single crystals better than the one used for X-ray analysis.
Bond parameters and bonding in [CoIII(HL)] Selected bond distances and bond angles are listed in Table 2. The metal ion is octahedrally coordinated by two sec-NH, two amido–N (deprotonated) and two phenolic oxygens one of which is deprotonated. The two transN-amido, O-(phenol), and a sec-NH bind the metal ion in the equatorial plane while, the O-(phenoxide) and the other sec-NH occupy the trans-axial positions. The two Co– N(amido) bond distances {Co–N1: 1.896(8), Co–N4: 1.903(7)} do not differ from each other and fall in the range of values observed for several Co–N(amido) complexes [20(a–c), 4(a, c)]. These are, however, shorter than the Co–
Transition Met Chem (2008) 33:39–53 ˚ and bond angles (°)a Table 2 Selected bond lengths (A Distances Co–O1
1.899(7)
Co–O3
1.887(7)
Co–N1
1.896(8)
Co–N4
1.903(7)
Co–N2
1.949(9)
Co–N3
1.937(9)
C1–O1
1.322 (11)
C7–O2
1.274(11)
N1–Co–N2
81.1(4)
N3–Co–N4
82.4(4)
N4–Co–O1
91.6(3)
O1–Co–N1
91.6(3)
N1–Co–N3
96.6(4)
N2–Co–O3
94.5(4)
O1–Co–N2
172.0(3)
O3–Co–N3
174.8(4)
N1–Co–N4
176.6(4)
C7–N1–Co
128.3(7)
N2–Co–N4
95.7(4)
C8–N1–Co
114.7(6)
Angles
a
Numbers in parentheses are standard deviations in the last significant digit
NH (amine) bond distances {Co–N2: 1.949 (9), Co–N3: ˚ . The observed Co–NH(amine) bond 1.937 (9)} by 0.04 A distances agree with those reported for CoIII amine complexes [20c, 21]. This lends support to the idea that the anionic amido-N is a stronger r-donor than the uncharged amine [20c]. The Co–O1 (phenol) {1.899(7)} and Co–O3 (phenolate){1.887(7)} distances are indistinguishable. The bond angles around cobalt reflect a small distortion from a regular octahedron. The maximum distortion is indicated by the bond angles N2–Co–N1 {81.1(4)}, N3–Co–N4 {82.4(4)}, N3–Co–O3{174.8(4)}, N1–Co–N4 {176.6(4)}, N2–Co–O1 {172.0(3)}.
Spectroscopic properties Mass spectra For collision energy (CE) 3 eV, only one intense peak at m/z 443 (100%) with a satellite at (m + 1)/z 444 (9%) was observed, which corresponded to [Co(H2L)]+ (F. wt. 442.9) {[Co(HL), 9H2O] – 9H2O + H} thus establishing the molecular formula of the complex. Interestingly at collision energy -7 eV, the spectrum displayed peaks at m/z 441(100%) ([Co(L)]-, F. wt. 440.9), 442 (16%), 882{38% for a dimer, [(Co(L)2]-, F. wt. 881.8) thus indicating that electron impact at low energy range can result in CoIII–L bond opening and dimerisation of the complex.
U.v.-vis spectra
Fig. 1 ORTEP diagram of [Co(HL)], 6H2O. Ellipsoids are drawn at 50% probability
123
Strikingly, the repetitive spectral scans for a given composition from rapid scan measurements overlapped with each other. Typical scans are presented in Fig. S2(a–c) (see supplementary material). The spectral parameters are
Transition Met Chem (2008) 33:39–53
43
Table 3 Spectral parameters from rapid scan spectral measurements k (nm) (e, dm3 mol-1 cm-1) Max.
Min.
328 (4.73 9 103)a; 312 (4.40 9 103)b 2 a
1028 cm-1 also appeared in the i. r. spectrum. The Raman bands at 485.6, 379 cm-1 are assignable to mstr.(CoIII–O) and mstr (CoIII–N) while the one at 152.5 cm-1 is due to the N–CoIII–N deformation [23].
477 (1.91 9 102)a 3 b
423 sh (4.87 9 10 ) ; 362 sh (1.0 9 10 ) 602 (3.41 9 102)a; 574 (3.69 9 102)b
464 (1.31 9 102)b
NMR
3 c
311 (4.86 9 10 )
379 sh (6.26 9 102)c 585 (3.10 9 102)c a
[HClO4]T (mol dm-3) = 0.1 (pH = 1)
b
[H2Asc]T (mol dm-3) = 0.03 (pH = 4.01)
c
[SIV]T = 0.025 mol dm-3 (pH = 6.60)
473 (0.99 9 102)c
collected in Table 3. These confirmed that there was no net reaction of the complex with SIV and ascorbic acid during 0 B t/s B 6 and the complex is stable to decomposition in moderately acidic medium in this time scale; a small change in the intensity and position of the absorption bands is ascribed to the fast protonation equilibrium of the complex. The u.v.-vis spectra recorded on a conventional spectrophotometer at several pHs within 50 s of mixing the reactants (25 °C) exhibited strong pH dependence in tune with the protonation equilibrium of the complex; kmax around 330 and 540 nm showed a small blue shift with enhancement of intensity with the increase of pH {kmax (nm) (e, dm3 mol-1 cm-1): 339 (6350), 556 (470) ([H+] = 0.05 mol dm-3); 326(6800), 539 (460)(pH 5.95); kmin (nm) (e, dm3 mol-1 cm-1): 304 (3900) ([H+] = 0.05 mol dm-3), 304 (5700) (pH 5.95)}.
