Tuning redox potentials of bis(imino)pyridine cobalt complexes: An experimental and theoretical study involving solvent and ligand effects

Article abstract of DOI:10.1039/c2dt12195f

The structure and electrochemical properties of a series of bis(imino)pyridine Co^II complexes (NNN)CoX₂ and [(NNN)₂Co][PF₆]₂ (NNN = 2,6-bis[1-(4-R- phenylimino)ethyl]pyridine, with R = CN, CF₃, H, CH₃, OCH₃, N(CH₃)₂; NNN = 2,6-bis[1-(2,6-(iPr) ₂-phenylimino)ethyl]pyridine and X = Cl, Br) were studied using a combination of electrochemical and theoretical methods. Cyclic voltammetry measurements and DFT/B3LYP calculations suggest that in solution (NNN)CoCl ₂ complexes exist in equilibrium with disproportionation products [(NNN)₂Co]²⁺ [CoCl₄]^2- with the position of the equilibrium heavily influenced by both the solvent polarity and the steric and electronic properties of the bis(imino)pyridine ligands. In strong polar solvents (e.g., CH₃CN or H₂O) or with electron donating substituents (R = OCH₃ or N(CH₃) ₂) the equilibrium is shifted and only oxidation of the charged products [(NNN)₂Co]²⁺ and [CoCl₄]^2- is observed. Conversely, in nonpolar organic solvents such as CH ₂Cl₂ or with electron withdrawing substituents (R = CN or CF₃), disproportionation is suppressed and oxidation of the (NNN)CoCl₂ complexes leads to 18e^- Co^III complexes stabilized by coordination of a solvent moiety. In addition, the [(NNN)₂Co][PF₆]₂ complexes exhibit reversible Co^II/III oxidation potentials that are strongly dependent on the electron withdrawing/donating nature of the N-aryl substituents, spanning nearly 750 mV in acetonitrile. The resulting insight on the regulation of redox properties of a series of bis(imino)pyridine cobalt(ii) complexes should be particularly valuable to tune suitable conditions for reactivity. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2dt12195f

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 3562
www.rsc.org/dalton
PAPER
Tuning redox potentials of bis(imino)pyridine cobalt complexes: an
experimental and theoretical study involving solvent and ligand effects†
C. Moyses Araujo,*a Mark D. Doherty,*b Steven J. Konezny,a Oana R. Luca,a Alex Usyatinsky,b
Hans Grade,b Emil Lobkovsky,c Grigorii L. Soloveichik,*b Robert H. Crabtreea and Victor S. Batista*a
Received 16th November 2011, Accepted 14th January 2012
DOI: 10.1039/c2dt12195f
The structure and electrochemical properties of a series of bis(imino)pyridine CoII complexes (NNN)
CoX2 and [(NNN)2Co][PF6]2 (NNN = 2,6-bis[1-(4-R-phenylimino)ethyl]pyridine, with R = CN, CF3, H,
CH3, OCH3, N(CH3)2; NNN = 2,6-bis[1-(2,6-(iPr)2-phenylimino)ethyl]pyridine and X = Cl, Br) were
studied using a combination of electrochemical and theoretical methods. Cyclic voltammetry
measurements and DFT/B3LYP calculations suggest that in solution (NNN)CoCl2 complexes exist in
equilibrium with disproportionation products [(NNN)2Co]2+ [CoCl4]2− with the position of the
equilibrium heavily influenced by both the solvent polarity and the steric and electronic properties of the
bis(imino)pyridine ligands. In strong polar solvents (e.g., CH3CN or H2O) or with electron donating
substituents (R = OCH3 or N(CH3)2) the equilibrium is shifted and only oxidation of the charged
products [(NNN)2Co]2+ and [CoCl4]2− is observed. Conversely, in nonpolar organic solvents such as
CH2Cl2 or with electron withdrawing substituents (R = CN or CF3), disproportionation is suppressed and
oxidation of the (NNN)CoCl2 complexes leads to 18e− CoIII complexes stabilized by coordination of a
solvent moiety. In addition, the [(NNN)2Co][PF6]2 complexes exhibit reversible CoII/III oxidation
potentials that are strongly dependent on the electron withdrawing/donating nature of the N-aryl
substituents, spanning nearly 750 mV in acetonitrile. The resulting insight on the regulation of redox
properties of a series of bis(imino)pyridine cobalt(II) complexes should be particularly valuable to tune
suitable conditions for reactivity.
provides the protons and electrons necessary to drive O2
Introduction
reduction to water.2 However, the development of electrocata-
Understanding the redox properties of earth-abundant transition
metal complexes is a challenge of great current interest, central
to a wide range of applications in electrocatalysis.1 An important
goal is to establish fundamental principles to tune redox poten-
tials and catalytic activity by modifying the ligands, or solvent
conditions. In particular, there is a lot of current interest in the
development of catalysts for direct organic proton exchange
membrane (PEM) fuel cells where oxidation of an organic fuel
lysts capable of oxidizing organic fuels such as isopropanol or
cyclohexane2c,d is hindered by a lack of understanding of the
conditions that regulate the redox properties of earth-abundant
transition metal complexes. This paper explores a series of Co
complexes with bis(imino)pyridine ligands (Chart 1), with
emphasis on the analysis of their electrochemical properties as
influenced by the solvent, or ligand substituents with electron
withdrawing/donating properties.
There is a sizable body of work documenting the utility of
various transition metal complexes of bis(imino)pyridine
ligands. Independently, Brookhart and Gibson have demon-
strated that such complexes, particularly those containing N-aryl
substituents with halides of iron(II) and cobalt(II), upon activation
with methylaluminoxane (MAO) serve as precatalysts for
α-olefin homopolymerization,3 dimerization,4 oligomeriza-
tion3c,4b,5 and co-polymerization with polar monomers.3c,f Deri-
vatization of the N-aryl groups resulted in numerous substitution
patterns3g and established the role that steric bulk at the ortho
positions plays in determining catalyst activity, selectivity and, in
the case of polymerization catalysts, polymer molecular
weight.3b,d,e,h,i Further studies have focused on preparing
reduced iron or cobalt alkyl complexes, particularly methyl
aDepartment of Chemistry, Yale University, New Haven, CT 06520.
E-mail: carlos.araujo@yale.edu, victor.batista@yale.edu
bGeneral Electric Global Research, 1 Research Circle, Niskayuna,
NY 12309. E-mail: doherty@ge.com, soloveichik@ge.com
cDepartment of Chemistry and Chemical Biology, Cornell University,
Ithaca, NY 14853
†Electronic supplementary information (ESI) available: The supporting
information includes X-ray crystallographic files (CIF) for [(1f)2Co]
[CoCl4] and 2g, a table containing the electrochemical reduction poten-
tials for compounds 2a–d, f, g and 3a, overlaid CV’s of 2f and 4f in
dichloromethane, 1H NMR spectra of paramagnetic complexes, gas
phase energies, Gibbs free energy upon solvation, and coordinates of the
optimized computational structural models. CCDC 854625 (2g) and
854626 ([(1f)2Co][CoCl4]). For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c2dt12195f
3562 | Dalton Trans., 2012, 41, 3562–3573
This journal is © The Royal Society of Chemistry 2012

View Online
Chart 1 NNN ligands 1a–g, (NNN)CoIIX2 complexes 2a–d, f, g and 3a, and [(NNN)2CoII][PF6]2 complexes 4b–g examined in this work.
complexes, as models of the propagating species present in the
olefin polymerization systems described above.6 Interest in these
reduced compounds has grown to include detailed investigations
into the electronic structure, redox properties and reactivity of
the iron and cobalt complexes and has established the redox
activity of these ligands.6e,7
Results and discussion
Synthesis and electrochemical characterization of (NNN)CoX2
complexes
The (NNN)CoX2 complexes were prepared in accordance with
literature procedures for similar compounds by reaction of equal
parts ligand and CoX2 and the compounds examined in this
study are shown in Chart 1 (see Experimental section for
details). As expected, complexes 2a–d, f and g are high spin and
More recently, several new applications of these bis(imino)
pyridine transition metal complexes have been reported includ-
ing a report by Chirik and co-workers in which they demon-
strated C–H bond activation via thermolysis and photolysis of
bis(imino) pyridine cobalt azide complexes.8 In addition, highly
efficient catalytic olefin hydrogenation has been achieved with
bis(imino)pyridine iron bis(dinitrogen) complexes.9 Related bis
(imino)pyridine iron(II) triflate complexes have also shown some
activity in the oxidation of cyclohexane with H2O2 to cyclohexa-
nol and cyclohexanone, presumably through a Fenton type free
radical mechanism.10 Therefore, it is natural to expect that a
wide range of applications could benefit from fundamental
understanding of the dependence of the redox properties of these
metal pincer complexes on the nature of the ligands, and on
solvent conditions. While previous studies have analyzed the
electronic structures associated with various possible oxidation
states of cobalt bis(imino)pyridine complexes,7c,e a systematic
study of the regulation of redox potentials as influenced by
changes in the ligand framework or solvent environment is
lacking.
In that context we report here an experimental and theoretical
examination of the influence of electrostatic, steric and electronic
factors on the redox potentials of a series of Co(NNN)X2 and
[(NNN)2Co][PF6]2 (NNN is a tridentate bis(imino)pyridine
ligand) complexes as shown in Chart 1. We examined ligands
with different electron withdrawing or donating substituents and
solvents of different polarity ranging from nonpolar to polar
protic solvents (e.g. CH2Cl2, CH3CN and H2O). While many
studies of these CoII complexes have dealt with their reduction
and subsequent chemical reactivity,6b,7b,c,8 our interest lies in
ultimately being able to oxidize liquid organic fuels for energy
storage purposes.2a Therefore we have chosen to focus on
the CoII/III redox couple and factors that contribute to one’s
ability to tune the oxidation potentials of these transition metal
complexes.
