Nitrogen-Nitrogen Bond Formation via a Substrate-Bound Anion at a Mononuclear Nickel Platform

Article abstract of DOI:10.1021/acs.organomet.7b00887

The nickel-C₄F₈ fragment coordinates an aminoaryl N-H ketimine to form a stable complex, which upon treatment with base and oxidant leads to an N-N bond-forming reaction and the release of indazole product. A key and previously unidentified intermediate in the formation of the indazole was a diimine complex of nickel bearing significant charge on the aryl ring that initially contained the amine substituent. The C₄F₈ coligand was key for the redox transformation and for stabilization of the intermediate for characterization.

Full text of DOI:10.1021/acs.organomet.7b00887

Communication
Cite This: Organometallics XXXX, XXX, XXX−XXX
Nitrogen−Nitrogen Bond Formation via a Substrate-Bound Anion at
a Mononuclear Nickel Platform
Mikhail D. Kosobokov,† Aaron Sandleben,‡ Nicolas Vogt,‡ Axel Klein,‡ and David A. Vicic*,†
†Department of Chemistry, Lehigh University, 6 East Packer Avenue, Bethlehem, Pennsylvania 18015, United States
‡Institut fur Anorganische Chemie, Department fur Chemie, Universitat zu Koln, Greinstraße 6, D-50939 Koln, Germany
̈
̈
̈
̈
̈
S
* Supporting Information
ABSTRACT: The nickel-C4F8 fragment coordinates an
aminoaryl N−H ketimine to form a stable complex, which
upon treatment with base and oxidant leads to an N−N bond-
forming reaction and the release of indazole product. A key
and previously unidentified intermediate in the formation of
the indazole was a diimine complex of nickel bearing signifi-
cant charge on the aryl ring that initially contained the amine
substituent. The C4F8 coligand was key for the redox transformation and for stabilization of the intermediate for characterization.
better understanding of how to control nitrogen−
Anitrogen bond-cleavage and bond-forming reactions at a
metal center is required for advances in applications spanning
the energy fields to the life sciences. Metal-mediated N−N
bond cleavage represents an important step in nitrogen fixation
chemistry, and much effort has been devoted to developing less
can be used to prepare indazoles bearing a variety of functional
energy intensive methods for ammonia production from N2.1−3
Fundamental studies on the reverse reaction, metal-mediated
N−N bond formation, not only can provide complementary
information related to the global nitrogen cycle as it relates to
fertilizer production but also can lead to useful synthetic
methodologies. Many natural products and pharmaceuticals
contain N−N bonds,4,5 yet methods to effect N−N bond
formation still often invoke highly reactive precursors or inter-
mediates such as azides or nitrenes to proceed.5 Much interest
has been focused on using copper for N−N bond-forming
reactions.5−7
groups and chiral centers.11 Unfortunately, no metal-containing
intermediates were isolated or studied, making an under-
standing of this reaction extremely limited.
The conversion outlined in eq 1 from the aminoaryl N−H
ketimine to an indazole is a two-electron oxidative process;
therefore, a nickel promoter for the reaction should be stable
under such conditions and possibly be able to shuttle between
its higher valent states. The nickel-C4F8 fragment was identified
as a promising platform for mediating the desired N−N bond-
forming reaction, given that it has recently been shown that the
C4F8 fragment enables multiple oxidations beyond the com-
monly encountered nickel(II) state.12
For instance, the square-planar complex [(tbpy)Ni(C4F8)]
(3, Chart 1, tbpy = 4,4′-di-tert-butyl-2,2′-dipyridyl) exhibits
two one-electron oxidations at −0.02 and +1.16 V versus
the ferrocene/ferrocenium couple in MeCN solvent (see the
Supporting Information). The related terpyridine complex
[(tpy)Ni(C4F8)] can likewise undergo chemical oxidation to
the nickel(III) species [(tpy)Ni(C4F8)(MeCN)]+ (4, Chart 1),
and this nickel(III) species exhibits an additional reversible
oxidation at +1.34 V.12 Given the accessibility of two additional
oxidations from the stable nickel(II) state upon ligation to a
C4F8 fragment, we explored whether these oxidations could be
coupled to a nitrogen−nitrogen bond-forming event in hopes
of developing a new methodology with nickel.
