Bis phenylene flattened 13-membered tetraamide macrocyclic ligand (TAML) for square planar cobalt(III)*

Article abstract of DOI:10.1080/00958972.2018.1487060

The preparation, characterization, and evaluation of a cobalt(III) complex (Figure presented.) with 13-membered tetraamide macrocyclic ligand (TAML) is described. This is a square-planar (X-ray) S?=?1 paramagnetic (¹H NMR) compound, which becomes an S?=?0 diamagnetic octahedral species in excess d₅-pyridine. Its one-electron oxidation at an electrode is fully reversible with the lowest E_? value (0.66?V vs SCE) among all investigated Co^III TAML complexes. The oxidation results in a neutral blue species which is consistent with a Co^III/radical-cation ligand. The ease of oxidation is likely due to the two benzene rings incorporated in the ligand structure (whereas there is just one in many other Co^III TAMLs). The oxidized neutral species are unexpectedly EPR silent, presumably due to the π-stacking aggregation. However, they display eight-line hyperfine patterns in the presence of excess of 4-tert-butylpyridine or 4-tert-butyl isonitrile. The EPR spectra are more consistent with the Co^III/radical-cation ligand formulation rather than with a Co^IV complex. Attempts to synthesize a similar vanadium complex under the same conditions as for cobalt using [V^VO(OCHMe₂)₃] were not successful. TAML-free decavanadate was isolated instead.

Full text of DOI:10.1080/00958972.2018.1487060

Journal of Coordination Chemistry
ISSN: 0095-8972 (Print) 1029-0389 (Online) Journal homepage: http://www.tandfonline.com/loi/gcoo20
Bis phenylene flattened 13-membered tetraamide
macrocyclic ligand (TAML) for square planar
cobalt(III)
W. Chadwick Ellis, Alexander D. Ryabov, Andreas Fischer, Joshua A. Hayden,
Longzhu Q. Shen, Emile L. Bominaar, Michael P. Hendrich & Terrence J.
Collins
To cite this article: W. Chadwick Ellis, Alexander D. Ryabov, Andreas Fischer, Joshua A. Hayden,
Longzhu Q. Shen, Emile L. Bominaar, Michael P. Hendrich & Terrence J. Collins (2018): Bis
phenylene flattened 13-membered tetraamide macrocyclic ligand (TAML) for square planar
cobalt(III), Journal of Coordination Chemistry, DOI: 10.1080/00958972.2018.1487060
To link to this article: https://doi.org/10.1080/00958972.2018.1487060
Accepted author version posted online: 11
Jun 2018.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=gcoo20

Publisher: Taylor & Francis
Journal: Journal of Coordination Chemistry
DOI: http://doi.org/10.1080/00958972.2018.1487060
Bis phenylene flattened 13-membered tetraamide macrocyclic ligand (TAML)
1
for square planar cobalt(III)
2
W. CHADWICK ELLIS†, ALEXANDER D. RYABOV*†, ANDREAS FISCHER‡, JOSHUA A. HAYDEN† ,
3
LONGZHU Q. SHEN† , EMILE L. BOMINAAR†, MICHAEL P. HENDRICH† and TERRENCE J. COLLINS†
†
‡
Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, USA
Inorganic Chemistry, Department of Chemistry, Royal Institute of Technology, S-100 44 Stockholm, Sweden
The preparation, characterization and evaluation of a cobalt(III) complex
-
[
Co{(OC) (o,o'-NC H NCO) CMe }(OH )] with 13-membered tetraamide macrocyclic
2
6
4
2
2
2
ligand
1
(
TAML) is described. This is a square-planar (X-ray) S = 1 paramagnetic ( H NMR) compound,
which becomes an S = 0 diamagnetic octahedral species in excess d -pyridine. Its one-electron
5
oxidation at an electrode is fully reversible with the lowest E value (0.66 V vs SCE) among all
½
III
investigated Co TAML complexes. The oxidation results in a neutral blue species which is
III
consistent with a Co /radical-cation ligand. The ease of oxidation is likely due to the two
benzene rings incorporated in the ligand structure (whereas there is just one in many other CoIII
TAMLs). The oxidized neutral species are unexpectedly EPR silent, presumably due to the π-
stacking aggregation. However, they display eight-line hyperfine patterns in the presence of
excess of 4-tert-butylpyridine or 4-tert-butyl isonitrile. The EPR spectra are more consistent with
III
IV
the Co /radical-cation ligand formulation rather than with a Co complex. Attempts to
synthesize a similar vanadium complex under the same conditions as for cobalt using
V
[
V O(OCHMe ) ] were not successful. TAML-free decavanadate was isolated instead.
2
3
Keywords: Macrocyclic complex; TAML; Cobalt; Radical cation; EPR; DFT
*
1
Corresponding author. Email: ryabov@andrew.cmu.edu
Dedicated to Professor Dan Meyerstein on occasion of his 80 birthday, who still enjoys chemistry and keeps his
th
smiling sense of humor!
2
Present address: Laboratory Medicine, 525 East 68th Street, Suite F-715, New York, NY 10065, USA
Present address: Department of Zoology, University of Cambridge, Downing Street, Cambridge CB2 3EJ, UK
3
1