1
display paramagnetically broadened H NMR spectra, with sol-
ution magnetic moments of 3.5–5.3 μB at 22 °C (Evans NMR
method11), which is consistent with a high spin d7 CoII
center.3d,12
One method of characterizing the redox activity of bis(imino)
pyridine ligands is X-ray crystallography where reduced ligands
exhibit bond distortions from neutral ligands. Black needle
shaped single crystals suitable for single-crystal X-ray diffraction
were obtained by vapor diffusion of diethyl ether into a saturated
acetonitrile solution of 2g. The molecular structure is shown in
Fig. 1 while the collection and refinement parameters are col-
lected in Table 6. Compound 2g exhibits approximate Cs sym-
metry about the plane containing the cobalt, pyridyl nitrogen and
two chlorine atoms. The geometry around the cobalt is best
described as distorted trigonal bipyramidal with the equatorial
plane formed by the pyridyl nitrogen and two chlorine atoms
with Npyridyl–Co–Cl angles of 130.79(5)° and 112.91(5)° and a
Cl–Co–Cl angle of 116.23(3)°. Two axial Co–Nimine bonds fill
out the cobalt(II) coordination sphere with an angle of 150.12
(7)°. A bond distance of 2.0344(18) Å for the Co–Npyridyl of 2g
is consistent with other para substituted complexes such as those
reported for 2c13 and 2d14 and shorter than those reported for
ortho substituted complexes such as 2a.3d As with 2c13 and 2d14
the Co–Nimine bonds are slightly longer at 2.2093(18) and
2.2289(18) Å, while the Cimine–Nimine distances of 1.289(3) and
1.288(3) Å are typical for CvN double bonds of neutral bis
(imino)pyridine ligands coordinated to first row transition
metals.7e In addition, Cimine–Cpy bond lengths of 1.488(3) and
1.489(3) Å and Cpy–Npy bond lengths of 1.345(3) and 1.339(3)
Å are consistent with the assignment of a neutral bis(imino)pyri-
dine ligand.7e The N-aryl rings of 2g exhibit dihedral angles of
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3562–3573 | 3563

View Online
Fig. 1 ORTEP diagram of 2g (40% probability ellipsoids). Hydrogen atoms have been removed for clarity. Selected bond distances (Å) and angles
(°): Co(1)–N(1) 2.2093(18), Co(1)–N(2) 2.0344(18), Co(1)–N(3) 2.2289(18), Co(1)–Cl(1) 2.2733(6), Co(1)–Cl(2) 2.2562(6), N(1)–C(2) 1.289(3), C
(2)–C(3) 1.488(3), C(3)–N(2) 1.345(3), N(2)–C(7) 1.339(3), C(7)–C(8) 1.489(3), N(3)–C(8) 1.288(2), N(2)–Co(1)–N(1) 75.03(7), N(2)–Co(1)–N(3)
75.55(3), (N1)–Co(1)–N(3) 150.11(7), N(2)–Co(1)–Cl(2) 130.79(5), N(1)–Co(1)–Cl(2) 94.46(5), N(3)–Co(1)–Cl(2) 100.75(5), N(2)–Co(1)–Cl(1)
112.91(5), N(1)–Co(1)–Cl(1) 99.53(5), N(3)–Co(1)–Cl(1) 96.58(5), Cl(2)–Co(1)–Cl(1) 116.23(3).
acetonitrile but irreversible in dichloromethane. Except for 2f
and 2g, changing from acetonitrile to dichloromethane induces
an anodic shift in the CoII/III oxidation potentials of +0.05 to
+0.55 V, which we attribute to decreased stabilization of the
developing cationic charge upon oxidation in a low polarity
solvent such as dichloromethane. The slight cathodic shift of
less than 50 mV observed for 2f and 2g upon moving to dichlo-
romethane was unexpected and prompted a more detailed
investigation.
Analysis of 2g by matrix assisted laser desorption ionization
time-of-flight (MALDI-TOF) mass spectrometry from an HBM–
acetonitrile matrix (see Experimental section for details) revealed
the presence of a species (ca. 15%) with m/z = 857 in addition to
the expected m/z = 493 corresponding to [2g ¬ Cl]+. This minor
component, which does not contain any chlorine atoms, corre-
sponds instead to [(1g)2Co]+. Similar analysis of 2f showed the
expected [2f ¬ Cl]+ (m/z = 467) ions as well as the high m/z
species with m/z = 805 corresponding to [(1f)2Co]+, which in
this case is the major component at ca. 60%. Mass spectrometry
analysis of 2d showed only a minor amount of the high m/z
species (<5% m/z = 685 [(1d)2Co]+, 95% m/z = 407 [2d ¬ Cl]+)
whereas 2c and 2b contained only the expected [2c ¬ Cl]+ (m/z
= 543) and [2b ¬ Cl]+ (m/z = 457), respectively. As expected,
complexes 2a and 3a exhibited only the anticipated [2a ¬ Cl]+
(m/z = 575) and [3a ¬ Cl]+ (m/z = 621) ions by mass spec-
trometry as a result of the steric bulk imparted by the ortho
i-propyl groups that prohibit formation of any bis-ligand species.
Table 1 Electrochemical data for 2a–d, f, g and 3a in acetonitrile and
dichloromethane (0.1 [Bu4N][BF4] supporting electrolyte) vs.
Cp2Fe0/+ at 0.1 V s−1 at a glassy carbon working electrode (d = 3 mm)
M
CH3CN
CH2Cl2
CoII/III (V)a ΔE (mV) CoII/III (V)a ΔE (mV) ΔCoII/III (V)b
2a +0.56
3a +0.35c
2b +0.44c
2c +0.40c
2d +0.11
120
N/A
N/A
N/A
80
70
60
+0.70
+0.69c
+0.82c
+0.81c
+0.67c
−0.04
−0.21
152
N/A
N/A
N/A
N/A
120
75
+0.14
+0.34
+0.38
+0.41
+0.56
−0.03
−0.02
2f
−0.01
2g −0.19
a Half wave potentials, E1/2, unless otherwise indicated. b ΔCoII/III
CoII/IIIDCM–CoII/III c Irreversible oxidation, EPA
=
.
.
MeCN
113.1 and 56.8° relative to the plane of the pyridine ring, which
is consistent with those observed for other complexes lacking
steric bulk at the ortho positions of the N-aryl ring 2d.14
Cyclic voltammograms of 1 mM solutions of 2a–d, f and g
and 3a exhibit CoII/III oxidation waves between −0.25 and +1.0
V vs. Cp2Fe0/+ in acetonitrile or dichloromethane (see Table 1).
Replacing the chloride ligands with bromide ligands results in a
CoII center that is more easily oxidized (2a vs. 3a), however,
whereas 2a exhibits a quasi-reversible oxidation wave in both
solvents, oxidation of 3a is electrochemically reversible in
3564 | Dalton Trans., 2012, 41, 3562–3573
This journal is © The Royal Society of Chemistry 2012

View Online
for the CoII/III oxidation wave at +0.03 V. For a one electron
transfer process with equimolar concentrations of 4f and 2f, the
ratio of peak current densities, i4f/i2f = 1. Ligand redistribution of
a 1 mM solution of 2f to form 0.5 mM [(1f)2Co][CoCl4] and
comparison with a 1 mM solution of 4f should give a ratio of
peak current densities, i4f/i2f = 2. As illustrated in Fig. 2a, the
ratio of peak current densities for 1 mM solutions of 4f (iPA
=
0.273 mA cm−2) and 2f (iPA = 0.135 mA cm−2), i4f/i2f is equal
to 2 and is therefore most consistent with ligand disproportiona-
tion of 2f to form [(1f)2Co][CoCl4] in acetonitrile solution. A
similar comparison of the CV of 2f and 4f in dichloromethane
with 0.1 M [Bu4N][BF4] as the supporting electrolyte indicates
this ligand disproportionation reaction also occurs in dichloro-
methane (see Fig. S1 in ESI†).
2ðNNNÞCoCl2 O ½ðNNNÞ2Coꢀ½CoCl4ꢀ
ð1Þ
Attempts to obtain single crystals of 2f by vapor diffusion of
diethyl ether into a concentrated acetonitrile solution of 2f
resulted instead in red plate like crystals of the ion pair com-
prised of [(1f)2Co]2+ and [CoCl4]2−, the molecular structure of
which is shown in Fig. 3 (see Table 6 for crystallographic data).