I.r and Raman spectra The complex displayed three strong bands at 1595, 1573, and 1526 cm-1 expected for mstr(CO) and mstr (C–N) of the deprotonated amide and mstr(C=C) of the aromatic groups. Similar observations were reported for several amido complexes of CoIII [20(a–c)] and FeIII [1(d, f), 4c]. The bands at 1259 (s) and 1249(sh) cm-1 are ascribed to the OH bending and mstr (C–O) of the coordinated phenol group and those at 774(s) and 764 (s) cm-1 may be assigned to the (1,2) di-substituted aromatic moiety [22]. The bands at 993, 946, 888, and 843 cm-1 (CH2 rocking) are in tune with the unsymmetrical arrangement of the trien (triethylenetetramine) spacer unit as reported for cis-b[Co(trien)Cl2]+ [23]. The mstr.(N–H of sec-amine) and those for H2O were observed at 3166 cm-1 and 3870, 3446, 3301 cm-1, respectively [22, 23]. The Raman spectrum of the complex displayed a band at 1595 cm-1 thus confirming the N-coordinated amido group. The other Raman bands observed at 1545, 1450, 1393, 1326, 1249, 1159,
The signals for 13C were sharp and observed at d (ppm, D2O, no added acid): 166.07, 169.46 (2C for the two amido units,–N–C–O); 130.55, 130.66, 131.87, 122.71, 123.37, 115.66, (12C, aromatic); 48.25, 51.37, 51.46, 52.02 (6C, aliphatic). The sharpness of the signals indicated that the chelate rings did not open up at least during the time of measurement. The two carbons of the amido functions (C7 and C14, see Fig. 1) are distinguishable by 13C NMR. The aromatic carbons essentially fall in to three groups (d ppm: 131, 122, 115) as also the aliphatic carbons (d ppm: 52, 51.4, 48.2). The 1H NMR lines (in D2O, no added acid) appeared at d (ppm): 2.10–2.20, 2.50–2.65, 3.65–3.80 (aliphatic CH2 groups); 2.85–2.91 (sec-NH), 4.50 (Ncoordinated–CO–NH–); 6.50–6.62, 6.75–6.81, 7.01–7.09, and 7.77–7.83 (aromatic protons). The free amide is reported to be an extremely weak acid as compared to phenol (DpK[(–CO(NH)–) –* (phenol)] [ 5 [1(e, f), 24]. It is reasonable to expect this trend to be maintained for the CoIII-bound amide and phenol functions. Interestingly, earlier works from this laboratory and elsewhere have shown that the coordinated phenol in [Co(en)2(Hsal)]+ in which Hsal- (salicylate monoanion) is chelated by the phenol and carboxylate groups is a moderately strong acid {pKOH = 1.0 ± 0.2 (30–40 °C, I = 0.3 mol dm-3) [25]; 1.0 ± 0.3 25 °C [26]}. The pH titration and spectral data (see later) also reveal that the complex retains its identity as [CoIII(HL)] at pH C 5 without undergoing further acid dissociation. It is, therefore, most likely that one of the coordinated amido –N is protonated (in contrast with the coordinated phenolate as in Fig. 1) yielding [CoIII(HL)] in solution as depicted by (I) and the signal at d (ppm) 4.50 is assigned accordingly.
O
H
O C
NH
N
Co
C=O N
O HN
(I) I Solution structure of [CoIII(HL)]
123
44
Transition Met Chem (2008) 33:39–53
This solid state–solution state difference in the location of the proton in [CoIII(HL)] (Fig. 1 versus I) is attributed to the fact that the coordinated amido function in solution is considerably more basic than its phenoxide analogue in the same ligand frame. Hereafter, we consider [CoIII(HL)] for all kinetics and equilibrium studies by its solution state structure (I). The 1H NMR spectra were also run in the presence of acid ([HClO4]T = 0.1, [CoIII(HL)]T = 3.06 9 10-3 mol dm-3, 27 °C) after 600 and 1200 s of the sample preparation. Under these conditions, the complex exists as [CoIII(H2L)]+ (both the amido–N sites are preferably protonated). The signals for CH2 groups, sec-NH and –CO–NH–Co are broadened and intensity varied with the increase of time. The salicylate ring in [Co(en)2(sal)]+/[Co(trien)(sal)]+ and the corresponding CoIII–O (phenol) bond are also reported to be retained under harsh acidic conditions [25, 26]. Hence, the observed time dependence of 1H NMR in acidic medium is attributed to the kinetic consequence of a slow intra molecular structural re-organisation of the complex (see later) and H/D exchange for sec-NH and CO–NH– (both N– coordinated to CoIII).
Cyclic voltammetery The c. v. scans of [CoIII(HL)] (50 mV s-1) in DMF displayed two irreversible responses: E (pa) = 0.41 V and E (pc) = -1.09 V (versus SCE) corresponding to the oxidation of CoIII to CoIV and reduction of CoIII to its CoII analogues, respectively. The possibility of the ligand oxidation by CoIV and high substitutional lability of the CoII [27] may be responsible for the irreversibility of the redox processes.
Fig. 2 pH dependence of the zero time absorbance of [CoIII(HL)]; [complex]T = 1.029 9 10-4, I = 0.5 mol dm-3, 25 °C, k (nm) = 360 (1), 306 (2); pH adjusted with HClO4 (0.30 B pH B 2.97), acetate buffer (3.52 B pH B 6.21, [NaOAc]T = 0.02 mol dm-3), NaOH (6.92 B pH B 12) (ca. pH = 12 for [NaOH]T = 0.01 mol dm-3)
(2.7 ± 0.3) 9 103 at 360 nm and 0.601 ± 0.005, 0.39 ± 0.11, (3.4 ± 0.6) 9 103 at 306 nm, respectively. The protonation equilibrium of the complex was also studied by pHtitration of an acidified solution of the complex with NaOH, the titration of a standard acid solution against the same NaOH served as the reference. Only one inflection was observed and the two curves virtually overlapped at pH [ 7 (see Fig. 3).
Equilibrium measurements Protonation equilibrium The pH dependence of the zero time absorbance (Aobs) of the complex at 360 and 306 (see Fig. 2) is interpreted in terms of the protonation equilibrium (1) III þ K1 Co (HL) + Hþ CoIII (H2 L) ð1Þ for which Aobs ¼ ðA1 þ A2 K1 [Hþ ]Þ=ð1 þ K1 ½Hþ Þ
ð2Þ III
where A1 and A2 denote the absorbances of [Co (HL)]T and [CoIII(H2L)+]T, respectively, when the complex exists exclusively in these forms. The data were fitted to Eq. 2 and yielded A1, A2 and K1 (dm3 mol-1) as 0.185 ± 0.005, 0.438 ± 0.08,
123
Fig. 3 pH versus v, cm3 (NaOH) (=0.1002 mol dm-3) plot; [CoIII(HL)] (=1.016 9 10-3) + [HClO4] (=1.21 9 10-3), I = 0.5 mol dm-3 (1); [HClO4] = 1.21 9 10-3 mol dm-3, I = 0.5 mol dm-3 (2); Vtotal = 40.0 cm-3, 30 °C
Transition Met Chem (2008) 33:39–53
45
Fig. 4 Successive spectral scans of [CoIII(HL)] in acid medium at 25.0 ± 0.1 °C. (1) [Complex]T = 2.0 9 104 + NaClO4 = 1.0 mol dm-3, pH 6.95; (2–8) [complex]T = 2.0 9 10-4 + [HClO4]T = 1.0 mol dm-3; 40 s (2), 360 s (3), 900 s (4), 1,200 s (5), 2,100 s (6), 3,000 s (7), 2.592 9 105 s (8)
The lack of deprotonation of [CoIII(HL)] is a clear indication of the very high pK of the amide proton in I. The H+ consumption by the complex was evident below pH 6 and the data were analyzed considering equilibrium (1). The pH data at 3.05–4.23 yielded K1 = (8.7 ± 0.9) 9 103 dm3 mol-1 (log K1 = 3.94 ± 0.05, 30 °C, I = 0.5 mol dm-3).