Copper-free and metal-catalyzed nitrogen−nitrogen bond-
forming reactions, however, are relatively rare.4,5,8−10 In early
studies, a nickel-catalyzed N−N bond-coupling reaction was
applied to the cyclotrimerizations of α-aminoketones.8 The first
example of a formal reductive elimination of a nitrogen−
nitrogen bond at nickel was recently reported,4 but the N−N
elimination proceeded at a dinickel complex through a mixed-
valent nickel(II)−nickel(III) intermediate that was accessed
using a strong oxidant such as PhICl2. Because the further
development of dinuclear metal complexes for such bond-
forming reactions can be challenging, we sought to understand
how we could construct a monomeric nickel complex for new
N−N bond-forming methodologies. As a starting point, we
wished to provide proof in principle that nickel could mediate
an intramolecular N−N coupling reaction. Chen and co-workers
recently reported that aminoaryl N−H ketimines such as 1 react
with copper salts under an aerobic environment to afford
indazoles such as 2 (eq 1).11 The reaction described in eq 1
formally involves the loss of two electrons and two protons and
The coordination chemistry of aminoaryl N−H ketimines
to nickel was initially explored in order to understand the
Received: December 13, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.7b00887
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
Chart 1. Electrochemical Potentials (vs Fc/Fc+ in MeCN) of
Bipyridine and Terpyridine Nickel-C4F8 Complexes
(1.970(2) Å), presumably due to steric interactions. Interest-
ingly, a weak contact between the amino-bound hydrogen and
the oxygen in a cocrystallized THF molecule was also observed
in the solid state (Figure 1).
The electrochemical features of complex 6 in MeCN solution
are highly dependent on the presence of base. In the absence of
base, 6 exhibits a reversible one-electron oxidation centered at
+0.11 V (vs the ferrocene/ferrocenium couple) and a second
less reversible oxidation whose anodic peak potential is +1.14 V
(Figure 2). Chemical oxidation of 6 in THF or MeCN using
stabilities of potential nickel intermediates in any N−N bond-
forming reactions. Reaction of the known [(MeCN)2Ni(C4F8)]13
with 2-(imino(phenyl)methyl)-N-methylaniline (5) led cleanly to
the formation of complex 6 in 78% yield (eq 2). Complex 6 is
Figure 2. Cyclic voltammogram of 6 in MeCN (20 mM). Working
and counter electrodes are platinum with a Ag pseudo reference
electrode. Conditions: electrolyte, [Bu4N][BF4], 100 mM; scan rate,
100 mV s−1.
surprisingly stable and can be purified by column chromatog-
raphy in the air. An X-ray crystal structure of 6 was obtained
(Figure 1), which reveals a roughly square planar geometry
around nickel. The imine nitrogen−nickel bond length
(1.893(2) Å) is shorter than the amino nitrogen−nickel bond
[(NH4)2][Ce(NO3)6] allowed a study of the radical cationic
complex 6•+ by EPR spectroscopy (spectra in the Supporting
Information). Interestingly, the isotropic g values at 298 K of
2.136 are not consistent with the three g values from the
rhombic spectra obtained at 120 K in glassy frozen solution and
T-dependent experiments revealed a marked shift of the g
values when the temperature was lowered. Obviously, the elec-
tron distribution within 6•+ changes slightly with T. We attri-
bute these spectral features to slightly varying contributions of
the two resonance forms [(L)NiIII(C4F8)]+ (A) ↔ [(L•+)-
NiII(C4F8)]+ (B). Overall, the g values and the large g aniso-
tropy Δg clearly indicate the largest contribution coming from
form A, a Ni(III) species, in agreement with related reports.13−16
When 6 was oxidized with (NH4)2[Ce(NO3)6] in THF in
the presence of the spin trap N-tert-butyl-α-phenylnitrone
(PBN), we observed the isotropic EPR signal of 6•+ alongside
with a minor signal at g = 2.006. This isotropic three-line signal
represents a PBN-trapped radical in line with an observed
hyperfine splitting (HFS) aN of 14 G. aH(α) and further HFS
could not be resolved. From this we can conclude that 6•+ is
quite stable but cleaves a small amount of a reactive, pre-
sumably organic radical under these nonbasic conditions.