1
. Introduction
Tetraamide macrocycle ligand (TAML) system offers a unique environment for transition metal
ions and, hence, uncommon polyhedra, oxidation states and coordination numbers of metal
complexes [1-3]. Particularly challenging is cobalt chemistry, when the metal in cobalt TAMLs
(chart 1) finds itself in the inner cavity of tetra deprotonated amide nitrogens. The oxidation state
of cobalt in these square planar complexes is three, i.e. both axial coordination sites are vacant.
Interestingly, their square planar configuration does not change after 1e oxidation of the
III
IV
cobalt(III) species. Square planar Co and Co complexes are less common than octahedral [4].
III
Oxidized Co TAML species are uncharged and therefore soluble in non-polar organic solvents.
The first square planar bis(n-propylbiuretato)cobalatate(III) complex was crystallographically
III
characterized by Bour et al. [5]; Co was surrounded by four deprotonated amide nitrogens.
III
Collins and Uffelman described the first macrocyclic square planar Co complex 1 with similar
donor centers, which is shown in chart 1 [6]. Similar coordinative polyhedra were later
confirmed crystallographically for 2 [7, 8], cyclohexyl variant of 2 [9], 3 [10, 11], 5 [12] and 6
[13]. No X-ray data were however provided for 4, the polyhedron of which is strongly distorted
due to the incorporation of the 2,2'-bipyridine unit in the macrocycle [14].
_
_
_
_
O
N
O
N
O
N
O
O
O
O
O
X
N
N
N
N
N
R
R
O
Co
Co
CoIII
CoIII
N
N
N
N
N
N
N
N
O
O
O
O
O
O
O
O
Co-TAMLs
1
2 (X = CH), 3 (X = N); R = Me (a), Et (b)
_
_
_
O
O
N
O
N
_
O
O
O
O
O
N
N
N
N
N
N
N
Co
Co
N
Co
Co
M
+
N
N
N
N
N
N
N
O
N
N
O
O
O
O
O
O
O
4
5
6
7
; M = Li (a), PPh (b)
4
III
Chart 1. Reported examples of Co TAMLs 1-6 and new complex 7 described in this work.
III
The square-planar geometry of both parent Co and oxidized forms of cobalt TAML
complexes makes them attractive as catalysts due to both the open axial coordination sites and
the zero charge of the oxidized species. Cobalt TAMLs have been used in the synthesis of
2

various cyclic carbonates from epoxides and CO with up to 100% yields and turnover frequency
2
-1
(
TOF) of 0.1 s [8], in oxygen atom transfer reactions [15] and in the oxidation of water [12, 13].
The reactivity of TAML complexes is controlled by the nature of their ligands [16].4
Understandably, new cobalt TAML complexes are potentially viable and herein we describe the
synthesis, characterization and some properties of the new cobalt(III) TAML complex 7 which
are compared with those of complexes 1-6. Previously, iron(III) complexes with similar ligands
have been reported [18, 19]. They belong to the fourth generation of iron TAML activators of
peroxides [20] and are very reactive, in fact the most reactive, in catalytic oxidations by H O in
2
2
the aqueous medium [16]. This noteworthy reactivity of the iron counterparts suggested that it
might be worthwhile to revisit cobalt-centered TAMLs by the example of 7. Though the
synthesis of the ligand of 7 and their ring-substituted derivatives was reported earlier [19], we
describe additional useful information regarding the preparation of this ligand system.
2
2
. Experimental
.1. Materials and methods
Syntheses were performed under argon with oven dried glassware using Schlenk techniques.
Tetrahydrofuran and pyridine were dried and distilled under argon prior to use according to
standard laboratory techniques [21]. Other chemicals used were purchased from Sigma Aldrich
1
Chemical or Fisher Scientific and used as received unless stated otherwise. H NMR spectra
were recorded on a Bruker Avance™ 300 spectrometer. Chemical shits in d -dmso () are
6
referenced to the residual solvent peak at  2.5. Mass spectra (ESI) were obtained using a
Thermo-Fisher LCQ ESI/APCI Ion Trap instrument with methanol as a matrix. Elemental
analyses were carried out by Midwest Microlabs. Electrochemical measurements were
performed on a PC-interfaced potentiostat-galvanostat AUTOLAB PGSTAT 12. A three-
electrode setup was used with a glassy carbon working electrode, SCE reference electrode, and
auxiliary Pt electrode. Before each measurement, the working electrode was polished with a
diamond paste and rinsed with acetone and distilled water.
4
Dan Meyerstein was among pioneers who paid special attention to ligand effects on reactivity of metal complexes.
More than fifty years ago he reported how ligands influence a speed of redox reactions of cobalt(III) complexes
[17].
3