Reaction of [CoIII(HL)] in acid medium The successive spectral scans in [HClO4]T = 1.0 mol dm-3 (see Fig. 4) over a prolonged time indicate that the complex undergoes a slow transformation. However, the spectral scans at pH C 5 at which the complex is exclusively speciated as [CoIII(HL] are virtually time independent. A comparison of the scans 1 and 2 is in tune with the fast protonation pre-equilibrium. The isosbestic points (284, 314, 443, and 624 nm) are maintained during the early stage of the reaction with a small red shift of kmax at
560–575 nm and kmin at 310–330, and 474–520 nm. At sufficiently long times (C3 days), however, the isosbestic points at 314 and 624 nm were lost thus indicating a further slow reaction. The kmax around 338 and 560 nm are also retained even after 13 days but with much reduced intensity and a peak intensified around 295 nm characteristic of the free ligand. A preliminary test for CoII was performed as follows. An acidified complex solution ([complex]T = 8.0 9 10-4, [HClO4]T = 0.05, I = 0.5 mol dm-3, 27 °C) was set aside for 48 h and then analyzed for CoII spectrophotometrically as [Co(NCS)4]2- by Kitson’s method [28] (e625 (dm3 mol-1 cm-1) = 2.0 9 103 for [Co(NCS)4]2-) [29]; the CoII yield was ca. 25% (ca. kred 1.6 9 10-6 s-1, 27 °C ). We are, therefore, led to believe that the complex undergoes two-phase reaction: (i) the acid dependent nonredox reaction which is relatively fast and (ii) a much slower H+ promoted redox reaction at very long time intervals. The latter reaction (t1/2 = 120 h at 27 °C)) does not interfere with the former (t1/2 \ 0.3 h). The rate constants (kobs) (see Table 4)
Table 4 Rate data for the reaction of [CoIII(HL)] in acid mediuma [HClO4]T (mol dm-3)
103 k (s-1)
[HClO4]T (mol dm-3)
Obs.
Cal.
0.002
7.69
7.67
0.005
3.97
4.11
0.0075
3.28
0.010 0.020
103 k (s-1) Obs.
Cal.
0.075
0.988
0.951
0.100
0.864,0.823
0.883
3.09
0.150
0.788
0.819
2.53
2.54
0.236
0.736
1.63
1.65
0.300
0.782, 0.733
0.050
1.07
1.08
0.470
0.739
A (s-1)c
23.3 ± 2.1
104 kf (s-1)
6.1 ± 1.1
B (dm3 mol-1 s-1)c C (dm3 mol-1)c
(7.61 ± 1.13) 9 102 (1.12 ± 0.13) 9 103
104 kr (s-1)d 102 kr0 (s-1)
0.73 ± 0.13 2.3 ± 0.2
0.770 b
0.752 0.728
a
370 nm, [Complex]T = 3.0 9 10-4 mol dm-3, 25.0 ± 0.1 °C; R[103 (kcal - kobs)]2 = 0.0622 9 103 kobs (s-1): 0.388 (20.0 °C), 2.98 (35.0 °C), 6.72 (45.0 °C) at [HClO4]T (mol dm-3) = 0.10
b
550 nm, [complex]T (mol dm-3) = 6.59 9 10-4
c
A = 103(kf + kr0 ), B = 103(kf + kr) K10 , C = K10
d
Based on Keq = kf/kr = 8.3
123
46 Scheme 1 Reactions of [CoIII(HL)] in acid medium
Transition Met Chem (2008) 33:39–53
K1 III
+
III
+
kf
[Co (H2L)]
[Co (HL)] + H
kr kr
decreased nonlinearly with the increase of [H+] in the range 0.002 B [H+]T (mol dm-3) B 0.47 and fitted Eq. 3 well. kobs ¼ ðA þ B½Hþ Þ=ð1 þ C½Hþ Þ
ð3Þ
Scheme 1 is proposed to interpret the rate data where IS+ is an intermediate product from [CoIII(H2L)]+ and (IS–H) is its ionized form (equivalent to [CoIII(HL)]). The spontaneous transformation of [CoIII(HL)] to (IS–H) was, however, insignificant as there was no spectral change of the complex with time at pH C 5. Accordingly, kobs is given by Eq. 4 kobs ¼ ½kf K1 ½Hþ =ð1 þ K1 ½Hþ Þ þ kr K10 ½Hþ þ kr0 =ð1 þ K10 ½Hþ Þ
ð4Þ
which also reduces to Eq. 3 {as K1[H+] [[ 1 at the lowest [H+]T for K1(av.) = 5 9 103 dm3 mol-1} such that A = (kf + kr0 ), B = (kf + kr)K10 and C = K10 . The calculated values of the parameters, A, B, C are collected in Table 4. A0 (see Experimental section) is in fact the equilibrium absorbance (A0 = Aeq). It decreased nonlinearly with the increase of [H+] (Fig. 5). On the contrary, the zero time absorbances from different runs for a given [CoIII(HL)]T were essentially [H+] independent (0.002 B [H+]T (mol dm-3) B 0.47) and averaged to a constant value (A0(370 nm) = 0.78 ± 0.01 for [complex]T = 3.0 9 10-4 mol dm-3) suggesting thereby that the protonation equilibrium of [CoIII(HL)] at zero time was virtually driven to completion. Hence for Scheme 1 A0 ¼ ðA0 þ AIS Keq f1 þ AISH Keq f2 Þ=ð1 þ Keq f3 Þ
'
[IS ]+
[IS-H] + H+ K1'
where f1 = K1[H+]/(1 + K1[H+]), f2 = f1/(K10 [H+]), f3 = f2(1 + K10 [H+]), f3= f2(1+K10 [H+]), and Keq = [IS+]eq/[CoIII(H2L)+]eq, AIS and AIS–H are the absorbances of IS+ and IS–H, respectively, for the same total concentration of the complex as for A0. With K10 = 1.12 9 103 (C = K10 ), K1 = 5.0 9 103 dm3 mol-1 and A0 = 0.78, the A0 values (see Eq. 5) yielded AIS = 0.535 ± 0.013, AIS–H = 1.022 ± 0.12 {e370nm (dm3 mol-1 cm-1): (1.78 ± 0.045) 9 103 and (3.40 ± 0.36) 9 103 for IS+ and IS–H, respecP tively} for Keq = 8.3( [10(Acal0 –Aobs0 )]2 = 1.505). It turned out that the data fitting was insensitive to the chosen value of Keq when it exceeded 8.3. The value of Keq (=kf/kr) enabled evaluation of the rate parameters, kf, kr and kr0 which are also collected in Table 4. The isotope effect (H/D) was studied at 25 °C under the condition at which the predominant term of kobs is kf. The values of kobs at [H+]T = 0.1 mol dm-3 in 50% (v/v) D2O + H2O and in 100% H2O are (7.89 ± 0.05) 9 10-4 s-1 and (8.23 ± 0.06) 9 10-4 s-1, respectively, which translates to [kobs (H2O)/kobs (D2O + H2O)] = 1.04 ± 0.01 thus indicating a small solvent isotope effect. The activation parameters for the kf path were calculated from the temperature dependence of kobs at [H+]T = 0.1 mol dm-3 (see footnote ‘‘a’’ of Table 4). Under this condition, correction to kf was small (kobs = 1.3 kf at 25 °C, see Eq. 4) and it was assumed to be constant in the temperature interval 20–45 °C {kobs = 1.3 (kBT/h) exp (-DH±/RT + DS±/R)} yielding DH± = 69.5 ± 6.0 kJ mol-1 and DS± = -71 ± 19 J K-1 mol-1.