Oxidative UV−vis spectroelectrochemical experiments were
also consistent with clean reversible oxidation of 6 in the
absence of base (see the Supporting Information). In the pres-
ence of a base such as pyridine, the CV shows an irreversible
oxidation at +0.66 V whose peak, on the basis of its
height, likely corresponds to an overall two-electron process
(Figure 2).
Figure 1. ORTEP diagram of 6·THF. All hydrogens except those
bound to nitrogen are omitted for clarity. Selected bond lengths (Å):
Ni−N1 1.970(2); Ni−N2 1.893(2); Ni−C4 1.899(3); Ni−C1
1.902(3); O1−H1 2.41(3); N1−C17 1.454(3). Selected bond angles
(deg): N2−Ni−C4 91.42(10); N2−Ni−C1 175.83(11); C4−Ni−C1
86.72(12); N2−Ni−N1 89.83(9); C4−Ni−N1 177.12(10); C1−Ni−N1
91.86(10).
B
DOI: 10.1021/acs.organomet.7b00887
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
Additionally, 9 already has 2 equiv of fluoride built in to the
system. Therefore, the ring-closing reaction described in eq 3
can proceed in good yield (75%) without the need for any CsF
salt additives. When the reaction described in eq 3 is monitored
by 19F NMR spectroscopy, the fate of the metal complex after
the reaction is [(pyr)2Ni(C4F8)]17 (97% after 18 h at room
temperature).
Table 1. Exploration of Conditions for Nitrogen−Nitrogen
Bond-Forming Reactions at Nickel
Because conversion of aminoaryl N−H ketimines to
indazoles ultimately involves proton transfers as well as
electron transfers, reaction of complex 6 with a Brønsted
base was studied. When 6 was stirred with sodium tert-butoxide
in THF solvent in the presence of 18-crown-6, conversion to
the deprotonated species 7 (eq 2) occurred in quantitative yield
by 19F NMR spectroscopy. The NMR features, presented in
Figure 3, reveal interesting details regarding the bonding of 6
entry
oxidant (2 equiv)
base/solvent/additive
yield (%)
1
[Ag][PF6]
[Ag][BF4]
[Ag][BF4]
AgOAc
pyridine or MeCN
pyridine or MeCN
KOAc (2 equiv) pyridine
pyridine
<5
<5
43
72
75
25
37
<5
68
61
88
2
3
4
5
AgF
pyridine
6
AgOAc
MeCN
7
AgF
MeCN
8
[Fc][BF4]
[Fc][BF4]
[Fc][BF4]
[Fc][BF4]
MeCN
9
pyridine/MeCN (1/1)
pyridine
10
11
pyridine/CsF (2 equiv)
Chemically, complex 6 reacts with a combination of oxidants
and bases to afford indazole product 8 (eq 4 and Table 1).
Silver salts containing weakly coordinating anions such as
Ag[PF6] and Ag[BF4] are known to oxidize nickel(II) com-
plexes to nickel(III),12 but reaction of those salts with 6 did not
afford indazole 8 (Table 1, entries 1 and 2). When silver salts
containing coordinating anions of Lewis and Brønsted base
character were employed, the yields of indazole rose dramat-
ically (Table 1, entries 4 and 5). We speculate that the enhanced
yields using AgF and AgOAc are due to interaction of the more
basic counterions with the N−H groups of 6, facilitating proton-
coupled electron transfer events. Indeed, the yields are modulated
by the presence of fluoride or acetate. For instance, the reaction
with Ag[BF4] oxidant could be “turned on” by the addition of
KOAc to give product in 43% yield (Table 1, entry 3). Oxidations
involving ferrocenium in MeCN solvent did not lead to significant
product formation (Table 1, entry 8), but when pyridine was
added, indazole was formed in over 60% yield (Table 1, entries 9
and 10). Finally, the combination of ferrocenium oxidant, pyridine
solvent, and cesium fluoride additive led to the product in 88%
yield (Table 1, entry 11).