Tetrahalometallate salts of [(NNN)2M]n+ have been observed
previously for iron(II) complexes bearing NNN ligands that lack
significant steric bulk in the 2,6 positions of the N-aryl rings
(e.g. 2,2′;6′,2′′-terpyridine15 or 2,6-bis(2-(2-fluorophenylimino)
ethyl)pyridine16). As expected the [CoCl4]2− anion exhibits a
slightly distorted tetrahedral geometry with nearly uniform Co–
Cl bond lengths of 2.2759(10)–2.2867(10) Å and Cl–Co–Cl
angles ranging from 105.74(4)° to 113.43(4)°. Although the
[(1f)2Co]2+ cation has been structurally characterized previously
as the hexafluorophosphate salt there are some significant differ-
ences in the metrical parameters of the cations in 4f7e and
[(1f)2Co][CoCl4] reported here as shown in Table 2. Both struc-
tures display distorted octahedral geometries with two meridion-
ally coordinated bis(imino)pyridine ligands with mutually trans
pyridine nitrogens. The Cpy–Npy, Cimine–Cpy and Cimine–Nimine
bond distances of the ligands in [(1f)2Co][CoCl4] and 4f do not
show any substantial deviations from one another and are in
agreement with the previous assignment of neutral bis(imino)
pyridine ligands in the dication.7e However, the bis(imino)pyri-
dine ligands are bound tighter in [(1f)2Co][CoCl4] than in 4f.
One of the Co–Npy and three of the Co–Nimine distances are
between 0.03 and 0.08 Å shorter in [(1f)2Co][CoCl4] (1.913(2),
1.996(2), 1.989(2) and 2.135(2) Å) than in 4f (1.991(2), 2.009
(2), 2.021(2) and 2.165(2) Å) while the remaining Co–Npy dis-
tance is equal.
Fig. 2 Cyclic voltammograms of 1 mM solutions of (a) 2f (—) and 4f
(- -) and (b) [Bu4N]2[CoCl4] (- -), 2f (—) and CoCl2 (— · ·) in aceto-
nitrile with 0.1 M [Bu4N][BF4] as supporting electrolyte and ν = 0.1
Vs−1 at a glassy carbon working electrode (d = 3 mm).
Close examination of the cyclic voltammogram (CV) of 2f
recorded in acetonitrile and comparison with literature data for
4f7e and the CV of an authentic sample of 4f (vide infra) shown
in Fig. 2a indicates the presence of the [(1f)2Co]2+ dication in
acetonitrile solutions of 2f. In addition, anodic scans to higher
potential reveal two electrochemically irreversible oxidations at
+0.68 V and +1.17 V vs. Cp2Fe0/+. Given the reaction stoichi-
ometry of 1 : 1 1f : CoCl2, the presence of a [(1f)2Co]2+ dication
suggests either outer sphere chloride anions and excess CoCl2 or
the formation of a tetrachlorocobaltate dianion, [CoCl4]2− (eqn
(1)). A cyclic voltammogram of [Bu4N]2[CoCl4] is shown in
Fig. 2b overlaid with a CV of 2f in acetonitrile, and shows two
electrochemically irreversible oxidation waves at +0.72 V and
+1.19 V vs. Cp2Fe0/+, which agrees with the assignment of a
[CoCl4]2− dianion. On the other hand, CoCl2 (— · · in Fig. 2b)
exhibits a sharp electrochemically irreversible oxidation at +1.54
V vs. Cp2Fe0/+ under similar conditions. Additional evidence for
the formation of [(1f)2Co][CoCl4] in acetonitrile solutions of 2f
can be gleaned from a comparison of the peak current densities
Whereas the intraligand Npy–Co–Nimine and Nimine–Co–Nimine
angles are in general only slightly larger (0.3 to 0.6°) in
[(1f)2Co][CoCl4] than in 4f, important differences between the
structures of these two cations are apparent upon comparison of
the interligand Npy–Co–Npy, Npy–Co–Nimine and Nimine–Co–
Nimine angles. The Npy–Co–Npy angle in [(1f)2Co][CoCl4]
(175.39(10)°) is farther from linearity than in 4f (178.48(8)°).
The average interligand Npy–Co–Nimine angles for 4f and
[(1f)2Co][CoCl4] are 100.97(8)° and 100.73(10)°, respectively,
however in 4f each ligand is tilted slightly toward the Co–Npy
bond of the other ligand and contains one large angle (N(2)–Co
(1)–N(6) at 103.46(8)° and N(5)–Co(1)–N(3) at 100.91(8)°) and
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3562–3573 | 3565

View Online
Fig. 3 ORTEP diagram of [(1f)2Co][CoCl4] (40% probability ellipsoids). Hydrogen atoms and a molecule of acetonitrile have been removed for
clarity. Metrical parameters for the cation are collected in Table 2 while selected bond distances (Å) and angles (°) for the anion follow: Co(2)–Cl(1)
2.2792(10), Co(2)–Cl(2) 2.2759(10), Co(2)–Cl(3) 2.2774(9), Co(2)–Cl(4) 2.2867(10), Cl(2)–Co(2)–Cl(3) 112.57(4), Cl(2)–Co(2)–Cl(1) 109.72(4), Cl
(3)–Co(2)–Cl(1) 106.79(4), Cl(2)–Co(2)–Cl(4) 105.74(4), Cl(3)–Co(2)–Cl(4) 108.70(4), Cl(1)–Co(2)–Cl(4) 113.43(4).
one small angle (N(2)–Co(1)–N(4) at 100.97(8)° and N(5)–Co
(1)–N(1) at 98.52(8)°). One ligand of [(1f)2Co][CoCl4] is tilted
more strongly toward the Co–Npy bond of the other ligand (N
(2)–Co(1)–N(4) at 106.61(10)° and N(2)–Co(1)–N(6) at 97.20
(10)°) while the imine nitrogens of the other ligand are virtually
equidistant from N(5) (N(5)–Co(1)–N(1) at 99.92(10)° and N
(5)–Co(1)–N(3) at 99.18(10)°). Finally, the Nimine–Co–Nimine
angles show the most significant structural deviations between 4f
and [(1f)2Co][CoCl4]. As with the Npy–Co–Nimine angles, there
are two large (N(1)–Co(1)–N(6) at 97.79(8)° and N(3)–Co(1)–N
(4) at 97.11(8)°) and two small (N(1)–Co(1)–N(4) at 87.60(8)°
and N(3)–Co(1)–N(6) at 85.73(8)°) Nimine–Co–Nimine angles in
4f. These angles in [(1f)2Co][CoCl4] differ from those in 4f by
5.4° to 5.9° and are much closer to right angles (N(1)–Co(1)–N
(4) at 93.47(9)°, N(1)–Co(1)–N(6) at 92.06(10)°, N(3)–Co(1)–N
(4) at 91.21(9)° and N(3)–Co(1)–N(6) at 91.11(9)°) suggesting
that the two bis(imino)pyridine ligands in [(1f)2Co][CoCl4] are
virtually orthogonal to one another and that the distortion from
an ideal octahedral geometry is less pronounced than in 4f.
corresponding [(NNN)2Co][PF6]2 complexes were prepared
using the method reported by de Bruin7e for 4f and are shown in
Chart 1. The addition of two equivalents of ligand to CoCl2 in
methanol followed by excess ammonium hexafluorophosphate
precipitates the product, which was collected by filtration,
washed with cold methanol and air dried. As with 2a–d, f and g
these cobalt(II) complexes exhibit paramagnetically shifted 1H
NMR spectra, however the solution magnetic moments at 22 °C
for 4b–g are considerably smaller than those measured for the
dichloride complexes and increase with decreasing electron
donating ability from 2.1 μB for 4g to 3.5 μB for 4b. These data
are consistent with a low spin d7 CoII center with S = ½ as pre-
viously observed for 4f7e except for 4b, which is better described
3
as having a high spin S = /2 ground state. This change in spin
state is most likely the result of the strongly electron withdrawing
4-CN substituent in 4b (σp = 0.66), which serves to stabilize the
eg (M–L σ*) orbitals and decrease the ligand field splitting. As
will be discussed in more detail later, this also causes the half
wave potential of 4b to shift to more positive potentials relative
to 4c–g as it becomes more difficult to remove an electron from
the stabilized σ* orbital.
As evidenced by the cyclic voltammograms shown in Fig. 4
and the data collected in Table 3, complexes 4b–g exhibit much
simpler redox chemistry. Each complex shows a reversible one-
electron metal centered oxidation (CoII/III) and one (4c and e) or
Synthesis and electrochemical characterization of [(NNN)2Co]
[PF6]2 complexes
In order to fully evaluate the contribution of the [(NNN)2Co]2+
ions to the electrochemical behavior of 2b–d, f and g, the
3566 | Dalton Trans., 2012, 41, 3562–3573
This journal is © The Royal Society of Chemistry 2012

View Online
two (4b, d, f and g) reversible one-electron reductions. In
accordance with previous electrochemical assignments for 4f7e,
we assign the first reduction between −0.6 and −1.1 V vs.
Cp2Fe0/+ as a metal centered reduction to [CoI(NNN0)2]+, and
the second reduction to a ligand centered process resulting in
[CoI(NNN0)(NNN−1)]0. The CoII/III oxidation potential and first
reduction potential (e.g. [ML2]+/2+) also show a strong corre-
lation with the Hammett parameter17 for the substituents in the
para position of the N-aryl groups on the ligands as shown in
Fig. 5. Replacing the p-H (σp = 0.00) on the N-aryl rings of 4d
with an electron withdrawing substituent induces an anodic shift
of the CoII/III oxidation potential by as much as +0.28 V for 4b
with a p-CN (σp = 0.66). As highlighted previously, this is the
result of stabilization afforded by the electron withdrawing sub-
stituents, which increases the potential required to remove an
electron from the singly occupied eg orbitals. Conversely, repla-
cing the p-H with a donating substituent such as p-NMe2 results
in a cathodic shift of the CoII/III oxidation potential of approxi-
mately the same magnitude (−0.31 V, σp = −0.83) for 4g. Here
the destabilization provided by the electron donating substituents
raises the energy of the singly occupied eg orbitals allowing the
oxidation to occur at much lower potentials.
