Biomimetic iron(iii) complexes of facially and meridionally coordinating tridentate 3N ligands: Tuning of regioselective extradiol dioxygenase activity in organized assemblies

Article abstract of DOI:10.1039/c3dt52350k

Four mononuclear iron(iii) complexes of the type [Fe(L)Cl₃] 1-4, where L is a tridentate 3N ligand such as (2-pyridin-2-ylethyl)(pyridin-2- ylmethyl)amine (L1), (methyl)(2-pyridin-2-ylethyl)(pyridin-2-ylmethyl)amine (L2), bis(pyridin-2-ylethyl)amine (L3), and (1-methyl-1H-imidazol-2-ylmethyl) (pyridin-2-ylethyl)amine (L4), have been isolated and studied as functional models for catechol dioxygenase enzymes. In [Fe(L2)Cl₃] 2, the ligand L2 is coordinated facially to iron(iii) whereas in [Fe(L1)Cl₃] 1 and [Fe(L4)Cl₃] 4 the ligands L1 and L4 are coordinated meridionally. In DCM, CH₃CN and aqueous SDS, CTAB and TX-100 micellar media, the positions of both the low and high energy catecholate-to-iron(iii) LMCT bands (465-530, 690-860 nm) observed for the 3,4-di-tert-butylcatecholate (DBC ^2-) adducts of the iron(iii) complexes vary in the order 2 > 1 > 3 > 4, which reflects the influence of the stereoelectronic factors, mode of coordination and the chelate ring size formed by the tridentate ligands. Spectral and electrochemical studies disclose the formation and location of the cationic adducts as solvated [Fe(L)(DBC)(H₂O)]⁺ species mostly in the aqueous micellar pseudophases of SDS and TX-100 and in the aqueous phase of CTAB micellar solution. The [Fe(L)(DBC)Cl] adducts of 1, 3 and 4, generated in situ, afford major amounts of intradiol cleavage products (17.0-70.0%) and smaller amounts of extradiol (1.2-4.2%) products with varying extradiol to intradiol cleavage product selectivity (E/I: 1, 0.08:1; 3, 0.02:1; 4, 0.3:1). On the other hand, interestingly, the adduct [Fe(L2)(DBC)Cl] of 2 generated in DCM yields a major amount of extradiol (54.0%) and a lower amount (18.3%) of the intradiol cleavage products (E/I, 3:1). Remarkably, in aqueous SDS micellar media, it shows exclusive extradiol cleavage products (79.4%) while all the other complexes show very low selectivity (E/I: 1, 0.03:1; 2, 79.4:0, 3, 0.06:1, 4, 0.06:1), suggesting the suitability of SDS medium for 2 to elicit exclusive extradiol cleavage. The TX-100 micellar medium also provides a suitable hydrophobic environment for 2 to elicit extradiol cleavage. However, in CTAB micellar medium, 2 shows cleavage selectivity lower than others. Also, the rate of dioxygenation is higher in SDS micellar medium than in DCM, and is dependent upon the chelate ring size. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c3dt52350k

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Biomimetic iron(III) complexes of facially and
meridionally coordinating tridentate 3N ligands:
tuning of regioselective extradiol dioxygenase
activity in organized assemblies†
Cite this: Dalton Trans., 2014, 43,
6828
Muniyandi Sankaralingam,a Natarajan Saravanan,a Natarajan Anitha,a
Eringathodi Sureshb and Mallayan Palaniandavar*a,c
Four mononuclear iron(III) complexes of the type [Fe(L)Cl3] 1–4, where L is a tridentate 3N ligand such as
(2-pyridin-2-ylethyl)(pyridin-2-ylmethyl)amine (L1), (methyl)(2-pyridin-2-ylethyl)(pyridin-2-ylmethyl)
amine (L2), bis(pyridin-2-ylethyl)amine (L3), and (1-methyl-1H-imidazol-2-ylmethyl)(pyridin-2-ylethyl)
amine (L4), have been isolated and studied as functional models for catechol dioxygenase enzymes. In
[Fe(L2)Cl3] 2, the ligand L2 is coordinated facially to iron(III) whereas in [Fe(L1)Cl3] 1 and [Fe(L4)Cl3] 4 the
ligands L1 and L4 are coordinated meridionally. In DCM, CH3CN and aqueous SDS, CTAB and
TX-100 micellar media, the positions of both the low and high energy catecholate-to-iron(III) LMCT bands
(465–530, 690–860 nm) observed for the 3,4-di-tert-butylcatecholate (DBC2−) adducts of the iron(III)
complexes vary in the order 2 > 1 > 3 > 4, which reflects the influence of the stereoelectronic factors,
mode of coordination and the chelate ring size formed by the tridentate ligands. Spectral and electroche-
mical studies disclose the formation and location of the cationic adducts as solvated [Fe(L)(DBC)(H2O)]+
species mostly in the aqueous micellar pseudophases of SDS and TX-100 and in the aqueous phase
of CTAB micellar solution. The [Fe(L)(DBC)Cl] adducts of 1, 3 and 4, generated in situ, afford major
amounts of intradiol cleavage products (17.0–70.0%) and smaller amounts of extradiol (1.2–4.2%) pro-
ducts with varying extradiol to intradiol cleavage product selectivity (E/I: 1, 0.08 : 1; 3, 0.02 : 1; 4, 0.3 : 1).
On the other hand, interestingly, the adduct [Fe(L2)(DBC)Cl] of 2 generated in DCM yields a major
amount of extradiol (54.0%) and a lower amount (18.3%) of the intradiol cleavage products (E/I, 3 : 1).
Remarkably, in aqueous SDS micellar media, it shows exclusive extradiol cleavage products (79.4%) while
all the other complexes show very low selectivity (E/I: 1, 0.03 : 1; 2, 79.4 : 0, 3, 0.06 : 1, 4, 0.06 : 1),
suggesting the suitability of SDS medium for 2 to elicit exclusive extradiol cleavage. The TX-100 micellar
medium also provides a suitable hydrophobic environment for 2 to elicit extradiol cleavage. However, in
CTAB micellar medium, 2 shows cleavage selectivity lower than others. Also, the rate of dioxygenation is
higher in SDS micellar medium than in DCM, and is dependent upon the chelate ring size.
Received 27th August 2013,
Accepted 16th February 2014
DOI: 10.1039/c3dt52350k
www.rsc.org/dalton
Introduction
Catechols are common intermediary catabolites in the aerobic
microbial degradation of aromatic compounds.1–3 The
aSchool of Chemistry, Bharathidasan University, Tiruchirappalli – 620 024,
Tamil Nadu, India
bAnalytical Science Discipline, Central Salt and Marine Chemicals Research Institute,
Bhavnagar – 364 002, India
cDepartment of Chemistry, Central University of Tamil Nadu, Thiruvarur – 610004,
enzymes that degrade them are the catechol dioxygenases,
which can be divided into two broad classes based on the posi-
tion of ring cleavage: intradiol and extradiol dioxygenases. The
intradiol dioxygenases utilize a high-spin ferric center to
cleave the catechol ring between the two adjacent hydroxylated
carbons with insertion of both atoms of dioxygen. In contrast,
extradiol dioxygenases utilize a high-spin ferrous center to
cleave the catechol substrate specifically on either the distal or
the proximal side of the hydroxyl groups. The differences in
Tamil Nadu, India. E-mail: palanim51@yahoo.com, palaniandavarm@gmail.com
†Electronic supplementary information (ESI) available: Crystallographic data in
CIF format for [Fe(L1)Cl3] 1 and [Fe(L4)Cl3] 4. The electronic copies of the
electrochemical data. CCDC 819011 and 819013 for the X-ray crystal structures
of 1 and 4. For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/c3dt52350k
6828 | Dalton Trans., 2014, 43, 6828–6841
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
iron oxidation state and ring cleavage positions are manifes- small amounts of extradiol cleavage products in DMF. In con-
tations of the different mechanisms utilized by the two classes trast, 2 elicits regioselective extradiol cleavage with 54.0% yield
of enzymes. Thus, in intradiol dioxygenases, substrate binding in dichloromethane and provides approximately equal
to the ferric center leads to activation of the substrate for amounts of both intra- and extradiol cleavage products in
direct attack by dioxygen, while in the extradiol dioxygenases water. It is remarkable that the same complex elicits exclusive
substrate binding to the iron activates the ferrous center for extradiol cleavage in SDS micellar medium (79.4%), but lower
dioxygen binding.3–6 Extradiol dioxygenases have been histori- extradiol cleavage product and selectivity in aqueous (30.0%)
cally less well characterized than intradiol dioxygenases due to and aqueous TX-100 (33.0%, E/I, 7 : 1) and CTAB (32.0%,
the lack of any pronounced spectral features of the enzymes. 1.4 : 1) media. Further, the N-methyl substitution and variation
However, in recent years, significant progress has been in chelate ring size of the complexes tune the cleavage product
made in gaining
dioxygenases.7–14
a
better understanding of extradiol and selectivity in micellar media. The rates of oxygenation are
several-folds higher in micellar media than in non-aqueous
In our laboratory, iron(III) complexes of several linear 3N,15 solvents illustrating the micellar effect.
3NO,16 N2O2 ligands have been investigated with an aim to
17
mimic the extradiol-cleaving enzymes and correlate the reac-
tion rate as well as extradiol cleavage yields of the functional
model complexes with ligand steric and environment demand.