Figure 3. 19F NMR spectra of 6 (top) and 7 (bottom) in THF-d8.
and 7. For complex 6, the tetrahedral geometry of the amino
nitrogen dictates that eight resonances be observed in the 19F
NMR spectrum for the fluoroalkyl ligand, as all fluorines are in
magnetically different environments. Once complex 6 is depro-
tonated to 7, the eight resonances collapse into four, indicating
the presence of symmetry.
Two possible interactions of the lone pair on the N-Me
nitrogen that could account for the NMR spectrum of 7 are
shown in Scheme 1. In one form (A, Scheme 1), there is a
Given that the optimal conditions for N−N bond formation
at a nickel(II) center included pyridine as solvent and 2 equiv
each of oxidant and CsF (Table 1, entry 8), we pursued the
synthesis and evaluation of the novel complex 9 (eq 3, see the
Scheme 1. Two Resonance Forms for the Core Structure of
7 Consistent with NMR Data
significant interaction of the lone pair on the amido nitrogen
to nickel to afford the formally 18-electron complex 7. This
form is consistent with the 19F NMR spectrum of 7 and may be
noteworthy in preventing the d electrons of nickel from
donating into a carbon-based orbital of the fluoroalkyl ligand
and eliminating fluoride in an irreversible manner, as has
been observed previously.18−20 Another resonance form (B,
Scheme 1) where the amido collapses to an imine and places
the negative charge on the arene, is also consistent with the
19F NMR spectrum.
Supporting Information for preparation). Notably, complex 9
is stable on the benchtop in air, and no changes could be
observed by NMR spectroscopy after a 2 week storage of the
solid in an open vial. Use of the nickel(IV) precursor 9 instead
of a nickel(II) analogue circumvented the need for 2 equiv
of oxidant in the ring-closing reactions described in Table 1.
C
DOI: 10.1021/acs.organomet.7b00887
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
The stability of 7, uniquely imparted by the C4F8 fragment,
permitted an X-ray crystallographic analysis of 7, and the data
(Figure 4) indeed confirms the planarity of the N-Me group
of ferrocenium oxidant, indazole is produced in yields that are
dependent on solvent. When Fc[BF4] is used as the oxidant in
THF solvent in the absence of base, indazole is produced
in 32% yield (Table 2, entry 1). When pyridine solvent is
Table 2. Exploration of Conditions for Nitrogen−Nitrogen
Bond Forming Reactions from Intermediate Anion 7
entry
oxidant (2 equiv)
solvent
yield (%)
1
2
3
4
[Fc][BF4]
[Fc][BF4]
[Ag][BF4]
AgF
THF
32
87
11
86
80
pyridine
pyridine
pyridine
pyridine
a
5
[Fc][BF4]
a
Reaction time was 15 min.
Figure 4. ORTEP diagram of 7·2THF. All hydrogens except those
bound to nitrogen are omitted for clarity. Selected bond lengths (Å):
Ni1−N2 1.867(2); Ni1−N1 1.911(2); N1−C2 1.344(3); N2−C8
1.302(3); C2−C3 1.441(4); C2−C7 1.443(4); C3−C4 1.358(4); C4−
C5 1.398(4); C5−C6 1.364(4); C6−C7 1.411(4); C7−C8 1.429(4);
C8−C9 1.498(4). Selected bond angles (deg): N2−Ni1−C18
87.15(11); N2−Ni1−N1 90.49(9); C2−N1−C1 114.4(2).
employed, the indazole production reaches 87% yield, which
compares favorably with the optimized yield reported in
Table 1, entry 11. Reaction of 7 with Ag[BF4] afforded the
product in 11% yield (Table 2, entry 3). Interestingly, no pro-
duct formation was observed when the neutral complex 6
was reacted with Ag[BF4] under similar conditions (Table 1,
entry 2). Reaction of 7 with AgF oxidant also gave higher yields
of product than did 6 (Table 2, entry 4 vs Table 1, entry 5).