Table 2 Selected bond distances (Å) and bond angles (°) for [(1f)2Co]
[CoCl4] and 4f and the minimum energy structure of 4f obtained in the
gas phase and at the B3LYP/lacvp-6-31g level of theory
[(1f)2Co][CoCl4]
4f7e
4f(DFT)
Bond distance
Co–Npy
Co(1)–N(2)
Co(1)–N(5)
1.852(2)
1.913(2)
Co–Nimine
1.996(2)
1.989(2)
2.191(2)
2.135(2)
1.852(2)
1.991(2)
1.899
1.929
Co(1)–N(1)
Co(1)–N(3)
Co(1)–N(4)
Co(1)–N(6)
2.009(2)
2.021(2)
2.145(2)
2.165(2)
2.146
2.149
2.236
2.228
C
imine–Nimine
N(1)–C(2)
N(3)–C(8)
N(4)–C(25)
N(6)–N(31)
1.297(4)
1.300(4)
1.294(4)
1.287(4)
1.298(3)
1.305(3)
1.289(3)
1.289(3)
1.311
1.311
1.306
1.307
C
imine–Cpy
Using the electrochemical data for 4b–g as a point of refer-
ence it is now possible to reexamine the data in Table 1 for evi-
dence of ligand redistribution to form [(NNN)2Co]2+ with 2b–d,
f and g. As with 2f and in agreement with the mass spectrometry
results (vide supra) the N,N-dimethylamino substituted 2g also
exhibits redox waves consistent with the formation of [(1g)2Co]
[CoCl4] in acetonitrile (cf. Table 1: CoII/III = (−0.19 V and
C(2)–C(3)
C(7)–C(8)
C(25)–C(26)
C(30)–C(31)
1.469(4)
1.476(4)
1.487(4)
1.487(4)
1.472(4)
1.476(3)
1.487(3)
1.485(3)
1.476
1.477
1.482
1.481
C
py–Npy
N(2)–C(3)
N(2)–C(7)
N(5)–C(26)
N(5)–C(30)
1.349(4)
1.347(4)
1.348(4)
1.346(4)
1.347(3)
1.350(3)
1.347(3)
1.350(3)
1.356
1.358
1.359
1.359
Bond Angle
N
py–Co–Npy (trans)
N(2)–Co(1)–N(5)
175.39(10)
178.48(8)
179.30
N
imine–Co–Nimine (trans)
N(1)–Co(1)–N(3)
N(4)–Co(1)–N(6)
160.88(10)
156.14(9)
160.56(8)
155.54(8)
156.22
158.42
N
py–Co–Nimine (cis-intraligand)
N(2)–Co(1)–N(1)
N(2)–Co(1)–N(3)
N(5)–Co(1)–N(4)
N(5)–Co(1)–N(6)
80.04(10)
80.86(10)
78.00(10)
78.19(10)
80.42(8)
77.99
78.27
79.32
79.10
80.42(8)
77.45(8)
78.16(8)
N
py–Co–Nimine (cis-interligand)
N(2)–Co(1)–N(4)
N(2)–Co(1)–N(6)
N(5)–Co(1)–N(1)
N(5)–Co(1)–N(3)
106.61(10)
97.20(10)
99.92(10)
99.18(10)
100.97(8)
101.36
100.20
102.05
101.68
103.46(8)
98.52(8)
100.91(8)
N
imine–Co–Nimine (cis)
N(1)–Co(1)–N(4)
N(1)–Co(1)–N(6)
N(3)–Co(1)–N(4)
N(3)–Co(1)–N(6)
93.47(9)
92.06(10)
91.21(9)
91.11(9)
87.60(8)
97.79(8)
97.11(8)
85.73(8)
88.42
96.02
94.73
89.10
Fig. 4 Overlay of cyclic voltammograms of 1 mM solutions of 4b–g
recorded in acetonitrile with 0.1 M [Bu4N][BF4] as supporting electro-
lyte and ν = 0.1 Vs−1 at a glassy carbon working electrode (d = 3 mm).
Table 3 Electrochemical data for 4b–g in acetonitrile (0.1 M [Bu4N][BF4] supporting electrolyte) vs. Cp2Fe0/+ at 0.1 V s−1 at a glassy carbon
working electrode (d = 3 mm)
[ML2]0/+
ΔE (mV)
[ML2]+/2+
ΔE (mV)
[ML2]2+/3+
ΔE (mV)
σp
a
4b
4c
4d
4e
4f
−1.59
−1.69c
−1.88
−1.52c
−1.92
−1.99
70
N/A
75
N/A
75
80
−0.65
−0.70
−0.87
−0.92
−0.94
−1.05
70
90
70
85
70
70
+0.39
+0.35
+0.11
+0.03
−0.01
−0.20
75
110
70
80
75
75
0.66
0.54
0.00
−0.17
−-0.27
−0.83
4gb
a See ref. 17. b This work and ref. 7e. c Irreversible reduction, EPC
.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3562–3573 | 3567

View Online
Table S1†: ERED = −1.13 V for 2g and Table 3: [ML2]2+/3+
=
−0.08 V for 2d vs. Cp2Fe0/+). While conversion of the (NNN)
CoCl2 complexes to the [(NNN)2Co][CoCl4] salts appears to be
more facile in acetonitrile than in dichloromethane, the electronic
effects exerted by the bis(imino)pyridine substituents are also
important and in the case of 2b and 2c the electronic effects are
strong enough to prevent this ligand disproportionation.
−0.20 V and [ML2]+/2+ = −1.05 V for 4g vs. Cp2Fe0/+) as well
as dichloromethane (cf. Table 1: CoII/III = −0.21 V and
Table S1†: ERED = −1.02 V for 2g vs. Cp2Fe0/+). In contrast, the
CVs of 2b and 2c bearing strong electron withdrawing substitu-
ents do not exhibit any significant contributions from
[(1b)2Co]2+ or [(1c)2Co]2+ in acetonitrile or dichloromethane, an
observation that is also consistent with the mass spectrometry
results. Interestingly, the CV of 2d, which does not contain
strongly donating or accepting substituents on the bis(imino)pyr-
idine ligand, shows complete conversion to [(1d)2Co]2+ in aceto-
Theoretical evaluation of electrochemical properties of (NNN)
CoX2 complexes
nitrile (cf. Table 1: CoII/III = +0.11 V and Table S1†: ERED
=
We have carried out a thorough analysis of the ligand binding
free energies and redox potentials at the DFT/B3LYP level
(Fig. 6), using the computational methodology based on implicit
solvation models (see Experimental section for details).18 We
investigated (NNN)CoIICl2 and [(NNN)CoIIICl2]+ complexes in
three solvents of increasing polarity (CH2Cl2: ε = 8.9, CH3CN:
ε = 37.5 and H2O: ε = 80.4) and considered various reaction
pathways involving solvent–ligand exchange and disproportiona-
tion. The results reported in Fig. 6 correspond to 2d as a rep-
resentative example for 2a-d, f and g. Table 4 compares the bond
lengths and angles obtained for the calculated minimum energy
structure of 2d to the corresponding X-ray structure obtained by
Gong et. al.14 The bond lengths and angles involving the bis
(imino)pyridine ligand and the cobalt center in the DFT opti-
mized structure of 2d agree very well with the reported struc-
ture14 (largest difference of ca. 0.03 Å and 0.1°, respectively)
and are consistent with a neutral bis(imino)pyridine ligand envir-
onment.7e Although the bond lengths and angles involving the
chloride ligands show substantial deviations from the reported
structure by ca. 0.09 Å and 5–20°, respectively, these are likely
the result of crystal packing forces not accounted for in the cal-
culated structure. In agreement with experimental data, 2d was
calculated to be a 17e− high spin CoII quartet ground state (d7,
−0.88 V for 2d and Table 3: [ML2]2+/3+ = +0.11 V and
[ML2]+/2+ = −0.87 V for 4d vs. Cp2Fe0/+) but not in dichloro-
methane (cf. Table 1: CoII/III = +0.67 V and Table S1†: ERED
=
Fig. 5 Hammett plot correlating the (a) [ML2]2+/3+ (e.g. CoII/III) and
(b) [ML2]+/2+ E1/2 with σp for the ligand N-aryl para substituents.
3
S = /2) with a trigonal bipyramidal geometry, while the singlet
Fig. 6 The reaction pathways calculated for 2d (L = 1d) in CH2Cl2 (DCM), CH3CN (AN), and H2O, respectively. Free-energy changes ΔG are
reported in kcal mol−1 and redox potentials, E° in V vs. Cp2Fe0/+
.
3568 | Dalton Trans., 2012, 41, 3562–3573
This journal is © The Royal Society of Chemistry 2012

View Online
Table 4 Selected bond distances (Å) and bond angles (°) for 2d and
the minimum energy structure of 2d obtained in the gas phase at the
DFT B3LYP/lacvp-6-31G level of theory (elements labeled according to
the reported structure)14
Table 5 Calculated free energies, ΔG° in kcal mol−1 for
disproportionation of 2b, d, f to [(1b, d, f)2Co][CoCl4]
Solvent
2f (R = 4-OCH3)
2d (R = 4-H)
2b (R = 4-CN)
2d14
2d(DFT)
H2O
CH3CN
CH2Cl2
−3.7
−0.9
16.9
−0.8
2.3
21.7
1.6
6.4
28.3
Bond distance
Co–Npy
2.031(18)
Co–Nimine
2.213(18)
2.222(18)
Co–Cl
Co(1)–N(2)
2.050
complexes react to form the ions [(NNN)2Co]2+ and [CoCl4]2−
that comprise 4d′. To investigate these pathways, we have calcu-
lated the Gibbs free energies and the results are shown in Fig. 6.