ð5Þ Reactions with H2Asc
Fig. 5 A0 (=Aeq) versus [H+]T (mol dm-3) plot. k = 370 nm, [complex]T = 3.0 9 10-4, I = 0.5 mol dm-3, 25 °C
123
The successive spectral scans at long reaction times displayed only one isosbestic point at 315 nm (Fig. 6) with the development of a peak at 298 nm. There was no evidence of build up of any other intermediate. The aged reaction mixtures gave positive test for CoII by Kitson’s method [28]. Qualitative test for dehydroascorbic acid by Roe’s method [30] in the spent reaction mixture, after ion exchange separation of the cations was positive. The complete reduction of the complex was also substantiated by the drastic change in its spectrum (see Fig. 6). The reaction stoichiometry was established by measuring CoII by Kitson’s method (loc. cit.) and the unreacted ascorbic acid iodimetrically [2c] after allowing the reaction
Transition Met Chem (2008) 33:39–53
47
Fig. 6 Successive spectral scans for reduction of [CoIII(HL)] by H2Asc, [complex]T (=1.0 9 10-4) + [H2Asc]T (=2.0 9 10-3) + [H+] (=0.05) mol dm-3, 25.0 ± 0.1 °C; 60 s (1), 300 s (2), 660 s (3), 1,200 s (4), 1,800 s (5), 2,400 s (6), 3,000 s (7), 18,000 s(8)
mixture ([CoIII]T/[H2Asc]T = 5) to undergo complete reduction (ca. 10 t1/2). The overall stoichiometry is given by Eq. 6. The CoII product under acidic conditions, however, dissociates reversibly to [Co(OH2)6]2+ and the free ligand. It may also be noted that HAsc- is too weak a ligand to bind [Co(OH2)6]2+ appreciably (log K[CoHAsc]+ = 1.4 at 25 °C, I = 0) [31]. 2[CoIII (HL)] þ H2 Asc ¼ 2[CoII (HL)] þ dehydroascorbic acid þ 2Hþ ð6Þ The zero time absorbance data (A0) from kinetic runs did not show any variation with [H2Asc]T but were pH sensitive in accord with the protonation equilibrium of
kobs
[CoIII(H2L)]+ and HAsc- is also equivalent to the reaction between [CoIII(HL)] and H2Asc due to the fast proton transfer equilibrium between the two preceding the rate limiting step. It is further interesting to note that the acid dependent reaction of the complex (as mentioned above) was kinetically insignificant in the presence of a large excess of ascorbic ([H2Asc]T/[complex]T C 50) and at pH [ 1. However, kobs values at pH \ 1 and varying [H2Asc]T are only marginally smaller than those for the reaction of the complex at the same acidity in absence of H2Asc (see kobs data in Table 3). This is plausible if the species, IS+ and IS–H, undergo fast reduction by H2Asc. Accordingly, Scheme 2 is proposed for which kobs is given by Eq. 7
2 k1red þ ðk3red Kd0 =½Hþ Þ Q1 þ k2red Q2 K1 ½Hþ f1 ½H2 AscT þ kf K1 ½Hþ ¼ 1 þ K1 [Hþ ] þ Q1 ð1 þ Kd0 =[Hþ ]Þ þ Q2 K1 [Hþ ] f1 [H2 Asc]T
[CoIII(HL)] (Eq. 1). The A0 data yielded (see Eq. 2) K1 = (1.9 ± 0.3) 9 103 dm3 mol-1, eCoH-1L = (2.22 ± 0.08) 9 103, eCoL = (4.6 ± 1.1) 9 103 dm3 mol-1 cm-1 (360 nm). The rate data are collected in Table 5. The observed rate constants (kobs) at a constant pH (=3.95 ± 0.06) increased with [H2Asc]T and approached a limiting value thus indicating equilibrium pre-association of ascorbic acid with the complex. kobs versus pH curve at constant [H2Asc]T (=0.05 mol dm-3) is interestingly bell shaped (see Fig. 7) and the rate constant tends to attain a virtually zero value at pH = 5.46. Ascorbic acid (pKd = 4.01) [2c] and the complex, [CoIII(H2L)]+ (pK = 3.7) have comparable pK values which fall in the range of pH used. Hence, this trend in the reactivity suggests that HAsc- is not an efficient reducing species for [CoIII(HL)]. The reaction between
ð7Þ
where the factor 2 stands for stoichiometry and f1 (=[H+]/ ([H+] + Kd) denotes the fraction of [total ascorbic acid] as [H2Asc]. It can be seen that Eq. 7 reduces to a limiting form at a constant pH and high [H2Asc]T. Neglecting the Kd0 and kf dependent terms, the initial estimates of k1red, Q1 and k2red, Q2 could be judged from the intercept and gradients of the plots of 1/kobs versus 1/[H2Asc]T under varying conditions of pH (0.33 B pH B 2.01) and [H2Asc]T and at constant pH (3.95 ± 0.05) and varying [H2Asc]T. All rate data were then iteratively fitted to Eq. 7 by varying K1, (k1redQ1), (k2redQ2), (k3redKd0 Q1), kf, Q1, Q2, and Kd0 . It turned out that k3red was not statistically significant and the association of H2Asc with [CoIII(HL)]/[CoIII(H2L)]+ was little controlled by the charge of the substrates (Q1 * Q2). The calculated values of the parameters based on Q1 = Q2 = Q and K1 = 1.9 9 103 dm3 mol-1 are collected in Table 5. A reasonably good
123
48
Transition Met Chem (2008) 33:39–53
Table 5 Rate data for the reaction of [CoIII(HL)] with H2Asca pHb
[H2Asc]T (mol dm-3)
104 k (s-1)
pHb
Obs.
Cal.
[H2Asc]T (mol dm-3)
104 k (s-1) Obs.
Cal.