Moreover, it was determined that conversion to indazole pro-
duct proceeded much more quickly from 7 than from 6.
Reaction of Fc[BF4]/pyridine with anionic 7 resulted in full
conversion and 80% yield of product (entry 5, Table 2), while
18 h of reaction time was typically required for complex 6.
These data are consistent with 7 as a plausible intermediate in
the metal-mediated conversion of aminoaryl N−H ketimines to
indazoles, which may proceed according to the mechanism
provided in Scheme 2. The mechanism involves an oxidation of
bound to nickel. Interestingly, the nitrogen−Carene bond in 7
was found to be rather short at 1.344(3) Å (versus 1.454(3) Å
in 6). Additionally, the bond lengths of the arene ring bearing
the N-Me group varied significantly, indicating a significant
contribution of resonance structure B to the total electronic
structure of the molecule in the solid state. To gain further
evidence for the electronic structure of 7, DFT calculations
were performed using the m6-31g* basis set.21 The structure
of 7 in the gas phase was similar to the solid-state structure in
that short nitrogen−Carene bond lengths (1.34 Å) were
observed for the N-Me nitrogen. The electronic structure
of 7 also displays large contributions from the extended π
system of the arene, particularly at the sites ortho and para to
the N-Me group (Figure 5). Taken together, these data show
Scheme 2. Possible Mechanism for the Cyclization Leading
to Indazole Formation from a Mononuclear Nickel Platform
Figure 5. Calculated HOMO (left, E = −1.553 eV) and HOMO-1
(right, E = −1.901 eV) of the anion 7.
that, upon deprotonation of 6, a key and previously unidentified
intermediate in the formation of indazoles is a metal complex
bearing significant diimine character (B, Scheme 1).
Stoichiometric reactions were performed on the isolated
anion in order to determine if 7 is a viable intermediate toward
indazole formation (Table 2). When 7 is reacted with 2 equiv
complex 7 to afford 10, whose electronic structure may have
contributions of the nickel(III) state as in 11. It is striking that
the one-electron difference between 7 and 10/11 triggers rapid
N−N bond formation. Diao and co-workers also found that the
D
DOI: 10.1021/acs.organomet.7b00887
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
(9) Fritsche, R. F.; Theumer, G.; Kataeva, O.; Knolker, H.-J. Angew.
̈
nickel(III) state was critical for N−N bond formation from
binuclear nickel complexes.4 Electrochemical data support the
first oxidation as being irreversible, which is also consistent with
the rapid formation of product 8 from 10/11.
Various conditions for such N−N bond forming reactions
through cyclizations of an amino group with an imine (or
oxime) partner have been known since the early work of
Hassner and Michelson in the 1960s.22 However, to our knowl-
edge, this work represents the first characterization of a metal-
bound intermediate in such a reaction. The oxidation of 7 to 10
involves to a large degree ligand-centered redox chemistry, and
we hope that such knowledge will be useful in developing more
sophisticated reactions involving N−N bond formations.
Chem., Int. Ed. 2017, 56, 549−553.
(10) Rosen, B. R.; Werner, E. W.; O’Brien, A. G.; Baran, P. S. J. Am.
Chem. Soc. 2014, 136, 5571−5574.
(11) Chen, C.-Y.; Tang, G.; He, F.; Wang, Z.; Jing, H.; Faessler, R.
Org. Lett. 2016, 18, 1690−1693.
(12) Yu, S.; Dudkina, Y.; Wang, H.; Kholin, K. V.; Kadirov, M. K.;
Budnikova, Y. H.; Vicic, D. A. Dalton Trans. 2015, 44, 19443−19446.
