Iron(III) chloride-benzotriazole adducts with trigonal bipyramidal geometry: Spectroscopic, structural and catalytic studies

Article abstract of DOI:10.1016/j.poly.2013.03.062

The reactions of FeCl₃ with benzotriazole (btaH), 1-methylbenzotriazole (Mebta), 5,6-dimethylbenzotriazole (5,6Me₂btaH) and 5-chlorobenzotriazole (5ClbtaH) were studied in non-polar solvents. The new solid complexes [FeCl₃(btaH)₂] (1), [FeCl ₃(Mebta)₂] (2), [FeCl₃(5,6Me ₂btaH)₂] (3) and [FeCl₃(5ClbtaH) ₂]·2(5ClbtaH) (4) have been isolated. The structures of the complexes have been determined by single-crystal, X-ray crystallography. The structures of 1-4 consist of mononuclear, high-spin 5-coordinate molecules; in addition, the crystal structure of 4 contains two lattice 5ClbtaH molecules per [FeCl₃(5ClbtaH)₂] unit. The coordinated benzotriazole molecules behave as monodentate ligands with their ligated atom being the nitrogen of the position 3 of the azole ring. The geometry at iron(III) is trigonal bipyramidal with the chlorido ligands occupying the equatorial sites. The crystal structures of the complexes are stabilized by stacking interactions and H bonds (for 1, 3 and 4 only). The new complexes were characterized by elemental analyses, magnetic susceptibilities at room temperature and spectroscopic (IR, far-IR, solid-state electronic UV/VIS/near-IR, ⁵⁷Fe-Mo?ssbauer, EPR only for complex 4) methods. All data are discussed in terms of the nature of bonding and the known structures. Complexes 1, 2 and 4 have been tested as homogeneous (MeCN) oxidation catalysts in the presence of the "green" H₂O₂ oxidant; they display moderate to high catalytic activity in the oxidation of several alkenes, cyclohexane and n-hexane, which is described in detail.

Full text of DOI:10.1016/j.poly.2013.03.062

Accepted Manuscript
Iron(III) chloride – benzotriazole adducts with trigonal bipyramidal geometry:
Spectroscopic, structural and catalytic studies
Nikolaos C. Anastasiadis, George Bilis, John C. Plakatouras, Catherine P.
Raptopoulou, Vassilis Psycharis, Christine Beavers, Simon J. Teat, Maria
Louloudi, Spyros P. Perlepes
PII:
S0277-5387(13)00275-1
DOI:
http://dx.doi.org/10.1016/j.poly.2013.03.062
POLY 10056
Reference:
To appear in:
Polyhedron
Please cite this article as: N.C. Anastasiadis, G. Bilis, J.C. Plakatouras, C.P. Raptopoulou, V. Psycharis, C. Beavers,
S.J. Teat, M. Louloudi, S.P. Perlepes, Iron(III) chloride – benzotriazole adducts with trigonal bipyramidal geometry:
Spectroscopic, structural and catalytic studies, Polyhedron (2013), doi: http://dx.doi.org/10.1016/j.poly.
2
013.03.062
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Iron(III) chloride – benzotriazole adducts with trigonal bipyramidal
geometry: Spectroscopic, structural and catalytic studies
a
b
b
Nikolaos C. Anastasiadis , George Bilis , John C. Plakatouras , Catherine P.
c
c
d
d
Raptopoulou , Vassilis Psycharis , Christine Beavers , Simon J. Teat , Maria
b,*
a,*
Louloudi , Spyros P. Perlepes
a
Department of Chemistry, University of Patras, 265 04 Patras, Greece
Department of Chemistry, University of Ioannina, 451 10 Ioannina, Greece
Institute of Advanced Materials, Physicochemical Processes, Nanotechnology and
b
c
Microsystems, Department of Materials Science, NCSR “Demokritos”, 153 10 Aghia
Paraskevi Attikis, Greece
d
Advanced Light Source, Lawrence Berkeley National Laboratory, 1 Cyclotron Road,
Berkeley, CA 94720, USA
ARTICLE INFO
Article history:
Keywords:
Dedicated to George Christou:
a great scientist, a precious friend
Benzotriazoles as ligands
Crystal structures
Iron(III)
III
Trigonal bipyramidal Fe complexes
Oxidation catalysis
*
Corresponding authors. Tel.: +30 26510 08418; fax: +30 26510 08786 (M.
Louloudi), tel.: +30 2610 996730; fax: +30 2610 997118 (S. P. Perlepes).
E-mail addresses: mlouloud@uoi.gr, perlepes@patreas.upatras.gr
1

ABSTRACT
The reactions of FeCl with benzotriazole (btaH), 1-methylbenzotriazole (Mebta), 5,6-
3
dimethylbenzotriazole (5,6Me btaH) and 5-chlorobenzotriazole (5ClbtaH) were
2
studied in non-polar solvents. The new solid complexes [FeCl (btaH) ] (1),
3
2
.
[
FeCl (Mebta) ] (2), [FeCl (5,6Me btaH) ] (3) and [FeCl (5ClbtaH) ] 2(5ClbtaH) (4)
3 2 3 2 2 3 2
have been isolated. The structures of the complexes have been determined by single-
crystal, X-ray crystallography. The structures of 1-4 consist of mononuclear, high-
spin 5-coordinate molecules; in addition, the crystal structure of 4 contains two lattice
5
ClbtaH molecules per [FeCl (5ClbtaH) ] unit. The coordinated benzotriazole
3 2
molecules behave as monodentate ligands with their ligated atom being the nitrogen
of the position 3 of the azole ring. The geometry at iron(III) is trigonal bipyramidal
with the chlorido ligands occupying the equatorial sites. The crystal structures of the
complexes are stabilized by stacking interactions and H bonds (for 1, 3 and 4 only).
The new complexes were characterized by elemental analyses, magnetic
susceptibilities at room temperature and spectroscopic (IR, far-IR, solid-state
5
7
electronic UV/VIS/near-IR, Fe-Mössbauer, EPR only for complex 4) methods. All
data are discussed in terms of the nature of bonding and the known structures.
Complexes 1, 2 and 4 have been tested as homogeneous (MeCN) oxidation catalysts
in the presence of the “green” H O oxidant; they display moderate to high catalytic
2
2
activity in the oxidation of several alkenes, cyclohexane and n-hexane, which is
described in detail.
2