Experimental
The complex [Fe(bpa)Cl3], where bpa is bis(pyrid-2-ylmethyl)- Syntheses of ligands
amine, with facial coordination of the 3N ligand,15 catalyzes
The ligands (2-pyrid-2-ylethyl)(pyrid-2-ylmethyl)amine (L1),
intradiol cleavage of catechols in DMF but extradiol cleavage
in dichloromethane. Also, interestingly, the same complex
yields 84.0% of extradiol cleavage products in aqueous SDS
micellar medium, and effects the oxidative cleavage with a rate
hundred-fold higher in SDS than that in DMF solution and
ten-fold higher than that in DCM solution.18 All these obser-
vations prompted us to vary the chelate ring sizes in the 1 : 1
iron(III) complexes of a few new bis(pyrid-2-yl) based 3N
ligands (Scheme 1) and study its effect on the rate of dioxy-
genation and the extradiol(E)-to-intradiol(I) cleavage selectivity
(E/I) in different solvents and micellar media. The single
crystal X-ray structures of [Fe(L1)Cl3] 1 and [Fe(L4)Cl3] 4 have
been obtained in this study while that of 2 has already been
reported.19 All of them contain three coordinated chloride
ions, which can be replaced by catechol substrate. The 3N
methyl(2-pyrid-2-ylethyl)(2-pyrid-yl-methyl)amine (L2) and bis-
(pyrid-2-ylethyl)amine (L3) were prepared by using the litera-
ture methods.19–21
(1-Methyl-1H-imidazol-2-ylmethyl)(pyrid-2-ylethyl)amine (L4).
This was synthesized by employing the procedure20 used for
preparing L2 and using 1-methylimidazol-2-carboxaldehyde
(1.10 g, 10 mmol). The pale yellow oil (1.77 g, yield 82%)
obtained was used for isolation of the iron(III) complex without
further purification. 1H NMR (CDCl3): δ 8.60–6.60 (6H, m,
aromatic), 3.90 (2H, s, NCH2), 3.75 (3H, s, NCH3), 2.94 (4H, s,
NCH2CH2), 2.0 (1H, s, NH).
Preparation of iron(III) complexes
[Fe(L1)Cl3] 1. To a solution of L1 (0.43 g, 2 mmol) in metha-
ligands in 1 and 4 are meridionally coordinated while that in 2 nol (10 mL) was added FeCl3 (0.32 g, 2 mmol) in methanol
is facially coordinated.19 Complexes 1, 3 and 4 elicit major (5 mL), stirred well and cooled. The yellow precipitate obtained
amounts of intradiol cleavage products (17.0–70.0%) and very was filtered off, washed with small amounts of cold methanol
and dried under vacuum. Yield: 0.29 g (77%). Anal. calcd for
C13H15N3Fe1Cl3: C, 39.87; H, 3.62; N, 11.63. Found: C, 39.80;
H, 3.7; N, 11.6. An acetonitrile solution of the complex on slow
evaporation yielded yellow crystals suitable for X-ray diffraction
studies.
[Fe(L2)Cl3] 2. This was prepared using the procedure
reported previously.19
[Fe(L3)(DBC)Cl] 3a. To a 2 mL solution (methanol–acetone,
3 : 2 v/v) of the ligand L3 (0.454 g, 2 mmol) was added a solu-
tion of FeCl3 (0.32 g, 2 mmol) in the same solvent mixture
(8 mL) with stirring. To this was then added a hot solution
(2 mL) of H2DBC (0.454 g, 2 mmol) dissolved in the same
solvent mixture. Triethylamine (0.40 g, 4.0 mmol) was then
added to the reaction mixture and the reaction mixture was
refluxed for a few seconds. The dark purple microcrystals of 3a
formed upon slow cooling and standing for over 12 h were fil-
tered off, washed with small amounts of cold methanol and
then dried under vacuum. Yield: 0.36 g (66%). Anal. calcd for
Scheme 1 Structures of tridentate 3N ligands.
C28H37N3Fe1O2Cl1: C, 62.4; H, 6.92; N, 7.80; O, 5.94. Found:
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6828–6841 | 6829

View Article Online
Paper
Dalton Transactions
C, 61.6; H, 6.89; N, 7.78; O, 5.92 and confirmed by ESI-MS, were located from the difference Fourier map and refined
(C19H17N3OFeCl)2+, m/z, 537.
isotropically.
[Fe(L4)Cl3] 4. The complex [Fe(L4)Cl3] was prepared by
adding L4 (0.40 g, 2 mmol) to a warm solution of FeCl3 (0.32 g,
Characterization of oxygenation products
2 mmol) in acetonitrile (8 mL), stirring the solution well and The three major cleavage products 3,5-di-tert-butyl-1-oxacyclo-
then cooling. The pale yellow colored microcrystalline product hepta-3,5-diene-2,7-dione (a), 3,5-di-tert-butyl-2-pyrone (c) and
was filtered off, washed with small amounts of cold methanol 4,6-di-tert-butyl-2-pyrone (d) were separated by column chrom-
and dried under vacuum. Yield: 0.30 g (78%). Anal. calcd for atography over silica gel (60–120 mesh) using 5–10% ethylace-
C12H16N4Fe1Cl3: C, 38.08; H, 4.26; N, 14.8. Found: C, 37.92; H, tate in n-hexane and identified using retention times of GC-MS
1
4.23; N, 14.23. A methanol–acetonitrile (6 : 1 v/v) solution of (EI) and H NMR spectroscopy. The other minor products 3,5-
the complex on refrigeration yielded yellow crystals suitable for di-tert-butyl-5-(carboxymethyl)-2-furanone (b), 4,6-di-tert-butyl-
X-ray diffraction studies.
oxepine-2,3-dione (e), 5,7-di-tert-butyloxepine-2,3-dione (f),
3,5-di-tert-butyl-[1,2]benzoquinone (g) and 3-tert-butyl-furan-
2,5-dione (h) were analyzed as a mixture and identified by
GC-MS (EI) analysis. The regioisomers e/f and c/d were distin-
guished by comparison of their retention times of GC (FID)
and GC-MS (EI) and the intensity of the fragmentation pattern
in the mass spectrum. All of the products (Scheme 2) were
quantified using GC (FID) with the following temperature
program: injector temperature 130 °C; initial temperature
60 °C, increasing at a heating rate of 10 °C min−1 to 130 °C,
then increasing at a rate of 2 °C min−1 to 160 °C, and then
increasing at a rate of 5 °C min−1 to 260 °C; FID temperature
280 °C. GC-MS analysis was performed under conditions iden-
tical to those used for GC analysis; retention times in GC-MS
(EI): 14.6 min for d, 15.9 min for e, 16.4 min for c, 18.1 min
for f and 21.2 min for a.
Crystallographic refinement and structure solution
A crystal of suitable size selected from the mother liquor was
immersed in paraffin oil, then mounted on the tip of a glass
fiber and cemented using epoxy resin. Intensity data for the
crystals were collected using MoKα (λ = 0.71073 Å) radiation on
a Bruker SMART APEX diffractometer equipped with a CCD
area detector at 273 and 293 K. The crystallographic data are
collected in Table 1. The SMART22 program was used for col-
lecting frames of data, indexing the reflections, and determi-
nation of lattice parameters; the SAINT22 program for
integration of the intensity of reflections and scaling; the
SADABS23 program for absorption correction, and the
SHELXTL24 program for space group and structure determi-
nation, and least-squares refinements on F2. The structure was
solved by the heavy atom method. Other non-hydrogen atoms
were located in successive difference Fourier syntheses. The
final refinement was performed by full-matrix least-squares
analysis. Hydrogen atoms attached to the ligand moiety
Kinetics and reactivity studies
Kinetic analyses15–17,25–35 of the catechol cleavage reactions
were carried out by time-dependent measurement of the dis-
appearance of the low energy catecholate-to-iron(III) LMCT
band at ambient temperature (25 °C) by exposing to molecular
oxygen the catecholate adducts prepared in situ. The water and
micellar solutions were equilibrated at the atmospheric
pressure of O2 at 25 °C and the solubility of O2 at 25 °C in
water is 8.1 × 10−4 M.36 Stock solutions of the adducts [Fe(L)-
(DBC)Cl] were generated in situ by treating 1–4 (1.0 × 10−2 M)
with an equivalent amount of H2DBC pretreated with two
equivalents of Et3N. Oxygenation was started by rapid delivery
of the calculated amount of the complex and the H2DBC from
the stock solution using a syringe to O2-saturated solvent/
micelle solution. Kinetic analyses of the catechol cleavage reac-
tions were carried out by time dependent measurement of the
disappearance of the low-energy DBC2−-to-iron(III) LMCT band
in the presence and absence of chloride ions. The complexes
[Fe(L)Cl3] were treated with three equivalents of AgClO4·H2O
in acetonitrile solution and the solution centrifuged to remove
the silver chloride precipitate. Stock solutions (6.0 × 10−3 M)
of the adducts [Fe(L)(DBC)(Sol)]+ were prepared by treat-
ing solutions of [Fe(L)(Sol)3]3+ prepared in situ with an equi-
valent amount of H2DBC pretreated with two equivalents
of Et3N.
