Secondary Sphere Hydrogen Bonding in Monocopper Complexes of Potentially Dinucleating Bis(carboxamide) Ligands

Article abstract of DOI:10.1002/ejic.201501060

Reaction of a macrocyclic ligand precursor comprising two bis(carboxamido)pyridine units (H₄L4) connected by ethylene linkers with NMe₄OH and CuX₂ (X = Cl, OAc, or OTf) yielded monocopper complexes [NMe₄][(H

Full text of DOI:10.1002/ejic.201501060

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
FULL PAPER
DOI:10.1002/ejic.201501060
Secondary Sphere Hydrogen Bonding in Monocopper
Complexes of Potentially Dinucleating Bis(carboxamide)
Ligands
Benjamin D. Neisen,[a] Pavlo V. Solntsev,[a]
Mohammad R. Halvagar,[a] and William B. Tolman*[a]
Keywords: Copper / Hydrogen bonds / Redox chemistry / Macrocycles
Reaction of a macrocyclic ligand precursor comprising two
bis(carboxamido)pyridine units (H4L4) connected by ethyl-
ene linkers with NMe4OH and CuX2 (X = Cl, OAc, or OTf)
yielded monocopper complexes [NMe4][(H2L4)Cu(X)] [X =
Cl (3), OAc (4), or OH (5)], in contrast to the results of pre-
vious work on a related ligand with ortho-phenylene linkers
wherein dicopper compounds were isolated. X-ray structures
of the complexes revealed hydrogen bonding from the free
carboxamide N–H groups in the doubly protonated form of
the ligand (H2L42–) to the monodentate fourth ligand coordi-
nated to the CuII ion. Similar secondary sphere hydrogen
bonding interactions were identified in multinuclear com-
pounds [NMe4]2[{(H2L4)Cu}n(CO3)] (n = 2 or 3) that were iso-
lated from exposure of 5 to air. Cyclic voltammetry revealed
oxidations of 3 and 5 at potentials about 300 mV higher than
those of analogous monocopper complexes of bis(arylcarbox-
amido)pyridine ligands, which lack the intramolecular
hydrogen bonds, consistent with removal of electron density
from the metal center by the hydrogen bonding array. An-
other ligand variant (H4L5) with ortho-phenylene linkers and
only one bis(carboxamido)pyridine moiety yielded mono-
copper complexes [NMe4][(H2L5)Cu(OAc)]·DMF (8) and
[NMe4][(H2L5)CuCl]·CH3CN (9), but the X-ray structures re-
vealed a different hydrogen bonding arrangement to the
solvate molecules. Nonetheless, a high redox potential for 9
was observed, consistent with intramolecular hydrogen
bonding interactions in solution.
Concerned about the potential redox noninnocence of
the ortho-phenylenediamine linkers in L34–, and with the
aim of further enhancing the electron-donating properties
of the carboxamide and thus further stabilizing oxidized
copper species to enable more complete characterization,
we turned to the analogous ligands L44–[10] comprising sim-
ple ethylene linkers. However, efforts to isolate dicopper
complexes of L44– have failed; instead we have discovered
a new class of monocopper complexes that feature hydrogen
bonding from the free carboxamide N–H groups in the
doubly protonated form of the ligand (H2L42–) to the
fourth ligand coordinated to the CuII ion. Such interactions
model secondary sphere hydrogen bonding in important
metalloenzyme active sites.[11] It is understood that
hydrogen-bond-donating and -accepting residues in the sec-
ond coordination sphere of active sites of enzymes have an
effect on the redox potential of the metal center. In pre-
viously reported enzymatic studies, the direction and mag-
nitude of a shift in potential is not always obvious, as there
are reports of second-sphere residues both raising and low-
ering the redox potential. For instance, in Fe superoxide
dismutase (SOD), a glutamine residue (Q69) acting as a
hydrogen-bond donor to the inner sphere of the iron has
been implicated in lowering the potential by –220 mV.[11b]
Conversely, increasing H-bond donation in the Fe–S pro-
tein rubredoxin has been shown to modulate the potential
Introduction
By understanding the properties and reactivity of cop-
per–oxygen intermediates implicated in reactions of metal-
loenzymes and other catalysts, progress toward the develop-
ment of new oxidation processes may be achieved.[1] Moti-
vated by intriguing proposals for such intermediates, includ-
ing [CuO]+ species in enzymes such as peptidylglycine
monooxygenase,[2] dopamine β-monooxygenase,[3] and lytic
polysaccharide monooxygenase[4] as well as [CuOCu]n+ (n
= 2–4) species in methane monooxygenase[5] and hetero-
geneous Cu-doped zeolites,[6] we and others have aimed to
synthesize and study in detail reactive mono- and dicopper
complexes containing these and related cores.[1,7] In recent
work,[8,9] we prepared compounds 1 and 2 proposed to con-
tain [CuIIIOH]2+ and [CuII/III(OH)CuIII 4+/5+
moieties using
]
the ligands L12–, L22–, and L34–, respectively (Figure 1).