Replacement of one chloride ligand by a solvent moiety to form
5d was found to be thermodynamically disfavored, demanding
Gibbs free-energy changes of 6.1 and 18.9 kcal mol−1 for H2O
and CH3CN, respectively. Similar ‘uphill’ free-energy changes
are predicted for the substitution of both chloride ligands by a
single solvent moiety in 6d, or exchange of both chloride
ligands by water in 7d. These results suggest that none of the
ligand-exchange reactions are thermodynamically favored,
leaving disproportionation to 4d′ as the most likely reaction in
water. It should also be mentioned that ion pairing was not con-
sidered for this calculation i.e. the free energy of the ions
[(L)2CoA]2+ and [CoBCl4]2− were evaluated separately.
Co(1)–N(1)
Co(1)–N(3)
2.241
2.242
Co(1)–Cl(1)
Co(1)–Cl(2)
2.268(6)
2.261(6)
imine–Nimine
2.356
2.353
C
C(14)–N(2)
C(6)–N(3)
1.281(3)
1.299
1.300
1.281(3)
C
imine–Cpy
1.494(3)
1.496(3)
C(1)–C(6)
C(5)–C(14)
1.492
1.487
C
py–Npy
1.347(3)
1.335(3)
C(1)–N(1)
C(5)–N(1)
1.350
1.348
Bond angle
N
imine–Co–Nimine
N(1)–Co(1)–N(3)
150.88(7)
151.40
We have also investigated the effect of solvent polarity on the
disproportionation reaction shown in Fig. 6 (2d → 4d′) in an
effort to further investigate such a reaction mechanism. The cal-
culated free-energy changes are collected in Table 5 and indicate
that such a disproportionation is spontaneous in water ΔG =
−0.8 kcal mol−1 and very likely to be observed to some extent
even in acetonitrile ΔG = 2.3 kcal mol−1 (Keq = 0.13). However,
in non-polar or weakly polar solvents (e.g., CH2Cl2) the reaction
is predicted to be non-spontaneous, with a significant free-
energy penalty of 21.7 kcal mol−1. Furthermore, the CoII/III oxi-
dation potential calculated for [(NNN)2CoII]2+ in 4d′ in water
(+0.13 V vs. Cp2Fe0/+) is only ca. 0.1 V higher than that deter-
mined experimentally (+0.02 V vs. Cp2Fe0/+) suggesting that in
water 2d disproportionates prior to oxidation. In contrast, oxi-
dation in CH2Cl2 must follow the pathway where the 18e− CoIII
cation is stabilized by coordination of a CH2Cl2 moiety upon
CoII/III oxidation (i.e. 2d → 2d(solv)+). Finally, in polar organic
solvents (e.g., CH3CN) although oxidation to the solvated CoIII
species (2d → 2d(solv)+) is calculated to be thermodynamically
more favored, both pathways are likely accessible at equilibrium.
These results agree nicely with the experimental observation of
disproportionation of 2d to 4d′ in acetonitrile but not in dichlo-
romethane (vide supra).
In addition to the analysis of the effect of solvent polarity on
the underlying reaction mechanism, we have explored the effect
of ligand substituents on the resulting free-energy profiles. As
can be seen from the data in Table 5 a similar effect of solvent
polarity on the disproportionation is observed for 2b and 2f,
with the reaction in water being the most favorable, while dispro-
portionation in dichloromethane is least favorable. The effect of
ligand substituents on the reaction is also in very good agree-
ment with experimental observations as the data in Table 5 indi-
cates that disproportionation is favored for electron donating
substituents and disfavored for electron withdrawing substituents
relative to the unsubstituted complex 2d. The electron donating
N
py–Co–Nimine
N(2)–Co(1)–N(1)
N(2)–Co(1)–N(3)
75.29(7)
75.64(7)
75.22
75.26
N
py–Co–Cl
N(2)–Co(1)–Cl(1)
N(2)–Co(1)–Cl(2)
118.36(5)
123.78(5)
imine–Co–Cl
97.85(5)
108.35
112.54
N
N(1)–Co(1)–Cl(1)
N(1)–Co(1)–Cl(2)
N(3)–Co(1)–Cl(1)
N(3)–Co(1)–Cl(2)
94.83
90.50
95.92
99.66
95.86(5)
97.31(5)
98.80(5)
Cl–Co–Cl
117.84(3)
Cl(1)–Co(1)–Cl(2)
138.70
state containing a low spin 16e− CoIII ion (d6, S = 0) was deter-
mined to be the ground state electronic configuration for 2d+.
Fig. 6 shows that the direct oxidation of 2d without changing
the coordination sphere at cobalt (Fig. 6, 2d → 2d+) would
require a significantly higher potential (+0.90, +0.99 and +0.93
V vs. Cp2Fe0/+ for CH2Cl2, CH3CN and H2O, respectively) than
observed experimentally (+0.67, +0.11 and +0.02 V vs. Cp2Fe0/+
for CH2Cl2, CH3CN and H2O, respectively). This is likely
because direct oxidation, without changing the coordination
sphere of the metal center, produces a relatively unstable 16-elec-
tron complex. In contrast, coordination of a solvent molecule
(e.g., CH3CN or CH2Cl2) upon oxidation of 2d to form the 18e−
low spin octahedral CoIII-species 2d(solv)+ at potentials of +0.06
V (CH3CN) and +0.54 V (CH2Cl2) vs. Cp2Fe0/+ (bold in Fig. 6)
is in good agreement with the experimental data reported in
Table 1. In H2O, however, the comparison is less favorable and
suggests that neither of the two reaction pathways discussed so
far are the predominant mechanisms for oxidation of 2d in water.
To analyze alternative reaction pathways for oxidation in
water, we have explored the feasibility of Cl−/H2O ligand
exchange (Fig 6, 2d → 5d, 6d, 7d) as well as the possibility of
disproportionation (Fig 6, 2d → 4d) where two neutral 2d
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3562–3573 | 3569

View Online
Table 6 Crystallographic data and structural refinement parameters for
[(1f)2Co][CoCl4] and 2g
coordination sphere of the metal to accommodate a solvent mol-
ecule. However, complexes lacking bulky substituents in the
ortho positions of the N-aryl rings (2b–d, f and g) are in equili-
brium with [(NNN)2Co][CoCl4] resulting from disproportiona-
tion, with the equilibrium constant heavily influenced by solvent
polarity and ligand electronic effects. In acetonitrile, complexes
bearing electron donating or neutral substituents (2d, 2f and 2g)
form the bis-ligand complexes almost exclusively, while those
bearing electron withdrawing substituents (2b and 2c) have equi-
libria predominantly displaced towards the (NNN)CoCl2 species,
with only minor amounts (∼5%) of [(NNN)2Co][CoCl4]. In
agreement with our DFT results, oxidation of the latter two com-
plexes follows a different reaction pathway which likely involves
coordination of a solvent molecule as in 2d(solv)+. These results
are also in agreement with the CV data of 4b–g, mass spec-
trometry data for 2b–d, f and g, and the single-crystal X-ray
structure of [(2f)2Co][CoCl4] (obtained from crystallizing
[(1f)2Co][CoCl4] from CH3CN/Et2O). In dichloromethane, the
disproportionation is observed only for 2f and 2g complexes
with strong electron donating substituents, while 2b–d do not
show any evidence for disproportionation.
Therefore, we conclude that although increasing the solvent
polarity can reduce the oxidation potential by as much as
600 mV (+0.67 V, +0.11 and +0.02 V vs. CpFe0/+ for CH2Cl2,
CH3CN, H2O) the stability of the complexes (NNN)CoX2 is jeo-
pardized by polar solvents such as H2O or CH3CN. This ulti-
mately leads to disproportionation into coordinatively saturated
[(NNN)2Co][CoCl4] salts, which may not have the desired cata-
lytic properties. A safer way of modulating the redox potentials
is by moving to more electron-donating groups in the substitu-
ents of the pincer ligands (e.g., R = NMe2 as in 2g), increasing
the electron-density at the metal center, thus generating a more
easily oxidized CoII center, while electron-withdrawing groups
(e.g., R = CN as in 2b) produce the opposite effect. A caveat to
such a strategy, however, is that steric effects (or lack thereof)
imparted by ligand substituents may affect any potential catalytic
activity.
[(1f)2Co][CoCl4]
2g
Formula
Formula weight
Crystal system
Space group
Cell dimensions
a (Å)
b (Å)
c (Å)
α (°)
C48H49Cl4Co2N7O4 C25H29Cl2CoN5
1047.60
Monoclinic
P21/n
529.36
Monoclinic
P21/n
17.8237(15)
15.3342(14)
19.2414(16)
90
12.3191(4)
15.1565(5)
14.1316(5)
90
β (°)
113.367(3)
90
4827.6(7)
4
1.441
0.960
107.0280(10)
90
2522.90(15)
4
1.394
0.915
γ (°)
V (Å3)
Z
Dcalc (Mg m−3
)
μ (Mo Kα) (mm−1
)
Crystal dimensions (mm3)
0.30 × 0.30 × 0.03
0.50 × 0.20 ×
0.05
T (K)
θ range (°)
No. of rflns
173(2)
1.76 to 27.10
42 880
173(2)
1.93 to 28.28
24 507
No. of indep reflns
R1 (I > 2σ(I))
wR2 (all data)
GOF
10 559
0.0480
0.1372
1.094
6267
0.0375
0.1094
1.003
Largest diff. peak, hole
1.643, −0.836
1.252, −0.358
(eÅ−3
)
effect of the methoxy substituent is strong enough in 2f to make
disproportionation spontaneous in water and acetonitrile, while
the electron withdrawing effect of the cyano group is strong
enough to discourage disproportionation in dichloromethane and
acetonitrile.