1.00
0.050
5.28
5.73
3.85
0.030
1.51
2.44
1.30
0.050
5.48
5.86
3.85
0.030
1.66
2.44
1.56
0.050
5.79
6.07
3.96
0.050
1.49
2.01
1.70
0.020
6.66
5.77
3.96
0.050
1.77
2.01
1.78
0.050
6.24
6.35
0.33
0.050
5.53
5.63
2.01
0.050
7.01
6.83
0.52
0.050
6.34
5.64
2.22 2.30
0.050 0.050
7.73 8.60
7.45 7.74
4.06 3.95
0.060 0.080
2.62 1.71
1.63 2.10
2.47
0.050
8.43
8.41
3.96
0.080
2.41
2.06
2.70
0.050
9.29
9.20
4.01
0.080
2.31
1.84
2.70
0.050
8.40
9.20
3.96
0.150
2.17
2.13
3.20
0.050
7.74
7.77
3.95
0.150
2.58
2.14
3.20
0.050
6.70
7.77
3.95
0.200
2.13
2.15
3.44
0.050
5.36
5.59
3.95
0.200
2.73
2.15
3.63
0.050
3.94
3.98
3.50
0.150
7.68
5.58
3.98
0.050
1.62
1.93
0.52
0.060
4.67
5.66
4.15
0.050
1.18
1.32
0.52
0.080
5.11
5.70
4.61
0.050
0.31
0.47
0.52
0.080
5.50
5.70
5.46
0.050
0.024
0.068
1.30
0.100
7.25
6.10
3.98
0.005
0.83
1.44
2.65
0.040
9.16
8.41
3.85
0.010
1.23
2.09
2.85
0.040
8.26
8.69
102 k1redQ (dm3 mol-1 s-1) 102 k2redQ (dm3 mol-1 s-1)
4.19 ± 0.21 0.208 ± 0.023
103 k1red (s-1) 103 k2red (s-1)
5.9 ± 0.3 0.293 ± 0.032
Q (dm3 mol-1)c
7.1
103 k3red (s-1)
0.00 ± 0.22
3
10 Kd0 -3 10
-3
(mol dm )
6.0
K1 (dm3 mol-1)
1.9
a
25.0 °C, [complex]T (mol dm-3) = (1-3) 9 10-4, k (nm) = 360
b
-Log [H+], pH adjusted with HClO4/NaOH
c
Q = Q1 = Q2. R[104(kcal - kobs)]2 = 17.3
fit could be achieved. The calculated value of kf (=5.5 9 10-4 s-1) also agreed well with that obtained for the nonredox reaction in acid medium in absence of H2Asc {(6.1 ± 1.2) 9 10-4 s-1 see Table 4} thus supporting the validity of Scheme 2.
Reaction with SIV The long duration repetitive spectral scans of a mixture of the complex and SIV at constant pH (see Fig. 8) displayed at the early phase isosbestic points at 314, 420 and 509, and 634 nm similar to (but not identical with) those observed in the acid mediated reaction (see Fig. 4). Also, these isosbestic points are completely lost ultimately at prolonged times with drastic change in the spectrum (not shown in Fig. 8) in tune with the reduction
123
4
-1
10 kf (s )
5.5
of CoIII (ca. 10% CoII yield after 4 days at 27 °C). The extremely slow reduction by SIV was not pursued further. The extrapolated zero time absorbance data (A0) from kinetic runs showed no dependence on [SIV]T (=0.005–0.35 mol dm-3, pH = 2.50–5.06, [CoIII(HL)]T (mol dm-3) = 1.0 9 10-4) in agreement with the rapid scan spectra and further confirmed that there was no fast SIV substitution at the metal centre. We would have expected significant spectral changes either for S- bonded or O-bonded sulfito complex formation [32]. On the other hand, A0 (from kinetic runs) was pH sensitive and fitted Eq. 2 well yielding A1 = 0.571 ± 0.033, A2K1 = (7.44 ± 0.11) 9 102, K1 (dm3 mol-1) = 1.02 9 103, and A2 = 0.729 ± 0.011 {eCo(HL) = (5.71 ± 0.033) 9 103, eCo(H2L) = (7.29 ± 0.11) 9 103 dm3 mol-1 cm-1, 340 nm, 40 °C}. The substantially low values of the long time absorbance, A0 , from kinetic runs were also [SIV]T and pH
Transition Met Chem (2008) 33:39–53
49
[[ k-1, is given by Eq. 8. With [HSOk2[HSO3][ 3] = + IV Kd1[H ] [S ]T/{[H+]2 + Kd1[H+] + Kd1Kd2} (pKd1 = 1.7, pKd2 = 6.5 for SO2 and HSO3 , respectively, I = 0.5 mol dm-3, 40.0 °C [33]), the rate constants were fitted to Eq. 8 and the calculated values of k1, k3, K1 and Q are collected in Table 6. kobs ¼
k1 K1 ½Hþ þ k3 Q½HSO 3 þ 1 þ K1 ½H ] þ Q HSO 3
ð8Þ
Mechanisms Reactions of [CoIII(HL)] in acid medium
Fig. 7 pH dependence of 104kobs (s-1) for reduction of [CoIII(HL)] by H2Asc, 25.0 °C, [H2Asc]T = 0.05, [complex]T = (1–3) 9 10-4 mol dm-3
III
[Co (HL)] Q2
+ H+
+ H2Asc
K1
HAsc- + H+
Kd ' (D-H) + + H
(IS)+ + H2Asc
(IS-H)
k1 red
D
kf
K1 '
II
[Co (HL/H2L)]
-/0
+
Q1
Kd
[CoIII(H2L)]+
k2 red
C
fast
k3 red
HAsc . + H+
+ H+ III
[Co (HL/H2L)] 0/+ + HAsc
.
fast
II
[Co (HL/H2L)]-/ 0 + DHA + H+
Scheme 2 Reduction of [CoIII(HL)] by H2Asc; C = {[CoIII(H2L)]+, (H2Asc)}, H2Asc)}, D = {[CoIII(HL)],(H2Asc)}, DHA = dehydroascorbic acid
independent indicating thereby that the sulfito complex formation was driven to completion {e (dm3 mol-1 cm-1): (3.86 ± 0.52) 9 103 for [CoIII(H2L)(HSO3)] at 340 nm}.