Table 1 Crystal data and structure refinement for [Fe(L1)Cl3] 1 and
[Fe(L4)Cl3] 4
1
4
Empirical formula
C
13H15Cl3FeN3
C12H16Cl3FeN4
378.49
Monoclinic
P21/n
Formula weight
Crystal system
Space group
Crystal size
a (Å)
b (Å)
c (Å)
375.48
Monoclinic
P21/c
0.37 × 0.13 × 0.08 0.29 × 0.12 × 0.08 nm
10.8426(11)
8.6483(8)
16.3104(16)
90
97.114(2)
90
1517.7(3)
4
1.643
7.4610(7)
15.9155(15)
13.6602(13)
90
103.724(2)
90
1575.8(3)
4
1.595
772
100(2)
0.71073
9373
3665
α (°)
β (°)
γ (°)
V (Å3)
Z
ρcalc (g cm−3
F(000)
)
764
Temperature (K)
100(2)
0.71073
8875
Radiation [MoKα] Å
No. reflections collected
No. unique reflections
3565
No. of refined parameters 221
186
0.0545/0.1017
0.0676/0.1062
R1/wR2 [I > 2σ(I)]a
R1/wR2 [all data]
0.0532/0.1025
0.0641/0.1066
The catechol cleavage activity of all the complexes toward
H2DBC was examined by exposing a solution of iron(III)–DBC2−
adduct generated in situ in dichloromethane (DCM) and water
a R1 = [∑(||Fo| − |Fc||)/∑|Fo|], wR2 = {[∑(w(Fo2 − Fc2)2)/∑(wFo4)]1/2}.
6830 | Dalton Trans., 2014, 43, 6828–6841
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Scheme 2 Cleavage products of H2DBC mediated by [Fe(L)Cl3] complexes using molecular oxygen: 3,5-di-tert-butyl-5-(carboxymethyl)-2-fura-
none (a). (2,4-Di-tert-butyl-5-oxo-2,5-dihydro-furan-2-yl)-N,N-dimethyl-acetamide (b), 3,5-di-tert-butyl-2-pyrone (c), 4,6-di-tert-butyl-2-pyrone
(d), 3,5-di-tert-butyl-[1,2]benzoquinone (e) and 3-tert-butyl-furan-2,5-dione (f).
aqueous micellar media (SDS, CTAB and TX-100) to molecular an atmosphere of nitrogen was maintained over the solutions
oxygen. The product analysis was carried out by stirring the during measurements. The E1/2 values were observed under
complex [Fe(L)3(Sol)3]3+ (0.1 mmol), H2DBC (0.1 mmol) and identical conditions for various scan rates. The instruments
triethylamine (0.2 mmol) in the dichloromethane (20 mL) utilized included an EG & G PAR 273 Potentiostat/Galvanostat
solvent under molecular oxygen over 4 h at room temperature. and Pentium-IV computer along with EG & G M270 software to
After the reaction was completed, the reaction mixture was carry out the experiments and to acquire the data. The product
concentrated under reduced pressure and extracted with analyses were performed using a HP 6890 GC series Gas Chro-
diethylether (3 × 5 mL). The remaining residue was acidified matograph equipped with an FID detector and an HP-5 capil-
with HCl to pH 3 to decompose the metal complexes and lary column (30 m × 0.32 mm × 2.5 μm) and GC-MS analysis
extracted with diethylether (3 × 5 mL). The combined extracts was performed on a Perkin Elmer Clarus 500 GC-MS instru-
were dried over Na2SO4 and then concentrated.
ment using a PE-5 with the previously15 reported temperature
program.
Physical measurements
Elemental analyses were performed on a Perkin Elmer Series II
CHNS/O Analyzer 2400. 1H NMR spectra were recorded on a
Bruker 400 MHz NMR spectrometer. The electronic spectra
were recorded on an Agilent diode array-8453 spectrophoto-
Results and discussion
Synthesis of ligands and iron(III) complexes and reactivity
studies
meter. The EPR spectra were recorded on a JEOL JES-TE 100 X- The linear 3N ligands L1 and L4 were synthesized by the reac-
band spectrometer. ESI-mass spectrometry was performed on tion of pyridine-2-aldehyde (L1)/N-methylimidazol-2-aldehyde
a Thermo Finnigan LCQ 6000 Advantage Max instrument. (L4) with 2-aminoethylpyridine to form a Schiff base and then
Cyclic voltammetry (CV) and differential pulse voltammetry reduction of the latter with sodium borohydride. The ligand
(DPV) were performed using a three electrode cell configur- L3 was synthesized by using the Michael addition of an excess
ation. A platinum sphere, a platinum plate and Ag(s)/Ag+ were of 2-vinylpyridine with ammonium chloride.21 The ligands L1,
used as working, auxiliary and reference electrodes respect- L2 and L4 were reacted with equimolar amounts of FeCl3 in
ively. The supporting electrolyte used was NBu4ClO4 (TBAP). methanol to obtain mononuclear iron(III) complexes in good
The temperature of the electrochemical cell was maintained at yields. Complexes 1, 2 and 4 are formulated as [Fe(L)Cl3] on
25.0 0.2 °C by a cryocirculator (HAAKE D8 G). By bubbling the basis of elemental analysis, which is supported by their
research grade nitrogen the solutions were deoxygenated and X-ray crystal structures.19 All the iron(III) complexes have
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6828–6841 | 6831

View Article Online
Paper
Dalton Transactions
magnetic moments in the range 5.9–6.0 BM at room tempera-
ture, which is characteristic of a high-spin ferric center. Con-
ductivity studies (ΛM, 5.2–7.9 Ω−1 cm2 mol−1
)
in
dichloromethane (DCM) solution suggest that the chloride
ions in all the complexes remain coordinated to iron(III).
However, in DMF solution (ΛM, 90–120 Ω−1 cm2 mol−1) one of
the chloride ions is not coordinated revealing that the labile
chloride ions can be replaced by bidentate catecholate anions.
The imidazole nitrogen donor in L4 would mimic the histidine
coordination in CTD and PCD enzymes. The catecholate
adducts of the complexes were prepared in situ in aqueous and
aqueous micellar media for spectral and electrochemical and
reactivity studies.
Description of structures of [Fe(L1)Cl3] 1 and [Fe(L4)Cl3] 4
The X-ray crystal structures of complexes 1 and 4 are shown in
Fig. 1 and 2 respectively and the selected bond lengths and
bond angles are listed in Tables 1 and 2. The iron(III) center in
1 is six-coordinate and the deviation of the bond angles at
iron(III) from the ideal angles of 90° and 180° indicates a dis-
torted octahedral coordination geometry. The ligand L1 is co-
ordinated meridionally to the iron(III) center and the
remaining coordination sites are occupied by three chloride
ions. The two Fe–Npy bond lengths (2.166(3); 2.170(3) Å) are
almost equal to the Fe–Namine bond (2.164(3) Å). This is inter-
esting because generally the Fe–Namine bond is expected to be
longer than the Fe–Npy bond due to sp3 and sp2 hybridizations
respectively of amine and imine nitrogens.15a,16,33 The meri-
dional coordination of the L1 ligand renders the lone pair
orbital on the amine nitrogen atom constrained to be oriented
exactly towards the iron(III) orbital and hence shortens the
Fe–Namine bond. Both the Fe–Cl1 (2.3659(9) Å) and Fe–Cl3
(2.3447(9) Å) bonds in 1 are longer than the Fe–Cl2 bond
(2.2974(9) Å) trans to the Fe–Namine bond. So, the Cl1 and Cl3
atoms are expected to be more labile than Cl2 and hence
would be displaced upon catecholate binding.
Fig. 2 Molecular structure of [Fe(L4)Cl3] 4 (50% probability factor for
the thermal ellipsoid). Hydrogen atoms have been omitted for clarity.
Table 2 Selected bond lengths [Å] and bond angles [°] for [Fe(L1)Cl3] 1
and [Fe(L4)Cl3] 4
1
4
Fe(1)–N(2)
Fe(1)–N(1)
Fe(1)–N(3)
Fe(1)–Cl(2)
Fe(1)–Cl(3)
Fe(1)–Cl(1)
N(2)–Fe(1)–N(1)
N(2)–Fe(1)–N(3)
N(1)–Fe(1)–N(3)
N(2)–Fe(1)–Cl(2)
N(1)–Fe(1)–Cl(2)
N(3)–Fe(1)–Cl(2)
N(2)–Fe(1)–Cl(3)
N(1)–Fe(1)–Cl(3)
N(3)–Fe(1)–Cl(3)
Cl(2)–Fe(1)–Cl(3)
N(2)–Fe(1)–Cl(1)
N(1)–Fe(1)–Cl(1)
N(3)–Fe(1)–Cl(1)
Cl(2)–Fe(1)–Cl(1)
Cl(3)–Fe(1)–Cl(1)
2.164(3)
2.166(3)
2.170(3)
2.2974(9)
2.3447(9)
2.3659(9)
74.99(11)
91.39(11)
166.27(11)
167.99(8)
93.03(8)
100.61(8)
86.85(9)
91.01(8)
86.44(8)
94.32(3)
86.35(9)
93.05(8)
87.70(8)
93.59(3)
170.90(4)
Fe(1)–N(4)
Fe(1)–N(1)
Fe(1)–N(2)
Fe(1)–Cl(2)
Fe(1)–Cl(1)
Fe(1)–Cl(3)
N(4)–Fe(1)–N(1)
N(4)–Fe(1)–N(2)
N(1)–Fe(1)–N(2)
N(4)–Fe(1)–Cl(2)
N(1)–Fe(1)–Cl(2)
N(2)–Fe(1)–Cl(2)
N(4)–Fe(1)–Cl(1)
N(1)–Fe(1)–Cl(1)
N(2)–Fe(1)–Cl(1)
Cl(2)–Fe(1)–Cl(1)
N(4)–Fe(1)–Cl(3)
N(1)–Fe(1)–Cl(3)
N(2)–Fe(1)–Cl(3)
Cl(2)–Fe(1)–Cl(3)
Cl(1)–Fe(1)–Cl(3)
2.092(3)
2.186(3)
2.198(3)
2.2951(9)
2.3365(9)
2.3781(9)
164.44(11)
75.24(10)
89.39(11)
93.70(8)
101.59(8)
168.87(8)
93.62(8)
87.96(7)
87.24(8)
94.84(3)
92.63(8)
83.62(7)
85.63(8)
93.76(3)
169.01(4)
The iron(III) center in [Fe(L4)Cl3] 4 is located in a distorted
octahedral environment composed of the three nitrogen atoms
of the meridionally coordinating L4 ligand and three chloride
ions. The Cl–Fe–Cl bond angles are higher than 90° and the
three Fe–Cl bond distances are different indicating that the
coordination geometry is rhombically distorted octahedral.