1
. Introduction
The continuous improvement of synthetic methodology in inorganic chemistry
has already allowed us to see coordination chemistry entering a phase of creative,
rather than investigative, character. The development of metal complexes for
applications in medicine and catalysis, and the synthesis of molecular clusters and
coordination polymers with exciting physical properties (magnetic, optical, …) are
clear examples of this new era.
The central “players” in coordination chemistry are the metal atoms/ions and
the ligands. The proper use of old ligands and the design of new ones underlies
much of modern coordination chemistry. Alfred Werner and his successors
-
2-
3
employed simple inorganic and organic ligands, e.g. NH , H O, Cl , CO , en, …
3
2
Later, inorganic chemists started to use varieties of complex, polydentate ligands
polypodal open-chain ligands, cyclic ligands, crown ethers, cryptands,
sepulchrates, ligands of unusual reach, calixarenes, biologically relevant ligands,
) to achieve specific purposes. Today, the design and use of such polydentate
ligands is an important frontier of research in contemporary inorganic chemistry
1].
(
…
[
A popular class of ligands consists of azoles and one family of them are
benzotriazole (btaH) and its derivatives (Fig. 1). Over the last thirty years or so, we
2-18] and other groups [19-51] have been exploring the coordination chemistry of
[
benzotriazoles. The first reason for this is the impressive anticorrosion activity of
btaH and its benzene ring-substituted derivatives towards certain metals,
particularly Cu and its alloys [52-55]. Chemical, physical and structural studies on
realistic model metal complexes [23, 24] are needed to understand the complicated
surface chemistry that leads to corrosion inhibition and to create better inhibitors
than btaH. The second reason is the observation that the reactions between metal
carboxylates or β-diketonates and btaH, its ring- (with non-donor groups) or its N-
substituted (with donor groups) derivatives lead to incomplete replacement of the
ancillary ligands by the deprotonated benzotriazoles and the formation of
heteroleptic homonuclear or heteronuclear coordination clusters with novel
structural motifs and interesting magnetic properties. For example, in previous
reports, we have shown that the synthetic investigation of the
.
[
Ni(acac) (H O) ]/btaOH and Ni(O CMe) 4H O/btaOH reaction systems (acac is
2
2
2
2
2
2
3

the anion of 2,4-pentanedione and btaH is 1-hydroxybenzotriazole, i.e. R = OH and
R’ = H in Fig. 1) lead to complexes [Ni (OH) (acac) (btaO) (H O) ] [13] and
7
2
8
4
2
2
[
Ni (OH) (O CMe) (btaO) (H O) (nPrOH) ] [18], respectively. The third reason
13 6 2 8 12 2 6 4
is the capability of deprotonated benzotriazoles to form coordination polymers of
various dimensionalities (depending on the benzotriazole coordination mode and
the nature of the ancillary ligand present) with interesting photophysical and
magnetic properties [10, 12, 15, 17, 34, 38, 40, 41, 43-45, 49]. Work from our
II
groups includes the synthetic investigation of the Cu /btaOH/aqueous NH
3
reaction system in MeOH which resulted in the solution of the 3D, diamond-like
complex [Cu(btaO) (MeOH)] , which is a soft molecular magnet exhibiting two
2
n
critical temperatures at 6.4 and 4.4 K [10] . Another reason for the intense interest
in the coordination chemistry of btaH stems from its use as a bridging ligand in
mixed organometallic/coordination compounds. For example, deprotonation of
btaH by tris(cyclopentadienyl)dysprosium(III) has led to the dinuclear complex
[
[
Dy (μ-bta) (Cp) ], a compound which exhibits single-molecule magnet behavior
2 2 4
56]. Finally, btaH is extremely useful in silver chemistry [37], because of its high
I
affinity for Ag ; this affinity has been the main reason of its successful use in
conventional photographic systems for many years [57]. Complex [Ag(bta)] has
n
been reported to be an efficient silver source in black and white
photothermographic (PTG) imaging systems [37]; more recently, it has been
investigated as the silver source to be used in colour PTG materials designed for
advanced photographic applications [58], compared to “conventional” PTG
materials. Despite the extensive use of neutral and anionic benzotriazoles in 3d-
metal chemistry, the number of Fe coordination complexes with benzotriazole or
benzotriazolate ligation remains limited [4, 26, 29, 31, 33, 36, 39, 50]. Of
III
1
particular interest are the Fe
clusters reported by McInnes, Brechin, Collison
4
and co-workers [29, 39]; these complexes contain deprotonated bridging
benzotriazoles, have high ground-state electronic spins of S ≤ 25 (among the
highest known) and exhibit very large magnetocaloric effects. Moreover, no metal
-
complexes of btaH, bta and its derivatives have been studied in catalysis to date.
This work aims at the amalgamation of these two areas and reports FeCl3-
benzotriazole adducts with remarkable activity as homogeneous catalysts for
hydrocarbon oxidation using H O as oxidant.
2
2
4

Fig. 1. The general structural formula of benzotriazoles. The parent member of this family of
ligands is 1H-benzo[d][1,2,3]triazole or simply benzotriazole (btaH; R = R’ = R’’ = H). R and
R’ can be non-donor (Me, Cl, NO , …) or donor (OH, COOH, CH OH, CH COOH, …)
2
2
2
groups. The ligands used in this work are btaH, 1-methylbenzotriazole (Mebta; R = Me, R’ =
R’’ = H), 5-chlorobenzotriazole (5ClbtaH; R = H, R’ = Cl, R’’ = H) and 5,6-
dimethylbenzotriazole (5,6Me btaH; R = H, R’ = R’’ = Me).
2
The development of efficient methods for hydrocarbon oxidation remains an
important challenge for synthetic chemistry. For example, the selective transformation
of inert C-H bonds of alkanes into useful functional groups has attracted much
attention because alkanes are less expensive and more readily available than the
current petrochemical feedstocks [59]. Moreover, the development of systems that
achieve high selectivity using environmentally friendly processes and inexpensive
oxidants represents a demand with direct ecological and financial benefits [60, 61].
Today there is an increased intense interest for the use of oxidants, like molecular O2
and H O , which are environmentally friendly and do not give rise to any waste
2
2
products [62]. It is also appreciated by industrial chemists that H O is miscible with
2
2
water and relatively easily to handle. Unfortunately, H O can undergo radical-
2
2
induced decomposition to H O and O by impurities and traces of metallic particles
2
2
[
63]. The direct oxidation of organic substrates by H O is rare as the energy barrier
2 2
for electron transfer from the substrate to the oxidant is usually high. However, nature
has found methods to make controlled aerobic oxidations under highly mild
conditions [64]. The remarkable increase in information on non-heme iron
oxygenases within the past decade [65] has spurred efforts to explore the use of non-
heme ligand scaffolds to facilitate such oxidation catalysis [66]. Detailed studies on
biomimetic catalysts is of key importance for the design and development of novel,
5

efficient catalysts [67]. The use of an iron salt or complex and H O traces back to the
2
2
th
1
9 century [68], but the major challenge has been to inhibit the homolytic cleavage
of the peroxo O-O bond that produces non-selective, and thus unwanted, hydroxyl
radicals [69]. Instead, the goal is to direct the metal-promoted cleavage towards the
generation of a metal-based oxidant that can carry out hydrocarbon oxidations with
high chemo-, regio- and stereoselectivity [69]. In this context, stereoselective
hydroxylation, epoxidation and cis-dihydroxylation by iron biomimetic catalysts have
been reported [70-78]. The most extensively studied oxidation catalysts thus far are
complexes containing tetradentate N-donor ligands with configurations that allow two
cis-orientated coordination sites to be available for peroxide binding and activation
[
72, 79, 80]. Involvement of macroacyclic N,O-donor ligands for the generation of
non-heme iron-based catalysts has also been attempted [81-83]. However, new,
efficient, low-cost and easy-to-prepare iron catalysts remain at high demand for
hydrocarbon oxidation by H O . We have thus decided to evaluate simple 5-
2
2
coordinate FeCl -benzotriazole complexes as homogeneous catalysts for hydrocarbon
3
oxidation with H O as oxidant; as it will be described, the complexes show
2
2
remarkable effectiveness towards several substrates at room temperature.
2
.
Experimental
2
.1. General information and physical measurements
All manipulations were performed under aerobic conditions using materials
(
reagent grade) and solvents as received. Elemental analyses (C, H, N) were
performed by the in-house facilities of the University of Patras (Greece) using an EA
108 Carlo Erba analyzer. Conductivity measurements were carried out at 25 °C with
a Metrohm-Herisau E-527 bridge and a cell of standard constant. IR spectra (4000-
1
-1
4
00 cm ) were recorded on a Perkin-Elmer 16 PC FT spectrometer with samples
-1
prepared as KBr pellets. Far-IR spectra (400-50 cm ) were recorded on a Brucker IFS
13v FT spectrometer with samples prepared as polyethylene pellets. Solid-state
diffuse reflectance) electronic spectra (1500-350 nm) were recorded on a Perkin-
1
(
Elmer 330 spectrophotometer using spectroscopic grade MgO as reference. Solution
electronic spectra (850-330 nm) were recorded on a Varian 634 spectrophotometer.
6