2
.2. EPR Data collection and simulation procedures
X-band (9 GHz) EPR spectra were recorded on a Bruker ESP 300 spectrophotometer with a
Bruker bimodal cavity. The modulation amplitude and frequency were 0.5 mT at 100 kHz for all
low temperature spectra. Room temperature measurements were done at 300 K. An Oxford ESR
9
10 cryostat was used for all low temperature measurements and the temperature was calibrated
with a carbon glass resistor (Lakeshore CGR-1-1000). A CuEDTA spin standard was used to
account for all relevant intensity factors. Simulations of EPR data were performed using the
SpinCount software [22].
2
.3. Density functional computations
Optimized geometries for cobalt TAML complexes 7, 13 and 14a,b were obtained using Becke’s
three-parameter hybrid functional (B3) [23, 24] along with the Lee−Yang−Parr correlation
functional (LYP) [25] and the 6-311G basis set. Spin state ½ and zero charge were assigned to
1
7
3 and 14a,b. Spin state 1 and charge ‒1 were assigned to 7. To obtain the simulated structure of
+
b, the cation PPh was first placed manually in the vicinity of anion 7 with a reference to the
4
X-ray structural data for 7b. The constructed complex was then subject to geometry optimization
at the DFT level (M06 density functional [26] and triple-ζ basis set 6-311G in the absence of
solvent using the Gaussian 09 computational package [27]. A stable structure was obtained upon
convergence to the default criteria set in Gaussian. The overlaid X-ray and DFT structures
revealed the root mean square difference of 0.36 Å, suggesting minute variation between them.
2
.4. Metalation of 10 by cobalt(II) chloride to form 7a
Tetraamide 10 [19] (100 mg, 366 mmol) was mixed with dry THF (50 mL) in a round bottom
flask under argon followed by addition of 4.4 equivalents of lithium bis(trimethylsilyl)amide
(1.22 mL of 1 M solution in THF). The mixture was stirred for ca. 5 min to obtain a
homogeneous solution. Then CoCl (43.3 mg, 334 mmol) was added and the reaction mixture
2
slowly turned dark blue to purple. After stirring overnight, the mixture was opened to air for
1
5 min. The solid formed was placed on a short silica gel column. The product turned yellow on
the silica. Elution by acetone yielded a dark blue to purple solution. The solvent was evaporated
to dryness to yield a solid of the same color in 25% yield. An additional portion of 7a, though of
4

lower purity, could be precipitated by pouring the THF filtrate (50 mL) into n-hexane (250 mL)
and recovered by filtration. ESI-MS negative mode (421 m/z).
2
.5. Metathesis of 7a to 7b
Complex 7a (5 mg, 0.01 mmol) was dissolved in a minimum amount of water (ca. 2 mL) to yield
a yellowish green solution, which was filtered through a 0.25 μm syringe filter. Then PPh Cl
4
(
40 mg, 0.1 mmol) dissolved in 2 mL H O was added dropwise to yield a purple precipitate. This
2
was collected by filtration, rinsed with water, and recrystallized from ethanol/heptane as small
1
pentagonal X-ray quality crystals. Mp 209-211 C. H NMR (CD CN): 27.95(s, 2H, ArH),
3
2
5.0(s, 2H, ArH), 7.2(s, 6H, CH ), 6.1(s, 2H, ArH), 0.2(s, 2H, ArH); 7.48.0(m, 20H, Ph).
3
ESI-MS (m/z) negative mode: 421 (7 anion); positive mode: 339 (PPh cation). Anal. Calcd. for
4
C H N O PCo: C, 67.90; H, 4.51; N, 7.37. Found: C, 68.30; H, 4.82; N, 7.29%.
4
3
34
4
4
2
.6. Single-crystal X-ray characterization
Single crystals of 7b were obtained from ethanol/heptane. C H CoN O P, M = 760.7,
4
3
34
4
4
r
monoclinic, P2 /c, red, a = 11.184 (2) Å, b = 10.9653 (3) Å, c = 28.742 (3) Å, β = 96.469 (10)°,
1
T = 299 K, Z = 4, R1 = 0.048, wR = 0.102, GOF = 1.02. Data collection was on a Bruker-
2
Nonius KappaCCD (Mo Kα radiation), structure solution with direct methods, structure
2
refinement on F with anisotropic thermal parameters for all non-H atoms. H atoms were placed
at calculated positions and refined using a riding model.
2
.7. Oxidation of 7 to a neutral species
IV
Purple 7a (1 mg) was placed into a mortar and an excess of (NH ) Ce (NO ) (5 mg) was added.
4
2
3 6
Neither of these compounds is soluble in dichloromethane, benzene, or toluene.
Dichloromethane (1.5 mL dried over MgSO ) was added to the mortar, and the heterogeneous
4
mixture was ground together under the solvent using a pestle. The solvent quickly turns deep
blue (max at 860 nm and slightly lesser peak at 700 nm), indicating the formation and dissolution
of neutral 13. When this color maximized, half of the solution was filtered through glass wool
t
into an EPR tube and frozen. The same is done to the other half, except 4- Bu-pyridine was
added to the EPR tube yielding a yellow/orange solution which was then frozen. Reaction yields
5