The Fe–Namine bond (Fe–N2, 2.198(3) Å) is in the same range
as the Fe–Npy (Fe–N1, 2.186(3) Å) and Fe–Nim (Fe–N3, 2.092(3)
Å) bonds, which is expected from the meridional coordination
Fig. 1 Molecular structure of [Fe(L1)Cl3] 1 (50% probability factor for
the thermal ellipsoid). Hydrogen atoms have been omitted for clarity.
6832 | Dalton Trans., 2014, 43, 6828–6841
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
of the ligand (cf. above). The observed Fe–Npy bond distance is capable of forming 5,5 or 6,6 membered chelate rings adopt
almost the same as the average Fe–Npy bond distance (2.184 Å) facial coordination. And N-alkyl substitution leads to cis-facial
and is slightly longer than those in 1 but it is lower than the coordination of the ligands.
Fe–Nim bond distance, obviously because the imidazole nitro-
Electronic absorption spectra
gen donor is a more strongly coordinating σ-donor than the
pyridine nitrogen [pKa (BH+): pyridine, 5.2; imidazole, 7.01].
The electronic absorption spectra of all the iron(III) complexes
The Fe–Cl3 bond (2.3781(9) Å) is significantly longer than the
other two bonds (Fe–Cl1, 2.3365(9), Fe–Cl2, 2.2951(9) Å).
in dichloromethane solution exhibit three intense bands in
the range 250–400 nm (Table S1†). In order to understand the
ability of the complexes to interact with catecholate substrates
and to probe the mode of micellar solubilisation of the cate-
cholate adducts, the complexes were treated with equimolar
amounts of 3,4-di-tert-butylcatechol (H2DBC), pretreated with
two moles of triethylamine in dichloromethane (DCM), aceto-
nitrile and aqueous SDS, CTAB and TX-100 micellar media,
and the significant spectral data obtained are collected in
Table 3. The two visible bands (465–530, 690–860 nm, Table 3,
It is interesting to compare the structures of 1 and 4 with
those of iron(III) complexes of the related 3N ligands L2 and
bpa. In [Fe(bpa)Cl3] the ligand bpa is facially coordinated,33
and the Fe–Npy bonds (Fe–N1py, 2.180, Fe–N2py, 2.183 Å) are
shorter than the Fe–Namine bond (2.194 Å), which is expected
from the sp2 and sp3 hybridizations of imine and amine nitro-
gens. Obviously, the linear 3N ligands L1 and L4 capable of
forming adjacent 5,6 chelate rings around iron(III) adopt meri-
dional coordination while those (bpa) capable of imposing
adjacent 5,5 chelate rings adopt facial coordination so as to
have effective overlapping of metal–ligand orbitals. Thus it is
expected that the ligand L3 capable of forming adjacent 6,6
chelate rings would coordinate to iron(III) meridionally. But it
is unfortunate that the complex could not be isolated and also
attempts to grow single crystals of its catecholate adduct 3a
were not fruitful. Also, the ligand L2 in [Fe(L2)Cl3] 2 with a dis-
torted octahedral coordination geometry is involved in facial
coordination to iron(III). This reveals that the incorporation of
the sterically hindering methyl group on the amine nitrogen of
L1 in 1 to obtain 2 leads to rendering the lone pair orbital on
the amine nitrogen atom not oriented exactly towards the
iron(III) orbital, rendering the Fe–Namine bond significantly
longer (2.254(6) Å). A similar elongation in Fe–Namine bond and
cis-facial coordination of the 3N ligand has been observed for
1 : 1 iron(III) complexes15a of N-alkyl-bis(pyrid-2-ylmethyl)amine
ligands. Further, the incorporation of the methyl group on the
amine nitrogen in [Fe(bpa)Cl3]33 to obtain [Fe(L5)Cl3] 5, where
L5 is N-methyl-bis(pyrid-2-ylmethyl)amine, leads to no change
in the cis-facial coordination of the bpa ligand but elongation
of the Fe–Namine bond (2.259(3) Å).15a,33 Thus in 1 : 1 iron(III)
complexes linear 3N ligands capable of forming 5,6 membered
chelate rings adopt meridional coordination while those
Fig.
3
and 4) that appear on adding DBC2− are
attributed12–17,25–32,34,35 to catecholate-to-iron(III) LMCT tran-
sitions involving two different catecholate frontier orbitals.
The positions of both the low and high energy catecholate-to-
iron(III) LMCT bands vary in the order 2 > 1 > 3 > 4 in all the
media. Upon replacing one of the pyridyl moieties in [Fe(L1)-
(DBC)Cl] by an N-methylimidazolyl moiety to obtain [Fe(L4)-
(DBC)Cl] both the bands are shifted to higher energies,
which is expected35,37 from the higher basicity of imidazole
[pKa (BH+): pyridine, 5.2; imidazole, 7.01]. Upon incorporat-
ing a methyl group in [Fe(L1)(DBC)Cl] to obtain [Fe(L2)-
(DBC)Cl] both the LMCT bands are shifted to lower energy
(798 to 894 nm in CH3CN, 773 to 942 nm in SDS), which is
expected of elongation of the Fe–Namine bond (2.164 to 2.254 Å,
cf. above).35,37 The replacement of the five-membered chelate
ring in [Fe(L1)(DBC)]+ with meridional coordination of the
ligand by a six-membered one as in [Fe(L3)(DBC)]+ with cis-facial
coordination enhances the electron density on iron(III) and shifts
the LMCT band to higher energy35 (674 to 660 nm in water, 773
to 671 nm in SDS). Thus the energy of the low energy LMCT
band reflects12–17,25–32,34,35 the Lewis acidity of the iron(III)
center, as modified by the stereoelectronic factors of ligands.
The wavelength of the higher energy catecholate-to-iron(III)
LMCT band decreases in the order ACN > DCM > SDS > water
Table 3 Electronic absorption spectral data of [Fe(L)Cl3] adductsa in acetonitrile, dichloromethane, water and micellar mediab
λmax in nm (εmax in M−1 cm−1
)
Complex
Acetonitrile
DCM
Water
SDS
CTAB
TX-100
[Fe(L1)Cl3]
+DBC2−
798 (1640)
521 (1240)
894 (1640)
531 (1240)
—
654 (1790)
849 (1760)
506 (1335)
708 (1580)
498 (1660)
855 (1640)
521 (1280)
—
622 (1240)
858 (1660)
525 (1240)
674 (1730)
529 (1734)
—
627 (1640)
—
660 (1650)
—
633 (1060)
773 (1830)
554 (1580)
942 (1060)
580 (870)
—
671 (1870)
856 (1960)
545 (1700)
—
675 (1160)
534 (1160)
—
610 (1230)
—
525 (1270)
—
566 (1400)
—
[Fe(L2)Cl3]
+DBC2−
[Fe(L3)Cl3]
+DBC2−
648 (1575)
656 (1780)
[Fe(L4)Cl3]
+DBC2−
596 (1110)
597 (1940)
a The iron(III) complexes were reacted with one equivalent of H2DBC pretreated with two equivalents of triethylamine. b Surfactant concentration:
SDS, 13.0 × 10−3 M; CTAB, 6.0 × 10−3 M and TX-100, 12.5 × 10−3 M.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6828–6841 | 6833

View Article Online
Paper
Dalton Transactions
and hence a higher reactivity18 for the adduct with O2 in non-
aqueous solvents. For the same adduct this difference
(∼220 nm) in SDS micellar medium is higher than that
(140 nm) in TX-100 micellar medium suggesting a higher reac-
tivity in SDS medium also. Also, this difference for the DBC2−
adduct of 1 is higher than those for 2 and 4 suggesting a
higher reactivity for the adduct in these media.37 All the above
observations support the presence of the catecholate adducts
as [Fe(L)(DBC)(H2O)]+ rather than as [Fe(L)(DBC)Cl] species in
aqueous solution and in the hydrophobic regions of aqueous
SDS micellar solutions, but mostly in the aqueous phase in
CTAB micelles.