Magnetic susceptibilities were measured at room temperature using the Faraday
method and employing a Cahn-Ventron RM-2 balance standardized with
HgCo(NCS) . Molar magnetic susceptibilities were corrected for the diamagnetism of
4
5
7
the constituent atoms using Pascal’s constants. Fe Mössbauer spectra were recorded
at room temperature with a conventional constant acceleration spectrometer; the γ-ray
5
7
source used was Co in a Rh matrix and the isomer shifts were referenced to metallic
.
iron. The EPR spectrum of a powdered sample of [FeCl (5ClbtaH) ] 2(5ClbtaH) was
3
2
recorded at 4.2 K using a Bruker ER 200D spectrometer operating at X-band
frequencies; the low temperature was obtained using an Oxford Instruments ESR
cryostat.
2
.2. Catalytic experiments
All substrates were purchased from Aldrich, in their highest commercial
purity, stored at 5 °C and purified by passage through a column of basic alumina prior
to use. A 30% aqueous solution of H O was used as oxidant. For the catalytic
2
2
evaluation, H O (30%) diluted in MeCN (1:10 v/v) was slowly added (within a
2
2
period of 10 min) to an MeCN solution of the iron(III) complex (the catalyst) and the
substrate under an Ar atmosphere at room temperature (26 °C). Acetophenone or
bromobenzene were used as internal standard. Catalytic reactions started by adding
the oxidant into the reaction mixture. The catalyst : oxidant : substrate ratio was 1 : 20
:
1000 μmol for cyclohexene, 1-methylcyclohexene, cyclooctene, limonene,
cyclohexane and n-hexane, and 1 : 50 : 1000 μmol for styrene, trans-β-methyl styrene
and cis-stilbene. The total volume of the catalytic reaction solution was 1 ml. The
large excess of the substrate was expected to minimize over-oxidation of oxidation
products and to protect the catalyst from oxidative destruction. The oxidant was added
slowly in order to minimize potential iron-catalyzed H O decomposition via catalase
2
2
activity.
The progress of the catalytic reaction was monitored using GC-MS by
removing small samples of the reaction mixture. GC analysis was performed using a
8
000 Fisons chromatograph with a flame ionization detector and a Shimadzu GC-17A
gas chromatograph coupled with a GCMS-QP5000 mass spectrometer. The yields
reported herein are based on the amount of H O oxidant converted into oxygenated
2
2
products. Reactions were complete within 6 h. In some cases, oxidations catalyzed by
7

.
[
FeCl (5ClbtaH) ] 2(5ClbtaH) were accomplished within 2 h (vide infra). To establish
3 2
the identity of the products unequivocally, the retention times and spectral data were
compared to those of the commercially available compounds. Black experiments
revealed that no oxidative reactions took place without the catalyst.
2
2
.3. Compound preparation
.3.1. [FeCl (btaH) ] (1)
3
2
To an orange solution of btaH (0.119 g, 1.00 mmol) in CH Cl (20 mL) was
2
2
slowly added solid FeCl (0.081 g, 0.50 mmol). Most of the solid soon dissolved. The
3
resulting slurry was kept under stirring for 10 min, filtered and the yellow filtrate was
stored at 5 °C in a closed flask. Well-formed, X-ray quality orange crystals of the
product were deposited after 2 days. The crystals were collected by filtration, washed
with cold CH Cl (1 mL) and Et O (2x3 mL), and dried in air. Typical yields were in
2
2
2
the 35-40% range. The yield can be increased up to ~70% using concentrated reaction
mixtures (e.g. 4.0 mmol of btaH and 2.0 mmol of FeCl in 25 mL of CH Cl ), but in
3
2
2
this case the product was isolated as a microcrystalline powder even at room
temperature. Anal. Calc. for C H FeN Cl : C, 36.0; H, 2.5; N, 21.0. Found: C, 35.7;
1
2
10
6
3
-1
H, 2.6; N, 20.8%. Selected IR data (KBr, cm ): ν = 3232s, 1622m, 1220m, 1112s,
-1
1
024m, 744s, 430m. Selected far-IR data (polyethylene, cm ): ν = 374s, 215s. Molar
-3
2
-1
-3
conductivity, Λ (MeCN, 10 M, 25 °C): 4 S cm mol . Λ (MeNO , 10 M, 25 °C):
M
M
2
2
-1
8
0 S cm mol . Effective magnetic moment, μ : 5.79 BM at 24 °C. Solid-state
eff
5
7
-1
electronic spectra (nm): λ = 415, 606, 837, 1123, 1385. Fe Mössbauer data (mm s ,
room temperature): δ = 0.37, ΔΕ = 0.56.
Q
2
.3.2. [FeCl (Mebta) ] (2)
3 2
To a colourless solution of Mebta (0.133 g, 1.00 mmol) in CHCl (15 mL) was
3
slowly added a green suspension of FeCl (0.081 g, 0.50 mmol) in the same solvent
3
(
15 mL). The resulting orange slurry was stirred for 30 min and filtered. Overnight
slow evaporation of the yellow filtrate at room temperature gave X-ray quality,
yellow crystals of the product, which were collected by filtration, washed with cold
CHCl (1 mL) and Et O (3x3 mL), and dried in air. Typical yields were in the range
3
2
5
3
5-60%. Anal. Calc. for C H FeN Cl : C, 39.2; H, 3.3; N, 19.6. Found: C, 39.0; H,
14 14 6 3
-1
.4; N, 19.4 %. Selected IR data (KBr, cm ): ν = 1618m, 1218s, 1136m, 750s, 432m.
8