These results are therefore consistent with oxidation of the
(NNN)CoX2 complexes in water following disproportionation to
[(NNN)2Co][CoCl4]. In contrast, 18e− solvated CoIII cations are
formed upon oxidation of the (NNN)CoX2 complexes in CH2Cl2
and coordination of a solvent moiety (e.g. formation of 2d
(solv)+ in Fig. 6). Finally, in polar organic solvents (e.g.,
CH3CN) both pathways are accessible via the equilibrium shown
in equation 1, although formation of the solvated CoIII species
[(NNN)CoX2(solv)]+ is predicted to be thermodynamically more
favored.
Experimental methods
General considerations
Acetonitrile and dichloromethane were dried by passage over
activated molecular sieves and degassed prior to use. Bu4NBF4
was recrystallized three times from ethanol and dried under
vacuum at 120 °C over P2O5 for 3 d. Elemental analyses were
performed by Galbraith Laboratories, Inc., in Knoxville, TN. All
1H and 13C{1H} NMR spectra were recorded on a Bruker
400 MHz Avance spectrometer and referenced to residual CHCl3
Conclusions
The electrochemical properties of a series of (NNN)CoX2 and
[(NNN)2Co][PF6]2 complexes have been investigated for ligands
with various electron donating/withdrawing substituents, in sol-
vents of low (CH2Cl2), high (CH3CN) and very high (H2O)
polarity. Experimental and theoretical results show clear evi-
dence of the regulation of CoII/III oxidation potentials using
ligand effects and/or solvent polarity, affecting not only the elec-
tronic properties but also the stability of the complexes. Introdu-
cing substituents in the para position of the N-aryl ring allows
one to vary the CoII/III oxidation potential of the [(NNN)2Co]
[PF6]2 complexes by up to 750 mV in acetonitrile.
1
(δ 7.27) or CHD2CN (δ 1.94) for H and CDCl3 (δ 77.16) for
13C{1H}. Spectra of the paramagnetic complexes can be found
in the ESI.† Solution magnetic moments were determined by
Evans NMR method11 using the residual solvent signal as the
reference and although these are generally the result of a single
experiment, in several cases the values obtained were checked
against multiple independent experiments and found to be in
agreement ( 0.1 μB). MALDI time-of-flight mass spectra were
acquired with an Applied Biosystems Voyager-DE STR mass
spectrometer utilizing a nitrogen laser (λ = 334 nm) to ionize a
Calculations show that oxidation of the (NNN)CoX2 com-
plexes in CH2Cl2 and CH3CN leads to expansion of the
3570 | Dalton Trans., 2012, 41, 3562–3573
This journal is © The Royal Society of Chemistry 2012

View Online
dried residue consisting of the analyte mixed with the HBM–
acetonitrile matrix (20 mg mL−1) from a polished stainless steel
plate (HBM = 4-hydroxybenzylidenemalononitrile). Electroche-
mical measurements were conducted on an IviumStat Electro-
the Cl− ion and the hydration of H2O, for which we have used
experimental results.19
X
hν
n
G ¼ Eelect þ Gsolv þ ZPVE þ
þ
kT
ehν=kT ꢁ 1
2
chemical Interface
& Impedance Analyzer from Ivium
ν
Technologies and were carried out under an inert Ar or N2
atmosphere using 0.1 M tetrabutylammonium tetrafluoroborate
as the supporting electrolyte in acetonitrile or dichloromethane.
Cyclic voltammetry experiments were performed in a beaker
type cell with a working volume of 5 mL using a standard three-
electrode cell with a glassy carbon disk (d = 3 mm) working
electrode, platinum wire counter electrode and a homemade
reference electrode consisting of Ag wire immersed in a 10 mM
AgNO3/0.1 M Bu4NBF4 electrolyte in acetonitrile separated
from the bulk solution by a “thirsty” VycorTM frit.
Single crystals suitable for X-ray diffraction were transferred
from a crystallization vessel into a drop of viscose organic oil,
transferred to a nylon loop and mounted on a Bruker X8 APEX
II diffractometer (Mo Kα radiation) and cooled to −100 °C. Data
collection and reduction were done using Bruker APEX2 and
SAINT + software packages and corrected for absorption using
SADABS. Structures were solved by direct methods and refined
on F2 by full matrix least-squares techniques using SHELXTL
software package. All non-hydrogen atoms were refined anisotro-
pically. Hydrogen atoms were found in a difference Fourier map
and refined isotropically (Table 6).
ꢁ TðSvib þ Srot þ Strans
Þ
ð4Þ
where n = 8 accounts for the internal energies of the translational
and rotational modes and the PV term. Calculated redox poten-
tials are reported relative to Cp2Fe0/+, calculated at the same
level of theory, to eliminate systematic differences due to the
nature of the solvent, electrolyte and working electrode
conditions.
All quantum chemistry calculations were carried out within
the framework of density functional theory, using the hybrid
density functional B3LYP20 as implemented in the Jaguar 7.7
software package.21 For geometry optimization and vibration cal-
culations, we have employed a hybrid basis set where Co ions
were described by LACVP effective core potential22 and basis
set (ECP), using the double-ζ contraction of valence functions,
while the other elements (Cl, C, N and H) were treated in the
level of 6-31G basis set.23 Single-point calculations were also
performed with a larger basis set cc-PVTZ(-f).24
Solvation energies were calculated using the Poisson–Boltz-
mann self-consistent reaction field method (PBF)25 to represent
the solvents with dielectric constant (effective radius) equal to
8.93 (2.33), 37.5 (2.19) and 80.37 (1.40) for dichloromethane
(CH2Cl2), acetonitrile (CH3CN) and water, respectively. Here,
we have employed B3LYP/LACVP_cc-PVTZ(-f) theory level.
Theoretical methods
The computational methods implemented in this study have been
described previously.18 Here, we outline the methodology only
briefly.
Synthetic procedures
Ligands 1a,3d,i 1c,13 1d,13,26 1e,27 and 1f26 and complexes 2a,3d,i
2c,13 2d,13–14,26 2f,26 3a3b and 4f7e were prepared as reported
previously while new ligands and complexes were prepared
using modifications of these published procedures, the details of
which are described below.
The standard reduction potentials are obtained, as follows:
ΔGðsolvÞðRedÞ
EW ¼ ꢁ
ð2Þ
nF
where F and n are the Faraday constant (23.06 kcal mol−1 V−1
)
and the number of electrons involved in the redox reaction,
respectively, by evaluating the reduction free energies ΔG(red)
using the Born–Haber thermodynamic cycle (illustrated in
Scheme 1), as follows:
2,6-Bis[1-(4-cyanophenylimino)ethyl]pyridine, (1b)
2,6-Diacetylpyridine (2.00 g, 12.26 mmol, 1.0 equiv.), p-amino-
benzonitrile (3.04 g, 25.75 mmol, 2.1 equiv.) and p-toluenesulfo-
nic acid (0.010 g, 0.053 mmol, 0.05 equiv.) were dissolved in
70 mL of toluene and heated to reflux under N2 with constant
stirring and removal of water using a Dean–Stark trap. After 2 d,
the reaction mixture was cooled to room temperature and the
volatiles were removed by evaporation to give an orange-red
residue. The addition of methanol and filtration of the resultant
pale yellow solid, followed by washing with ethyl ether and pet-
roleum ether gave 1.20 g (3.31 mmol, 27%) of a pale yellow
powder. 1H NMR (CDCl3, 400.13 MHz, 24.4 °C) δ 8.36 (d, 2H,
ΔGðsolvÞðRedÞ ¼ ΔGðgÞðRedÞ þ ΔGðsÞðRedÞ ꢁ ΔGðsÞðOxÞ ð3Þ
with ΔG(g)(Red) = ΔU(g)(Red) + PV − TΔS(g)(Red). The
vibrational, rotational and translational contributions (statistical
mechanics contributions) were included in the calculation of
internal energy and entropy. The solvation free energies of the
oxidized and reduced species, ΔG(s)(Ox) and ΔG(s)(Red), were
calculated using the PBF method as described below, except for
3
3
py-Hmeta, JHH = 7.8 Hz), 7.93 (t, 1H, py-Hpara, JHH = 7.8 Hz),
3
7.69 (d, 4H, N-Ar-Hmeta, JHH = 8.5 Hz), 6.92 (d, 4H, N-Ar-
Hortho, 3JHH = 8.5 Hz), 2.41 (s, 6H, C(Me) = N). 13C{1H} NMR
(CDCl3, 100.6 MHz, 22.0 °C) δ 167.9, 155.3, 154.7, 137.2,
133.4, 123.1, 119.8, 119.2, 107.1, 16.6. MALDI MS, m/z: 364
[1b + H]+.
Scheme 1 Born–Haber thermodynamic cycle.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3562–3573 | 3571

View Online
2,6-Bis[1-(4-dimethylaminophenylimino)ethyl]pyridine, (1g)
to give 0.382 g (0.35 mmol, 100%) of 4b. Elemental analysis
calcd for C46H34CoF12N10P2: C, 51.36; H, 3.19; N, 13.02.