The powerful electron-donating ability of the carboxamide
donors in the ligands was critical for stabilizing the reactive
oxidized cores of these complexes.
[a] Department of Chemistry and Center for Metals and
Biocatalysis, University of Minnesota,
207 Pleasant St. SE, Minneapolis, Minnesota 55455, USA
E-mail: wtolman@umn.edu
www.chem.umn.edu/groups/tolman
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201501060.
Eur. J. Inorg. Chem. 2015, 5856–5863
5856
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
www.eurjic.org
FULL PAPER
complexes 3–5. A more purposeful synthesis was developed
by using 2 equiv. of base and 1 equiv. of CuX2 (Scheme 1).
The complexes were isolated as crystalline solids, in modest
yields, and characterized by electrospray ionization mass
Scheme 1. Synthesis of complexes of H2L42–. (i) NMe4OH
(2 equiv.), CuX2 (X = Cl, OAc, or OTf). (ii) air, DMF.
Figure 1. Ligands and complexes discussed herein.
up to +126 mV.[11a] A number of studies aimed at under-
standing how such hydrogen bonding influences the reactiv-
ity and redox properties of metal complexes have also been
published.[12] In model complex work, there is a much more
obvious trend of second-sphere H-bonding interactions
raising the redox potential of the metal as compared to the
analogous complex with little to no second-sphere interac-
tions. Particularly relevant in the present context is recent
work in which hydrogen bonding groups were introduced
as unconstrained appendages in bis(carboxamido)pyridine
ligands to evaluate their effects on the redox potentials and/
or the catalytic activity of CuII complexes.[13] Herein, we
evaluate the structures of a series of monocopper complexes
of H2L42– and the new purposefully mononucleating li-
gands H2L52– (RЈ = H or tBu) in which the carboxamide
N–H groups are held in place by the macrocycle. Different
secondary sphere interactions were observed, and their role
in influencing the redox properties of a subset of the com-
pounds prepared was examined.
Results and Discussion
Complexes of H2L42–
Ligand precursor H4L4 was prepared by using a modi-
fied version of the published method.[10] Initial attempts to
generate dicopper complexes of L44– by treating H4L4 with
an excess amount of base in the presence of CuII salts
Figure 2. Representations of the anionic portions of the X-ray
structures of (a) [NMe4][H2L4Cu(Cl)] (3), (b) [NMe4]-
[H2L4Cu(OAc)] (4), and (c) [NMe4][H2L4Cu(OH)] (5), showing all
non-hydrogen atoms as 50% thermal ellipsoids and hypothesized
(greater than or equal to 2 equiv.) only yielded monocopper hydrogen bonds as dashed lines.