-1
-3
Selected far-IR data (polyethylene, cm ): ν = 378s, 196s. Λ (MeCN, 10 M, 25 °C):
M
2
-1
-3
2
-1
6
S cm mol . Λ (MeNO , 10 M, 25 °C): 77 S cm mol . μ : 6.06 BM at 20 °C.
M 2 eff
57
Solid-state electronic spectra (nm): λ = 436, 604, 837, 1137, 1345, 1379. Fe
-1
Mössbauer data (mm s , room temperature): δ = 0.24, ΔΕ = 0.27.
Q
2
.3.3. [FeCl (5,6Me btaH) ] (3)
3 2 2
To an orange solution of 5,6Me btaH (0.216 g, 1.50 mmol) in CHCl (10 mL)
2
3
was added a green suspension of FeCl (0.081 g, 0.50 mmol) in the same solvent (10
3
mL). The resulting slurry dissolved upon stirring and addition of MeOH (3 mL) to
give a dark red solution, which was layered with n-hexane (40 mL). Slow mixing
gave well-formed, X-ray quality red crystals of the product over a period of 3 days.
The crystals were collected by filtration, washed with cold CHCl (1 mL) and Et O
3
2
(
4x3 mL), and dried in air. Typical yields were in the 65-70% range (based on the
III
Fe available). Anal. Calc. for C H FeN Cl : C, 42.1; H, 4.0; N, 18.4. Found: C,
1
6
18
6
3
-
1
4
1
2.4; H, 3.9; N, 18.1%. Selected IR data (KBr, cm ): ν = 3244sb, 1586m, 1226s,
-1
100s, 1024m, 770m, 474w. Selected far-IR data (polyethylene, cm ): ν = 378s, 214s.
-3
2
-1
-3
2
Λ (MeCN, 10 M, 25 °C): 3 S cm mol . Λ (MeNO , 10 M, 25 °C): 80 S cm
M
M
2
-1
mol . μ : 6.01 BM at 26 °C. Solid-state electronic spectra (nm): λ = 411, 582, 895,
eff
5
7
-1
1
097, 1369. Fe Mössbauer data (mm s , room temperature): δ = 0.25, ΔΕ = 0.30.
Q
.
2
.3.4. [FeCl (5ClbtaH) ] 2(5ClbtaH) (4)
3
2
To a colourless solution of 5ClbtaH (0.230 g, 1.50 mmol) in a solvent mixture
comprising 20 mL CHCl and 1.5 mL EtOH was added solid FeCl (0.081 g, 0.50
3
3
mmol). The solid soon dissolved. The resulting dark orange solution was kept under
stirring for 20 min and then layered with n-hexane (35 mL). Slow mixing gave well-
formed, X-ray quality red crystals of the product over a period of 3 days. The crystals
were collected by filtration, washed with cold EtOH (1 mL) and Et O (2x3 mL), and
2
dried in air. The yield was ~70% (based on the ligand available). Anal. Calc. for
C H FeN Cl : C, 37.1; H, 2.1; N, 21.7. Found: C, 36.8; H, 2.1; N, 21.3%. Selected
24
16
12
7
-1
IR data (KBr, cm ): ν = 3260-2800sb, 1264m, 1202s, 1112m, 1070m, 942m, 804m,
-
1
-3
4
2
2
32w. Selected far-IR data (polyethylene, cm ): ν = 380s, 213s. Λ (MeCN, 10 M,
M
2
-1
-3
2
-1
5 °C): 8 S cm mol . Λ (MeNO , 10 M, 25 °C): 81 S cm mol . μ : 6.05 BM at
M
2
eff
5
7
4 °C. Solid-state electronic spectra (nm): λ = 394, 610, 838, 1385. Fe Mössbauer
-1
data (mm s , room temperature): δ = 0.28, ΔΕ = 0.28. EPR data (solid state, 4.2 K):
Q
9

g = 5.83, 1.89. The EPR spectrum of a frozen solution (4.2 K) of the complex in
MeNO is complicated; six transitions with g values in the 9.22-2.01 range are
2
observed.
2
.4. Single-crystal X-ray crystallography
A yellow crystal of 2 (0.05 x 0.14 x 0.30 mm) and red crystals of 3 (0.11 x
.12 x 0.48 mm) and 4 (0.11 x 0.41 x 0.54 mm) were taken directly from the mother
0
liquid and immediately cooled to 160 K. Diffraction measurements were performed
on a Rigaku R-Axis SPIDER Image Plate diffractometer using graphite-
monochromated Cu Kα radiation. Data collection (ω-scans) and processing (cell
refinement, data reduction and empirical absorption correction) were carried out using
the CrystalClear program package [84]. Data for a selected orange needle of 1 (0.05 x
0
.05 x 0.15 mm) were collected at Station 11.3.1 of the Advanced Light Source at
Lawrence Berkeley National Laboratory, using a Bruker 08 diffractometer equipped
with an ApexII CCD detector (synchrotron radiation of 0.77490 Å, silicon 111
monochromator, 100 K). The data were intergrated using the program SAINT
v7.60A. A multi-scan correction for absorption was applied using the program
SADABS-2008/1.
The structures of 2-4 were solved by direct methods using SHELXS-97 [85]
2
and refined by full-matrix least-squares techniques on F with SHELXL-97 [86]. The
H-atoms were located by difference maps and refined isotropically. All non-H atoms
were refined anisotropically. In the structure of 4, the chlorine atom of the
coordinated ligand was found disordered and refined over two orientations with
occupation factors summing one. The structure of 1 was solved by direct methods
[
85] and refined using the SHELX-TL suite of programs [87]. All carbon bonded H-
atoms were affixed geometrically and refined as riding atoms with C-H = 0.95-0.99 Å
and U (H) = 1.2 times U (C). The azole H-atom was located and refined freely.
iso
eq
Important crystal data and parameters for data collection and refinement are listed in
Table 1.
1
0

Table 1
Crystallographic data for complexes 1-4.
Parameter
1
2
3
4
Empirical formula
C H FeN Cl
C H FeN Cl
C H FeN Cl
C H FeN Cl
1
2
10
6
3
14 14
6
3
16 18
6
3
24 16
12
7
-
1
M (g mol )
Crystal system
Space group
a (Å)
400.46
428.51
456.56
776.49
monoclinic
C2/c
monoclinic
C2/c
monoclinic
P2/n
orthorhombic
P22121
7.5066(1)
8.0738(1)
25.0627(4)
90
18.385(2)
6.7661(7)
14.3770(16)
90
15.0523(2)
7.3248(1)
17.1315(3)
90
7.8454(1)
7.9152(1)
16.1391(2)
90
b (Å)
c (Å)
α (°)
β (°)
121.667(3)
90
112.030(1)
90
100.633(1)
90
90
γ (°)
90
3
V (Å )
1522.1(3)
4
1750.93(5)
4
985.00(2)
2
1518.97(4)
2
Z
-
3
pcalc (g cm )
Radiation, λ (Å)
T (K)
1.748
1.626
1.539
1.698(2)
synchrotron, 0.77490 Cu Kα, 1.54178
Cu Kα, 1.54178
160(2)
Cu Kα, 1.54178
160(2)
100(2)
69.1
160(2)
130.0
2
θmax (°)
130.0
130.0
-
1
μ (mm )
1.919
11.200
868
9.992
9.997
F (000)
804
466
778
Total reflections
Unique reflections (Rint)
Data with I > 2σ(I)
Parameters refined
10613
9217
8343
13094
2415 (0.0543)
2215
1409 (0.0997)
1261
1654 (0.0552)
1565
2524 (0.0786)
2338
105
126
155
233
-
3
(
Δρ)max / (Δρ)min (e Å )
0.493 / -0.305
1.018
0.546 / -1.313
1.113
0.376 / -0.439
1.091
0.475 / -0.695
1.049
2
Goodness-of-fit (GOF)(on F )
a
R (I> 2σ(I))
0.0277
0.0751
0.0495
0.1132
0.0351
0.0441
1
b,c
wR2 (I> 2σ(I))
0.0912
0.1010
a
b
c
R = Σ(|F | - |F |) / Σ(|F |).
1
o
c
o
2
o
2 2
2 2 1/2
o
wR = {Σ[w(F – F ) ] / Σ[w(F ) ]} .
2
c
2
2
o
2
2
o
2
w = 1 / [σ (F ) + (aP) + bP] and P = (max(F ,0) + 2F ) / 3. a = 0.0378, b = 1.3396 for 1; a =
c
0
.0416, b = 5.5658 for 2; a = 0.0442, b = 0.5039 for 3; a = 0.0448, b = 0.4603 for 4.
1
1