Rate data The rate data are collected in Table 6. The observed trend in the variation of kobs on [SIV]T at constant pH (=4.3 ± 0.05) is in tune with the fast association of the reactants followed by a slow transformation of the associated species. At constant [SIV]T (=0.05 mol dm-3) kobs also increased nonlinearly with the increase of [H+]. These facts are consistent with Scheme 3 for which kobs, based on the steady state approximation for [INT] and
A careful examination of Fig. 4 clearly reveals a slow and significant red shift of the absorption maxima and minima of [CoIII(HL)] subsequent to its rapid protonation to the form [CoIII(H2L)]+. A considerable reduction of intensity of the absorption extrema with retention of the isosbestic points at the early phase of the reaction and also a visual color change of the reaction mixture from purple to light green are also characteristic features. These changes are also rapidly reversed if the reaction is arrested in the middle by raising pH of the mixture quickly to C4. Gillard et al. [26] have already demonstrated that the CoIII–O bonds (phenolate/phenol) in the chelated salicylato complexes, [Co(en)2(sal)]+ and [Co(trien)(sal)]+ are extremely inert to cleavage under very harsh acidic conditions. The CoIII–O bond of the sal moiety does not open up at the phenolate/phenol coordination site. We, therefore, believe that an isomer (or an aqua product) is generated as a consequence of the opening up of a chelate ring at the amide site when the binding efficiency of the amido group is weakened due to protonation (see Scheme 4a, b). Consequently, there is a change in the coordination environment in IS+/IS–H ([CoIII(N, N, N)(O, O, O)]). However, the seven membered ring in IS+/IS-H exerts steric strain (see Scheme 4a), so that these species revert to their congener, the process being much faster for its amido form (i.e., IS–H). No substantial (H/D) kinetic isotope effect (loc. cit.) further suggests that this intramolecular isomerisation is not controlled by a rate-limiting proton transfer. Further, the values of the activation parameters (DH± = 69.5 ± 6.0 kJ mol-1 and DS± = -71 ± 19 J K-1 mol-1) suggest that the transition state is assembled via a low energy path involving some degree of ordering. The rate-limiting opening up of the chelate ring by Co–NH (amide) bond cleavage followed by the rapid capture of the C=O (amide) function by the CoIII center is a reasonable possibility (see Scheme 4a). This, however, demands minimum adjustment of the axial-equitorial dispositions of the coordinated phenolates and also the sec-NHs with retention of the stereochemistry at these sites despite the
123
50
Transition Met Chem (2008) 33:39–53
Fig. 8 Successive spectral scans for complex + SIV solution: [complex]T = 2.17 9 10-4, [SIV]T = 0.10 mol dm-3, pH = 4.07, 40.0 ± 0.1 °C: 60 s (1), 300 s (2), 600 s (3), 900 s (4), 1,200 s (5), 1,800 s (6), 2,700 s (7), 3,900 s (8)
Table 6 Rate constants for the reaction of [CoIII(HL)] with SIVa [SIV]T (mol dm-3)
pHb
103 k (s-1)
[SIV]T (mol dm-3)
Obs.
Cal.
pHb
103 k (s-1) Obs.
Cal.
0.005
4.30
0.48
0.48
0.050
2.50
4.15
4.37
0.010
4.18
0.55
0.62
0.050
2.64
4.13
4.13
0.020
4.35
0.72
0.51
0.050
3.02
3.78
3.26
0.050
4.32
0.92
0.66
0.050
3.02
3.70
3.26
0.050
4.18
0.92
0.78
0.050
3.50
1.57
1.96
0.070
4.30
0.84
0.74
0.050
3.50
1.46
1.96
0.070
4.31
0.92
0.74
0.050
3.54
2.20
1.86
0.100
4.30
0.91
0.84
0.050
3.69
1.77
1.52
0.100
4.35
0.93
0.81
0.050
4.65
0.32
0.47
c
0.100
4.35
0.92
0.81
0.050
4.92
0.37
0.39
0.150
4.31
0.92
0.96
0.050
5.06
0.20
0.36
0.180
4.31
0.96
1.03
0.010
4.53
0.38
0.33
0.200
4.35
0.83
1.05
0.020
3.20
2.10
2.80
0.100
3.46
2.04
2.06
0.020
4.10
0.49
0.76
0.100
3.45
1.94
2.08
0.020
3.53
2.11
1.87
0.100 0.150
3.82 3.48
1.31 1.93
1.35 2.02
0.080 0.250
3.70 3.63
1.30 1.89
1.53 1.78
0.200
3.26
3.14
2.51
0.350
3.18
2.28
2.61
k1K1 (dm3 mol-1 s-1)d
9.91 ± 1.15
103 k3Q (dm3 mol-1 s-1)d
7.11 ± 1.16
-3
10
3
-1 d
K1 (dm mol )
1.93 ± 0.32
Q (dm3 mol-1)d
3.4
103 k1 (s-1)
5.1 ± 2.1
103 k3 (s-1)
2.1 ± 0.3
a
40.0 ± 0.1 °C, [Co (HL)]T (mol dm ) = 1.0 9 10 , k (nm) = 340
b
pH = -log [H+]
c
k (nm) = 540 nm, [CoIII(HL)]T = 6.59 9 10-4 mol dm-3 R[(103(kcal - kobs)]2 = 2.64
d
III
123
-3
-4
Transition Met Chem (2008) 33:39–53
III
[Co (HL)]
k1
K1
+ H+
51
III
+
+
(a)
k -1
+ HSO3-
HSO3-
H
O
O N
C
NH
k2
+ H+
III
Co O
N
C
- H+
N (-)
O
H
O
C=O
Co
HN
Q III
[INT]
[Co (H2L)]
[CoIII(HL)]
+
H III
[Co (H2L)(SO3)]
C=O HN
HN
fast
k3
III
O
[CoIII(H2L)(HSO3)]
-
{[Co (HL)],(HSO3 )}
NH
[CoIII(H2L)] +
kf
kr
kr'
Scheme 3 Reactions of [CoIII(HL)] with SIV H H
steric constraints imposed by the bond angle expansion as stated above. This might be realized if IS and IS–H assume a strongly distorted configuration having an elongated HN– C=O–CoIII bond for the isomerisation process, CoIII–N(H)– C=O) ? CoIII–O=C–NH. An alternative mechanism is presented in Scheme 4b where the rate-limiting opening up of the CoIII–NH (amide) bond results in the formation of the corresponding aqua complex. The reported data on the acidity of the aqua ligand for some CoIII complexes, [Co(NH3)5(OH2)]3+/(cis/ trans)[Co(en)2(X)(H2O)]2+ (X = NO2) (pKH2O = 6.6 – 6.4, 25 °C) [34] leads us to believe that the pK of IS+, if it is an aqua complex, ought to be [6 which, however, is not the case (considering pKðISþ Þ = log K10 = 3.05). Further considering that the substantially high rate of transformation of (IS–H) to [CoIII(HL)] arises due to an internal conjugate base mechanism involving the deporonated aqua ligand and the vicinal –CO–NH– (see Scheme 4b), we obtain kr0 = kcb 9 KCB where KCB = Kd(–CO–NH)/KdðISþ Þ (=10-15/10-3 = 10-12, KdðISþ Þ = 1/K10 ) and kcb is the rate constant for the transformation of the aqua-amido conjugate base (CB) to [CoIII(HL)]. Using the value of kr0 and the estimated value of KCB we get kcb & 2 9 1010 s-1 (25 °C), a value exceedingly too high for the charge delocalized amido conjugate base (i.e., the negative charge on the amido N- is delocalized to the carbonyl). Hence, Scheme 4a is preferred and the relatively high rate of transformation: (IS–H) ? [Co(HL)] (kr0 = 2.3 9 10-2 s-1, 25 °C) is attributed to the labilizing influence of the trans-amido group as well as the distorted configuration of IS–H. The very slow reduction of the complex in acidic medium ([H+] C 0.05 mol dm-3) is tentatively attributed to the species, IS+ with [CoIII(N,N,N)(O,O,O)] coordination or its partially aquated product. Reduction of CoIII by coordinated solvent (H2O) (and ligand) may be involved. The 298 nm band is intensified in the long run, due to the release of the ligand from the CoII product, more likely from the redox path involving the solvent. This extremely slow redox reaction was not pursued further.