were difficult to estimate due to the heterogeneous nature of the reaction and difficulty of
isolation of the highly oxidizing product.
2
.8. Oxidation of 7 by Br2
Complex 7a (1 mg) was dissolved in 1 mL water followed by addition of dichloromethane
1.5 mL) and a drop of bromine water (0.1 mL Br in 1 mL water). The mixture was shaken
(
2
vigorously until the organic layer was blue. The organic layer was pipetted off and dried over
MgSO . The blue color was lost slowly and began turning green.
4
2
.9. Attempted metalation of 10 by vanadium(V)
Macrocycle 10 (50 mg, 0.136 mmol) was placed into a 50 mL dried round bottom flask under
argon. Distilled THF (25 mL) was added. Lithium bis(trimethylsilyl) amide (0.2 mL, 1 M in
THF, 0.2 mmol) was added (0.367 eq relative to amide protons). Vanadium(V) triisopropoxide
oxide (40 mg, 0.164 mmol) was added and stirred. After 24 h, the reaction mixture was poured
into 100 mL heptane, and over the course of a few hours, a green solid precipitate formed on the
walls of the beaker as heptane evaporated. This was collected (approximately 5 mg), dissolved in
water, precipitated with tetraphenylphosphonium chloride, and collected by filtration on filter
paper. The precipitate was then dissolved in 2 mL MeCN and 1.5 mL water was added. This
mixture was heated at 70 C in a water bath until approximately 0.5 mL of solvent evaporated,
and then left to stand for 2 weeks at room temperature. Small yellow/green crystals that formed
were studied by X-ray crystallography [28].
3
3
. Results and discussion
.1. Comments of the ligand synthesis
An important step in the ligand synthesis is the condensation of two molecules of semi-protected
,2-diaminobenzene 8 with dimethyl malonyl dichloride to form 9, which after deprotection with
1
HCl undergoes macrocyclization into 10 reacting with oxalyl dichloride [19] (scheme 1). In
principle, a similar result could be achieved through the macrocyclization of 11 (produced from
8
and oxalyl dichloride) with dimethyl malonyl dichloride.
6

O
O
O
O

NH HN
N
N
N
H H
(
COCl)2
Me C(COCl)2
2
H H
NH HN

N
NH2
O
HCl
H
H
9
O
1
0
O
N
H
O
HCl
H
H
Me C(COCl)
2
2
(
COCl)2
O
8
NH HN
NH HN
O
N
H N
2
NH
O
HN
O
O
O
11
12
Scheme 1. Two possible syntheses of 10 macrocycle of which the 8 → 9 → 10 sequence is
effective.
Searching for improving the yield of 10 (its overall yield via the 8 → 9 → 10 sequence
was 24% [19]), we have also explored the 8 → 11 → 10 sequence with a negative result. The
macrocyclization of 11 to form 10 did not occur and the imide 12 was produced instead
1
according to the H NMR data. This was rather unexpected because dialkyl malonyl dichloride
has previously been used for the macrocyclization of the ligands of complexes 1-3 [8, 29, 30]
and 6 [16]. Compounds such as 12 were also observed in the attempted macrocyclization to form
ligands of 2 using dimethyl malonyl dichloride in the presence of very strong bases [31]. As
another example, using dimethyl malonyl dichloride to couple two equivalents of o-nitroaniline
resulted in the imide when triethylamine was used, while the coupled product formed when
pyridine was the base. In this case, weaker bases did promote the formation of the nitro-
derivative of 9.
To find out why 9 and 11 differed so much in their reactivity towards macrocyclization,
theoretical studies were undertaken for 11. First, the energy of the latter was analyzed as a
function of the torsion angle at the oxalamide C‒C bond, which was varied from 0.87 to 359.13°.
-1
The transoid conformation of 11 with the torsion angle of 180° was by ca. 71 kJ mol , more
energetically advantageous than the cisoid conformation with the torsion angle of 0°. As a result,
the two free amino groups of transoid 11 are too far away from each other making it very
7

problematic for dimethylmalonyl dichloride to "find" them to form macrocycle 10 as opposed to
forming imide 12.
Further, a separation between the free amino groups sufficient for the macrocyclization of
1
1 was investigated as a function of the same torsion angle. Our data suggest that the angle
optimal for the macrocyclization lies in the range from ‒55 to +55°. When the absolute value of
the torsion angle is outside this range, the macrocyclization becomes unlikely. Compound 9
lacks the transoid stability of 11 and thus can attain an optimal conformation for
macrocyclization via reaction with oxalyl chloride.
3
.2. Metalation of 10
The standard procedure for metalation of TAMLs has been applied for 10. The ligand was first
treated with a strong base (lithium bis(trimethylsilyl)amide) in THF followed by reacting the
deprotonated form with cobalt(II) chloride. Exposure of the reaction mixture to air resulted in the
formation of the cobalt(III) derivative isolated as a lithium salt 7a. The higher oxidation state 3+
is effectively stabilized by four negatively charged deprotonated amide donor centers and 7a is
very stable in the air. Compound 7a is poorly soluble in organic solvents and therefore the
tetraphenylphosphonium salt 7b was prepared via cation metathesis.
The ease of metalation of 10 encouraged us to make the corresponding TAML complex
V
of vanadium, which was never reported previously [2]. Thus, [V O(OCHMe ) ] was reacted
2
3
with 10 in THF in the presence of lithium bis(trimethylsilyl)amide. The product, the structure of
which was established by X-ray crystallography [28], did not contain 10 and turned out to be an
6
-
anionic decavanadate (figure 1), presumably V O as characterized, for example, by Evans
1
0
28
[32]. It is worth noting that decavanadates are intensively studied due their activity in biological
systems [33, 34].
3
.3. X-ray Structural and DFT studies of 7b
The crystal of 7b suitable for an X-ray study was grown from ethanol/heptane. The structural
data have been preliminarily reported elsewhere [28]. The central metal is surrounded by four
deprotonated amide nitrogens (figure 2a), which are located in the apices of the square with
III
cobalt sitting in the center. The coordination of Co by 10 makes the entire complex almost
perfectly planar and only the Co-containing six-membered ring exhibits minor distortion; C16
8