3
. Results and discussion
3
.1. Synthetic comments in brief
Complexes 1-4 are prepared readily by the direct reactions between almost
anhydrous iron chloride(III) and the ligands in non-polar solvents or predominantly
non-polar solvent mixtures in moderate to good yields. Their formation is summarized
in Eqs. (1) and (2), where L = btaH, Mebta and 5,6Me btaH (Fig. 1).
2
FeCl + 2L
[FeCl L ]
(1)
3
3 2
1
-3
.
FeCl + 4(5ClbtaH)
[FeCl (5ClbtaH) ] 2(5ClbtaH)
(2)
3
3
2
4
The use of non-polar organic solvents, i.e. CH Cl and CHCl , is reflected on
2
2
3
the molecular (i.e. non-ionic) nature of the products. Employment of more polar
solvents (e.g. alcohols) in FeCl /monodentate organic ligand reaction systems often
3
leads to ionic complexes of the general formula [FeCl (monodentate ligand) ][FeCl ]
2
4
4
[
88,89]. The ligand to metal reaction molar ratio does not affect the identity of the
products in the solvents used. Adducts 1-4 are the only products from the 1:1, 2:1, 3:1
and 4:1 reaction mixtures. For example, we wondered whether we could prepare a
compound with the stoichiometry [FeCl (5ClbtaH) ], similar to that in 1-3. However,
3
2
all our efforts failed; it seems that the incorporation of two lattice 5ClbtaH molecules
is favoured by thermodynamic reasons.
3
.2. Description of structures
Selected bond distances and angles for complexes 1-4 are listed in Tables 2-5.
Partially labeled plots and aspects of the crystal (supramolecular) structures of the
complexes are shown in Figs. 2-9. The structures of 1-3 consist of mononuclear
[
FeCl L ] molecules (L = btaH, Mebta, 5,6Me btaH, while the structure of 4 consists
3 2 2
of [FeCl (5ClbtaH) ] species and lattice 5ClbtaH molecules in an 1:2 ratio. Since the
3
2
molecules
[FeCl (btaH) ],
[FeCl (Mebta) ],
[FeCl (5,6Me btaH) ]
and
3
2
3
2
3
2
2
1
2

[
FeCl (5ClbtaH) ] have similar structures, a general description will be given. The
3 2
crystal structures of the compounds will be analyzed separately.
Fig. 2. Partially labeled plot of the molecular structure of 1. Only the H atom of the azole ring
is shown. Primes are used for symmetry-related atoms, see the footnote of Table 2. Note that
the donor atom N(1) [and its symmetry equivalent] is the nitrogen atom of the position 3 in
III
the IUPAC numbering of btaH (Fig. 1). Colour scheme: Fe , orange; Cl, green; N, blue; C,
grey.(Colour online).
Table 2
a
Selected interatomic distances (Å) and angles (°) for complex 1.
Bond lengths
Fe(1) – Cl(1)
Fe(1) – Cl(2)
Fe(1) – N(1)
2.240(1)
2.220(1)
2.169(1)
N(1) – N(2)
N(2) – N(3)
1.314(1)
1.335(2)
Bond angles
Cl(1)-Fe(1)-Cl(1’)
Cl(1)-Fe(1)-Cl(2)
Cl(1)-Fe(1)-N(1)
114.1(1)
122.9(1)
91.9(1)
89.8(1)
Cl(2)-Fe(1)-N(1)
N(1)-Fe(1)-N(1’)
N(1)-N(2)-N(3)
88.5(1)
177.0
108.1(1)
Cl(1)-Fe(1)-N(1’)
a
Symmetry operation to generate equivalent atoms: (’) –x, y, 1/2-z.
1
3

The molecules [FeCl L ] (L = btaH, Mebta, 5,6Me btaH, 5ClbtaH) possess a
3
2
2
III
crystallographically imposed 2-fold axis of symmetry passing through the Fe and
III
Cl(2) atoms in 1-3 (Figs. 2-4), and through the Fe and Cl(1) atoms in 4 (Fig. 5). The
III
Fe center is surrounded by three terminal chloro (or chlorido) ligands and two
nitrogen atoms from two planar monodentate benzotriazole ligands; the latter use the
nitrogen of the position 3 of the azole ring (Fig. 1) as the donor atom [this atom is
labeled as N(1) in the structure of 1, see Fig.2]. The five donor atoms define a trigonal
bipyramid around the metal center with the three chloro ligands in the equatorial
plane and the two L nitrogens in the axial positions. Analysis of the shape-
determining angles using the approach of Reedijk, Addison and co-workers [90] yield
values of trigonality index, τ, of 0.90, 0.85, 0.91 and 0.99 for 1, 2, 3 and 4,
respectively (τ = 0 and 1 for perfect square pyramidal and trigonal bipyramidal
geometries, respectively). Thus, the tbp geometries in 1-3 are slightly distorted, the
distortions from the idealized D3h symmetry occurring both in the three Cl-Fe-Cl
angles and in the nonlinearity of the N-Fe-N group.
Fig. 3. Partially labeled plot of the molecular structure of 2, with H atoms omitted. Primes are
III
used for symmetry-related atoms, see the footnote of Table 3. Colour scheme: Fe , orange;
Cl, green; N, blue; C, grey. (Coulor online).
1
4

Table 3
a
Selected interatomic distances (Å) and angles (°) for complex 2.
Bond lengths
Fe – Cl(1)
Fe – Cl(2)
Fe – N(3)
2.212(1)
2.215(2)
2.178(3)
N(1) – N(2)
N(2) – N(3)
1.340(4)
1.321(5)
Bond angles
Cl(1)-Fe-Cl(1’)
Cl(1)-Fe-Cl(2)
Cl(1)-Fe-N(3)
114.9(1)
122.5(1)
90.7(1)
92.9(1)
Cl(2)-Fe-N(3)
N(3)-Fe-N(3’)
N(1)-N(2)-N(3)
86.7(1)
173.4(2)
108.0(4)
Cl(1)-Fe-N(3’)
a
Symmetry operation to generate equivalent atoms: (’) 1–x, y, 1/2-z.
III
The average Fe -Cl distances are 2.233(1), 2.213(2), 2.227(1) and 2.212(2) Å
in 1, 2, 3 and 4, respectively. These values compare well with the average values of
2
.216(1) and 2.221(2) Å reported for complexes [FeCl (4-cpy) ](4-cpy= 4-
3 2
cyanopyridine) [91] and [FeCl (NMe ) ] [92], respectively; these two compounds are
3
3 2
the only structurally characterized mononuclear tbp complexes of the general formula
trans-[FeCl L ], where L is a monodentate N-donor (the term trans is used to indicate
3
2
III
that the L ligands occupy axial sites in the tbp arrangement). The lengths of the Fe -
N bonds [2.169(1)-2.185(3) Å] do not differ significantly in the four complexes. The
relatively long bond distances may be due [91] to atom-atom repulsions between the
…
donor nitrogen atoms and the chloro ligands; the donor N Cl intramolecular distances
vary from 3.015 to 3.183 Å and they are similar to the values observed in [FeCl (4-
3
cpy) ] [91] for which a long Fe-N bond distance was also found.
2
1
5