Found: C, 50.90; H, 3.34; N, 12.55.28 MALDI MS, m/z: 785
[4b – 2 PF6]+. Magnetic susceptibility (acetonitrile-d3, 297 K):
μeff = 3.5 μB.
2,6-Diacetylpyridine (2.00 g, 12.26 mmol, 1.0 equiv.), N,N-
dimethyl-1,4-phenylene diamine (3.51 g, 25.75 mmol, 2.1
equiv.) and p-toluenesulfonic acid (0.010 g, 0.053 mmol, 0.05
equiv.) were dissolved in 70 mL of toluene and heated to reflux
under N2 with constant stirring and removal of water using a
Dean–Stark trap. After 1 d, the reaction mixture was cooled to
room temperature and the volatiles were removed by evaporation
to give an orange-red residue. Addition of methanol and filtration
of the resultant dark yellow solid, followed by washing with
ethyl ether and petroleum ether gave 2.88 g (7.22 mmol, 59%)
of a dark yellow powder. 1H NMR (CDCl3, 400.13 MHz,
Bis{2,6-Bis[1-(4-trifluoromethylphenylimino)ethyl]pyridine}
cobalt(II) hexafluorophosphate, 4c
2c (0.30 g, 0.67 mmol, 2.04 equiv.) and CoCl2 (0.042 g,
0.33 mmol, 1.00 equiv.) were dissolved in 25 mL of methanol
and the mixture was stirred at room temperature for 30 minutes
before excess NH4PF6 (0.81 g, 4.99 mmol, 15.3 equiv.) was
added. The solution was dried under reduced pressure and the
resulting red residue was taken up in dichloromethane, washed
with water and the organic phase was dried with Na2SO4 and
filtered. The filtrate was dried under reduced pressure and 0.38 g
(0.30 mmol, 92%) of a red powder was obtained after trituration
from a 1 : 1 diethyl ether : hexane mixture. Elemental analysis
calc. for C46H34CoF24N6P2: C, 44.28; H, 2.75; N, 6.74. Found:
C, 44.13; H, 2.84; N, 6.39. MALDI MS, m/z: 957 [4c – 2 PF6]+,
1104 [4c – PF6]+. Magnetic susceptibility (acetonitrile-d3,
297 K): μeff = 2.7 μB.
3
24.4 °C) δ 8.32 (d, 2H, py-Hmeta, JHH = 7.8 Hz), 7.84 (t, 1H,
3
3
py-Hpara, JHH = 7.8 Hz), 6.87 (d, 4H, N-Ar-Hmeta, JHH = 9.1
3
Hz), 6.81 (d, 4H, N-Ar-Hortho, JHH = 9.1 Hz), 2.98 (s, 12H, N
(Me2)), 2.49 (s, 6H, C(Me) = N). 13C{1H} NMR (CDCl3,
100.6 MHz, 22.0 °C) δ 166.5, 156.0, 147.7, 141.0, 136.6, 121.8,
121.4, 113.2, 41.1, 16.3. MALDI MS, m/z: 400 [1g + H]+.
2,6-Bis[1-(4-cyanophenylimino)ethyl]pyridine cobalt(II) chloride,
(2b)
1b (0.75 g, 2.07 mmol, 1.001 equiv.) and CoCl2 (0.27 g,
2.06 mmol, 1.0 equiv.) were combined in 65 mL of THF and
stirred under N2 overnight at room temperature. The reaction
mixture was then poured into 125 mL of ethyl ether and the
resulting precipitate was isolated by vacuum filtration, washed
with ethyl ether and petroleum ether and air dried to give 1.02 g
(2.06 mmol, 99%) of a green powder. Elemental analysis calcd
for C23H17Cl2CoN5: C, 56.00; H, 3.47; N, 14.20. Found: C,
55.19; H, 3.49; N, 13.66.28 MALDI MS, m/z: 457 [2b – Cl]+.
Magnetic susceptibility (acetonitrile-d3, 297 K): μeff = 5.1 μB.
Bis{2,6-Bis[1-(phenylimino)ethyl]pyridine}cobalt(II)
hexafluorophosphate, 4d
1d (0.14 g, 0.45 mmol, 2.3 equiv.) and CoCl2 (0.025 g,
0.19 mmol, 1.00 equiv.) were dissolved in 20 mL of methanol
and the mixture was stirred at room temperature for 30 minutes
before excess NH4PF6 (0.52 g, 3.17 mmol, 16.5 equiv.) was
added to precipitate the product. The brown crystalline solid was
collected by filtration, washed with ice cold methanol and air
dried to give 0.143 g (0.15 mmol, 76%) of 4d. Elemental analy-
sis calcd for C42H38CoF12N6P2: C, 51.70; H, 3.93; N, 8.61.
Found: C, 51.99; H, 3.98; N, 8.66. MALDI MS, m/z: 685
[4d – 2 PF6]+. Magnetic susceptibility (acetonitrile-d3, 297 K):
μeff = 2.6 μB.
2,6-Bis[1-(4-dimethylaminophenylimino)ethyl]pyridine cobalt(II)
chloride, (2g)
1g (1.00 g, 2.50 mmol, 1.0 equiv.) and CoCl2 (0.33 g,
2.50 mmol, 1.0 equiv.) were combined in 70 mL of THF and
stirred under N2 overnight at room temperature. The reaction
mixture was then poured into 150 mL of ethyl ether and the
resulting precipitate was isolated by vacuum filtration, washed
with ethyl ether and petroleum ether and air dried to give 1.18 g
(2.23 mmol, 89%) of a brown powder. Elemental analysis calcd
for C25H29Cl2CoN5: C, 56.72; H, 5.52; N, 13.23. Found: C,
57.00; H, 5.63; N, 13.01. MALDI MS, m/z: 493 [2g – Cl]+, 857
[(1g)2Co]+. Magnetic susceptibility (acetonitrile-d3, 297 K):
μeff = 3.5 μB.
Bis{2,6-Bis[1-(4-methylphenylimino)ethyl]pyridine}cobalt(II)
hexafluorophosphate, 4e
1e (0.30 g, 0.88 mmol, 2.24 equiv.) and CoCl2 (0.051 g,
0.39 mmol, 1.00 equiv.) were dissolved in 20 mL of methanol
and the mixture was stirred at room temperature for 30 minutes
before excess NH4PF6 (1.00 g, 6.17 mmol, 15.7 equiv.) was
added to precipitate the product. The brick-red solid was col-
lected by filtration, washed with ice cold methanol and air dried
to give 0.258 g (0.25 mmol, 64%) of 4e. Elemental analysis
calcd for C46H46CoF12N6P2: C, 53.55; H, 4.49; N, 8.15. Found:
C, 53.04; H, 4.50; N, 8.02. MALDI MS, m/z: 741 [4e – 2 PF6]+.
Magnetic susceptibility (acetonitrile-d3, 297 K): μeff = 2.7 μB.
Bis{2,6-Bis[1-(4-cyanophenylimino)ethyl]pyridine}cobalt(II)
hexafluorophosphate, 4b
1b (0.26 g, 0.71 mmol, 2.01 equiv.) and CoCl2 (0.046 g,
0.35 mmol, 1.00 equiv.) were dissolved in 20 mL of methanol
and the mixture was stirred at room temperature for 30 minutes
before excess NH4PF6 (0.78 g, 4.82 mmol, 13.6 equiv.) was
added to precipitate the product. The brick-red solid was col-
lected by filtration, washed with ice cold methanol and air dried
Bis{2,6-Bis[1-(4-dimethylaminophenylimino)ethyl]pyridine}
cobalt(II) hexafluorophosphate, 4g
1g (0.39 g, 0.98 mmol, 2.02 equiv.) and CoCl2 (0.063 g,
0.48 mmol, 1.00 equiv.) were dissolved in 20 mL of methanol
3572 | Dalton Trans., 2012, 41, 3562–3573
This journal is © The Royal Society of Chemistry 2012

View Online
24, 6298; (e) A. M. Tondreau, C. Milsmann, A. D. Patrick, H. M. Hoyt,
E. Lobkovsky, K. Wieghardt and P. J. Chirik, J. Am. Chem. Soc., 2011,
132, 15046–15059.
and the mixture was stirred at room temperature for 30 minutes
before excess NH4PF6 (1.14 g, 7.03 mmol, 14.5 equiv.) was
added to precipitate the product. The brick-red solid was col-
lected by filtration, washed with ice cold methanol and air dried
to give 0.346 g (0.30 mmol, 62%) of 4g. Elemental analysis
calcd for C50H58CoF12N10P2: C, 52.31; H, 5.09; N, 12.20.
Found: C, 51.31; H, 5.14; N, 11.98.28 MALDI MS, m/z: 857
[4g ¬ 2 PF6]+, 1002 [4g ¬ PF6]. Magnetic susceptibility (aceto-
nitrile-d3, 297 K): μeff = 2.1(4) μB.
7 (a) S. C. Bart, K. Chlopek, E. Bill, M. W. Bouwkamp, E. Lobkovsky,
F. Neese, K. Wieghardt and P. J. Chirik, J. Am. Chem. Soc., 2006, 128,
13901–13912; (b) A. C. Bowman, C. Milsmann, C. C. H. Atienza,
E. Lobkovsky, K. Wieghardt and P. J. Chirik, J. Am. Chem. Soc., 2010,
132, 1676–1684; (c) A. C. Bowman, C. Milsmann, E. Bill,
E. Lobkovsky, T. Weyhermuller, K. Wieghardt and P. J. Chirik, Inorg.
Chem., 2010, 49, 6110–6123; (d) P. M. Budzelaar, B. de Bruin,
A. W. Gal, K. Wieghardt and J. L. van Lenthe, Inorg. Chem., 2001, 40,
4649–4655; (e) B. de Bruin, E. Bill, E. Bothe, T. Weyhermuller and
K. Wieghardt, Inorg. Chem., 2000, 39, 2936–2947.
8 C. C. H. Atienza, A. C. Bowman, E. Lobkovsky and P. J. Chirik, J. Am.
Chem. Soc., 2010, 132, 16343–16345.
Acknowledgements
9 (a) H. Nakazawa and M. Itazaki, in Topics in Organometallic Chemistry:
Iron Catalysis Fundamentals and Applications, ed. B. Plietker,
Springer, New York, NY, Editon edn, 2011, vol. 33, pp. 27–81;
(b) R. J. Trovitch, E. Lobkovsky, E. Bill and P. J. Chirik, Organometal-
lics, 2008, 27, 1470–1478; (c) R. J. Trovitch, E. Lobkovsky and P.
J. Chirik, J. Am. Chem. Soc., 2008, 130, 11631–11640; (d) S. C. Bart,
E. Lobkovsky and P. J. Chirik, J. Am. Chem. Soc., 2004, 126, 13794–
13807.
This material is based upon work supported as part of the Center
for Electrocatalysis, Transport Phenomena, and Materials
(CETM) for Innovative Energy Storage, an Energy Frontier
Research Center funded by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences under Award
Number DE-SC00001055.
10 G. J. P. Britovsek, J. England, S. K. Spitzmesser, A. J. P. White and D.
J. Williams, J. Chem. Soc., Dalton Trans., 2005, 945–955.
11 (a) D. F. Evans, J. Chem. Soc., 1959, 2003–2005; (b) S. K. Sur, J. Magn.
Reson., 1989, 82, 169–173.
References
12 R. S. Drago, Physical Methods for Chemists, Surfside Scientific Publish-
ers, Gainesville, FL, 2nd edn, 1992.
13 D. Gong, B. Wang, H. Cai, X. Zhang and L. Jiang, J. Organomet. Chem.,
2011, 696, 1584–1590.
14 D. Gong, B. Wang, C. Bai, J. Bi, F. Wang, W. Dong, X. Zhang and
L. Jiang, Polymer, 2009, 50, 6259–6264.
15 W. M. Reiff, N. E. Erickson and W. A. Baker Jr, Inorg. Chem., 1969, 8,
2019–2021.
1 G. Jerkiewicz, Electrocatalysis, 2010, 1, 1.
2 (a) R. H. Crabtree, Energy Environ. Sci., 2008, 1, 134–138;
(b) G. L. Soloveichik, J. P. Lemmon, J. -C. Zhao (General Electric
Company), USA Pat., 2008/0248339, 2008; (c) N. Kariya, A. Fukuoka
and M. Ichikawa, Chem. Commun., 2003, 690–691; (d) N. Kariya,
A. Fukuoka and M. Ichikawa, Phys. Chem. Chem. Phys., 2006, 8, 1724–
1730; (e) D. L. DuBois and R. M. Bullock, Eur. J. Inorg. Chem., 2011,
1017–1027; (f) A. D. Wilson, R. H. Newell, M. J. McNevin, J.
T. Muckerman, M. R. DuBois and D. L. duBois, J. Am. Chem. Soc.,
2006, 128, 358–366; (g) B. R. Galan, J. Schoffel, J. C. Linehan, C. Seu,
A. M. Appel, J. A. S. Roberts, M. L. Helm, U. J. Kilgore, J. Y. Yang,
D. L. DuBois and C. P. Kubiak, J. Am. Chem. Soc., 2011, 133, 12767–
12779.
3 (a) C. C. H. Atienza, C. Milsmann, E. Lobkovsky and P. J. Chirik,
Angew. Chem., Int. Ed., 2011, 50, 8143–8147; (b) A. M. A. Bennett
(DuPont), WO Pat., 98/27124, 1998; (c) C. Bianchini, G. Giambastiani,
I. G. Rios, G. Mantovani, A. Meli and A. M. Segarra, Coord. Chem.
Rev., 2006, 250, 1391–1418; (d) G. J. P. Britovsek, M. Bruce,
V. C. Gibson, B. S. Kimberley, P. J. Maddox, S. Mastroianni, S.
J. McTavish, C. Redshaw, G. A. Sloan, S. Stromberg, A. J. P. White and
D. J. Williams, J. Am. Chem. Soc., 1999, 121, 8728–8740;
(e) G. J. P. Britovsek, V. C. Gibson, B. S. Kimberley, P. J. Maddox, S.
J. McTavish, G. A. Solan, A. J. P. White and D. J. Williams, Chem.
Commun., 1998, 849–850; (f) G. J. P. Britovsek, V. C. Gibson, S.
K. Spitzmesser, K. P. Tellmann, A. J. P. White and D. J. Williams,
J. Chem. Soc., Dalton Trans., 2002, 1159–1171; (g) V. C. Gibson,
C. Redshaw and G. A. Solan, Chem. Rev., 2007, 107, 1745–1776; (h) B.
L. Small and M. Brookhart, Macromolecules, 1999, 32, 2120–2130;
(i) B. L. Small, M. Brookhart and A. M. A. Bennett, J. Am. Chem. Soc.,
1998, 120, 4049–4050.
16 Y. Chen, R. Chen, C. Qian, X. Dong and J. Sun, Organometallics, 2003,
22, 4312–4321.
17 (a) D. H. McDaniel and H. C. Brown, J. Org. Chem., 1958, 23, 420–427;
(b) W. G. Herkstroeter, J. Am. Chem. Soc., 1973, 95, 8686–8691.
18 (a) T. Wang, G. Brudvig and V. S. Batista, J. Chem. Theory Comput.,
2010, 6, 755–760; (b) T. Wang, G. Brudvig and V. S. Batista, J. Chem.
Theory Comput., 2010, 6, 2395–2401.
19 (a) C. P. Kelly, C. J. Cramer and D. G. Truhlar, J. Phys. Chem. B, 2007,
111, 408–422; (b) J. Sefcik and W. A. Goddard III, Geochim. Cosmo-
chim. Acta, 2001, 65, 4435–4443.
20 (a) A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098–
3100; (b) C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785–
789; (c) A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
21 Jaguar, Schrodinger, LLC, New York, NY, Editon edn., 2010.
22 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 299–310.
23 (a) W. J. Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys., 1972, 56,
2257–2261; (b) M. M. Francl, W. J. Pietro, W. J. Hehre, J. S. Binkley,
M. S. Gordon, D. J. DeFrees and J. A. Pople, J. Chem. Phys., 1982, 77,
3654–3665.
24 (a) T. H. Dunning Jr, J. Chem. Phys., 1989, 90, 1007–1023;
(b) R. A. Kendall, T. H. Dunning Jr and R. J. Harrison, J. Chem. Phys.,
1992, 96, 6796–6806; (c) D. E. Woon and T. H. Dunning Jr, J. Chem.
Phys., 1994, 100, 2975–2988.
4 (a) B. L. Small, Organometallics, 2003, 22, 3178–3183;
(b) K. P. Tellmann, V. C. Gibson, A. J. P. White and D. J. Williams, Orga-
nometallics, 2005, 24, 280–286.
5 (a) B. L. Small and M. Brookhart, J. Am. Chem. Soc., 1998, 120, 7143–
7144; (b) B. L. Small, R. Rios, E. R. Fernandez and M. J. Carney, Orga-
nometallics, 2007, 26, 1744–1749.
25 (a) D. J. Tannor, B. Marten, R. Murphy, R. A. Friesner, D. Sitkoff,
A. Nicholls, M. Ringnalda, W. A. Goddard III and B. Honig, J. Am.
Chem. Soc., 1994, 116, 11875–11882; (b) B. Marten, K. Kim, C. Cortis,
R. A. Friesner, R. B. Murphy, M. N. Ringnalda, D. Sitkoff and B. Honig,
J. Phys. Chem., 1996, 100, 11775–11788.
26 D. A. Edwards, S. D. Edwards, W. R. Martin, T. J. Pringle and
P. Thornton, Polyhedron, 1992, 11, 1569–1573.
27 J. Granifo, S. J. Bird, K. G. Orrell, A. G. Osborne and V. Sik, Inorg.
Chim. Acta, 1999, 295, 56–63.
28 Despite repeated attempts to obtain better elemental analysis data, the low
oxidation potential of this compound resulted in lower than expected C
and N values.
6 (a) M. W. Bouwkamp, S. C. Bart, E. J. Hawrelak, R. J. Trovitch,
E. Lobkovsky and P. J. Chirik, Chem. Commun., 2005, 3406–3408;
(b) M. J. Humphries, K. P. Tellmann, V. C. Gibson, A. J. P. White and D.
J. Williams, Organometallics, 2005, 24, 2039–2050; (c) T. M. Kooistra,
Q. Knijnenburg, J. M. M. Smits, A. D. Horton, P. M. Budzelaar and A.
W. Gal, Angew. Chem., Int. Ed., 2001, 40, 4719–4722; (d) J. Scott,
S. Gambarotta, I. Korobkov and P. M. Budzelaar, Organometallics, 2005,
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3562–3573 | 3573