N O
N O
O
C
O
C
NH
+H
Co III
C=O
C=O
HN
O
- H+
N (-)
O
NH
Co III
+
HN
HN
+
(IS ) (IS-H)
(b) O
H
O
N
C
Co
NH
N (-)
O
III
C =O
HN
[C o III (H 2 L)] +
O N
C
III
HN
Co HN
N
C
III
O
H
O
NH
O
O
kf
kr
(-H 2 O)
H
O
NH
O
(H L)]
k cb
Co
- H+
HN
[C o
N
C
C =O
O
H
O
+ H+
III
Co
H C =O
III
C =O
OH 2
O HN
H
NH
NH
N (-)
(IS + )
(C B)
- H+ K CB
O
H
O C
N
Co
NH
III
C=O
OH
O HN
NH
(IS-H )
Scheme 4 Structural aspects of the reactions of [CoIII(HL)]+ in acid medium involving: (a) intra molecular isomerisation, (b) aquation
Redox reaction with H2Asc There was no evidence of direct coordination of ascorbic acid to the CoIII center. However, the outer sphere association of [CoIII(HL)] and [CoIII(H2L)]+ with H2Asc was kinetically identified with little disparity for the charge of the complex. The value of the association constant is
123
52
Transition Met Chem (2008) 33:39–53
higher than the value predicted from Fuoss theory on the basis of hard sphere model (0.3 dm3 mol-1 for +1, 0 or 0, 0 charge type association for a distance of closest approach ´˚ of 5 A ) [35]. It appears that factors other than simple columbic interaction play a dominant role in stabilizing these outer sphere complexes. It is interesting to note that the value of Kd0 (=6 9 10-3 mol dm-3, see Table 5) is ca. 60 times higher than the same for the first acid dissociation constant of H2Asc. Hydrogen bond formation by H2Asc with the amido (N-) site is not likely to favor the acid ionization of ascorbic acid in the outer sphere complex, D {[CoIII(HL)], (H2Asc)}. On the other hand, intra molecular proton transfer in D can give rise to the species, {[CoIII(H2L)]+, (HAsc-)}, and Kd0 can be reconciled with the H+ dissociation equilibrium of the coordinated ligand. The observed trend in the redox activities, k1red [ k2red [[ k3red, reflected that protonation of [CoIII(HL)] had substantial rate retarding effect (k1red/k2red = 20) for the same reductant, i.e., H2Asc. It could possibly arise due to the transformation of the outer sphere complex, {[CoIII(HL)] (H2Asc)}, to a more reactive form {[CoIII(H2L)]+(HAsc-)}. The vanishingly small value of k3red, however, stands in good support that reduction of CoIII by HAsc- in D–H, i.e., {[CoIII(HL)], (HAsc-)} is strongly inhibited by the coordinated amido function. Since there was no direct binding of H2Asc/HAsc- at the CoIII center, the mechanism of electron transfer (ET) was considered to be outer sphere bridging type. Applying the simplified Marcus relationship [36] k12 ¼ (k11 k22 K12 f )1=2
ð9Þ
where k11 and k22 are the rate constants of the self exchange reactions of the couples H2Asc/H2Asc.+ and [CoIII(HL)]/ [CoII(HL)]-, respectively, K12 is the overall equilibrium constant of the reaction, K12 [Co (HL)] + H2 Asc [CoII (HL)] + H2 Ascþ III
ð10Þ
and Log f = {Log K12)2/[4 Log(k11k22/Z2)], the value of k22 could be calculated on the basis, k11 = 1.0 9 105 dm3 mol-1 s-1, and E01(H2Asc+/H2Asc) = 0.95 V [37]. From the c. v. data, the value of E02 ([CoIII(HL)]/[CoII(HL)]-) was taken to be 0.849 V (NHE with no solvent correction). Taking k12 = k1redQ (see Table 5), and Z = 1011 s-1, k22 and f were calculated iteratively as 3.7 9 10-10 dm3 -1 mol s-1 and 0.938, respectively. This showed that the ligand HL3- enhanced the kinetic stability of the couple, [CoIII(HL)]/[CoII(HL)]- towards self exchange by three orders of magnitude greater than EDTA4- for the corresponding EDTA complex {k22 = 2 9 10-7 dm3 mol-1 s-1 for [CoIII(EDTA)]-/[CoII(EDTA)]2- [38]}, a fact which might reflect the influence of the N-coordinated amido function.
123
Reaction with SIV The rapid scan spectra indicated that there was no instan2taneous bonding interaction of HSO3 /SO3 directly with III the Co center. The kinetic data does reveal diffusion limited ion pairing of [CoIII(HL)] with HSO3 with a value of Q substantially higher than predicted from Fuoss theory [35]. This large difference might be at least partly due to the hydrogen bonding between the associating species. The stretching of one of the CoIII–NH (amide) bonds to the III limit of breaking limited the entry of HSO3 at the Co center; the relevant transition states for both k1 and k3 paths, however, recognized SIV nucleophile in preference to the amido –C=O (or H2O) resulting in the formation of the sulfito complex exclusively. The comparable values of k1 and k3 (see Table 6), which also compare excellently with the value of kf for the isomerisation (ca. kf = 2.8 9 10-3 s-1 at 40 °C based on DH= = 70 kJ mol-1) clearly point to the fact that the transition states for the sulfito complex formation and the isomerisation are very similar indeed. The protonated sulfito complex, [CoIII(H2L)(HSO3)] is inert to SO2 elimination reaction. An estimated value of k-2 (i.e., for the reverse reaction, see Scheme 3) may be taken to be \10-4 s-1 (at 40 °C). This is at least seven orders of magnitude smaller than what is expected for an oxygen bonded sulfito complex of CoIII [39]. On the contrary the S-bonded mono sulfito complexes of CoIII are relatively more stable to H+ catalysed SIV elimination [29, 32, 39]. This comparison enables us to suggest that [CoIII(H2L)(HSO3)] must be an inner sphere S-bonded sulfito complex.
Supplementary material CCDC 635886 contains the supplementary crystallographic data for this paper. These data can be obtained at www.ccdc.cam.ac.uk/deposit {or from the Cambridge Crystallographic Data Center 12, Union Road Cambridge CB2 1EZ, UK; Fax: (internet) +44-1223/336-033; E. mail:
[email protected]). Supporting informations are submitted as follows. Figures S1(a–c) and S2(a–c) are the representative pseudofirst order kinetic plots in the presence of HClO4, H2Asc, SIV and rapid scan spectra of the complex in the presence of HClO4, H2Asc, and SIV, respectively. Figures S3–S7 present cyclic voltammetric plot, 13C NMR, mass and 1H NMR spectra, respectively. Acknowledgements This work was supported by a research grant (Emeritus Scientist Scheme) awarded to A C D by the C S I R, New Delhi (Grant No 21 (0502)/01/EMR II). S N thanks the C S I R for award of a Senior Research Fellowship in this scheme. We thank Professor A K Mishra, Department of Chemistry, Indian Institute of
Transition Met Chem (2008) 33:39–53 Technology, Madras for micro analysis, and mass spectral measurements; Dr. B. B. Nayak, Dr. B Nanda, Scientists, Regional Research Laboratory, Bhubaneswar, India for Raman and NMR spectroscopy; Professor S. Mazumdar and Mr. R. K. Behera, Tata Institute of Fundamental Research, Mumbai, India for stopped flow Rapid Scan Spectral measurements and Dr. G. S. Brahma, ICFAI, Hyderabad, India for assistance
53
17. 18.