and C19 deviate slightly above and below the four nitrogen planes, respectively. The CoN bond
lengths are 1.8161.844 Å. The two within the five-membered ring (1.816 and 1.824 Å) are
slightly shorter than those of the six-membered ring (1.835 and 1.844 Å). Inverse sequence could
be anticipated because N3 and N4 should be more electron-donating due to two methyl groups
attached to C16.
+
Counter ions are usually not shown for transition metal cations or anions but PPh of 7b
4
is included in figure 2a. The reason is that three phenyl rings could be compared with three piano
stool legs leaning on the imaginary plane of the anion. Though the Co···P separation is large
(5.335 Å), some atoms of the cation approach rather close to those of the transition metal anion.
For example, the C9···C29 and C2···C36 separations of 3.227 and 3.668 Å, respectively, may be
indicative of -stacking interactions between phenyl rings of the cation and the anion of 7b. It is
worth noting that the DFT optimization of 7b ended up in a quite remarkable similarity of the
experimental and theoretical portraits (figure 2b). The similarity spreads even on the
phosphonium counterion with its closest separations with the anion of 3.406 and 3.604 Å.
1
3
.4. H NMR spectra of free ligand 10 and cobalt(III) complex 7b
The spectrum of 10 is simple reflecting the existence of the symmetry plane. The amide
hydrogens appear as two broadened singlets at δ 9.56 and 9.66 in d -dmso. Presumably the
6

former arises from the H hydrogens of 10 (scheme 1) because of the electron donating effect
from the adjacent methyl groups. The methyl groups resonate as one singlet at δ 1.55 at room
temperature suggesting a dynamic behavior of 10 in solution. The spectrum of 7b in
d3-acetonitrile (figure 3) reveals paramagnetism of the S = 1 complex typical of other square
III
planar Co compounds of this family [7, 9, 35]. All resonances of 7b cover the δ range from 22
to 8 and only the signals from tetraphenylphosphonium cation are not paramagnetically shifted.
The methyl groups appear as one singlet at δ 7.3. Two of the four aromatic hydrogens are
stronger shifted to the higher field (δ 27.9 and 25.0) as compared with those at δ 6.2 and

0.1.
1
The H NMR spectrum of 7b in less coordinating CDCl solvent is noticeably broader
3
than that in CD CN though chemical shifts of the anionic part of the complex do not change. The
3
+
counter-cation PPh , however, reveals itself at δ 4.5-5.5 which is presumably due some ion-
4
paring between the diamagnetic cation and the paramagnetic anion similar to that observed in the
9

solid state (see figure 2). The paramagnetic character of the spectrum of 7b in CDCl practically
3
vanishes on addition of excess d -py (ca. 0.06 mL), which is also accompanied by color change
5
from green to yellow (figure 4). A sharp singlet from the two methyl groups at δ 0.0 is indicative
of the octahedral composition of the C product formed in accord to eq. 1. A new spectral
2
v
pattern shown in figure 4 suggests a spin change from S = 1 to S = 0 on axial coordination of two
pyridine ligands.
_
L
Co
L
_
Co
+ 2 L
(1)
3
.5. Electrochemistry of 7b
-1
Cyclic voltammograms of 7b obtained in acetonitrile at scan rates of 50200 mV s show a
Nernstian behavior at the electrode (figure 5). There is a separation between anodic and cathodic
peaks of ca. 60 mV which is unaffected by the scan rate and consistent with a reversible 1e
transfer [36]. It is tempting to formally assign this feature observed at 0.66 V vs SCE to the
IV
III
Co /Co transition but it might be misleading considering the data in table 1 which collects
III
similar information for related Co TAML complexes. The key question here, which has been
raised before [7, 29], is whether this is metal- or ligand-centered oxidation. We believe that the
electrochemical data presented here supports the ligand-based model.
First, the electrochemical oxidation at ca. 0.8-1.0 V vs SCE is observed clearly when the
III
TAML of Co complex contains a benzene unit. No oxidation was reported for 1, which is an
arene-free species, for the "switch" complex 3b, in which benzene ring is replaced by the
pyridine fragment, and for the 'beheaded' [38, 39] complex 6 (table 1). Second, the value of E½
for 7b with two benzene rings is much lower than that for all other complexes in table 1, which
have just one ring.
IV
IV
3
.6. Oxidation of 7 by Ce and Br and EPR studies of the Co complexes produced
2
The value of E for 7b obtained by cyclic voltammetry suggests that strong oxidants should
½
readily convert monoanionic cobalt(III) species 7 into the corresponding neutral derivatives. This
happens under the action of cerium(IV) or dibromine. Complex 7a was oxidized by
IV
[
NH ] [Ce (NO ) ] in a slurry using benzene or CH Cl where both the reductant and oxidant
4 2 3 6 2 2
1
0