Fig. 4. Partially labeled plot of the structure of complex 3, also showing the intermolecular H
bonds (see text for details). Only the H atoms of the azole rings are shown, Symmetry
III
operations: (’) 3/2-x, y, 1/2-z; (’’) 1-x, 1-y, z. Colour scheme: Fe , orange; Cl, green; N, blue;
C, grey. (Colour online).
Table 4
a
Selected interatomic distances (Å) and angles (°) for complex 3.
Bond lengths
Fe – Cl(1)
Fe – Cl(2)
Fe – N(3)
2.235(1)
2.210(1)
2.162(2)
N(1) – N(2)
N(2) – N(3)
1.333(3)
1.312(3)
Bond angles
Cl(1)-Fe-Cl(1’)
Cl(1)-Fe-Cl(2)
Cl(1)-Fe-N(3)
120.8(1)
119.6(1)
90.2(1)
92.2(1)
Cl(2)-Fe-N(3)
N(3)-Fe-N(3’)
N(1)-N(2)-N(3)
87.6(1)
175.2(1)
107.8(2)
Cl(1)-Fe-N(3’)
a
Symmetry operation to generate equivalent atoms: (’) 3/2–x, y, 1/2-z.
Complexes 1-4 join a small family of structurally characterized tbp iron(III)
complexes of the general formula [FeCl (monodentate ligand) ] [91-93], in which the
3
2
monodentate ligands are in axial positions. In only two members of this family the
monodentate ligands are N-donors [91,92], the two other members being complexes
[
FeCl (PMe ) ] and [FeCl (PPh ) ] [93].
3 3 2 3 3 2
1
6

Fig. 5. Partially labeled plot of the structure of complex 4, also showing the intermolecular H
bonds (see text for details). Only the H atoms of the azole rings are shown. Symmetry
III
operations: (’) x, 2-y, -z; (’’) 1-x, 1/2+y, -1/2-z; (’’’) x, 1+y, z. Colour scheme: Fe , orange;
Cl, green; N, blue; C, grey. (Colour online).
Table 5
a,b
Selected interatomic distances (Å) and angles (°) for complex 4.
Bond lengths
Fe – Cl(1)
Fe – Cl(2)
Fe – N(3)
N(1) – N(2)
2.204(2)
2.216(1)
2.185(3)
1.336(4)
N(2) – N(3)
1.316(4)
1.346(4)
1.308(5)
N(11) – N(12)
N(12) – N(13)
Bond angles
Cl(1)-Fe-Cl(2)
Cl(2)-Fe-Cl(2’)
Cl(1)-Fe-N(3)
120.6(1)
118.7(1)
90.0(1)
90.6(1)
Cl(2)-Fe-N(3’)
N(3)-Fe-N(3’)
89.4(1)
179.9(2)
107.3(3)
107.9(3)
N(1)-N(2)-N(3)
N(11)-N(12)-N(13)
Cl(2)-Fe-N(3)
a
Symmetry operation to generate equivalent atoms: (’) x, 2-y, -z.
b
Atoms N(11), N(12) and N(13) belong to the uncoordinated, i.e. lattice 5ClbtaH
molecule.
The crystal structure of 1 is stabilized by intermolecular H bonds and π-π
stacking interactions. The [FeCl (btaH) ] molecules form 1D chains as a consequence
3
2
…
of the N(3)-H(N3) Cl(1’) H-bonding scheme; the dimensions of this intermolecular
…
…
…
H bond are: N(3) Cl(1’) = 3.246 Å, H(N3) Cl(1’) = 2.397 Å, N(3)-H(N3) Cl(1’) =
1
7

1
58.2 °. The π-π interactions (Fig. 6) in this compound concern the stacking of the
btaH molecules, which occurs in the direction of the b axis. Due to the 2-fold
symmetry that generates the full coordination sphere of Fe(1), the benzotriazole
ligands are not co-planar. They are canted by 6.45(1) °, with a centroid-to-centroid
distance of 3.768 Å. Due to the canting of the ligands, the shift distance alternates
between 1.553 and 1.776 Å. When the two rings that make up the btaH ligand are
considered separately, it is clear that the triazole ring makes one closer interaction
with the benzene ring (3.546 Å) and one longer (3.789 Å).
Fig. 6. The stacking interactions in complex 1.
The lattice structure of 2 is built through intermolecular π-π stacking
interactions between the Mebta ligands (Fig. 7). Each [FeCl (Mebta) ] molecule
3
2
interacts with four of its neighbors. The mean planes defined by atoms N(1)-C(9) and
N(1’)-C(9’) [(’) = 1-x, 1-y, -z] are parallel with an interplanar distance of 3.794 Å
(
green dashed lines) and these defined by N(1)-C(9) and N(1’’)-C(9’’) [(’’) = 1-x, -y, -
z] are also parallel with an interplanar distance of 3.782 Å (red dashed lines). The so-
formed 2D network of 2 extends parallel to the bc plane.
1
8

Fig. 7. A small part of the 2D lattice structure of 2 due to π-π stacking interactions (dashed
lines); see text for details. (Colour online).
The crystal structure of complex 3 is stabilized by intermolecular π-π stacking
interaction between the 5,6Me btaH ligands of neighboring molecules (Fig. 8a) as
2
well as H bonding interactions (Figs. 4 and 8b). The ligands defined by atoms N(1)-
C(9) and their symmetry equivalents through the operation 2-x, 1-y, -z are parallel
with an interplanar distance of 3.530 Å (red dashed lines). Moreover, the benzene
rings defined by atoms C(4)-C(9) and their symmetry equivalents through the
operation 2-x, -y, -z are also parallel to each other with an interplanar distance of
4
.322 Å (green dashed lines). These two types of π-π interactions (one strong, the
other weak) build the 2D structure of the complex which extends parallel to the (101)
plane (Fig. 8a). The layers are further linked through H bonds, creating the 3D lattice
structure of 3. Each [FeCl (5,6Me btaH) ] unit links to two neighboring molecules
3
2
2
…
…
through two weak N(1)-H N(2) H bonds as shown in Fig. 4 [N(1) N(2)(1-x, 1-y, -z)
…
…
=
2.886 Å, H(N1) N(2)(1-x, 1-y, -z) = 2.357 Å, N(1)- H(N1) N(2)(1-x, 1-y, -z) =
1
24.0°]. In Fig. 4 only one neighboring molecule is shown; the other has been omitted
because it is obviously created at the other side of the molecule by the 2-fold axis
…
possessed by the molecule. Fig. 8b shows three layers linked through N-H N
interactions (orange lines) which build the 3D network of the complex.
1
9