19.
References 1. (a) Martin DS, Mascharak PK (2000) Chem Soc Rev 29:69; (b) Kostka KL, Fox BG, Hendrich MP, Collins TJ, Rickard CEF, Wright LJ, Mu¨nek E (1993) J Am Chem Soc 115:6746; (c) Dominguez-Vera JM, Sua´rez-Varela J, Maimoun IB, Colacio E (2005) Eur J Inorg Chem 2005:1907; (d) Singh AK, Mukherjee R (2005) Inorg Chem 44:5813; (e) Nayak S, Dash AC, Nayak PK, Das D (2005) Trans Met Chem 30:917; (f) Nayak S, Dash AC (2006) Trans Met Chem 31:813 2. (a) Chandra SK, Chakravorty A (1992) Inorg Chem 31:760; (b) Nayak S, Dash AC (2005) Trans Met Chem 30:560; (c) Nayak S, Dash AC (2006) Trans Met Chem 31:316 3. Collins TJ, Santarsiero BD, Spies GH (1983) J Chem Soc Chem Commun 681 4. (a) Collins TJ, Richmond TG, Santarsiero BD, Treco BGRT (1986) J Am Chem Soc 108:2088; (b) Anson FC, Collins TJ, Coots RJ, Gipson SL, Richmond TG (1984) J Am Chem Soc 106:5037; (c) Rotthaus O, LeRoy S, Tomas A, Barkigia KM, Artaud I (2004) Eur J Inorg Chem 2004:1545 5. Kiss T, Petrohan K, Buglyo P, Sanna D, Micera G, Pessoa JC, Madeira C (1998) Inorg Chem 37:6389 6. Anson F, Collins TJ, Gipson SL, Keech JT, Krafft TE, Peake GT (1986) J Am Chem Soc 108:6593 7. Comba P, Gavrish SP, Lampeka YD, Lightfoot P, Peters A (1999) J Chem Soc Dalton Trans 4099 8. Serratrice G, Baret P, Boukhalfa H, Gautier-Luneau I, Luneau D, Pierre J-L (1999) Inorg Chem 38:840 9. Nagashima S, Nakasako M, Dohmae N, Tsujimura M, Takio K, Odaka M, Yohda M, Kamiya N, Endo I (1998) Nat Struct Biol 5:347 10. Mascharak PK (2002) Coord Chem Rev 225:201 11. Neupane KP, Shearer J (2006) Inorg Chem 45:10552 12. Wuerges J, Lee J-W, Yim Y-I, Yim H-S, Kang S-O, Carugo KD (2004) Proc Natl Acad Sci USA 101:8569 13. Wolkenberg SE, Boger DL (2002) Chem Rev 102:2477 14. Peters JW, Stowell MHB, Soltis SM, Finnegan MG, Johnson MK, Rees DC (1997) Biochemistry 36:1181 15. Nayak S, Dash AC (2006) Transit Met Chem 31:930 16. (a) Nayak S, Dash AC (2003) Indian J Chem 42A:2427; (b) Sahu M, Pradhan GC, Mohanty P, Dash AC (2004) Indian J Chem
20.
21. 22. 23. 24. 25. 26. 27.
28. 29. 30. 31.
32. 33. 34. 35. 36. 37. 38. 39.
43A:1228; (c) Rath H, Dash AC (2002) Indian J Chem 41A:1600; (d) Rath H, Pradhan GC, Dash AC (2001) Indian J Chem 40A:437 Dash AC, Nanda RK (1973) Inorg Chem 12:2024 (a) Irving HM, Miles MG, Pettit LD (1967) Anal Chim Acta 38:475; (b) Das A, Dash AC (2000) J Chem Soc Dalton Trans 1949 Sheldrick GM (1997) Program for crystal structure solution and refinement. Universita¨t Go¨ttingen, Go¨ttingen (a) Sams CK, Somoza F, Bernal I, Toftlund H (2001) Inorg Chim Acta 318:45; (b) Afshar RK, Bhalla R, Rowland JM, Olmstead MM, Mascharak PK (2006) Inorg Chim Acta 359:4105; (c) Angus PM, Elliott AJ, Sargeson AM, Willis AC (1999) J Chem Soc Dalton Trans 1131 Snow MR (1972) J Chem Soc Dalton Trans 1627 Dyer JR (2005) Application of absorption spectroscopy of organic compounds. Prentice Hall of India, New Delhi, pp 33–38 Nakamoto K (1997) Infrared and Raman spectra of inorganic and coordination compounds part B, 5th edn. Wiley, New York, p 22 Sigel H, Martin RB (1982) Chem Rev 82:385 Dash AC, Nanda RK, Ray N (1982) Indian J Chem 21A:1068 Gillard RD, Lyons JR, Mitchell PR (1973) J Chem Soc Dalton Trans 233 Margerum DW, Cayley GR, Weatherburn DC, Pagenkopf GK (1978) In: Martell AE (ed) Coordination chemistry, ACS Monograph 174, vol 2, p 118 Kitson RE (1950) Anal Chem 22:664 Dash AC, Jena KC, Roy A, Mukherjee D, Aditya S (1997) J Chem Soc Dalton Trans 2451 Roe JH (1945) Methods of biochemical analysis, vol 1. Interscience, New York, p 115 Davies MB, Austin J, Partridge DA (1991) Vitamin C: its chemistry and biochemistry, The Royal Society of Chemistry (Paper backs), Cambridge, p133 Dash AC, Patnaik AK, Acharya A (1998) Trans Met Chem 23:45 Martell AE, Smith RM (eds) (1976) Critical stability constants, vol 4. Plenum, New York, p 78 Basolo F, Pearson RG (1973) Mechanisms of inorganic reactions, 2nd edn. Wiley, Eastern New Delhi, p 32 Fuoss RM (1958) J Am Chem Soc 80:5059 (a) Marcus RA (1963) J Phys Chem 67:853; (b) Pennington DE In ref. (27) p 484 Kagayama N, Sekiguchi M, Inada Y, Takagi HD, Funahashi S (1994) Inorg Chem 33:1881 Ref. 27 Table 3–2, p 487 (a) Dash AC, El-Awady AA, Harris GM (1981) Inorg Chem 20:3160; (b) Dash AC, Dash N, Aditya S, Roy A (1988) Indian J Chem 27A:398
123