are insoluble (eq. 2). Upon oxidation, the neutral complex formed dissolves in the solvent
resulting in its color change from colorless to bright blue, which is typical of oxidized square
planar TAML complexes of cobalt [7, 9]. The solution was then filtered and the filtrate was used
for EPR measurements.
_
L
Co
L
CeIV or Br2
+2 L
2 L
Co
Co
(2)
-
7
13
14: L = 4-tert-butylpyridine (a)
-tert-butyl isonitrile (b)
4
The oxidation of 7 by Br in a 1:1 (v/v) water-CH Cl mixture (eq. 2) looks spectacular.
2
2
2
Charged 7a is insoluble in non-polar organic solvents and therefore the aqueous layer is purple.
Addition of a trace of Br to the binary mixture followed by its vigorous shaking moves the
2
colored species into the organic phase, which turns bright sky blue due to the formation of the
neutral product. The aqueous layer becomes colorless indicative of a quantitative oxidation
of 7a.
The neutral complex 13 is EPR silent, presumably due to aggregation upon freezing
which provides anti-ferromagnetic interactions. The aggregation does not occur in the presence
of excess N- or C-ligand, viz. 4-tert-butylpyridine and 4-tert-butyl isonitrile, respectively (eq. 2),
which were added prior to freezing. The EPR spectra of 14a,b are shown in figure 6. The spectra
5
9
7
display Co (I = / ) eight-line hyperfine patterns at low temperature with a = 5.2 and 4.1 mT for
2
1
4a,b, respectively; for 14a thawed at room temperature, a = 2.2 mT. The spectra in figure 6
were simulated using g = (1.990, 2.008, 2004) for 14a, g = (1.998, 2.002, 2.006) for 14b; and
axial hyperfine tensors A = (15, 15 147) MHz for 14a and A = (9, 9, 115) MHz for 14b. The low
temperature isotropic values of g (2.000) and A (59 MHz) for 14a agree with those obtained
from simulation of the room temperature data. The EPR data presented above indicate S = ½ for
both 14a and 14b. Furthermore, the very low isotropic value for A and isotropic value g of 2.00
indicate that the unpaired electron is primarily ligand centered. Therefore, the EPR data confirm
III
a Co -ligand radical electronic configuration for the 1 electron oxidized state of both complexes.
1
1

3
.7. DFT Analysis of 7 and 13/14
A goal of the computational work was to gain theoretical support for the cobalt(III)-ligand
radical-cation formulation of the neutral species 13/14, the experimental evidence for which was
deduced from the electrochemical and EPR data. An obstacle here is that the electrochemical
data report on the 7 ⇄ 13 redox feature but neutral, square-planar 13 did not yield an EPR signal.
Instead, it is octahedral 14 that is a source of relevant information.
The orbital energy diagram for square-planar anion 7 is shown in figure 7A. It differs
noticeably from idealized square-planar complexes due to π interactions with four N-donors
which raise the energy of the d and d orbitals. The d orbital becomes the lowest. The net
xz
yz
2
z
6
1
result is that d 7 is S = 1 high spin as illustrated by its paramagnetism (as observed by H NMR),
which practically vanishes on addition of excess of pyridine ligand. The square-planar species
turns into octahedral as in eq. 1, the energy of the d orbital increases expectedly making the
2
z
1
complex low spin (S = 0) (figure 7B) as observed by H NMR.
The primary product of 1e oxidation of 7 is 13, and one might assume that its d orbital
diagram would match that of 7 (figure 7A) with one fewer electron. However, provided the
electron is removed from the ligand and not from the metal, it is hypothesized that the metal
retains its S = 1 configuration with an antiferromagnetically coupled electron residing in a
delocalized fashion across the ligand as in figure 7C. This diagram for a square-planar species
corresponds to the cobalt(III)/radical cation ligand. While direct theoretical modeling of the case
in figure 7C is problematic due to the lack of relevant EPR data for 13, the electrochemical
behavior in combination with the color and solubility changes seem to be rather convincing.
The model such as in figure 7C, which is consistent with the cobalt(III)/radical cation
formulation, could be built for octahedral species 14 (figure 7D). In this case, just as in the
reduced state, the axial ligands boost the energy of the d orbital thereby reducing the spin on
2
z
the metal to S = 0 while leaving the ligand radical character intact. It should be taken into
account that axial ligands, which raise the energy of the d orbital, are added after the electron
2
z
transfer. Hence, it could be inferred that the electronic assignment for square-planar 13 is similar
and shows that one electron oxidation of cobalt(III) TAMLs is strongly favored by the presence
of benzene rings in the ligand system (table 1).
1
2

4
. Conclusion
The TAML ligand 10, due to its symmetry, provides interesting insight into the synthesis of
macrocycles regarding the nature of synthetic route to be used, which in this case involved final
macrocyclization by oxalyl dichloride (scheme 1). The ease of one electron oxidation of 7
III
compared to all previously reported Co TAMLs containing a benzene ring in the ligand
structure forces one to think that a radical-cation ligand is formed with the oxidation state of
cobalt unchanged. Theoretical evidence for this was obtained only for the octahedral product of
1
e oxidation 14 generated after formation of neutral species 14. The experimental data are
1
consistent with a ligand centered radical for both species. The H NMR data for 7 obtained in the
absence and in the presence of excess of d -pyridine indicates the S = 1 → S = 0 spin transition
5
on going from square-planar to octahedral species (eq. 1), which suggests that a similar
phenomenon could occur in the oxidized states of 13 and 14.
Supplementary material
Cif and checkcif files for 7b.
Acknowledgements
NMR instrumentation at CMU was partially supported by NSF (CHE-0130903 and CHE-
1
039870). M. P. H. thanks the National Institutes of Health GM49970 for funding. The EPR
spectrometer was funded with NSF CHE1126268.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
Operating support is acknowledged from the Heinz Endowments (T.J.C.), the Institute for Green
Science (T.J.C.), and the Environmental Protection Agency (grant RD 83 to T.J.C.).
References
[1]
A. Chanda, D.-L. Popescu, F. Tiago de Oliveira, E.L. Bominaar, A.D. Ryabov, E.
Münck, T.J. Collins. J. Inorg. Biochem., 100, 606 (2006).
1
3

[
2]
D.-L. Popescu, A. Chanda, M. Stadler, F. Tiago de Oliveira, A.D. Ryabov, E.L.
Bominaar, E. Münck, T.J. Collins. Coord. Chem. Rev., 252, 2050 (2008).
T.J. Collins, A.D. Ryabov. Chem. Rev., 117, 9140 (2017).
[
3]
4]
[
P.V. Bernhardt, G.A. Lawrance. In Compr. Coord. Chem. II, J.A. McCleverty, T.J.
Meyer (Eds.), pp. 1-145. Elsevier Ltd., Oxford, UK (2004).
[5]
[6]
[7]
J.J. Bour, P.T. Beurskens, J.J. Steggarda. J. Chem. Soc., Chem. Commun., 221 (1972).
T.J. Collins, E.S. Uffelman. Angew. Chem., Int. Ed. Engl., 28, 1509 (1989).
T.J. Collins, R.D. Powell, C. Slebodnick, E.S. Uffelman. J. Am. Chem. Soc., 113, 8419
(1991).
[
[
[
[
[
[
[
[
[
8]
A. Ghosh, P. Ramidi, S. Pulla, S.Z. Sullivan, S.L. Collom, Y. Gartia, P. Munshi, A.S.
Biris, B.C. Noll, B.C. Berry. Catal. Lett., 137, 1 (2010).
9]
R.E. Patterson, S.W. Gordon-Wylie, C.G. Woomer, R.E. Norman, S.T. Weintraub, C.P.
Horwitz, T.J. Collins. Inorg. Chem., 37, 4748 (1998).
10] S.W. Gordon-Wylie, High valent manganese and cobalt complexes of oxidatively robust
N and O donor ligands, PhD Thesis. Carnegie Mellon University, Pittsburgh (1995).
11] C.P. Horwitz, S.W. Gordon-Wylie, Y. Leychkis, D.M. Flynn, S.T. Weintraub, G.R.
Clark, T.J. Collins. J. Phys. Chem. B, 105, 8821 (2001).
12] D. Das, S. Pattanayak, K.K. Singh, G.S. Sen, B. Garai. Chem. Commun., 52, 11787
(
2016).
13] H.-Y. Du, S.-C. Chen, X.-J. Su, L. Jiao, M.-T. Zhang. J. Am. Chem. Soc.,
0.1021/jacs.1028b00032 (2018).
14] R.O. Saavedra Diaz, R. Le Lagadec, L.Q. Shen, A.D. Ryabov. J. Coord. Chem., 67, 3909
2014).
1
(
15] S. Hong, F.F. Pfaff, E. Kwon, Y. Wang, M.-S. Seo, E. Bill, K. Ray, W. Nam. Angew.
Chem., Int. Ed., 53, 10403 (2014).
16] M.A. DeNardo, M.R. Mills, A.D. Ryabov, T.J. Collins. J. Am. Chem. Soc., 138, 2933
(2016).
[
17] M. Anbar, D. Meyerstein. Nature, 206, 818 (1965).
18] W.C. Ellis, C.T. Tran, M.A. Denardo, A. Fischer, A.D. Ryabov, T.J. Collins. J. Am.
Chem. Soc. 131, 18052-18053 (2009).
[
1
4

[
[
[
19] W.C. Ellis, C.T. Tran, R. Roy, M. Rusten, A. Fischer, A.D. Ryabov, B. Blumberg, T.J.
Collins. J. Am. Chem. Soc., 132, 9774 (2010).
20] G. Warner, M. Mills, C. Enslin, S. Pattanayak, C. Panda, S. Sen Gupta, A.D. Ryabov,
T.J. Collins. Chem. Eur. J., 21, 6226 (2015).
21] W.L.F. Armarego, C.L.L. Chai, Purification of laboratory chemicals, Elsiever, USA
(2003).
[
[
[
[
[
[
22] D.T. Petasis, M.P. Hendrich. Methods Enzymol., 563, 171 (2015).
23] A.D. Becke. J. Chem. Phys., 98, 5648 (1993).
24] A.D. Becke. J. Chem. Phys., 109, 2092 (1998).
25] C. Lee, W. Yang, R.G. Parr. Phys. Rev. B, 37, 785 (1988).
26] Y. Zhao, D.G. Truhlar. Theor. Chem. Acc., 120, 215 (2008).
27] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, G.
Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J.A. Montgomery, J.E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd,
E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J.M. Millam, M.
Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E.
Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, R.L. Martin,
K. Morokuma, V.G. Zakrzewski, G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich,
A.D. Daniels, O. Farkas, J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox. In Gaussian
0
9, Revision D.01, Gaussian, Inc., Wallingford, CT (2009).
[
28] W.C. Ellis, Green design, synthesis, characterization, and application of a novel family
of Fe(III)-TAML peroxide activating homogeneous catalysis. PhD Thesis, Carnegie
Mellon University, Pittsburgh (2010).
[
[
[
29] T.J. Collins. Acc. Chem. Res., 27, 279 (1994).
30] T.J. Collins. Acc. Chem. Res., 35, 782 (2002).
31] A. Ghosh, Design, synthesis and mechanistic studies of iron-TAML catalytic activators of
hydrogen peroxide and a new activation chemistry of dioxygen by iron, PhD Thesis,
Carnegie Mellon University, Pittsburgh (2004).
1
5