(
a)
(
b)
Fig. 8. (a) A small part of the 2D network of 3 created by the combination of strong and weak
π-π stacking interactions (red and green dashed lines). (b) A small part of the 3D network of
…
the complex formed by N-H N interactions (orange lines) between molecules of different
layers. See text for details. (Colour online).
The lattice structure of 4 is built through two types of π-π stacking interactions
(
Fig. 9). The coordinated 5ClbtaH ligand defined by atoms N(1)-C(9) and the lattice
ClbtaH molecule defined by atoms N(11)-C(19) are connected through π-π
5
interactions; the two mean planes form an angle of 3.37 ° with an interplanar distance
of 3.544 Å (red dashed lines). A second type of π-π interactions between the benzene
rings defined by atoms C(4)-C(9) and C(14)-C(19) [through the symmetry operation -
1
+x, y, z] is responsible for the 1D lattice structure which extends parallel to the a
axis; the two mean planes form an angle of 3.27 ° with an interplanar distance of
2
0

3
.750 Å (green dashed lines). The chains further link through H bonding interactions
to generate the overall 3D network of 4. The dimensions of the H bonds are:
…
…
N(1) N(13) [1-x, 1/2+y, -1/2-z] = 2.826 Å, H(N1) N(13) [1-x, 1/2+y, -1/2-z] = 2.005
…
…
Å, N(1)-H(N1) N(13) [1-x, 1/2+y, -1/2-z] = 144.4 °; N(2) N(11) [x, 1+y, z] = 3.054
…
…
Å, H(N11) [x, 1+y, z] N(2) = 2.323 Å, N(11) [x, 1+y, z]-H(N11) [x, 1+y, z] N(2) =
1
5
5
43.4 Å. Each [FeCl (5ClbtaH) ] molecule interacts through H bonds with four lattice
3 2
ClbtaH molecules belonging to neighboring chains. In Fig. 5 only two lattice
ClbtaH molecules belonging to a neighboring chain and lying at one side of
[
FeCl (5ClbtaH) ] are shown; the other have been omitted because they are obviously
3 2
created at the other side of the molecule by the 2-fold axis possessed by
FeCl (5ClbtaH) ].
[
3
2
Fig. 9. A small part of the 1D lattice structure of complex 4 created by π-π stacking
interactions; see text for details. (Colour online).
3
.3. Physical and spectroscopic characterization
Data have been given in Part 2.3 (vide supra). A brief discussion will be
-3
presented here. Conductivity measurements for ~10 M solutions of 1-4 at 25 °C
2
-1
indicated the complexes to be non-electrolytes in MeCN (Λ = 3-8 S cm mol ) [94].
M
2
-1
Somewhat to our surprise, the Λ values in MeNO (77-81 S cm mol ) suggest that
M
2
the compounds behave as 1:1 electrolytes in this solvent [94].
2
1

-1
The IR spectra of 1, 3 and 4 exhibit a band at 3232, 3244 and 3260-2800 cm ,
respectively, assignable [21] to ν(NH); the broadness and relatively low wavenumber
of this band are both indicate of hydrogen bonding. In the spectra of 1-4 the bands due
-1
to ν(N=N) and ν(N-N) appearing at 1218-1264 and 1100-1136 cm , respectively, are
shifted to higher frequencies compared with those of the free ligands (1197-1212 and
-1
1
078-1110 cm , respectively). For example, the IR spectrum of free Mebta exhibits
-1
two bands at 1197 and 1110 cm associated with the ν(N=N) and ν(N-N) modes of
-1
vibration, respectively [14]. These bands are shifted to 1218 and 1136 cm ,
respectively, in the spectrum of 2. Such shifts to higher frequencies are compatible
with N(3) [Fig. 1] being the donor atom to the metal ion [2,3,9,14,21]. The
-1
-1
appearance of two ν(N=N) [1264, 1202 cm ] and two ν(N-N) [1112, 1070 cm ]
modes in the IR spectrum of 4 reflects the presence of both coordinated and
uncoordinated 5ClbtaH molecules in its structure.
III
The far-IR spectra of 1-4 show one strong, terminal Fe -Cl stretch, ν(Fe-Cl) ,
t
-1
III
at 374-380 cm [91,93] and one weaker Fe -N stretching mode, ν(Fe-N), at 196-215
-1
III
cm [95]. A reasonable approximation of the molecular symmetry around Fe in the
complexes is D , and thus [91,95] one ν(Fe-Cl) (E’) and one ν(Fe-N) (A ’’) would be
3h
t
2
expected. The wavenumbers of the ν(Fe-Cl) vibration in 1-4 compare very well with
t
the wavenumber of this mode reported for the tbp complex [FeCl (4-cpy) ] [91].
3
2
The experimental μ values of 1-4 at room temperature are in the range 5.79-
eff
1
1
1
6
.06 BM, indicating a high-spin (S = 5/2) state and supporting a d , d < d , d
xz yz xy x2-
1
<
dz2 electronic structure [93].
The solid-state electronic (diffuse reflectance) spectra of 1-4 exhibit four to
y2
five bands in the 550-1500 nm region, and a band at 394-436 nm; the wavelength of
the latter and the fairly oxidizing nature of Fe(III) are suggestive of ligand-to-metal
charge transfer character [96]. The bands in the 550-1500 nm region have their origin
in spin-forbidden d-d transitions from the ground sextet state to the quartet levels
[
91,97]. The solution spectra (MeCN) of the complexes, recorded in the 350-800 nm
region, are identical with the solid-state spectra.
Zero-field Mössbauer spectra of polycrystalline samples of 1-4 were measured
at room temperature. The spectrum of 3 is shown in Fig. 10. The Mössbauer spectral
parameters listed in the experimental section have been obtained from least-squares fit
to the experimental points assuming Lorentzian absorption lines. The spectra consist
of a single quadrupole-split doublet with no evidence of iron contaminants. The
2
2

-1
isomer shift value δ (0.24-0.37 mm s ) is within the range expected for high-spin
iron(III) in a non-sulfur environment [98-100]. The relatively low δ values may
suggest 5-coordinate iron(III) complexes [100-102]. A relatively narrow half-line
-1
width (ca. 0.2 mm s ) is observed indicating that only one type of iron site is present
103]. The asymmetry in peak intensity at room temperature is typical of high-spin
[
ferric complexes that have a large positive zero-field splitting parameter, D [104]; this
phenomenon has been interpreted by Blume in terms of a temperature-dependent
spin-spin relaxation time [105].
Fig. 10. Mössbauer spectrum of [FeCl (5,6Me btaH) ] (3) in zero applied field at room
3
2
2
temperature.
.
The X-band EPR spectrum of complex [FeCl (5ClbtaH) ] 2(5ClbtaH) (4) has
3
2
been recorded at 4.2 K in the solid state (Fig. 11α) and in frozen MeNO (Fig. 11β).
2
The solid-state spectrum exhibits a signal at g = 5.83, typical of the perpendicular
┴
component in a tetragonal high-spin Fe(III) [93]. For a pure axial symmetry (E = 0)
and a D value much greater than the microwave quantum, signals with g ≈ 6 and g
┴
≈
2
are expected [93]. We do not observe a signal at g ≈ 2 that can be attributed to a
parallel component, but this probably overlaps with the stronger signal at geff = 1.89.
Considering the axial symmetry of this compound, proved by its structure, we can
assume that E will be close to 0. According to the classical study of Dowsing and
Gibson [106], five transition should be observed for small D; the 1/2 → -1/2 transition
is expected to be nearly isotropic, but the other two (two at higher field, two at lower)
2
3

should be anisotropic with observable perpendicular and parallel components. Since
we do not observe signals at g < 1.89 (up to a magnetic field of 12000 G), an
-
1
examination of Figure 1 in ref. [106] suggests a D value larger than 0.2 cm for 4.
The spectrum of 4 in MeNO at 4.2 K is different compared with that in the
2
solid state, suggesting that the tbp structure is not retained in solution. Six transitions
are observed with g values in the 9.22-2.01 range. The high-field signal at g = 2.01 is
-
assigned to the tetrahedral [FeCl ] ion, while the remaining spectrum may be due to
4
an octahedral species with small D and axial symmetry [93]. We tentatively propose
that the decomposition of 4 takes place according to Eq. (3). This hypothesis is
consistent with molar conductivity data which show that complex 4 is an 1:1
electrolyte in MeNO2.
Fig. 11. X-band EPR spectrum of complex 4 in the solid state (α) and in MeNO (β) at 4.2
2
K.
2
4