Eur. J. Inorg. Chem. 2015, 5856–5863
5857
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
www.eurjic.org
FULL PAPER
Table 1. Selected interatomic distances [Å] and angles [°] for the spectrometry (ESI-MS), EPR spectroscopy, CHN analysis,
indicated X-ray crystal structures.
and X-ray crystallography (Figure 2). The X-ray structures
reveal a common motif featuring a single CuII ion bound
[NMe4][(H2L4)Cu(Cl)] (3)
to one bis(carboxamido)pyridyl portion of the macrocyclic
Cu1–N1 1.986(2)
Cu1–N2 1.926(2)
Cu1–N3 1.984(2)
Cu1–Cl1 2.2408(8)
N4···Cl1 3.393(3) N2–Cu1–Cl1
N6···Cl1 3.390(2) N2–Cu1–N1
H4···Cl1 2.602
H6···Cl1 2.628
178.75(7)
80.40(9)
99.99(7)
80.78(10)
98.77(7)
161.01(10)
ligand. The fourth chloride (in 3), acetate (in 4), or hydrox-
ide (in 5) ligand is hydrogen bonded to two carboxamido
N–H groups, as indicated by the appropriate NH–X and
N···X distances [X = Cl, O–C(O)Me, OH; Table 1]. For ex-
ample, in 4, O1···H4 is 2.202 Å, O1···H6 is 2.197 Å, O1···N4
is 3.003(2) Å, and O1···N6 is 3.002(2) Å. These parameters
compare favorably to related ones reported previously for
CuII complexes of other carboxamido complexes.[13] There
is a slight elongation of the Cu–X (X = OH, Cl) distance
in 3 and 5 compared to those in analogous systems without
hydrogen bonding, as in the case for 1b compared to 5, for
which the Cu–OH distances are 1.863 and 1.884 Å, respec-
tively.[8] Beyond the elongation of the Cu–X distances, the
geometry around the CuII center is nearly identical to those
of non-hydrogen-bonded systems. The retention of the indi-
cated overall formulations for the complexes in solution was
indicated by EPR spectroscopy (for compounds 3–5) and
negative-ion-mode ESI-MS (for compounds 3 and 4; Fig-
ures S2–S4, Table S1); axial EPR signals and the appropri-
ate parent ions and isotope patterns in the mass spectra are
consistent with the respective monocopper species.
Similar hydrogen bonding patterns were observed in the
products of CO2 fixation by the hydroxide complex 5. Thus,
exposure of solutions of 5 in DMF to air followed by dif-
fusion of Et2O led to the formation of purple crystals,
which were identified as a mixture of [NMe4]2[{(H2L4)
Cu}2(CO3)] (6) and [NMe4]2[{(H2L4)Cu}3(CO3)] (7) on the
basis of X-ray crystallography and ESI-MS (7; Figure S5)
performed on crystals selected randomly from the batch
(Figures 3 and 4, Table 1). In both complexes, [(H2L4)Cu]
units surround a carbonate ion. In 6, two such units are
present, and the CuII–O3 distance [2.5425(3) Å] is longer
than that for the equatorial O1 [1.9602(2) Å]. This type of
coordination mode has been observed in other examples of
dicopper–carbonate complexes.[14] Internal hydrogen bond-
ing is evident between the macrocyclic N–H groups to the
two O atoms of the carbonate group that occupy the equa-
torial coordination positions (H4 and H6 to O1; H10 and
H12 to O2), as reflected in the N–H···O distances (Table 1).
The weak axial interaction Cu–O3 is broken in 7, which
features carbonate bound to three CuII ions in η1 fashion,
a precedented tricopper–carbonate motif.[15] Again, intra-
molecular hydrogen bonding between the macrocycle NH
groups and the carbonate O atoms occurs, here resulting in
complete sequestration of all of the available electron lone
pairs of the carbonate ion in the complex.