[32] H.T. Evans, Jr. Inorg. Chem., 5, 967 (1966).
[
33] M.P.M. Marques, D. Gianolio, S. Ramos, L.A.E. Batista de Carvalho, M. Aureliano.
Inorg. Chem., 56, 10893 (2017).
[
34] M. Aureliano, C.A. Ohlin. J. Inorg. Biochem., 137, 123 (2014).
35] J.C. Brewer, T.J. Collins, M.R. Smith, B.D. Santarsiero. J. Am. Chem. Soc., 110, 423
[
(1988).
[
36] A.J. Bard, L.R. Faulkner, Electrochemical Methods. Fundamentals and Applications,
John Wiley & Sons, New York (1980).
[
37] E.S. Uffelman, Design, synthesis, and structure of a macrocyclic tetraamide that
stabilizes high-valent middle and later transition metals, PhD Thesis, California Institute
of Technology, Pasadena (1992).
[
[
[
38] M.R. Mills, L.Q. Shen, D.Z. Zhang, A.D. Ryabov, T.J. Collins. Inorg. Chem., 56, 10226
(2017).
39] M.R. Mills, A.C. Weitz, D.Z. Zhang, M.P. Hendrich, A.D. Ryabov, T.J. Collins. Inorg.
Chem., 55, 12263 (2016).
40] L.Q. Shen, Experimental and theoretical studies of TAML ® activators : pharmaceuticals
degradation, nuclear tunneling and electronic structure analysis, PhD Thesis, Carnegie
Mellon University, Pittsburgh (2013).
1
6

Figure 1. Structure of decavanadate produced by attempted metalation of 10 by
V
[
V O(OCHMe ) ] (see text for details) [28].
2 3
a
b
Figure 2. Structure of 7b elucidated by X-ray crystallography (a) and its overlay with a model
created by DFT (b). Displacement ellipsoids in a are drawn at the 50% probability level.
Hydrogens are omitted for clarity. Selected bond lengths (in Å): CoN(1) 1.824(2), CoN(2)
1
.816(2), CoN(3) 1.835(3), CoN(4) 1.844(2). DIAMOND release 3.1e, Crystal Impact GbR,
Bonn, Germany.
1
7

1
Figure 3. H NMR spectrum of 7b in CD CN. Assignments (δ): ‒27.9, ‒25.0, ‒6.2 and ‒0.1 (2H,
3
phenylene hydrogens); ‒7.3 (6H, CH3); 7.5-8.0 (20H, PPh4).
1
Figure 4. H NMR spectrum of 7b in CDCl recorded in the excess of d -py. The top inset
3
5
highlights the absence of resonances at δ < 0; the bottom inset details resonances of aromatic
protons.
1
8

1.5e-5
1.0e-5
5.0e-6
0.0
-
-
5.0e-6
1.0e-5
0.3 0.4 0.5 0.6 0.7 0.8
E / V (vs SCE)
Figure 5. Cyclic voltammograms of 7b (0.001 M) at a glassy carbon electrode and scan rates 50,
-1
n
1
00, and 200 mV s in MeCN in the presence of 0.1 M [ Bu N][PF ].
4 6
Figure 6. EPR spectra (9.625 GHz) of 14a (A, B) and 14b (C) in CH Cl . The black traces are
2
2
simulations (see text for parameters). Sample temperatures 10 K (B, C); 300 K (A).
1
9

z2
xy
z2
xy

R
xz
yz
xy
xy
xz
yz
x2-y2
R
x2-y2
xz
yz
xz
yz
z2
x2-y2
z2
x2-y2
Ligand
Ligand
Metal
Metal
A
B
C
D
Figure 7. DFT supported d orbital diagrams for anionic cobalt(III) TAML complexes 7 (S = 1)
IV
(
A) and 7a (S = 0) (B), neutral (formally Co ) complexes 13 (S = ½) (C) and 14 (S = ½) (D).
Abstracted in part from ref. [40].
Table 1. Comparison of the reduction potentials (vs SCE) for one-electron
III
oxidation of Co complexes of various TAMLs.
Complex
Solvent
CH Cl
Behavior type
a)
E½ / V
Reference
[6, 37]
1
2
2
2
2
2
3
4
5
6
7
a
MeCN
CH Cl
Reversible
Reversible
1.00
0.98
0.846
[15]
b b)
b
[9]
2
2
CH Cl
Reversible
[7]
2
2
c)
b
MeCN
MeCN
MeCN
MeCN
MeCN
[11]
Irreversible
0.825
0.846
[14]
Reversible
[12]
d)
[13]
b
Reversible
0.66
This work
a)
No electrochemical activity up to 2.0 V vs NHE.
b)
c)
d)
2 2
Variant of 2b with two C(cyclohexyl) instead of two CMe groups.
No electrochemical features reported at potentials above 0.0 V.
Only irreversible feature at 1.2 V was reported.
2
0