3
.4. Catalytic studies
3
.4.1. Hydrocarbon oxidation by complexes 1, 2 and 4
The catalytic properties of 1, 2 and 4 have been evaluated for hydrocarbon
oxidation with H O as oxidant. The substrates are shown in Fig. 12. The oxidation
2
2
reactions were carried out at room temperature in an Ar atmosphere as detailed in the
experimental section (Part 2.2). The reactions were usually complete within 6 h.
.
Interestingly, when [FeCl (5ClbtaH) ] 2(5ClbtaH) (4) was used as catalyst, 1-methyl-
3
2
cyclohexene and cyclohexane oxidations were accomplished within 2 h. The obtained
catalytic results are summarized in Table 6. Fig. 13 provides a histogram plot of the
data included in Table 1, while the time-depended cyclohexene oxidations catalyzed
by the three complexes are shown in Fig. 14.
trans-β-methyl-styrene
Fig. 12. The substrates used for oxidation reactions catalyzed homogeneously by complexes
1
, 2 and 4 with H O as oxidant.
2
2
2
5

(
(
(
) [FeCl (Mebta) ]
3
2
●) [FeCl (btaH) ]
3
2
.
○) [FeCl (5ClbtaH) ] 2(5ClbtaH)
3
2
Fig. 13. Distribution of oxidation products from reactions catalyzed by complexes 1, 2 and 4
in MeCN in the presence of H O . See Table 6 for further details. (Colour online).
2
2
2
6

Table 6
Hydrocarbon oxidations catalyzed by complexes 1, 2 and 4 using H O in MeCN under Ar.
2
2
.
2
Substrate
Products
[FeCl (btaH) ] (1)
[FeCl (Mebta) ] (2)
[FeCl (5ClbtaH) 2(5ClbtaH) (4)
3
2
3
2
3
Yield (%)
Total yield (%)
Yield (%)
Total yield (%)
Yield (%)
Total yield (%)
a
Cyclohexene
cis-epoxide
22.4
27.0
22.0
5.0
6.0
26.0
16.0
-
20.0
26.0
25.0
5.0
2-cyclohexenol
2-cyclohexenone
cyclohexanediol
c
c
d
c
76.4
48.0
76.0 (80.0)
a
1
-methyl-cyclohexene
cis-epoxide
31.7
25.9
23.6
18.4
23.7
27.0
19.2
20.6
32.0
27.5
16.3
24.2
1-methyl-2-cyclohexen-1-ol
3-methyl-2-cyclohexen-1-ol
3-methyl-2-cyclohexen-1-one
c
c
c
d
9
9.6
90.5
100.0
a
Cyclooctene
cis-epoxide
-cyclooctenone
60.0
4.0
52.0
-
42.0
-
2
c
d
c
64.0
52.0
42.0 (48.5)
a
Limonene
cis-1,2-epoxide
trans-1,2-epoxide
cis-8,9-epoxide
trans-8,9-epoxide
limonene alcoholi
limonene ketone
14.8
12.9
4.0
3.5
11.2
10.3
9.2
2.9
2.6
10.8
5.1
16.6
10.0
5.0
4.0
14.3
e
f
g
h
5.0
10.0
c
c
c
51.4
40.9
57.4
b
Styrene
epoxide
phenyl acetaldehyde
benzaldehyde
8.0
2.0
36.8
6.2
1.0
36.0
11.8
2.3
25.5
c
c
c
46.8
43.2
39.6
b
Trans-β-methyl-styrene
trans-epoxide
methyl-benzyl alcohol
methyl-benzyl ketone
30.0
-
1.9
14.5
-
5.0
32.0
3.0
-
2
7

benzaldehyde
45.5
30.8
31.0
c
c
c
77.4
50.3
66.0
b
Cis-stilbene
cis-epoxide
5.2
10.0
-
3.3
2.0
3.7
14.3
6.7
13.0
2.4
trans-epoxide
stilbene ketone
benzaldehyde
24.5
30.0
c
c
c
c
3
4
9.7
4.3
23.3
33.8
52.1
a
Cyclohexane
cyclohexanol
cyclohexanone
26.3
15.0
3.0
13.7
12.9
7.2
27.0
14.0
1.8
4-hexen-1-ol
c
d
c
42.8 (42.8)
a
n-hexane
3-hexanone
4.5
4.0
4.8
-
2.3
2.9
3.0
2.0
-
5.8
5.6
6.4
4.8
3.3
2-hexanone
3-hexanol
2-hexanol
1-hexanol
8.2
c
c
d
c
21.5
10.2
25.9 (29.7)
a
b
c
d
e
f
Conditions: ratio of catalyst : H O : substrate = 1 : 20 : 1000 in MeCN.
2
2
Conditions: ratio of catalyst : H O : substrate = 1 : 50 : 1000 in MeCN.
2
2
The reaction was completed within 6 h.
The reaction was completed within 2 h.
Limonene alcohols were found to be a mixture of 1-ol and 6-ol.
The 11.2 % yield corresponds to 2.5% for 1-ol and 8.7% for 6-ol.
The 10.8 % yield corresponds to 3.8% for 1-ol and 7.0% for 6-ol.
The 14.3 % yield corresponds to 4.3% for 1-ol and 10.0% for 6-ol.
The only observed ketone is the 6-one.
g
h
i
2
8

80
70
60
50
40
30
20
10
0
[
FeCl (Mebta) ]
3 2
[
FeCl (btaH) ]
3 2
[FeCl (5ClbtaH) ].2(5ClbtaH)
3 2
0
5
10
15
20
time (h)
Fig. 14. Time dependence of cyclohexene oxidation catalyzed by complexes 1 (red line), 2 (black line)
and 4 (green line) in MeCN with H O as oxidant. (Colour online).
2
2
Based on Table 6, the present non-heme iron(III) catalysts are efficient in alkene
oxidations providing significant yields. More specifically, cyclohexene oxidation catalyzed by
complexes 1, 2 and 4 provided oxidation products with combined yields of 76.4, 48.0 and 80.0
%, respectively within 6 h. It is worth mentioning that cyclohexene oxidation by 4 showed a
7
6.0 % yield only after 2 h (Figs. 14 and 15). Cyclohexene undergoes mainly allylic oxidation
forming 2-cyclohexene-1-ol and 2-cyclohexene-1-one. However, cyclohexene epoxidation was
also observed with low to moderate yields. In cyclohexene oxidation catalyzed by 1 and 4,
cyclohexanediol was also detected (5.0 %).
2
9