Cl1–Cu1–N1
N2–Cu1–N3
Cl1–Cu1–N3
N1–Cu1–N3
[NMe4][(H2L4)Cu(OAc)] (4)
Cu1–N1 1.980(2) N4···O1 3.003(2)
Cu1–N2 1.920(2) N6···O1 3.002(2)
Cu1–N3 1.991(2) H4···O1 2.202
Cu1–O1 1.937(1) H6···O1 2.197
Cu1–O2 2.636(2)
N2–Cu1–O1
N2–Cu1–N1
O1–Cu1–N1
N2–Cu1–N3
O1–Cu1–N3
N1–Cu1–N3
171.46(6)
80.97(7)
99.43(6)
80.73(7)
98.11(6)
161.31(7)
[NMe4][(H2L4)Cu(OH)] (5)
Cu1–N1 2.005(2) N4···O1 2.890(2)
Cu1–N2 1.931(2) N6···O1 2.865(2)
Cu1–N3 2.002(2) H4···O1 2.038
Cu1–O1 1.884(1) H6···O1 2.055
N2–Cu1–O1
N2–Cu1–N1
O1–Cu1–N1
N2–Cu1–N3
O1–Cu1–N3
N1–Cu1–N3
172.85(7)
80.08(7)
99.69(6)
80.27(7)
99.59(7)
160.25(7)
[NMe4]2[{(H2L4)Cu}2(CO3)] (6)
Cu1–N1 1.982(2) N4···O1 2.957(4)
Cu1–N2 1.931(2) N6···O1 3.024(4)
Cu1–N3 1.982(2) H4···O1 2.106
Cu1–O1 1.960(2) H6···O1 2.170
Cu1–O3 2.543(3) N10···O2 3.1939(48)
Cu2–N7 1.991(2) N12···O2 3.1276(50)
Cu2–N8 1.919(2) H10···O2 2.019
Cu2–N9 2.019(3) H12···O2 1.908
Cu2–O2 1.908(2)
N2–Cu1–O1
N2–Cu1–N1
O1–Cu1–N1
N2–Cu1–N3
O1–Cu1–N3
N1–Cu1–N3
N8–Cu2–O2
N8–Cu2–N7
O2–Cu2–N7
N8–Cu2–N9
O2–Cu2–N9
N7–Cu2–N9
169.46(8)
80.38(9)
99.96(8)
80.32(9)
98.30(8)
160.32(9)
177.96(9)
80.84(10)
99.76(10)
80.50(10)
98.68(10)
160.35(10)
[NMe4]2[{(H2L4)Cu}3(CO3)] (7)
Cu1–N1 1.979(3)
Cu1–N2 1.923(3)
Cu1–N3 1.991(3)
Cu1–O1 1.945(3)
Cu2–N7 1.979(3)
Cu2–N8 1.919(3)
Cu2–N9 1.975(3)
Cu2–O2 1.949(3)
Cu3–N13 1.987(4)
Cu3–N14 1.931(3)
Cu3–N15 1.977(4)
Cu3–O3 1.942(3)
N4···O1 3.045(4) N2–Cu1–O1
N6···O1 3.167(4) N2–Cu1–N1
H4···O1 2.219
H6···O1 2.395
N10···O2 3.108(4) O1–Cu1–N3
N12···O2 3.140(4) N1–Cu1–N3
H10···O2 2.440
H12···O2 2.421
N16···O3 3.090(5) O2–Cu2–N7
N18···O3 2.984(4) N8–Cu2–N9
H16···O3 2.503
H18···O3 2.398
171.37(12)
80.68(14)
100.31(12)
80.35(14)
97.95(12)
160.75(14)
168.19(12)
80.36(13)
99.71(12)
80.87(13)
98.46(12)
161.21(13)
173.86(13)
O1–Cu1–N1
N2–Cu1–N3
N8–Cu2–O2
N8–Cu2–N7
O2–Cu2–N9
N7–Cu2–N9
N14–Cu3–O3
N14–Cu3–N13 80.08(16)
O3–Cu3–N13 98.16(13)
N14–Cu3–N15 81.17(16)
O3–Cu3–N15 99.94(13)
N13–Cu3–N15 160.54(14)
[NMe4][(H2L5)Cu(OAc)]·DMF (8, RЈ = H)
Cu1–N1 2.005(2)
N4···O7 3.096(5) N2–Cu1–O1
N5···O7 3.164(3) N2–Cu1–N1
H4···O7 2.183
H5···O7 2.691
177.84(6)
80.55(6)
100.68(6)
81.00(7)
97.65(6)
161.14(7)
Cu1–N2 1.915(2)
Cu1–N3 2.018(2)
Cu1–O1 1.912(1)
Cu1–O2 2.819(2)
O1–Cu1–N1
N2–Cu1–N3
O1–Cu1–N3
N1–Cu1–N3
In an effort to evaluate the effects of the hydrogen bond-
ing interactions on the reactivity of complexes of H2L42–,
we compared the cyclic voltammograms of the chloride and
hydroxide complexes 3 and 5 to those of the previously re-
ported[8] complexes [(L2)CuX]– (X = Cl or OH) that lack
such interactions (Figure 5, Table 2). For the hydroxide
complexes in DMF (Figure 5a, i and ii), a pseudo-reversible
[NMe4][(H2L5)CuCl]·CH3CN (9, RЈ = tBu)
Cu1–N1 2.043(2)
Cu1–N2 1.918(2)
Cu1–N3 2.021(2)
Cu1–Cl1 2.1852(5)
N4···N6 3.348(3) N2–Cu1–Cl1