(13) Zhang, C.-P.; Wang, H.; Klein, A.; Biewer, C.; Stirnat, K.;
Yamaguchi, Y.; Xu, L.; Gomez-Benitez, V.; Vicic, D. A. J. Am. Chem.
Soc. 2013, 135, 8141−8144.
(14) Zhou, W.; Watson, M. B.; Zheng, S.; Rath, N. P.; Mirica, L. M.
Dalton Trans. 2016, 45, 15886−15893.
(15) Zheng, B.; Tang, F.; Luo, J.; Schultz, J. W.; Rath, N. P.; Mirica,
L. M. J. Am. Chem. Soc. 2014, 136, 6499−6504.
(16) Spasyuk, D. M.; Zargarian, D.; van der Est, A. Organometallics
2009, 28, 6531−6546.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.organomet.7b00887.
■
S
(17) Giffin, K. A.; Korobkov, I.; Baker, R. T. Dalton Trans. 2015, 44,
19587−19596.
(18) Lee, G. M.; Korobkov, I.; Baker, R. T. J. Organomet. Chem. 2017,
847, 270−277.
Cartesian coordinates of the calculuated structures
(MOL)
Experimental procedures, characterization data, NMR
(19) Harrison, D. J.; Daniels, A. L.; Korobkov, I.; Baker, R. T.
Organometallics 2015, 34, 5683−5686.
(20) Giffin, K. A.; Harrison, D. J.; Korobkov, I.; Baker, R. T.
Organometallics 2013, 32, 7424−7430.
spectra, and X-ray and computational data (PDF)
(21) Mitin, A. V.; Baker, J.; Pulay, P. J. Chem. Phys. 2003, 118, 7775−
7782.
Accession Codes
CCDC 1811727−1811728 and 1817446 contain the supple-
mentary crystallographic data for this paper. These data can be
obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif,
or by emailing data_request@ccdc.cam.ac.uk, or by contacting
The Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
(22) Hassner, A.; Michelson, M. J. J. Org. Chem. 1962, 27, 298−301.
AUTHOR INFORMATION
Corresponding Author
*E-mail for D.A.V.: vicic@lehigh.edu.
■
ORCID
Aaron Sandleben: 0000-0002-4141-7406
Axel Klein: 0000-0003-0093-9619
David A. Vicic: 0000-0002-4990-0355
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
D.A.V. thanks the Office of Basic Energy Sciences of the U.S.
Department of Energy (DE-SC0009363) for support of this
work. A.K. and A.S. are grateful for support by Deutsche
Forschungsgemeinschaft DFG KL1194/15-1, N.V. acknowl-
edges the PROMI program of the University of Cologne.
REFERENCES
■
(1) Tanaka, H.; Nishibayashi, Y.; Yoshizawa, K. Acc. Chem. Res. 2016,
49, 987−995.
(2) McWilliams, S. F.; Holland, P. L. Acc. Chem. Res. 2015, 48, 2059−
2065.
(3) Hoffman, B. M.; Lukoyanov, D.; Yang, Z.-Y.; Dean, D. R.;
Seefeldt, L. C. Chem. Rev. 2014, 114 (8), 4041−4062.
(4) Diccianni, J. B.; Hu, C.; Diao, T. Angew. Chem., Int. Ed. 2016, 55,
7534−7538.
(5) Guo, Q.; Lu, Z. Synthesis 2017, 49, 3835−3847.
(6) Dai, G.; Yang, L.; Zhou, W. Org. Chem. Front. 2017, 4, 229−231.
(7) Chen, C.-y.; He, F.; Tang, G.; Ding, H.; Wang, Z.; Li, D.; Deng,
L.; Faessler, R. Eur. J. Org. Chem. 2017, 2017, 6604−6608.
(8) Tang, R.-Y.; Guo, X.-K.; Hu, M.; Wang, Z.-Q.; Xiang, J.-N.; Li, J.-
H. Synlett 2013, 25, 64−68.
E
DOI: 10.1021/acs.organomet.7b00887
Organometallics XXXX, XXX, XXX−XXX