Fig. 3 Electronic absorption spectra of [Fe(L1)Cl3]-DBC2− adducts
(2.0 × 10−4 M) generated in situ by treating complexes with DBC2−
(2 equivalents of Et3N) in aqueous solution (a); SDS (b); CTAB (c) and
TX-100 (d).
Electron paramagnetic resonance spectra
The frozen solution EPR spectra of 1–4 and their DBC2−
adducts display (Fig. 5, Table 4) high-spin27,31,32,40,41 (S = 5/2)
rhombic ferric signals27,40,41 around g = 8.00, 4.00 and 2.00,
Fig. 4 Electronic absorption spectra of [Fe(L2)Cl3]-DBC2− adducts
(2.0 × 10−4 M) generated in situ by treating complexes with DBC2−
(2 equivalents of Et3N) in aqueous solution (a); SDS (b); CTAB (c) and
TX-100 (d).
∼ TX-100 ∼ CTAB. The LMCT band energies in aqueous SDS
micellar solution are lower than those in aqueous solutions
but are higher than those in DCM and CH3CN solutions. This
suggests18 that the cationic adduct species [Fe(L)(DBC)(H2O)]+
rather than the neutral species [Fe(L)(DBC)Cl] interact with
the anionic SDS head groups and that they are docked on the
anionic micellar surface in the first Helmholtz layer with the
ligand moieties located in the hydrophobic regions of the SDS
micelles. The spectral data obtained in CTAB and
TX-100 micellar solutions are close to those in aqueous solu-
tions, revealing that most of the catecholate adducts are
located in the aqueous rather than in the hydrophobic micellar
pseudophase and that the species present in the Stern layer
are possibly the neutral [Fe(L)(DBC)Br] rather than the cationic
[Fe(L)(DBC)(H2O)]+ species. The higher wavelength observed
for the adduct of 2 in SDS micelles may be due to the weaker
interaction of the cationic adduct with the anionic SDS
micellar surface.13,18,31,38,39 Also, the difference between the
positions of the high and low energy bands observed in
aqueous solution for the adduct of 1 is around 145 nm while it
is higher (∼310 nm) in DCM and ACN (∼280 nm) solvents
suggesting a large asymmetry in the iron–catecholate bonds
Fig. 5 EPR spectra of [Fe(L2)Cl3] 2 in aqueous solution (a), [Fe(L1)-
(DBC)]+ in aqueous solution (b); [Fe(L1)(DBC)]+ in SDS (c); [Fe(L1)(DBC)]+
in CTAB (d) and [Fe(L1)(DBC)]+ in Triton X-100 (e) at 77 K.
6834 | Dalton Trans., 2014, 43, 6828–6841
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Table 4 Electron paramagnetic resonance spectral data of iron(III) complexesa in water and micellar mediab
DBC2− adduct
Complex
Water
Water
SDS
CTAB
TX-100
[Fe(L1)Cl3]
[Fe(L2)Cl3]
[Fe(L3)Cl3]
[Fe(L4)Cl3]
g values
E/D
g values
E/D
g values
E/D
g values
E/D
7.87, 5.65, 4.01, 2.01
0.068
7.82, 4.51, 4.00, 2.01
0.074
8.83, 5.64, 4.25, 2.00
0.085
8.20, 4.48, 4.01, 2.01
0.085
8.34, 6.93, 5.16, 2.05
0.085
8.67, 5.85, 4.03, 2.03
0.079
9.01, 5.68, 4.24, 2.02
0.086
8.34, 4.58, 4.05, 2.01
0.085
8.38, 5.98, 5.21, 2.06
0.078
8.76, 5.87, 4.05, 2.05
—
8.33, 5.64, 4.21, 2.02
0.086
8.30, 4.56, 4.00, 2.04
—
7.93, 5.93, 4.23, 2.03
0.073
8.37, 5.72, 4.03, 2.02
0.099
8.30, 5.56, 4.21, 2.02
0.085
7.84, 4.57, 4.01, 2.01
0.074
7.91, 5.73, 4.30, 2.03
0.073
8.38, 5.75, 4.02, 2.01
0.093
—
8.61, 5.56, 4.07, 2.00
0.093
a Spectra were recorded at liquid nitrogen temperature. b Surfactant concentration: SDS, 13.0 × 10−3 M; CTAB, 6.0 × 10−3 M and TX-100,
12.5 × 10−3 M.
which are associated with |5/2, −1/2> → |5/2, +1/2> transitions.
The intense signal appearing at the g = 2.02 region is assigned
to the iron(III)-bound semiquinone radical.41 For all the com-
plexes except 4, a sharp increase in E/D values is observed
when DBC2− is added to them in aqueous solution revealing
an increase in rhombicity of the iron(III) coordination sphere
upon replacing the chloride ions by DBC2− ions. For 4, a sharp
decrease in E/D value from 0.093 to 0.079 is observed revealing
a decrease in rhombicity upon adduct formation. However, no
appreciable change in E/D values is observed in SDS, CTAB
and TX-100 micelles for all the complexes upon adduct for-
mation, revealing small or negligible changes in coordination
Fig. 6 Differential pulse voltammograms of 1 mM [Fe(L4)(DBC)2− +
aqueous solution (a); SDS (b); CTAB (c) and TX-100 (d) micellar solution
at 25 °C. Supporting electrolyte: 0.1 M NaCl, scan rate 5 mV s−1
] in
geometry upon adduct formation. However, the observation of
appreciable differences in the E/D values of complexes and
their respective adducts in aqueous solution shows that the
rhombicity is variable27,31,32,40,41 in aqueous medium.
.
On adding one equivalent of DBC2− (H2DBC pretreated
with two equivalents of triethylamine) to 1, 2 and 4 (Fig. 6) the
cathodic peak current of the FeIII/FeII redox couple completely
vanishes, and the redox couple is possibly shifted to enor-
mously more negative potentials due to chelation of DBC2− to
iron(III).15,18 Also, the DBSQ/DBC2− couple appears to be a new
redox wave at potentials (0.190 to −0.092 V) more positive than
the FeIII/FeII redox couple. The redox potentials of this couple
are considerably more positive than that of the free DBSQ/
DBC2− couple (−1.434 V)42 reflecting the significant stabiliz-
ation of DBC2− towards oxidation upon coordination to
iron(III). In aqueous medium for 1 and 3 the E1/2 values of the
couple decrease to different extent on micellization while for 2
and 4 they increase. This shows that the reactivity of the sub-
strate is modified depending upon the micellar medium and
the ligand donor atoms in the complex.
Electrochemical behavior
The electrochemical features of 1–4 and their DBC2− adducts
generated in situ were investigated in water and in aqueous
SDS, CTAB and TX-100 micellar solutions by employing cyclic
(CV) and differential pulse voltammetry (DPV) on a stationary
platinum sphere electrode. All the present complexes exhibit a
coupled pair of redox waves at positive potentials, which are
assigned to the FeIII/FeII redox couple (Fig. 6; Fig. S1–S3;†
Table 5; Tables S2–S4†). The E1/2 values of the FeIII/FeII redox
couple of the complexes with the 5,6 chelate ring systems
follow the trend 1 (0.042) < 2 (0.126) > 4 (0.024 V) in aqueous
solution and in all aqueous micellar media. Upon incorporat-
ing a N-methyl group in 1 to obtain 2, the Fe–Namine bond is
weakened (cf. above) and hence the reduction of iron(III)
becomes facile. Upon replacing the pyridyl nitrogen in 1 by the
more Lewis basic imidazole nitrogen as in 4 (cf. above) the
reduction of iron(III) becomes less facile. Interestingly, the E1/2
Dioxygenase activity of the complexes
values of 1 and 4 in CTAB medium are higher than those in The dioxygenase activities of the present complexes were deter-
SDS medium, possibly due to solubilization–micellization of mined using a known procedure with modifications.7,27,42–44
[Fe(L)(DBC)Br] species as they are located in the hydropohobic The catecholate adducts [Fe(L)(DBC)Cl] were generated in situ
region of the micelles. Also, the E1/2 values of 2 in all the by treating the complexes [Fe(L)Cl3] 1, 2 and 4 with equimolar
micelles are higher than those in aqueous solution, possibly quantities of H2DBC pretreated with two equivalents of Et3N
because the cis-facial coordination of the L2 ligand encourages in DCM, water and aqueous micellar media and then exposed
formation of [Fe(L)(DBC)(H2O)]+ species.
to molecular oxygen. In control reactions performed in the
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6828–6841 | 6835

View Article Online
Paper
Dalton Transactions
Table 5 Electrochemical dataa of [Fe(L3)(DBC)Cl] complexes and its DBC2− adducts in aqueous and aqueous micellarb (SDS, CTAB, TX-100) solu-
tions at 25.0 0.2 °C at a scan rate of 50 mV s−1
E1/2 (V)
Complex
Epc (V)
Epa (V)
ΔE (mV)
CV
DPVd
Redox processc
Water
[Fe(L3)(DBC)Cl]
0.210
−0.650
0.032
—
78
—
0.121
—
0.180
−0.472
DBSQ → DBC2−
FeIII → FeII + LR
SDS
[Fe(L3)(DBC)Cl]
0.226
−0.660
0.126
—
100
—
0.176
—
0.101
−0.576
DBSQ → DBC2−
FeIII → FeII + LR
CTAB
[Fe(L3)(DBC)Cl]
0.076
−0.568
−0.082
—
158
—
0.031
—
0.080
−0.438
DBSQ → DBC2−
FeIII → FeII + LR
TX-100
[Fe(L3)(DBC)Cl]
−0.036
−0.830
0.058
—
94
—
0.047
—
0.068
−0.508
DBSQ → DBC2−
FeIII → FeII + LR
a Potential measured ( 0.002 V) vs. calomel electrode (0.01 M, 0.1 M NaCl). Add 0.244 V to convert to NHE. b Surfactant concentration: SDS,
8.2 × 10−3 M. CTAB, 1.2 × 10−3 M and TX-100, 3.1 × 10−3 M. c Reduction of dissociated species, LR = ligand reduction. d Differential pulse
voltammetry; scan rate, 5 mV s−1
.
absence of complexes under the same reaction conditions, no
The product analysis performed after 12 h of reaction
cleavage but oxidized products were observed. For 3, the revealed that in DCM solvent the catecholate adducts of all the
complex was generated in situ and then treated with DBC2− to present complexes except 2 afford major amounts of intradiol
generate the adduct species [Fe(L3)(DBC)Cl]. The disappear- (17.0–70.0%) and smaller amounts of extradiol (1.2–4.2%) clea-
ance of the low-energy catecholate-to-iron(III) LMCT band vage products with the product selectivity (E/I) being 0.08 : 1
(Fig. 7) on oxygenation exhibits pseudo-first order kinetics as (1), 0.02 : 1 (3) and 0.30 : 1 (4). The highest yield of intradiol
judged from the linearity of the plot [1 + log(Absorbance)] vs. cleavage products (70%) was observed for 3 while major
time, and the values of kobs were obtained from the slopes of amounts of extradiol (54.0%; E/I, 3.0 : 1) products were
these plots. The second order rate constants and then t1/2 observed for 2. In aqueous solution, 1, 3 and 4 show higher
values were calculated (Table 6) using the equations: t1/2
=
yields of intradiol cleavage products (E/I: 1, 0.31 : 1; 3, 0.02 : 1,
4, 0.27 : 1) while 2 shows a higher yield of extradiol cleavage
products (30%; E/I: 1.0 : 1). It is remarkable that in aqueous
SDS micellar medium 2 shows exclusive extradiol cleavage pro-
ducts (79.4%) while all the other complexes show very low
selectivity (E/I: 1, 0.03 : 1; 2, 79.4 : 0, 3, 0.06 : 1, 4, 0.06 : 1). Also,
the TX-100 micellar medium provides a more suitable hydro-
phobic environment for 2 to elicit extradiol cleavage (E/I: 1,
0.04 : 1; 2, 7 : 1; 3, 0.04 : 1; 4, 0.1 : 1) than CTAB micellar
medium (E/I: 1, 1.2 : 1; 2, 1.4 : 1; 3, 0.08 : 1; 4, 0.8 : 1).
15–18,32,34,35
0.693/kobs
.
kO ¼ kobs=½O2ꢀ
2
To illustrate the ability of the complexes to elicit extradiol
cleavage and also the observed trend in rates of cleavage we
propose a O2 activation mechanism,26,27,38–41 which involves
the key substrate-alkylperoxo-Fe3+ reaction intermediate10,27,42
(Scheme 3). The chloride ion or the solvent molecule in the
coordination site in [Fe(L)(DBC)Cl] is displaced by O2 to form
the superoxide adduct [Fe(L)(DBC)(O2)]+ and the bound super-
oxide interacts with the semiquinone (SQ) radical, followed by
rearrangement, to give the proximal peroxy intermedi-
ate.10,37,44,45 On alkenyl migration the peroxo intermediate
gives extradiol cleavage products and on acyl migration intra-
diol cleavage products. Lipscomb et al.10 have established the
formation of such a substrate-alkylperoxo-Fe2+ intermediate
upon the simultaneous binding of the substrate and O2
to the iron site in the extradiol-cleaving enzyme
Fig. 7 Progress of reaction of [Fe(L4)(DBC)]+ (2 × 10−4 M) with O2 in
SDS micellar solution monitored disappearance of the DBC2−-to-Fe(III)
LMCT band at 856 nm and 545 nm.
6836 | Dalton Trans., 2014, 43, 6828–6841
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Table 6 Kinetic dataa, oxidative cleavage of H2DBC, identification of cleavage products, and % yield of [Fe(L)Cl3] in DCM, aqueous and aqueous
micellar solutionsb
Complex
Medium
Extradiol %
Intradiol %
Others %
kO (M−1 s−1
)
2
[Fe(L1)Cl3]
DCM
H2O
SDS
CTAB
TX-100
4.2(c,d)
14.0(c,d,e)
1.8(c,d)
37.0(c,d)
—
52.3(a)
45.4(a,b)
56.0(a)
29.6(a)
28.0(a)
7.6(g,h)
0.0219 0.0020
0.9082 0.0006
0.8710 0.0013
0.1299 0.0011
0.4156 0.0018
—
—
—
—
[Fe(L2)Cl3]
[Fe(L3)Cl3]
[Fe(L4)Cl3]
DCM
H2O
SDS
CTAB
TX-100
54.0(c,d)
30.0(c,d,e)
79.4(c,d)
32.0(c,d)
33.0(c,d)
18.3(a)
33.0(a,b)
—
23.3(a)
5.0(a)
28.1(g,h)
2.2(g,h)
11.0(g,h)
3.0(h)
0.4873 0.0028
1.6623 0.0001
1.3658 0.0032
0.1430 0.0017
0.0425 0.0015
8.0(h)
DCM
H2O
SDS
CTAB
TX-100
1.2(c,d)
1.0(c,d)
4.0(c,d)
2.0(c,d)
1.1(c,d)
70.0(c,d)
53.0(c,d)
62.5(c,d)
22.5(c,d)
22.5(c,d)
7.0(h)
—
3.0(h)
9.0(h)
9.0(h)
0.1132 0.0002
0.0972 0.0013
0.2695 0.0028
0.8152 0.0008
0.6705 0.0098
DCM
H2O
SDS
CTAB
TX-100
5.0(c,d)
10.0(c,d,e)
1.7(c,d)
22.0(c,d)
—
16.8(a)
6.6(h)
—
—
—
—
0.2536 0.0028
0.4612 0.0002
1.9317 0.0003
0.4497 0.0002
0.2238 0.0032
37.0(a,b)
30.0(a,b)
27.4(a)
23.5(a,b)
a kO = kobs/[O2]. The solubility of O2 in water at 25 °C is 8.2 × 10−3 M. The kinetic data were obtained by monitoring the disappearance of the
2
lower energy LMCT band. b Surfactant concentration: SDS, 13.0 mM; CTAB, 6.0 mM and TX-100, 12.5 mM a, b, c, d, e, f, g and h are the cleavage
products obtained as per Scheme 2.
Scheme 3 Proposed mechanism for the reaction of iron(III) complexes of facially coordinating ligands with O2 leading to extra- and intradiol clea-
vage products.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6828–6841 | 6837

View Article Online
Paper
Dalton Transactions
homoprotocatechuate-2,3-dioxygenase (2,3-HPCD). Using this products while for the adducts of 3 and 4 acyl rather than
mechanism we have successfully illustrated the enhancement alkenyl migration is preferred to result in more of intradiol
in yields of extradiol cleavage products for the 1 : 1 iron(III) rather than extradiol cleavage products. So it emerges from the
complexes of N-alkyl substituted bis(pyrid-2-ylmethyl)amine15 present study that the N-methyl group of the facially coordinat-
and certain sterically hindering 4N ligands.35 Now we propose ing ligand in 2 with a 5,6-chelate ring system facilitates alkenyl
that the higher regioselectivity observed in SDS micellar migration while the meridionally coordinating 3N ligands,
medium for 2 is ascribed to the formation of the adduct which form the 5,6 chelate ring system (1,4) and the
species [Fe(L2)(DBC)(H2O)]+ (cf. above)46 and its location in 6,6-chelate ring system (3), favor mainly acyl migration
large amounts in the hydrophobic regions of the micelles, (Scheme 4) leading to regioselective cleavages. The six-mem-
which facilitates the attack of molecular oxygen (cf. above). bered chelate ring would cause the Fe–Ocat bonds to be
This is very similar to our earlier finding that the complex [Fe- different in length (cf. above) and the longer Fe–Ocat bond
(bpa)Cl3] and its N-alkyl substituted analogues with facial would dissociate leading to monodentate coordination of cate-
coordination of the 3N ligands18 yield intradiol cleavage pro- chol, which has been proposed18,43 to facilitate extradiol clea-
ducts exclusively (89%) in DMF as a solvent and, interestingly, vage. So it is clear that apart from the bulkiness of N-alkyl
yield higher amounts of extradiol cleavage products (84%: E/I: groups of the facially coordinating 3N ligands, the presence of
61.0 : 1) in SDS micellar medium. Also, the adduct species the six-membered chelate ring composed of py nitrogens
[Fe(L)(DBC)(H2O)]+ would interact strongly with the anionic favours regioselective formation of cleavage products.
head group of SDS leading to efficient micellization and stron-
The trend in the second order reaction rate constants
ger interaction with both the substrate and dioxygen while the observed (kO , 0.02–1.93 M−1
s
−1, Table 6) for the adducts
2
species [Fe(L)(DBC)Br] weakly with the cationic CTAB micelles [Fe(L)(DBC)]+ derived from 1–4 may be accounted for by invok-
illustrating the higher cleavage yields in the former. Also, both ing the oxygen activation mechanism (Scheme 3) in which the
the alkenyl and acyl migrations compete with each other in rate determining step is O2 attack on the substrate.13,14,27,42,43,47
CTAB micelles to form higher amounts of extradiol cleavage A two state reaction pathway,8,9,48 which involves a spin-inver-
products, as also observed in DCM. The neutral sion process from the spin-sextet state (S = 5/2) in (L)FeIII-Cat2−
TX-100 micelles would be expected to solubilize the adduct to the upper lying spin-doublet (S = 1/2) state in (L)FeII-SQ
species much better than the cationic micelles leading to upon O2 approaching the iron(III) center, has been envisaged.
higher amounts of extradiol cleavage products (E/I, 7.0 : 1) and As O2 approaches the iron(III)–catecholate adduct, the ligand
also alkenyl migration is preferred over acyl migration.24,42,43,47
In contrast to the adduct of 2, which shows the highest
yield of extradiol cleavage products (54%; E/I: 3 : 1), those of 1,
3 and 4 show a lower selectivity in DCM (E/I, 0.08 : 1(1),
0.02 : 1(3), 0.30 : 1(4)) because the meridional coordination of
the 3N ligands in the latter does not favor the formation of the
substrate-alkylperoxo-Fe3+ intermediate responsible10 for extra-
diol cleavage to occur, but provides intradiol cleavage products
due to the preference of acyl rather than alkenyl migration2b,
c,18
in the adduct. In aqueous solution the product selectivity
for 1, 3 and 4 is lower (E/I, 0.02–0.31 : 1; 1 : 1 (2)) suggesting
that alkenyl rather than acyl migration is preferred for 2. In
SDS micellar solution also exclusive alkenyl migration is pre-
ferred for 2 while for others acyl migration is preferred. The
bulky N-methyl group at the central amine nitrogen donor in
[Fe(L2)(DBC)Cl] allows both the substrate and molecular
oxygen to occupy the opposite face of the iron(III) coordination
sphere thereby bringing them in close proximity and in proper
orientation to form the substrate-alkylperoxo-Fe3+ intermedi-
ate. Also, the higher Lewis acidity of the iron(III) center in 2
(cf. above) encourages catecholate binding and enhances the
O2 attack. Further, the N-methyl group in [Fe(L2)(DBC)Cl] is
oriented away from the sp3 hybridized carbon atom of the cate-
cholate substrate and thus facilitates alkenyl migration
(Scheme 3) leading to regioselective extradiol cleavage with
higher product selectivity.
In CTAB and TX 100 micelles for the adducts of 1 and 2,
Scheme 4 Proposed mechanism for the reaction of Fe(III) complexes of
both alkenyl and acyl migrations compete with each other
meridionally coordinating ligands with O2 leading to extradiol (a) and
equally to afford equal amounts of extra- and intradiol intradiol cleavage (b) products.
6838 | Dalton Trans., 2014, 43, 6828–6841
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
field strength increases causing spin-pairing leading to spin- meridionally coordinated ligands rearrange to coordinate
inversion and formation of the Fe(II)-SQ state, the energy for facially to iron(III) upon aquation in SDS micellar medium and
which is equivalent to the energy of photoactivation. For elicit regioselective extradiol cleavage through the displace-
[Fe(L2)(DBC)(H2O)]+ the strong iron(III)–catecholate interaction ment of bound water molecules by dioxygen.
(cf. above) and the facile approach of O2 taking place on the
Interestingly, the rates of dioxygenase reactions of the 3N
face opposite to the N-methyl group facilitate the formation of ligand complexes are several-fold higher in aqueous and
the upper lying (L)Fe(II)-SQ spin state, which is primed to react aqueous micellar solutions than in non-aqueous solvents.
with O2, and hence the highest rate observed for 2. The LMCT They are dependent upon the nature of micelles, the rate of
band energy for the adduct of 2, which corresponds to the dioxygenation being higher in SDS micellar medium than in
energy needed for the spin-inversion, varies as SDS (942) < the dichloromethane solvent, and very much upon the ligand
DCM (855) < TX-100 (610) < CTAB (566 nm), which parallels stereoelectronic factors like chelate ring size and steric bulk of
the trend in the observed reaction rates. The catecholate ligand N-alkyl subtituent. Thus organized assemblies provide
adducts of 1, 3 and 4 undergo mainly intradiol cleavage with restricted geometries and suitable biomimetic microenviron-
rates depending upon the reaction medium: for 1, SDS > TX ments for iron(III) complexes to effect coordinative rearrange-
100 > CTAB and for 3 and 4, SDS > CTAB > TX 100. Interest- ment and also to alter the reaction pathways and enhance the
ingly, for [Fe(L)(DBC)(H2O)]+ the reaction rates in SDS are rate of catechol cleavage reaction by enhancing the encounter
2.5–44.0 fold higher than those in DCM, those in SDS are probability and by decreasing the free energies of activation
1.0–4.2 fold higher than those in aqueous solution, and those for oxygen insertion.
in aqueous solution are 1–45 fold higher than those in DCM.
Thus SDS micellar medium is more suitable than others to
effect higher reaction rates. Also, the trend in reaction rates in
DCM and SDS solution is 4 > 3 > 1, which is in agreement with
Acknowledgements
the higher E1/2 value of the Fe(III)/Fe(II) couple of 4. The We sincerely thank the Council of Scientific and Industrial
increase in bulkiness of the ligand moieties and chelate ring Research, New Delhi, for a Senior Research Fellowship to M.S.
size along the above series would enhance the reaction rates.
Professor M. Palaniandavar is a recipient of a Ramanna Fellow-
Thus for the present complexes stereoelectronic factors like ship [Scheme No. SR/S1/RFIC-01/2010] from DST, New Delhi.
Lewis acidity of the iron(III) centre and ligand steric hindrance We thank Professor K. Natarajan, Bharathiar University, Coim-
determine the regioselectivity of catechol cleavage while both batore for CHN analysis. We thank Professor P. Sambasiva
the steric and electronic factors and the chelate ring size deter- Rao, Pondicherry University for providing EPR facility.
mine the rates of cleavage reaction. Also, the rate of regioselec-
tive extradiol cleavage of catechol is controlled by the medium
of reaction.
References
1 (a) D. T. Gibson, Microbial Degradation of Organic Molecules,
Conclusions and relevance to iron
dioxygenases
Marcel Dekker, New York, 1984, vol. 1, p. 535; (b) S. Dagley,
Essays Biochem., 1975, 11, 81; (c) D. H. Ohlendorf,
J. D. Lipscomb and P. C. Weber, Nature, 1988, 336, 403;
(d) Microbial Degradation of Organic Molecules, ed.
D. T. Gibson, Marcel Dekker, New York, 1984, 535;
(e) W. Reineke and M. J. Knackmuss, Annu. Rev. Microbiol.,
1988, 42, 263; (f) O. Hayaishi, M. Katagiri and S. Rothberg,
J. Am. Chem. Soc., 1955, 77, 5450.
2 (a) J. D. Lipscomb and A. M. Orville, Met. Ions Biol. Syst.,
1992, 28, 243; (b) M. P. Valley, C. K. Brown, D. L. Burk,
M. W. Vetting, D. H. Ohlendorf and J. D. Lipscomb, Bio-
chemistry, 2005, 44, 11024; (c) M. Y. M. Pau, M. I. Davis,
A. M. Orville, J. D. Lipscomb and E. I. Solomon, J. Am.
Chem. Soc., 2007, 129, 1944; (d) R. W. Frazee, A. M. Orville,
K. B. Dolbeare, H. Yu, D. H. Ohlendorf and J. D. Lipscomb,
Biochemistry, 1998, 37, 2131; (e) S. Han, L. D. Eltis,
K. N. Timmis, S. W. Muchmore and J. T. Bolin, Science,
1995, 270, 976.
A few iron(III) complexes of the type [Fe(L)Cl3], where L is a tri-
dentate 3N ligand, have been isolated and studied in non-
aqueous, aqueous and different aqueous micellar media as
structural and functional models for catechol dioxygenase
enzymes. It is interesting that one of the tridentate 3N ligand
is facially bound while two other such ligands are meridionally
bound to iron(III). Spectral and electrochemical studies dis-
close the formation and location of the cationic adducts as sol-
vated [Fe(L)(DBC)(H2O)]+ species mostly in the aqueous
micellar pseudophases of SDS and TX-100 and in the aqueous
phase of CTAB micellar solution. It is remarkable that the
iron(III) complex with the facially coordinated ligand cleaves
H2DBC regioselectively using O2 to produce extradiol cleavage
products exclusively (79.4%) in SDS micellar medium but in
lower amounts in dichloromethane solvent. In contrast, all the
other complexes with meridionally coordinated 3N ligands
afford major amounts of intradiol cleavage products in
dichloromethane but, interestingly, afford major amounts of
extradiol cleavage products in SDS micellar medium. The
3 L. Que Jr. and T. M. Loehr, Iron Carriers and Iron Proteins,
VCH, New York, 1989, p. 467.
4 (a) L. Que Jr. and R. Y. N. Ho, Chem. Rev., 1996, 96, 2607;
(b) T. D. H. Bugg and C. J. Winfield, Nat. Prod. Rep., 1998,
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6828–6841 | 6839

View Article Online
Paper
Dalton Transactions
15, 513; (c) A. L. Feig and S. J. Lippard, Chem. Rev., 1994, 22 SMART & SAINT Software Reference manuals, version 5.0,
94, 759; (d) O. Hayaishi, M. Katagiri and S. Rothberg, J. Am.
Chem. Soc., 1955, 77, 5450.
5 (a) M. W. Vetting, D. A. D’ Argenio, L. N. Ornston and
D. H. Ohlendorf, Biochemistry, 2000, 39, 7943;
Bruker AXS Inc., Madison, WI, 1998.
23 G. M. Sheldrick, SADABS Software for Empirical Absorption
Correction, University of Göttingen, Göttingen, Germany,
2000.
(b) T. E. Elgren, A. M. Orville, K. A. Kelly, J. D. Lipscomb, 24 SHELXTL Reference Manual, version 5.1, Bruker AXS Inc.,
D. H. Ohlendorf and L. Que Jr., Biochemistry, 1997, 36, Madison, WI, 1998.
11504; (c) A. Kita, S. I. Kita, I. Fujisawa, K. Inaka, T. Ishida, 25 R. Viswanathan, M. Palaniandavar, T. Balasubramanian
K. Horiike, M. Nozaki and K. Miki, Structure, 1998, 7, 25;
(d) M. W. Vetting and D. H. Ohlendorf, Structure, 2000, 8,
and P. T. Muthiah, J. Chem. Soc., Dalton Trans., 1996,
2519.
429; (e) K. Sugimoto, T. Senda, H. Aoshima, E. Masai, 26 R. Viswanathan, M. Palaniandavar, T. Balasubramanian
M. Fukuda and Y. Mitsui, Structure, 1999, 7, 953. and P. T. Muthiah, Inorg. Chem., 1998, 37, 2943.
6 D. H. Ohlendorf, A. M. Orville and J. D. Lipscomb, J. Mol. 27 G. Lin, G. Reid and T. D. H. Bugg, J. Am. Chem. Soc., 2001,
Biol., 1994, 244, 586. 123, 5030.
7 P. C. A. Bruijnincx, M. Lutz, A. L. Spek, W. R. Hagen, 28 L. Que Jr. and M. Ito, Angew. Chem., Int. Ed. Engl., 1997, 36,
B. M. Weckhuysen, G. Koten and R. J. M. Klein Gebbink,
J. Am. Chem. Soc., 2007, 129, 2275.
8 Y. Hitomi, M. Yoshida, M. Higuchi, H. Minami, T. Tanaka
and T. Funabiki, J. Inorg. Biochem., 2005, 99, 755.
9 M. Y. M. Pau, J. D. Lipscomb and E. I. Solomon, Proc. Natl.
Acad. Sci. U. S. A., 2007, 104, 18355.
1342.
29 L. Que Jr. and D.-H. Jo Jr., Angew. Chem., Int. Ed., 2000, 39,
4284.
30 J. H. Lim, T. H. Park, H.-J. Lee, K.-B. Lee and H. G. Jang,
Bull. Korean Chem. Soc., 1999, 20, 1428.
31 Y.-M. Chiou and L. Que Jr., Inorg. Chem., 1995, 34,
3577.
10 E. G. Kovaleva and J. D. Lipscomb, Science, 2007, 316, 453.
11 X. Wang, S. Wang, L. Li, E. B. Sundberg and G. P. Gacho, 32 R. Mayilmurugan, H. S. Evans and M. Palaniandavar, Inorg.
Inorg. Chem., 2003, 42, 799. Chem., 2008, 47, 6647.
12 M. Velusamy, R. Mayil Murugan and M. Palaniandavar, 33 K. R. JustinThomas, M. Velusamy and M. Palaniandavar,
J. Inorg. Biochem., 2005, 99, 1032.
Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 1998, 54,
13 M. Costas, M. P. Mehn, M. P. Jensen and L. Que Jr., Chem.
Rev., 2002, 104, 939.
741.
34 T.
Dhanalakshmi,
M.
Bhuvaneshwari
and
14 M. Merkel, F. K. Müller and B. Krebs, Inorg. Chim. Acta,
2002, 337, 308.
15 (a) K. Visvaganesan, R. Mayilmurugan, E. Suresh and
M. Palaniandavar, J. Inorg. Biochem., 2006, 100, 1527.
35 R. Mayilmurugan, E. Suresh and M. Palaniandavar, Inorg.
Chem., 2007, 46, 6038.
M. Palaniandavar, Inorg. Chem., 2007, 46, 10294; 36 U. S. Geological Survey, TWRI book 9, 4/98 [from R. F. Weiss
(b) K. Sundaravel, T. Dhanalakshmi, E. Suresh and
M. Palaniandavar, Dalton Trans., 2007, 7012.
16 (a) M. Velusamy, R. Mayilmurugan and M. Palaniandavar,
(1970)].
37 L. Que Jr., R. C. Kolanczyk and L. S. White, J. Am. Chem.
Soc., 1987, 109, 5373.
Inorg. Chem., 2004, 43, 6284; (b) R. Mayilmurugan, 38 S. V. Pavlova, H. L. To, E. S. H. Chan, H. W. Li,
K. Visvaganesan, E. Suresh and M. Palaniandavar, Inorg.
Chem., 2009, 48, 8771; (c) K. Sundaravel,
T. C. W. Mak, H. K. Lee and S. I. Chan, Dalton Trans., 2006,
2232.
M. Sankaralingam, E. Suresh and M. Palaniandavar, Dalton 39 J. W. Bradley and J. D. Lipscomb, Chem. Rev., 1996, 96,
Trans., 2011, 40, 8444; (d) K. Sundaravel, E. Suresh, 2625.
K. Saminathan and M. Palaniandavar, Dalton Trans., 2011, 40 M. Costas, M. P. Mehn, M. P. Jensen and L. Que Jr., Chem.
40, 8092. Rev., 2004, 104, 939.
17 (a) M. Velusamy, M. Palaniandavar, R. Srinivasagopalan 41 S. Yoon, H. J. Lee and H. G. Jang, Bull. Korean Chem. Soc.,
and G. U. Kulkarni, Inorg. Chem., 2003, 42, 8283; 2000, 21, 923.
(b) R. Mayilmurugan, M. Sankaralingam, E. Suresh and 42 (a) R. B. Lauffer, R. H. Heistand II and L. Que Jr., J. Am.
M. Palaniandavar, Dalton Trans., 2010, 39, 9611.
18 (a) N. Anitha and M. Palaniandavar, Dalton Trans., 2010,
39, 1195; (b) N. Anitha and M. Palaniandavar, Dalton
Trans., 2011, 40, 1888.
19 V. Balamurugan, W. Jacob, J. Mukherjee and R. Mukherjee,
CrystEngComm, 2004, 6(66), 396.
Chem. Soc., 1981, 103, 3947; (b) R. H. Heistand II,
A. L. Roe and L. Que Jr., Inorg. Chem., 1982, 21, 676;
(c) R. H. Heistand II, R. B. Lauffer, E. Fikrig and L. Que
Jr., J. Am. Chem. Soc., 1982, 104, 2789; (d) R. B. Lauffer,
R. H. Heistand II and L. Que Jr., Inorg. Chem., 1983, 22,
50.
20 J. Mukherjee, V. Balamurugan, R. Gupta and R. Mukherjee, 43 D. D. Cox, S. J. Benkovic, L. M. Bloom, F. C. Bradley,
Dalton Trans., 2003, 3686.
21 S. A. Leaver, M. Palaniandavar, C. A. Kilner and
M. J. Nelson, L. Que Jr. and D. E. Wallick, J. Am. Chem. Soc.,
1988, 110, 2026.
M. A. Halcrow, Dalton Trans., 2003, 4224 and references 44 D. D. Cox and L. Que Jr., J. Am. Chem. Soc., 1988, 110,
therein.
8085.
6840 | Dalton Trans., 2014, 43, 6828–6841
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
45 J. W. Pyrz, A. L. Roe, L. Stern and L. Que Jr., J. Am. Chem.
Soc., 1985, 107, 614.
46 T. Ogihara, S. Hikichi, M. Akita and Y. Moro-oka, Inorg.
Chem., 1998, 37, 2614.
Mallart, G. Blondin, A. Nivorojkine, J. Guilhem,
L. Tchertanova, M. Cesario, N. Ravi, E. Bominaar,
J.-J. Girerd and E. Münck, Inorg. Chim. Acta, 1997, 263,
367.
47 (a) S. Mendel, A. Arndt, D. Timothy and H. Bugg, Bio- 48 (a) M. Higuchi, Y. Hitomi, H. Minami, T. Tanaka and
chemistry,
2004,
43,
13390;
(b)
P.
Mialane,
T. Funabiki, Inorg. Chem., 2005, 44, 8810; (b) A. Dei, Inorg.
Chem., 1993, 32, 5730; (c) K. P. Jensen and U. Ryde, J. Biol.
Chem., 2004, 279, 14561.
L. Tehertanov, F. Banse, J. Sainton and J. Girerd, Inorg.
Chem., 2000, 39, 2440; (c) P. Mialane, E. Anxolabéhère-
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6828–6841 | 6841