N5···N6 4.013(3) N2–Cu1–N1
H4···N6 2.602
H5···N6 2.628
178.05(5)
80.56(7)
99.81(5)
80.23(6)
99.62(5)
159.62(7)
Cl1–Cu1–N1
N2–Cu1–N3
Cl1–Cu1–N3
N1–Cu1–N3
wave is observed for [(L2)CuOH]– (ipc/ipa ≈ 1, E1/2
=
Eur. J. Inorg. Chem. 2015, 5856–5863
5858
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
www.eurjic.org
FULL PAPER
Figure 3. (a) Representation of the anionic portion of the X-ray
structure of [NMe4]2[{(H2L4)Cu}2(CO3)] (6), showing all non-
hydrogen atoms as 50% thermal ellipsoids and hypothesized
hydrogen bonds as dashed lines. (b) Expanded view of the core,
with all carbon atoms (except C36) omitted.
Figure 4. (a) Representation of the anionic portion of the X-ray
structure of [NMe4]2[{(H2L4)Cu}3(CO3)] (7), showing all non-
hydrogen atoms as 50% thermal ellipsoids and hypothesized
hydrogen bonds as dashed lines. (b) Expanded view of the core,
with all carbon atoms (except C63) omitted.
–67 mV vs. Fc+/Fc, ΔEp = 134 mV) similar to that reported
previously in acetone or 1,2-difluorobenzene,[8] but a return
wave in the CV of 5 only becomes apparent at high scan
rates, indicating high reactivity for the oxidized species. Im-
portantly, the E1/2 for 5 is about 350 mV higher than that
for [(L2)CuOH]–. Similarly, reversibility is greater for [(L2)
CuCl]– (Figure 5b) than that for 3, which shows only a
small reverse scan wave at high scan rates, and the approxi-
mated E1/2 for 3 is about 300 mV higher than that for [(L2)
CuCl]–. The significant redox potential increases for 3 and
5 relative to the complexes of L22– are similar to those re-
ported previously for systems having analogous secondary-
sphere motifs, for which it was proposed that the shifts were
induced by removal of electron density from the metal cen-
ter by the hydrogen bonding array.[12] This rationale seems
likely for our systems as well. However, the comparison be-
tween the systems supported by H2L42– and L22– is compli-
cated by the different nature of the carboxamide substitu-
ents (alkyl vs. aryl). To evaluate this influence and replicate
the aryl substituents of [(L2)CuX]– in a macrocyclic envi-
ronment that would enable hydrogen bonding interactions
but avoid the possibility of dinuclear complex formation
that occurs with L34–, we targeted complexes of ligand
H2L52–.
Figure 5. Cyclic voltammograms (0.1 m Bu4NPF6, Pt electrode) of
(a) the complexes (i) [(L2)CuOH]– and (ii) [(H2L4)CuOH]– (5) in
DMA, and (b) the complexes (iii) [(L2)CuCl]–, (iv) [(H2L4)CuCl]–
(3), and (v) [(H2L5)CuCl]– (9, RЈ = tBu) in CH3CN. Scan rates:
(i) 500 mVs–1, (ii) 500 (black), 2000 (blue), 4000 (red) mVs–1, (iii)
100 mVs–1, (iv, v) 500 (black), 2000 (blue), 4000 (red) mVs–1.
Eur. J. Inorg. Chem. 2015, 5856–5863
5859
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim