Chemical Oxidation of a Coordinated PNP-Pincer Ligand Forms Unexpected Re-Nitroxide Complexes with Reversal of Nitride Reactivity

Article abstract of DOI:10.1021/acs.inorgchem.9b01075

Because of the thermodynamic demands of N₂ cleavage, N₂-derived nitride complexes are often unreactive. The development of multistep N₂ functionalization reactions hinges on methods for modulating nitride reactivity with s

Full text of DOI:10.1021/acs.inorgchem.9b01075

Article
pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
Chemical Oxidation of a Coordinated PNP-Pincer Ligand Forms
Unexpected Re−Nitroxide Complexes with Reversal of Nitride
Reactivity
Gannon P. Connor, Brandon Q. Mercado, Hannah M. C. Lant, James M. Mayer,*
and Patrick L. Holland*
Department of Chemistry, Yale University, New Haven, Connecticut 06511, United States
S
* Supporting Information
ABSTRACT: Because of the thermodynamic demands of N2 cleavage, N2-
derived nitride complexes are often unreactive. The development of
multistep N2 functionalization reactions hinges on methods for modulating
nitride reactivity with supporting ligands. Here, we describe the reactions of
N2-derived Re−nitride complexes, including the first Re nitrides supported
by a nitroxide-containing pincer ligand, and unusual examples of Re6+−
nitride complexes. The previously reported N2-derived complex (PNP)Re-
(N)(Cl) (PNP = N(CH2CH2PtBu2)2) can be oxidized by O atom transfer to
the backbone amide to form a novel nitroxide-pincer complex or by 1e− to
form a rare S = 1/2 Re6+−nitride complex. The Re−nitrido interaction in a
series of Re- and ligand-oxidized complexes is characterized using 15N NMR
spectroscopy, IR spectroscopy, and DFT calculations, and shows changes in
the Re−N bond order from both ligand- and metal-centered oxidations. Chemical oxidation of the supporting ligand to form a
nitroxide-pincer ligand results in subtle electronic changes at Re and a more electron-deficient nitride ligand. Combined ligand-
and metal-centered oxidation to form a Re6+−nitroxide complex results in a reversal of reactivity at the nitride ligand from
nucleophilic to electrophilic. These systematic electronic structure and reactivity studies demonstrate methods for inducing
reactivity in N2-derived nitride complexes.
Metal-bound nitrides are known to range from strongly
nucleophilic to strongly electrophilic,22−24 including some
complexes in which the nitride ligand can react in an
ambiphilic fashion.25 Here, we study the reactivity of
(PNP)Re(N)(Cl) (1, PNP = N(CH2CH2PtBu2)2), which
forms from the reductive cleavage of N2 by (PNP)ReCl2.8,11
Schneider has reported that complex 1 reacts with H+ at the
supporting ligand to protonate the backbone amide but reacts
with carbon electrophiles at the nitride ligand to form
alkylimido complexes, suggesting a fine balance between
nitride and supporting ligand reactivity.8,21 Thus, we were
interested to explore other electrophiles in efforts toward net
N2 functionalization to other products.
INTRODUCTION
■
A growing number of systems can reductively cleave dinitrogen
(N2) to form metal−nitride complexes.1−13 To take advantage
of these discoveries, it is now important to learn how to control
the reactivity of metal−nitride complexes. Functionalization of
N2-derived nitrides is particularly challenging because these
systems tend to have strong metal−nitride bonds that are
formed through thermodynamically favorable N2 splitting.
Despite this challenge, there are some examples where
reductive functionalization of N2-derived, metal−nitride
complexes have been reported,13−20 including some that
feature pincer-type supporting ligands.8−10,21 However, in
other cases, supporting ligands react preferentially to the
nitride ligand.5,8 These disparate results highlight the difficulty
of predicting the reactivity of N2-derived metal−nitrido
complexes and show the need for understanding how nitrides
respond to the supporting ligand. Understanding how
supporting ligands control reactivity at the nitride ligand of
these complexes is therefore key to developing systems that
can achieve both the difficult reductive cleavage of N2 as well
as subsequent functionalization of the nitride ligand to useful
products. In addition, chemists need to contend with the
selectivity of attack on the supporting ligand versus attack on
the nitride, as described below.
The Re6+/5+ potential in tetrahydrofuran (THF) for 1 is
−0.08 V vs Cp2Fe+/0,8 which suggests that its formally Re5+
center is surprisingly electron rich. We reasoned that 1 could
display reactions toward O atom donors that would form new
N−O bonds to the N2-derived nitride, giving interesting
functionalized products such as nitrosyl or nitro complexes.
The transformation of electrophilic nitride ligands to nitrosyl
ligands has been reported in metal−nitride complexes,26−30
and oxidation of nitrosyl to nitro ligands is also well
Received: April 12, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
precedented,31−35 but neither has been observed in nitrides
derived from N2. To assess these possibilities, as well as to
explore new methods to quantify and modulate the reactivity
of metal−nitrido complexes, we experimentally evaluated the
oxidative reactivity of 1 with O atom transfer reagents and 1e−
chemical oxidants. These studies not only serve to help expand
the understanding of supporting-ligand versus nitride-ligand
reactivity in systems that can cleave N2, but could also address
new pathways of transition-metal-mediated N2 functionaliza-
tion.
knowledge. Thus, this reaction represents a distinctive example
of chemical oxidation of a coordinated pincer ligand to form a
nitroxide ligand.
In contrast, 2 is not formed when nucleophilic O atom
transfer reagents are employed. For example, reaction of 1 with
excess trimethylamine-N-oxide, pyridine-N-oxide, or iodoso-
1
benzene gave no conversion to 2 by H and 31P{1H} NMR
spectroscopies, even after prolonged heating at 115 °C.
Instead, these reactions gave an intractable mixture of multiple
products, if any (see Figures S6−8). IR spectra of the crude
reaction products showed no evidence of nitrosyl formation.
These results suggest that oxidized products of these reactions
involve neither O atom transfer to the amide backbone nor the
nitride ligand. The reactivity of 1 with an oxygen electrophile
(mCPBA) at the backbone amide to form 2 differs from the
reactivity of 1 with carbon electrophiles, which react at the
nitride ligand to form a new N−C bond.8,21 However,
formation of 2 in this reaction is more reminiscent of ligand-
based reactivity of 1 with hydrochloric or triflic acid (HOTf)
to form cationic complexes with the amide protonated
(Scheme 1).8
RESULTS AND DISCUSSION
■
Synthesis and Characterization of a Novel Nitroxide-
Pincer Complex. Complex 1 reacts with 1.5 equiv of 3-
chloroperbenzoic acid (mCPBA) to give a new diamagnetic
complex 2 at ambient temperature in THF or diethyl ether
(Et2O). This reaction proceeds in >95% yield when monitored
by 1H and 31P{1H} NMR spectroscopy. The oxidation of 1 to
2 leads to a slight shift in the 31P resonance of the phosphine
arms from δ 84.6 ppm for 1 to δ 64.6 ppm for 2 in the 31P{1H}
NMR spectrum collected in THF-d8. The solid-state crystal
structure of 2 reveals that the oxidation occurs at the backbone
amide of the pincer ligand rather than at the nitride ligand,
forming a nitroxide moiety that binds η2 to the metal center
(Scheme 1). In the crystallographic structure of 2, the nitroxide
While the amide backbones in PNP-pincer ligands40−48 are
strong σ- and π-donors, it is not obvious how to view the
bonding nature of the new “PONP”-pincer ligand that features
a nitroxide backbone. The solid-state structure of 2 exhibits a
long Re−O interaction of 2.306(6) Å, and the Re−amide bond
length elongates slightly from 2.033(6) to 2.080(7) Å going
from 1 to 2, respectively (Figure 1). There is also a slight
Scheme 1. Oxidative Reactivity of Re−Nitride Complex 1
and Formation of Novel Nitroxide Complexes 2 and 4
O atom has a hydrogen-bond to a molecule of 3-chlorobenzoic
Figure 1. Solid-state structures of parent nitride complex 1 (left,
matches previously reported DFT structure8), Re5+−nitroxide
complex 2 (mCBA molecule omitted), Re6+−nitroxide complex 4
(PF6 ion omitted), and Re6+−nitride complex 5 (PF6 ion omitted)
with thermal ellipsoids at 50% probability. Hydrogen atoms are
omitted for clarity.
acid (mCBA) in the solid state (see Figure S39), and 2 is
1
isolated with one equivalent of mCBA as evident from the H
NMR spectrum of crystals of 2 (see Figure S1). The H atom
involved in the hydrogen-bond was not resolved by X-ray
crystallography, but DFT calculations, structural parameters,
and reduction potentials all suggest that the mCBA molecule is
protonated rather than the Re-bound nitroxide (see below).
Reaction of 1 with 1.5 equiv of H2O2 (as a 30% w/w solution
in H2O) also forms 2 in 23% yield according to 31P{1H} NMR
spectroscopy in THF-d8 (see Figure S5). Because of the low
spectroscopic yield, we did not attempt to isolate 2 from the
reaction of 1 with H2O2. Additionally, attempts to isolate 2
without a hydrogen-bonded mCBA equivalent were not
successful. Although numerous transition metal complexes
with nitroxide or hydroxylamine ligands have been re-
ported,36−38 only one other example of forming a nitroxide
ligand via oxidation of a coordinated amide ligand has been
reported with any transition metal.39 No examples of a
nitroxide coordinated to Re have been reported to our
shortening of the Re−Cl bond length from 2.441(2) Å to
2.409(2) Å upon oxidation of the supporting ligand, which is
consistent with a loss of electron density at the metal center.
Otherwise, the coordination geometry around Re deviates little
from 1 to 2 (Figure 2a). To explore the change in electron
density at Re, cyclic voltammetry (CV) experiments on 1 and
2 were performed in THF. These experiments show an anodic
shift of the Re6+/5+ couple of 80 mV from −0.08 V vs Cp2Fe+/0
for 1 to 0.00 V vs Cp2Fe+/0 for 2 (Figure 3).8 This represents
quite a small shift in electron-richness between 1 and 2 despite
chemical oxidation of the supporting ligand. This observation
agrees with the lack of change in the coordination geometry
B
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Selected Bond Lengths (Å) and Bond Angles (deg)
of Complexes 1, 2, 4, and 5
bond/angle
1
2
4
5
Re1−N1
Re1−N2
Re1−Cl1
Re1−O1
Re1−P1
Re1−P2
O1−N2
N1−Re1−N2
N1−Re1−Cl1
N2−Re1−Cl1
N1−Re1−P1
N1−Re1−P2
O1−N2−Re1
O1−Re1−N2
O1−Re1−Cl1
1.643(6)
2.033(6)
2.441(2)
1.638(7)
2.080(7)
2.409(2)
2.306(6)
2.443(2)
2.460(2)
1.436(9)
105.3(3)
105.3(2)
149.5(2)
96.1(3)
1.665(7)
2.097(8)
2.295(3)
2.200(6)
2.516(2)
2.507(2)
1.433(1)
105.0(4)
103.1(3)
151.9(2)
95.8(3)
1.650(5)
2.054(5)
2.339(2)
Figure 2. (a) Overlay of solid-state structures for Re5+−nitride (1,
gray) and Re5+−nitroxide (2, red). (b) Overlay of solid-state
structures for Re5+−nitroxide (2, red) and Re6+−nitroxide (4, blue).
2.443(2)
2.435(2)
2.476(2)
2.479(2)
105.8(3)
106.5(2)
147.7(2)
100.4(2)
99.9(2)
103.4(2)
105.6(2)
151.0(1)
98.9(2)
97.6(3)
79.7(4)
37.8(2)
111.7(2)
96.5(3)
74.5(4)
38.8(3)
113.1(2)
100.8(2)
Figure 3. Cyclic voltammograms of 1 (gray, dashed), 2 (red), and 3-
OTf (blue) demonstrating effect of functionalizing the backbone
amide on the Re6+/5+ redox potential. Details: 100 mV/s scan rate,
scanning anodically in 0.1 M tetrabutylammonium hexafluorophos-
phate solution in THF with glassy carbon disk working, Pt wire
auxiliary, and Ag wire pseudoreference electrodes. An additional redox
feature in the voltammogram of 2 may be from a minor adventitious
impurity.
around Re. For comparison, protonation of 1 with HOTf to
form the cationic amide-protonated complex [(PHNP)Re(N)-
(Cl)][OTf] (3-OTf) has a much larger effect, shifting the Re5+
to Re6+ reduction potential to 0.60 V vs Cp2Fe+/0. These data
show that there are only subtle electronic changes between 1
and 2, supporting the conclusion that the ligand in 2 is a
nitroxide anion with a hydrogen-bonded carboxylic acid, rather
than a hydroxylamine ligand. The small changes between 1 and
2 indicate that oxidation of the amide of a PNP-pincer complex
to a nitroxide may offer an opportunity to “fine-tune” the
electronics of the complex to improve reactivity without
changing its overall charge or requiring the synthesis of a
separate ligand.
To elucidate the changes in bonding interactions arising
from formation of the nitroxide, complexes 1 and 2 were
modeled using DFT calculations. Complex 2 was modeled
with and without a coordinated mCBA molecule, and the
presence of mCBA did not significantly affect any of the
computational results (see SI). Natural bond order (NBO)
analysis of 2 (without mCBA) identifies a weak Re−O
interaction with a Wiberg bond order49 of 0.22, which is
consistent with the long Re−O bond observed in the solid-
state structure (Figure 4). The Wiberg bond order of the Re−
amide decreases from 0.65 in 1 to 0.43 in 2, suggesting loss of
π-donation from the amide. Second-order perturbation theory
(SOPT) analysis50 identifies several donor−acceptor inter-
actions between the nitroxide and Re-containing molecular
orbitals. This analysis estimates the energy lowering in a
molecule achieved by donation of electrons in a filled (bonding
or lone pair) orbital into a vacant or antibonding orbital. The
dominant interaction identified by SOPT analysis is the
Figure 4. Calculated Wiberg bond orders of (a) Re5+−nitroxide
complex (2) and (b) Re6+−nitroxide complex (4) and (c)
representation of donor−acceptor interaction from O atom LP to
antibonding d−π* Re−nitride orbital. For the purpose of clarity, the
supporting ligand is truncated.
donation of lone pairs (LPs) on the O atom into antibonding
d−π* Re−nitride orbitals (Figure 4c). These interactions
replace analogous donation from the amide LP in 1 into the
same antibonding d−π* Re−nitride orbitals. However, the
donor−acceptor interactions from the O atom in 2 are
stronger than the analogous interactions from the amide in 1,
as evident by an approximately 30% increase in the calculated
exchange energies for these interactions in 2 versus 1 (see SI
for more details). Additionally, a weak donor−acceptor
interaction between the N−O σ-bond and the overlapping
antibonding Re−nitride orbital was identified in 2.
Overall, the change from 1 to 2 decreases the reactivity of
the complex. Nitroxide complex 2 does not react with mCPBA,
trimethylamine-N-oxide, pyridine-N-oxide, or iodosobenzene.
In addition, 2 does not react with potential nucleophiles PMe3
or PPh3. Thus, it is apparent that introduction of the nitroxide
moiety does not affect the strong Re−nitrido interaction
enough to increase reactivity at the nitride ligand.
Synthesis of Re6+−Nitride Complexes. The reversible
anodic waves observed in CV experiments of 2 indicate that
the (PONP)Re6+ complex could be accessible. Oxidation of 2
at −78 °C in dry CH2Cl2 with tris(4-bromo-phenyl)aminium
C
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 2. Re−15N Stretching Frequencies, Bond Lengths, and 15N NMR Chemical Shifts for Re−Nitride Complexes
a
complex
Re−15N stretch (cm−1
)
Re−N bond length (Å)
nitride 15N NMR chemical shift (ppm)
(PNP)Re(N)(Cl) (1-15N)
1037
1042
1075
1024
1035
1.643(6)
1.638(7)
1.624(7)
1.665(7)
1.650(5)
371
393
387
[(PONP)Re(N)(Cl)][mCBA] (2-15N)
[(PHNP)Re(N)(Cl)][OTf] (3-15N)
[(PONP)Re(N)(Cl)][PF6] (4-15N)
[(PNP)Re(N)(Cl)][PF6] (5-15N)
b
a
b
Chemical shift reported versus CH3NO2. This distance has been previously reported as 1.659(2) Å.3 We remeasured this distance as 1.624(7) Å
from a crystal with a different unit cell, and the shorter bond distance more closely corresponds to the high Re−N stretching frequency in this
complex, as well as its higher calculated Re−N Wiberg bond order.
hexafluorophosphate affords [(PONP)Re(N)(Cl)][PF6] (4), a
rare example of an isolable paramagnetic Re6+−nitride species.
Complex 4 shows several broad paramagnetic proton
subsequently prepared from 1-15N. Isotopic labeling of the
nitride ligand enables direct spectroscopic comparison of the
Re−nitride stretching frequencies in the IR spectra of these
complexes, and the 15N resonance for the nitride is easily
observed in the 15N NMR spectra of the diamagnetic Re5+
complexes 1-15N, 2-15N, and 3-15N. Additionally, the Re−
nitride bond lengths are available from the solid-state
structures of the complexes. The data in Table 2 show that
Re−nitride interaction varies throughout the series of
complexes, demonstrating small changes in Re−nitride bond
lengths and notable changes in Re−nitride stretching
frequencies. For example, the Re−nitride stretching frequency
ranges from 1024 cm−1 in 4-15N to 1075 cm−1 in 3-15N, a
difference of 51 cm−1. In general, we see that functionalization
of the backbone amide results in slightly higher Re−nitride
stretching frequencies, while oxidation of the metal center
results in slightly lower frequencies. However, the Re−nitride
bond length ranges only from 1.624(7) Å in 3-OTf to
1.665(7) Å in 4. Excluding the protonated complex 3-OTf, the
complexes adhere well to Badger’s rule, which predicts a
correlation between bond length and stretching frequency
(Figure 5a).54 It is worth noting that the backbone- and metal-
oxidized complex 4-15N demonstrates the lowest Re−nitride
stretching frequency, likely resulting from stronger donation
from O atom LPs into the antibonding d−π* Re−nitride
orbitals. Assuming that changes in stretching frequency parallel
changes in bond strength, this implies that 4-15N has the
weakest Re−nitride interaction.
1
resonances in its H NMR spectrum between δ 5.5 and δ
34.0 ppm in CD2Cl2 (see Figure S2), and its EPR spectrum is
reminiscent of the few other reported S = 1/2 Re6+−nitride
complexes (see Figure S12).51−53 The solid-state structure of 4
shows a shortening of the Re−O interaction by 0.1 Å to
2.200(6) Å (Figure 2b). The Re−nitride and Re−amide bond
distances remain similar between 2 and 4, but the Re−Cl bond
distance shortens by about 0.1 Å from 2.409(2) to 2.295(3) Å
upon oxidation of the Re metal center. NBO analysis of the
DFT model of 4 calculates a larger Wiberg bond order of 0.27
for the Re−O bond and diminished value of 0.41 for the Re−
amide bond in agreement with the shortening of the Re−O
bond and slight lengthening of the Re−amide bond,
respectively (Figure 4). SOPT analysis identifies the same
donor−acceptor interactions in 4 as in 2; however, oxidation of
the metal center in 4 results in a 60% increase in exchange
energies of the donation from O atom LPs into the
antibonding d−π* Re−nitride orbitals.
CV analysis of parent nitride 1 shows a reversible Re6+/5+
couple in dry CH2Cl2,8 suggesting that the PNP-Re6+ species is
accessible in this case as well. This redox event becomes less
reversible in dry acetonitrile and is only quasi-reversible in dry
THF, becoming more irreversible with added aliquots of H2O,
indicating the instability of the Re6+ oxidation product (see
Figure S11). Oxidizing 1 in dry CH2Cl2 at −78 °C with
[Cp2Fe][PF6] results in a color change from orange to red, and
the 1e− oxidized complex [(PNP)Re(N)(Cl)][PF6] (5) was
isolated from the reaction as a crystalline solid in 76% yield.
EPR spectroscopy of the isolated crystals provides a spectrum
The chemical shifts of the 15N nitride resonances in the 15
N
NMR spectra of 1-15N, 2-15N, and 3-15N suggest the relative
electron density of the nitride ligands, at least to a first
approximation. In organic compounds, 15N NMR chemical
shifts have been shown to be correlated to various chemical
properties, including Lewis basicity.55−59 We wondered
whether the 15N NMR chemical shifts of the nitride ligands
could thus be an estimate of the relative electron richness of
these ligands and, by extension, be used to predict reactivity.
Functionalization of the amide in 1-15N to form a nitroxide
similar to that of 4 (see Figure S13).51−53 However, H and
1
31P{1H} NMR spectroscopies indicate that the crystals contain
approximately 25% of a diamagnetic impurity, identified as
[(PHNP)Re(N)(Cl)][PF6] (3-PF6), where PHNP indicates the
pincer ligand with protonated nitrogen. Upon further
investigation, 5 was found to be extremely sensitive to
decomposition in the presence of traces of moisture, forming
3-PF6 as the major product (see Figures S4 and S11). Complex
3-PF6 is structurally and spectroscopically analogous to the
related triflate salt 3-OTf (see Figures S4 and S43).8
(2-15N) or protonated amide (3-15N) both shift the nitride 15
N
resonance downfield, suggesting a decrease in Lewis basicity.
This can be rationalized partially from first-principles, since
oxidation or protonation of the complex should give a more
electron-deficient Re center, which in turn should facilitate
stronger donation from the nitride ligand. However, the
changes in the nitride 15N chemical shift from 1-15N to 2-15N
(22 ppm) and from 1-15N to 3-15N (16 ppm) are very close.
Despite the similar 15N-nitride chemical shifts, there is a
significant change in Re6+/5+ reduction potential of +520 mV
for 3 versus 2. Indeed, 2, which has a more electron-rich Re
center than 3 based on their relative Re6+ reduction potentials,
is predicted to have the most electron-deficient nitride by 15N
Characterizing Re−Nitrido Interactions via 15N Label-
ing. With this series of Re−nitride complexes oxidized at the
supporting ligand or metal center, we sought to quantitatively
study the effects of these oxidations on the Re−nitrido bond
via isotope labeling experiments. The 15N-labeled parent
nitride (1-15N) was synthesized from reaction of (PNP)ReCl2
8
15
with terminally labeled [PPN][NN N], resulting in
50% labeling of the nitride ligand with 15N. The 50% 15N-
labeled nitride complexes 2-15N, 3-15N,8 4-15N, and 5-15N were
D
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 5. Re−nitride bond length vs 1/νRe−N2/3, with Badger’s rule
demonstrated as a linear fit for complexes 1, 2, 4, and 5. (b)
Calculated Wiberg bond order of Re−nitride bonds vs observed Re−
nitride stretching frequencies in complexes 1−5.
Figure 6. Calculated Wiberg bond order of Re−nitride bonds in
complexes 1−5.
spectrum of the solution shows two singlets in a 2:1 ratio at δ
71.6 and 54.6 ppm, respectively. Formation of this product is
accompanied by loss of the PPh3 signal, which is matched by
an analogous change in the PPh3 resonances in the
NMR spectroscopy, assuming that the chemical shifts show the
same correlations with electron density in 1−3 as in organic
compounds. This suggests that the electron-richness of the Re
center may not be the only factor affecting the electron-
richness of the nitride ligand. However, it is worth noting that
difference in charge between 3 and 1−2 may also play a large
role in the differences in redox potentials between the
complexes; furthermore, since 1−3 are generally unreactive
at the nitride, we could not experimentally test the Lewis
basicity trends hinted by the 15N NMR chemical shifts.
To supplement our spectroscopic studies, we compared the
DFT models of complexes 1−5. These models demonstrate
that oxidation of the amide to form a nitroxide has an effect on
the distribution and relative energy of the frontier molecular
orbitals. The perturbation of molecular orbitals by replacing
Re−amide interactions with Re−nitroxide interactions may be
responsible for the decreased electron-richness of the nitride
ligand in complex 2 versus complex 3. The redox trends
between 1, 2, and 3 are reproduced with DFT calculations
using these models, which validates the accuracy of the
computational method (see SI). NBO analysis of the series
predicts changes in the Re−nitride Wiberg bond orders that
match the experimentally observed trends in Re−nitride
stretching frequencies (Figure 5b and 6). These trends
indicate that oxidation of the Re metal center results in a
slight decrease in Re−nitride partial bond order, with 4 having
the lowest Re−nitride bond order of 2.38.
1
corresponding H NMR spectra. These data are consistent
with formation of a diamagnetic Re3+−phosphinimide
complex, which would require an additional 1e− reduction of
the complex formed from initial 2e− reduction of the nitride to
phosphinimide. To determine whether 6 is indeed a
phosphinimide complex, PPh3 was reacted with 4-15N. This
resulted in splitting of the 31P resonance at δ 54.6 ppm into a
doublet (J = 35.1 Hz), demonstrating that PPh3 binds to the
15N-labeled nitride (Figure 7, inset).
Formation of a phosphinimide from 4 with phosphines can
involve the nitride acting either as an electrophile24 or, less
commonly, a nucleophile.60 To assess the nature of the
transition state during the reaction of the nitride ligand in 4
with phosphines, we varied the phosphines with substituents of
differing electron richness. When 4 is mixed with stoichio-
metric tris(p-tolyl)phosphine (P(p-tolyl)3) or tris(4-
fluorophenyl)phosphine (P(4-FPh)3), formation of phosphini-
mide complexes 6-P(p-tolyl)3 and 6-(4-FPh)3, respectively, is
1
observed by H and 31P{1H} NMR spectroscopy (Figures
1
S22−S25). Since H and 31P{1H} NMR spectra indicate that
the phosphinimide complex is the first product observed from
reaction of 4 with phosphines, following the disappearance of
the distinct UV−vis absorbance feature of 4 with λmax = 668
nm allowed us to qualitatively compare the rates of
phosphinimide complex formation. To determine the kinetic
preference of 4 for phosphines of different electron-richness,
we compared the consumption of 4 in the presence of
Oxidation-Induced Change in Nitride Reactivity. At
ambient temperature in CD2Cl2, addition of 1.2 equiv of PPh3
to 4 leads to slow formation over the course of several hours of
a new product (Scheme 2, drawn as 6). The 31P{1H} NMR
E
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 2. Formation of Phosphinimide Complex 6 upon Addition of PPh3 to 4
Figure 7. 31P{1H} NMR spectrum of 4 with PPh3 in CD2Cl2 after 15 h showing formation of a phosphinimide intermediate 6. Inset: Difference in
splitting of 31P resonance at δ 54.6 ppm depending on N isotope used.
Scheme 3. Competition Reaction between P(p-tolyl)3 and P(4-FPh)3 with 4
P(4-FPh)3, PPh3, and P(p-tolyl)3 under pseudo-first order
conditions (10 equiv phosphine). Under identical reaction
conditions, 4 reacts more quickly in the presence of P(p-tolyl)3
than PPh3, which in turn reacts more quickly than P(4-FPh)3
(Figure S26). Since consumption of 4 corresponds to
formation of the phosphinimide complex 6, the rate of P−N
bond formation is more rapid for more nucleophilic
phosphines. This observation indicates that the nitride ligand
in 4 behaves as an electrophile. Furthermore, in a competition
experiment containing 1 equiv of both P(p-tolyl)3 and
P(4-FPh)3, 4 reacts preferentially with P(p-tolyl)3 to form 6-
P(p-tolyl)3 (Scheme 3), and no formation of 6-P(4-FPh)3 is
observed (Figures S27 and S28). The addition of excess P(p-
tolyl)3 to solution containing 6-P(4-FPh)3 results in conversion
of the phosphinimide complex to 6-P(p-tolyl)3 within 10 min,
suggesting a labile N−P interaction that favors the more
electron-rich phosphinimide (Figure S29). This is evidence of
a thermodynamic preference for formation of the more
electron-rich phosphinimide due to a stronger N−P inter-
action.
with carbon electrophiles.8,21 Importantly, umpolung, which
makes the nitride electrophilic, appears to require both
oxidation of the backbone to form a nitroxide and 1e−
oxidation of the metal center to form a Re6+ complex, as
neither 2 (just oxidized at ligand) nor 5 (just oxidized at Re)
react with phosphine nucleophiles (Figures S30 and 31). Since
2 and 5 do not react with phosphines, it suggests that neither
1e− reduction of 4 (to form 2) nor O atom abstraction from
the nitroxide in 4 (to form 5) occurs prior to phosphinimide
formation.
The phosphinimide product 6 is unstable under the reaction
conditions and decomposes to 3-PF6 (31P resonance at δ 69
1
ppm), a paramagnetic complex (which resembles 5 by H
NMR spectroscopy), and OPPh3, among other unidentified
phosphorus-containing products. The concentration of 6
during the reaction reached a maximum 15 h into the reaction
at approximately 31% spectroscopic yield by integration versus
an internal standard under these conditions (Figure 7). The
instability of 6 has precluded its isolation as a pure material.
The formation of 3-PF6, the paramagnetic species, and OPPh3
is observed only after 6 is formed, which suggests that OPPh3
is formed by net O atom abstraction from the nitroxide to a
PPh3 (see SI). This could either be from the phosphinimide or
free PPh3 in the reaction. However, in phosphinimide-forming
reactions that contain both P(p-tolyl)3 and P(4-FPh)3,
phosphine oxides of both phosphines are observed as
decomposition products (see Figure S28). The observation
It is significant that only complex 4, in which both the
backbone amide and metal center of 1 have been oxidized, was
found to react with phosphines to form transiently observed
phosphinimide products. This demonstrates that the nitride in
4 is electrophilic enough to undergo reaction with
nucleophiles, representing a reversal of reactivity at the nitride
ligand relative to 1, wherein the nitride reacts as a nucleophile
F
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 8. Top: DFT model of complex 4 and visualizations of LUMO and LUMO+1 molecular orbitals (isosurface value = 0.05). Bottom:
Drawings of the LUMO and LUMO+1 orbitals for 1−5 with atomic contributions of the nitride (blue percentages) showing that 4 demonstrates
the highest combined nitride contribution to the lowest unoccupied frontier orbitals (energy difference between molecular orbitals denoted in
italics).
of OP(4-FPh)3 indicates that one mechanism of decomposition
is likely the abstraction of the nitroxide O atom by free
phosphine in solution, since 6-P(4-FPh)3 is not observed
under these conditions.
DFT analysis of the atomic contributions to the lowest-
unoccupied molecular orbitals (LUMOs) of 1−5 provides a
rationalization for this reversal in reactivity (see Tables S15−
19). In complex 4, the LUMO and LUMO+1 are more
localized at the nitride ligand than in the rest of the series,
which is consistent with increased electrophilic behavior at the
nitride (Figure 8). Thus, the cooperative effect of ligand-
centered and metal-centered oxidations has a notable effect on
nitride reactivity despite the strong Re−nitrido interaction.
coupled with 1e− oxidation of the metal center, cause the
nitride ligand to become electrophilic. This umpolung in
reactivity at the nitride ligand is achieved without exchanging
any of the supporting ligands, potentially offering a new
strategy for “fine-tuning” metal-center electronics via oxidation
of a supporting amide ligand to a nitroxide.
EXPERIMENTAL SECTION
■
General Considerations. All manipulations were performed
under an inert atmosphere of N2 gas in a M. Braun glovebox or on a
Schlenk line unless otherwise specified. Tetrahydrofuran (THF) was
distilled under argon from potassium benzophenone ketyl and stored
over molecular sieves prior to use. Unless otherwise noted, all other
solvents were dried via passage through Q5 columns from Glass
Contour Co., and stored over molecular sieves prior to use.
Deuterated solvents were degassed and dried over calcium hydride
before storing over molecular sieves prior to use, and THF-d8 was
dried additionally with potassium benzophenone ketyl. Ammonium
perrhenate (Strem, 99+%), bis-[2-(ditertbutylphosphino)ethyl)]-
amine (Strem, 10% w/w in hexanes), hydrogen peroxide (Fisher
Chemicals, 30% in H2O), trimethylamine-N-oxide (Sigma-Aldrich,
98%), pyridine-N-oxide (Sigma-Aldrich, 95%), iodosobenzene (TCI,
95+%), ferrocenium hexafluorophosphate (Santa Cruz Biotechnology,
97%), nitrosonium hexafluorophosphate (Alfa Aesar, 96%), tris(4-
bromophenyl)amine (Sigma-Aldrich, 98%), bis(triphenylphosphine)-
iminium chloride (Strem, 97%), sodium azide (terminal 15N-labeled,
Cambridge Isotope Laboratories, 98+%), trifluoromethanesulfonic
acid (Sigma-Aldrich, 99+%), and trimethylphosphine (Sigma-Aldrich,
1.0 M in THF) were purchased and used without further purification.
Triethylamine (Sigma-Aldrich, 99+%, distilled), triphenylphosphine
(Aldrich, 99%, sublimed), tris(p-tolyl)phosphine (Aldrich, 98%, dried
under high vacuum), tris(4-fluoro-phenyl)phosphine (Strem, 99%,
dried under high vacuum), ferrocene (Aldrich, 98%, recrystallized),
tetrabutylammonium hexafluorophosphate (Sigma-Aldrich, 99%,
recrystallized ×3), and m-chloroperbenzoic acid (Aldrich, < 77%,
washed with pH 8.0 buffer and extracted with Et2O) were purified
prior to use. ReOCl3(PPh3)2,61 (ReCl3(MeCN)(PPh3)2,62 (PNP)-
CONCLUSIONS
■
This work describes the oxidative reactivity of an N2-derived
metal−nitrido complex, wherein two kinds of oxidations have
been observed. Outer-sphere agents oxidize the metal center,
forming unusual Re6+−nitrido complexes. Electrophilic O
atom transfer agents, however, transfer O to the amide portion
of the pincer ligand to form a nitroxide. Cooperative changes
in oxidation state and ligand oxidation are required to induce
electrophilic reactivity in the normally unreactive nitride
complex. Our studies also uncover the unusual formation of
Re complexes supported by nitroxide-containing pincer ligands
that provide a new variant on a well-characterized PNP pincer.
The O atom transfer to the supporting ligand suggests that
careful consideration of the supporting ligand is required for
conversion of N2 to oxidized N species.
Formation of the nitroxide moiety causes electronic changes
at the metal center by replacing amide LP interactions with O
atom LP interactions, resulting in a redistribution of molecular
orbitals. These changes have a surprisingly small effect on the
strong Re−nitride interaction in these complexes but, when
G
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
ReCl2,8 and bis(triphenylphosphine)iminium azide63 were synthe-
sized and purified according to literature procedures.
(m), 846 (s), 808 (s), 750 (s), 710 (s), 698 (m), 675 (m), 651 (m),
616 (m), 592 (w), 572 (w), 529 (m), 484 (m), 427 (m). Anal. Calcd
(found) for C27H49Cl2N2O3P2Re (%): C, 42.18(43.04); H,
6.42(6.43); N, 3.64(3.33). We were unable to obtain elemental
analysis with closer agreement in carbon.
Syntheses. 15N-Labeled Bis(triphenylphosphine)iminium Azide.
Following a literature procedure,64 200 mg Na(NN N) (3.03 ×
15
10−3 mol, 1 equiv) and 1.74 g (PPN)Cl (3.03 × 10−3 mol, 1 equiv)
were dissolved in 50 mL of EtOH and stirred at ambient temperature
for 6 h. The resulting mixture was filtered to remove NaCl, and the
precipitate was washed with 15 mL of EtOH. The filtrate was
concentrated to dryness under vacuum to give a white solid. The
residue was triturated with 3 × 10 mL Et2O to give a white powder,
which was dried in the vacuum oven at 120 °C for 24 h. Caution:
Azide salts can be explosive. Use extreme caution and allow the azide to
cool to ambient temperature before handling after or during drying. We
isolated 1.56 g of fine white powder (88% yield). IR stretches
matched reported values.64
[(PONP)Re(15N)(Cl)][mCBA] (2-15N). The synthesis of 2-15N was
completed analogously to the synthesis of 2 using 1-15N as the starting
1
material. H and 31P{1H} NMR spectra were identical to those of 2.
15N NMR (51 MHz, C6D6): δ 393 (s). FTIR (solid, cm−1): matches 2
except for a new band at 1042 cm−1 (see Figure S16).
[(PHNP)Re(N)(Cl)][OTf] (3-OTf). Compound 3-OTf was synthesized
from 1 according to a literature procedure.8 1H and 31P{1H} NMR
spectra matched previously reported data. FTIR (solid, cm−1): 3048
(br), 2987 (m), 2967 (m), 2911 (m), 2874 (m), 1480 (m), 1472 (m),
1395 (m), 1371 (m), 1334 (m), 1289 (s), 1241 (s), 1224 (s), 1154
(s), 1108 (m), 1027 (s), 990 (m), 940 (w), 824 (m), 809 (m), 776
(w), 757 (w), 742 (w), 727 (w), 691 (m), 636 (s), 613 (m), 575 (m),
539 (w), 517 (m), 497 (m), 435 (m), 426 (m).
(PNP)Re(N)(Cl) (1). Compound 1 was synthesized using a modified
literature procedure.8 Crystals suitable for X-ray diffraction were
1
grown from a concentrated solution of 1 in pentane at −40 °C. H
and 31P{1H} NMR spectra matched previously reported data. FTIR
(solid, cm−1): 2983 (m), 2942 (m), 2922 (m), 2897 (m), 2861 (m),
2814 (m), 1467 (m), 1390 (m), 1368 (m), 1326 (m), 1281 (w), 1261
(w), 1213 (m), 1180 (m), 1161 (m), 1092 (m), 1083 (m), 1056 (m),
1020 (m), 966 (m), 939 (m), 918 (m), 874 (w), 809 (s), 116 (m),
102 (m), 612 (m), 584 (m), 544 (m), 503 (m), 478 (s), 424 (m).
(PNP)Re(15N)(Cl) (1-15N). This compound was synthesized
analogous to 1 via a modified literature procedure,8 in which 20.0
mg of (PNP)ReCl2 (3.24 × 10−5 mol, 1.00 equiv) dissolved in 5 mL
[(PHNP)Re(15N)(Cl)][OTf] (3-15N). The synthesis of 3-15N was
completed analogously to the synthesis of 3 using 1-15N as the
starting material. 1H and 31P{1H} NMR spectra were identical to
those of 3. 15N NMR (51 MHz, C6D6): δ 387 (s). FTIR (solid,
cm−1): matches 3 except for new bands at 1037 and 1074 cm−1 (see
Figure S17).
Tris(4-bromophenyl)aminium Hexafluorophosphate. This com-
pound was synthesized according to a modified literature procedure,65
in which 1.00 g of tris(4-bromophenyl)amine (2.08 × 10−3 mol, 1.06
equiv) was dissolved in 8 mL of CH2Cl2 and added via pipet to a
slurry of 345 mg [NO][PF6] (1.97 × 10−3 mol, 1.00 equiv) in 30 mL
of CH2Cl2 at −78 °C while stirring vigorously. Addition of the amine
solution was accompanied by a distinct color change to deep blue.
The reaction was stirred at −78 °C for 1 h, then allowed to warm to
ambient temperature. Removal of the solvent under vacuum gave a
dark indigo solid. This residue was slurried in 15 mL of Et2O and
filtered. The solid was washed with 2 × 10 mL of Et2O and dried
under vacuum for 30 min, resulting in 1.11 g of the compound as an
indigo-purple solid (89% yield).
15
of THF was added to 20.2 mg (PPN)(NN N) (3.47 × 10−5
mol, 1.07 equiv) slurried in 5 mL of THF at −40 °C. The reaction
was allowed to warm to ambient temperature and stirred for 12 h,
accompanied by a color change from purple to orange and formation
of a white precipitate. The solvent was removed under vacuum, and
the product was extracted with 2 × 5 mL of Et2O and filtered through
Celite. Removal of the solvent under vacuum gave an orange residue.
To further purify the product, the residue was extracted with 4 × 5
mL of pentane and filtered through Celite. The combined filtrates
1
were evaporated to give 14.7 mg of orange powder (76% yield). H
and 31P{1H} NMR spectra were identical to those of 1. 15N NMR (51
MHz, C6D6): δ 371 (s). FTIR (solid): matches 1 except for new
bands at 1036 and 1074 cm−1 (see Figure S15).
[(PONP)Re(N)(Cl)][PF6] (4). In separate vials, 29.9 mg of 2 (3.89 ×
10−5 mol, 1.0 equiv) was dissolved in 3 mL of CH2Cl2 to give a yellow
solution, and 23.0 mg tris(4-bromo-phenyl)aminium hexafluorophos-
phate (3.67 × 10−5 mol, 0.94 equiv) was dissolved in 5 mL of CH2Cl2
to give a dark blue solution. Both solutions were cooled to −78 °C,
and the [Re] solution was added to the aminium dropwise via pipet
while stirring. The reaction was stirred at −78 °C for 1 h, after which
it had become bright green. The reaction was allowed to warm to
ambient temperature, and the solvent was removed under vacuum to
give a bright green residue. The product was washed with 4 × 5 mL of
Et2O, redissolved in 4 mL of CH2Cl2, layered with 12 mL of Et2O,
and left at −40 °C overnight. Decanting the mother liquor and
evaporation of the residual solvent under vacuum gave 25.0 mg of 4 as
green feathery crystals (85% yield). The 1H NMR spectrum of 4
demonstrates several broad paramagnetic resonances (Figure S2). 1H
NMR (500 MHz, CD2Cl2): δ 34.00 (2H, CH2), δ 12.12 (18H,
P(tBu)(tBu’)), δ 5.58 (20H, superimposed CH2 and P(tBu)(tBu′)).
UV−vis (CH2Cl2 solution): λ = 307 nm (ε = 5000 M−1 cm−1), 350
nm (shoulder, ε = 2800 M−1 cm−1), 417 nm (shoulder, ε = 580 M−1
cm−1), 668 nm (ε = 390 M−1 cm−1). FTIR (solid, cm−1): 2971 (m),
2939 (m), 2907 (m), 2873 (m), 1478 (m), 1468 (m), 1445 (w), 1397
(m), 1374 (m), 1245 (w), 1209 (w), 1175 (m), 1058 (m), 1023 (m),
992 (w), 936 (w), 874 (m), 832 (s), 740 (m), 709 (m), 612 (w), 589
(w), 557 (m), 485 (m), 424 (m). Anal. Calcd (found) for
C20H44ClF6N2OP3Re (%): C, 31.73(31.75); H, 5.86(5.86); N,
3.70(3.65).
[(PONP)Re(N)(Cl)][mCBA] (2). In separate 20 mL vials, 105.6 mg 1
(1.77 × 10−4 mol, 1.0 equiv) was suspended in 5 mL of Et2O to form
an orange slurry, and 45.8 mg of m-chloroperbenzoic acid (2.65 ×
10−4 mol, 1.5 equiv) was dissolved in 3 mL of Et2O to give a colorless
solution. The perbenzoic acid solution was quantitatively added to 1
dropwise via pipet while stirring at ambient temperature, resulting in a
color change to a gold solution. The reaction was stirred at ambient
temperature for 1 h, after which the reaction was filtered through
Celite to remove a brown precipitate. The gold filtrate was
concentrated under vacuum to give a sticky yellow solid. This solid
was triturated and washed with 4 × 3 mL pentane, then dried under
vacuum to yield 114.5 mg 2 as a yellow powder (84% yield), which
was pure by 1H and 31P{1H} NMR spectroscopy. Crystals suitable for
X-ray diffraction were grown from a concentrated solution of 2 in
1
Et2O at −40 °C. H NMR (400 MHz, THF-d8): δ 12.03 (br s, 1H,
OH), δ 7.98 (s, 1H, o-H of mCBA), δ 7.92 (d, J = 7.7 Hz, 1H, o-H of
mCBA), 7.56 (d, J = 8.0 Hz, 1H, p-H of mCBA), 7.43 (t, J = 7.9 Hz,
1H, m-H of mCBA), δ 3.39 (m, 2H, N(CH2CH2)2), δ 3.32
(superimposed m, 2H, N(CH2CH2)2), δ 2.07 (m, 2H, N-
(CH2CH2)2), δ 1.58 (superimposed m, 2H, N(CH2CH2)2), δ 1.54
(vt, 18H, P(tBu)(tBu’)), δ 1.34 (vt, 18H, P(tBu)(tBu′)). 13C{1H}
NMR (126 MHz, THF-d8): δ 165.55, 133.95, 133.20, 132.18, 129.67,
129.31, 127.74, 66.37, 36.85, 34.64, 28.32, 24.29, 18.99. 31P{1H}
NMR (162 MHz, THF-d8): δ 64.60 (s). UV−vis (CH2Cl2 solution):
λ = 332 nm (ε = 1300 M−1 cm−1), 393 nm (shoulder, ε = 290 M−1
cm−1), 458 nm (shoulder, ε = 160 M−1 cm−1). FTIR (solid, cm−1):
3073 (w), 2997 (m), 2981 (m), 2954 (m), 2922 (m), 2899 (m), 2868
(m), 1731 (br), 1607 (m), 1574 (m), 1470 (m), 1429 (m), 1394 (m),
1367 (m), 1278 (m), 1263 (m), 1248 (m), 1210 (m), 1179 (m),
1136 (m), 1077 (s), 1061 (m), 1026 (m), 990 (m), 939 (m), 911
[(PONP)Re(15N)(Cl)][PF6] (4-15N). Synthesis of 4-15N was completed
analogously to the synthesis of 4 using 2-15N as the starting material.
FTIR (solid, cm−1): matches 4 except for a new band at 1024 cm−1
(see Figure S18).
[(PNP)Re(N)(Cl)][PF6] (5). In a 20 mL vial, 30.0 mg of 1 (5.03 ×
10−5 mol, 1.0 equiv) was dissolved in 5 mL of CH2Cl2 to give an
orange solution. The solution was cooled to −78 °C, at which point
H
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
15.8 mg of [Cp2Fe][PF6] (4.78 × 10−5 mol, 0.95 equiv) was slowly
added as a solid while stirring. The solution darkened from orange to
bright red. The solvent was removed under vacuum while the solution
warmed to ambient temperature. The resulting residue was washed
with 3 × 5 mL cold (−40 °C) Et2O to remove Cp2Fe and residual 1,
leaving behind a pink-red solid. The pink-red solid was dissolved in 3
mL of cold (−40 °C) CH2Cl2 to give a red solution, which was
layered with 10 mL of cold (−40 °C) Et2O and placed in a −40 °C
freezer overnight. Decanting the mother liquor and evaporating
residual solvent under vacuum gave 28.3 mg of pink crystals suitable
for X-ray diffraction. 1H NMR spectroscopy of the pink crystals
showed that 5 was isolated with 26% 3-PF6 as an impurity, giving an
overall yield of 56% for 5. The 1H NMR spectrum of 5 demonstrates
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.9b01075.
Analytical data and spectra for novel complexes,
electrochemical data, IR spectra, NMR spectra of
reactions, crystallographic data for 1−5, DFT methods,
DFT-optimized structures, orbital decomposition, NBO,
and SOPT analysis (PDF)
1
several broad paramagnetic resonances (Figure S3). H NMR (500
Accession Codes
MHz, CD2Cl2): δ 6.91 (18H, P(tBu)(tBu′)), δ 4.39 (18H,
P(tBu)(tBu′)), δ 4.33 (2H, superimposed CH2), δ −0.52 (2H,
CH2). UV−vis (CH2Cl2 solution with impurity of 3-PF6): λ = 291 nm
(shoulder, ε = 2500 M−1 cm−1), 325 nm (shoulder, ε = 1700 M−1
cm−1), 516 nm (ε = 290 M−1 cm−1), 627 nm (ε = 90 M−1 cm−1).
FTIR (solid with impurity of 3-PF6, cm−1): 3208 (w), 2987 (m),
2969 (m), 2908 (m), 2874 (m), 1480 (m), 1471 (m), 1397 (m),
1372 (m), 1327 (m), 1290 (w), 1281 (w) (m), 1218 (m), 1178 (m),
1109 (w), 1076 (m), 1052 (m), 1021 (m), 988 (w), 968 (w), 943
(w), 832 (s), 782 (s), 740 (m), 716 (m), 693 (m), 611 (m), 583 (w),
557 (s), 519 (m), 479 (s), 443 (m), 422 (m). Anal. Calcd (found) for
C20H44ClF6N2P3Re (%): C, 32.41(33.19); H, 5.98(5.99); N,
3.78(3.61). We were unable to obtain elemental analysis with closer
agreement in carbon.
CCDC 1893092−1893097 contain the supplementary crys-
tallographic 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.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: patrick.holland@yale.edu.
*E-mail: james.mayer@yale.edu.
ORCID
[(PNP)Re(15N)(Cl)][PF6] (5-15N). Synthesis of 5-15N was completed
analogously to synthesis of 5 using 1-15N as the starting material.
FTIR (solid, cm−1): matches 5 except for a new band at 1035 cm−1
(see Figure S19).
Gannon P. Connor: 0000-0002-9660-5666
James M. Mayer: 0000-0002-3943-5250
Patrick L. Holland: 0000-0002-2883-2031
Instrumentation and Methods. NMR spectra were acquired on
Notes
an Agilent 400 or 500 MHz spectrometer. 1H spectra were referenced
The authors declare no competing financial interest.
1
to residual H signals from the deuterated solvent with which the
sample was prepared,66 and 13C{1H}, 31P{1H}, and 15N spectra were
1
ACKNOWLEDGMENTS
absolute referenced to the corresponding H spectra. Line fitting of
■
paramagnetic resonances in 1H NMR spectra was done using
MestReNova 10.0.1. IR spectra were obtained using a Bruker Alpha
spectrometer containing a diamond ATR unit with 2 cm−1 resolution.
Electrochemical measurements were collected using a CH
Instruments 600D potentiostat in a N2 glovebox or in solutions
purged with dry N2 gas. Cyclic voltammetry (CV) experiments were
conducted in 0.1 M tetrabutylammonium hexafluorophosphate (TBA-
PF6) solutions in dry solvent using a glassy carbon disc working
electrode (3.0 mm diameter), a platinum wire auxiliary electrode, and
a silver wire pseudoreference. Wire electrodes were sanded and rinsed
with dry solvent prior to use, and glassy carbon disc electrodes were
polished using 0.05 μm alumina powder and rinsed with deionized
H2O, then dry solvent prior to use. Measurements conducted in THF
or CH2Cl2 were compensated for internal resistance prior to
collecting data. CV experiments were referenced to ferrocene
(Cp2Fe) as an internal standard after completion.
The authors thank Professors Sven Schneider, Alex Miller, and
Alan Goldman and their research groups for support and
valuable discussions. G.P.C. gratefully acknowledges H. Lant
and S. Lee for assistance with EPR experiments, as well as S.
Kolmar, A. Speelman, D. Kim, and J. Agarwal for assistance
with DFT studies. This work was funded by the U.S. National
Science Foundation (CHE-1665135, with initial support from
CHE-1205189 via the Center for Enabling New Technologies
through Catalysis).
REFERENCES
■
(1) Klopsch, I.; Yuzik-Klimova, E. Y.; Schneider, S., Functionaliza-
tion of N2 by Mid to Late Transition Metals via N−N Bond Cleavage.
In Nitrogen Fixation; Nishibayashi, Y., Ed.; Springer International
Publishing: Cham, 2017; pp 71−112.
Elemental analyses were obtained from the CENTC Elemental
Analysis Facility at the University of Rochester. Microanalysis samples
were weighed on a PerkinElmer Model AD-6 Autobalance, analyzed
on a PerkinElmer 2400 Series II Analyzer, and handled in a VAC
Atmospheres argon glovebox.
Differential EPR spectra were measured in derivative mode on a
Bruker ELEXSYS E500 spectrometer utilizing a superhigh Q
resonator and an Oxford ESR-900 helium flow cryostat at 8−11 K.
Instrument parameters: microwave frequency, 9.37 GHz; microwave
power, 10 mW; modulation frequency, 100 kHz; modulation
amplitude, 19.49 G; conversion time, 5.12 ms; time constant,
1.28−2.56 ms. Samples of approximately 0.2 mM were dissolved in
mixtures of 1:1 CH2Cl2: toluene and frozen at 77 K. Simulations were
performed using the “pepper” function in EasySpin, version 5.0.5.
Line broadening was simulated using gStrain and AStrain to account
for Gaussian distributions in g principal values and corresponding A
values.
(2) Laplaza, C. E.; Cummins, C. C. Dinitrogen Cleavage by a Three-
Coordinate Molybdenum(III) Complex. Science 1995, 268, 861.
(3) Laplaza, C. E.; Johnson, M. J. A.; Peters, J.; Odom, A. L.; Kim,
E.; Cummins, C. C.; George, G. N.; Pickering, I. J. Dinitrogen
Cleavage by Three-Coordinate Molybdenum(III) Complexes: Mech-
anistic and Structural Data. J. Am. Chem. Soc. 1996, 118, 8623−8638.
(4) Bezdek, M. J.; Chirik, P. J. Expanding Boundaries: N2 Cleavage
and Functionalization beyond Early Transition Metals. Angew. Chem.,
Int. Ed. 2016, 55, 7892−7896.
(5) Hebden, T. J.; Schrock, R. R.; Takase, M. K.; Mueller, P.
Cleavage of Dinitrogen to Yield a (t-BuPOCOP)molybdenum(IV)
Nitride. Chem. Commun. 2012, 48, 1851−1853.
(6) Silantyev, G. A.; Foerster, M.; Schluschass, B.; Abbenseth, J.;
Wuertele, C.; Volkmann, C.; Holthausen, M. C.; Schneider, S.
Dinitrogen Splitting Coupled to Protonation. Angew. Chem., Int. Ed.
2017, 56, 5872−5876.
I
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(7) Ishida, Y.; Kawaguchi, H. Nitrogen Atom Transfer from a
Dinitrogen-derived Vanadium Nitride Complex to Carbon Monoxide
and Isocyanide. J. Am. Chem. Soc. 2014, 136, 16990−3.
(8) Klopsch, I.; Finger, M.; Wuertele, C.; Milde, B.; Werz, D. B.;
Schneider, S. Dinitrogen Splitting and Functionalization in the
Coordination Sphere of Rhenium. J. Am. Chem. Soc. 2014, 136,
6881−6883.
(9) Liao, Q.; Cavaille, A.; Saffon-Merceron, N.; Mezailles, N. Direct
Synthesis of Silylamine from N2 and a Silane: Mediated by a
Tridentate Phosphine Molybdenum Fragment. Angew. Chem., Int. Ed.
2016, 55, 11212−11216.
(27) Coia, G. M.; Demadis, K. D.; Meyer, T. J. Oxidation of
Ammonia in Osmium Polypyridyl Complexes. Inorg. Chem. 2000, 39,
2212−2223.
(28) Crevier, T. J.; Lovell, S.; Mayer, J. M.; Rheingold, A. L.; Guzei,
I. A. Chalcogen Atom Transfer to a Metal Nitrido. The First
Transition Metal Selenonitrosyl Complex. J. Am. Chem. Soc. 1998,
120, 6607−6608.
(29) Walstrom, A.; Pink, M.; Fan, H.; Tomaszewski, J.; Caulton, K.
G. Radical (NO) and Nonradical (N2O) Reagents Convert a
Ruthenium(IV) Nitride to the Same Nitrosyl Complex. Inorg.
Chem. 2007, 46, 7704−7706.
(10) Arashiba, K.; Eizawa, A.; Tanaka, H.; Nakajima, K.; Yoshizawa,
K.; Nishibayashi, Y. Catalytic Nitrogen Fixation via Direct Cleavage of
Nitrogen−Nitrogen Triple Bond of Molecular Dinitrogen under
Ambient Reaction Conditions. Bull. Chem. Soc. Jpn. 2017, 90, 1111−
1118.
(30) Scheibel, M. G.; Askevold, B.; Heinemann, F. W.; Reijerse, E. J.;
de Bruin, B.; Schneider, S. Closed-shell and Open-shell Square-planar
Iridium Nitrido Complexes. Nat. Chem. 2012, 4, 552−558.
(31) Clarkson, S. G.; Basolo, F. Reaction of Some Cobalt Nitrosyl
Complexes with Oxygen. Inorg. Chem. 1973, 12, 1528−34.
(32) Ugo, R.; Bhaduri, S.; Johnson, B. F. G.; Khair, A.; Pickard, A.;
Benn-Taarit, Y. Oxygenation of the Nitrosyl Ligand in Complexes of
Nickel and Platinum. J. Chem. Soc., Chem. Commun. 1976, 694−5.
(33) Tovrog, B. S.; Diamond, S. E.; Mares, F. Oxygen Transfer from
Ligands: Cobalt Nitro Complexes as Oxygenation Catalysts. J. Am.
Chem. Soc. 1979, 101, 270−2.
(11) Lindley, B. M.; van Alten, R. S.; Finger, M.; Schendzielorz, F.;
Wurtele, C.; Miller, A. J. M.; Siewert, I.; Schneider, S. Mechanism of
̈
Chemical and Electrochemical N2 Splitting by a Rhenium Pincer
Complex. J. Am. Chem. Soc. 2018, 140, 7922−7935.
(12) Rodriguez, M. M.; Bill, E.; Brennessel, W. W.; Holland, P. L.
Reduction and Hydrogenation to Ammonia by a Molecular Iron-
Potassium Complex. Science 2011, 334, 780−783.
(34) Pipes, D. W.; Meyer, T. J. Comparisons Between Polypyridyl
Nitrosyl Complexes of Osmium(II) and Ruthenium(II). Inorg. Chem.
1984, 23, 2466−72.
(13) Miyazaki, T.; Tanaka, H.; Tanabe, Y.; Yuki, M.; Nakajima, K.;
Yoshizawa, K.; Nishibayashi, Y. Cleavage and Formation of Molecular
Dinitrogen in a Single System Assisted by Molybdenum Complexes
Bearing Ferrocenyldiphosphine. Angew. Chem., Int. Ed. 2014, 53,
11488−11492.
(35) Fernandez, B. O.; Lorkovic, I. M.; Ford, P. C. Mechanisms of
Ferriheme Reduction by Nitric Oxide: Nitrite and General Base
Catalysis. Inorg. Chem. 2004, 43, 5393−5402.
(36) Zhang, J.-Y.; Liu, C.-M.; Zhang, D.-Q.; Gao, S.; Zhu, D.-B.
Syntheses, Crystal Structures, and Magnetic Properties of two Cyclic
Dimer M2L2 Complexes Constructed from a New Nitronyl Nitroxide
Ligand and M(hfac)2 (M = Cu2+, Mn2+). Inorg. Chim. Acta 2007, 360,
3553−3559.
(37) Okamura, Y.; Ishii, N.; Nogami, T.; Ishida, T. Hard Magnets
after Freezing of Spin Dynamics of Soft Magnets in Cobalt(II)-radical
Chain Compounds. Bull. Chem. Soc. Jpn. 2010, 83, 716−725.
(38) Boreen, M. A.; Bogart, J. A.; Carroll, P. J.; Schelter, E. J.
Rearrangement in a Tripodal Nitroxide Ligand To Modulate the
Reactivity of a Ti-F Bond. Inorg. Chem. 2015, 54, 9588−9593.
(39) Chen, S.-J.; Zhang, X.-H.; Yu, X.; Qiu, H.; Yap, G. P. A.; Guzei,
I. A.; Lin, Z.; Wu, Y.-D.; Xue, Z.-L. Reaction of Ta(NMe2)5 with O2:
Formation of Aminooxy and Unusual (Aminomethyl)amide Oxo
Complexes and Theoretical Studies of the Mechanistic Pathways. J.
Am. Chem. Soc. 2007, 129, 14408−14421.
(40) Graham, T. W.; Tsang, C.-W.; Chen, X.; Guo, R.; Jia, W.; Lu,
S.-M.; Sui-Seng, C.; Ewart, C. B.; Lough, A.; Amoroso, D.; Abdur-
Rashid, K. Catalytic Solvolysis of Ammonia Borane. Angew. Chem., Int.
Ed. 2010, 49, 8708−8711.
(41) Nielsen, M.; Kammer, A.; Cozzula, D.; Junge, H.; Gladiali, S.;
Beller, M. Efficient Hydrogen Production from Alcohols under Mild
Reaction Conditions. Angew. Chem., Int. Ed. 2011, 50, 9593−9597.
(42) Han, Z.; Rong, L.; Wu, J.; Zhang, L.; Wang, Z.; Ding, K.
Catalytic Hydrogenation of Cyclic Carbonates: A Practical Approach
from CO2 and Epoxides to Methanol and Diols. Angew. Chem., Int. Ed.
2012, 51, 13041−13045.
(14) MacLeod, K. C.; McWilliams, S. F.; Mercado, B. Q.; Holland,
P. L. Stepwise N−H Bond Formation from N2-derived Iron Nitride,
Imide and Amide Intermediates to Ammonia. Chem. Sci. 2016, 7,
5736−5746.
(15) Lee, Y.; Sloane, F. T.; Blondin, G.; Abboud, K. A.; García-
Serres, R.; Murray, L. J. Dinitrogen Activation Upon Reduction of a
Triiron(II) Complex. Angew. Chem., Int. Ed. 2015, 54, 1499−1503.
(16) Nikiforov, G. B.; Vidyaratne, I.; Gambarotta, S.; Korobkov, I.
Titanium-Promoted Dinitrogen Cleavage, Partial Hydrogenation, and
Silylation. Angew. Chem., Int. Ed. 2009, 48, 7415−7419.
(17) Akagi, F.; Matsuo, T.; Kawaguchi, H. Dinitrogen Cleavage by a
Diniobium Tetrahydride Complex: Formation of a Nitride and Its
Conversion into Imide Species. Angew. Chem., Int. Ed. 2007, 46,
8778−8781.
(18) Shima, T.; Hu, S.; Luo, G.; Kang, X.; Luo, Y.; Hou, Z.
Dinitrogen Cleavage and Hydrogenation by a Trinuclear Titanium
Polyhydride Complex. Science 2013, 340, 1549.
(19) Curley, J. J.; Sceats, E. L.; Cummins, C. C. A Cycle for Organic
Nitrile Synthesis via Dinitrogen Cleavage. J. Am. Chem. Soc. 2006,
128, 14036−14037.
(20) Guru, M. M.; Shima, T.; Hou, Z. Conversion of Dinitrogen to
Nitriles at a Multinuclear Titanium Framework. Angew. Chem., Int. Ed.
2016, 55, 12316−12320.
(21) Klopsch, I.; Kinauer, M.; Finger, M.; Wurtele, C.; Schneider, S.
Conversion of Dinitrogen into Acetonitrile under Ambient Con-
ditions. Angew. Chem., Int. Ed. 2016, 55, 4786−9.
(22) Man, W.-L.; Lam, W. W. Y.; Lau, T.-C. Reactivity of Nitrido
Complexes of Ruthenium(VI), Osmium(VI), and Manganese(V)
Bearing Schiff Base and Simple Anionic Ligands. Acc. Chem. Res.
2014, 47, 427−439.
(43) Choi, J.-H.; Prechtl, M. H. G. Tuneable Hydrogenation of
Nitriles into Imines or Amines with a Ruthenium Pincer Complex
under Mild Conditions. ChemCatChem 2015, 7, 1023−1028.
(44) Schiwek, C.; Meiners, J.; Foerster, M.; Wuertele, C.;
Diefenbach, M.; Holthausen, M. C.; Schneider, S. Oxygen Reduction
with a Bifunctional Iridium Dihydride Complex. Angew. Chem., Int.
Ed. 2015, 54, 15271−15275.
(23) Berry, J. F. Terminal Nitrido and Imido Complexes of the Late
Transition Metals. Comments Inorg. Chem. 2009, 30, 28−66.
(24) Smith, J. M. Reactive Transition Metal Nitride Complexes.
Prog. Inorg. Chem. 2014, 58, 417−470.
(25) Schendzielorz, F. S.; Finger, M.; Volkmann, C.; Wurtele, C.;
Schneider, S. A Terminal Osmium(IV) Nitride: Ammonia Formation
and Ambiphilic Reactivity. Angew. Chem., Int. Ed. 2016, 55, 11417−
20.
(26) Williams, D. S.; Meyer, T. J.; White, P. S. Preparation of
Osmium(II) Nitrosyls by Direct Oxidation of Osmium(VI) Nitrides.
J. Am. Chem. Soc. 1995, 117, 823−4.
(45) Fu, S.; Chen, N.-Y.; Liu, X.; Shao, Z.; Luo, S.-P.; Liu, Q.
Ligand-Controlled Cobalt-Catalyzed Transfer Hydrogenation of
Alkynes: Stereodivergent Synthesis of Z- and E-Alkenes. J. Am.
Chem. Soc. 2016, 138, 8588−8594.
(46) Fu, S.; Shao, Z.; Wang, Y.; Liu, Q. Manganese-Catalyzed
Upgrading of Ethanol into 1-Butanol. J. Am. Chem. Soc. 2017, 139,
11941−11948.
J
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(47) Affan, M. A.; Jessop, P. G. Catalytic Formylation of Primary
and Secondary Amines with CO2 and H2 Using Abundant-Metal
Catalysts. Inorg. Chem. 2017, 56, 7301−7305.
Organics, and Gases in Deuterated Solvents Relevant to the
Organometallic Chemist. Organometallics 2010, 29, 2176−2179.
(48) Nguyen, D. H.; Trivelli, X.; Capet, F.; Paul, J.-F.; Dumeignil, F.;
Gauvin, R. M. Manganese Pincer Complexes for the Base-Free,
Acceptorless Dehydrogenative Coupling of Alcohols to Esters:
Development, Scope, and Understanding. ACS Catal. 2017, 7,
2022−2032.
(49) Wiberg, K. B. Application of the Pople-Santry-Segal CNDO
Method to the Cyclopropylcarbinyl and Cyclobutyl Cation and to
Bicyclobutane. Tetrahedron 1968, 24, 1083−1096.
(50) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Intermolecular
Interactions from a Natural Bond Orbital, Donor-acceptor Viewpoint.
Chem. Rev. 1988, 88, 899−926.
(51) Voigt, A.; Abram, U.; Kirmse, R. The Existence of [ReNF4]− −
An EPR Study. Inorg. Chem. Commun. 1998, 1, 141−142.
(52) Voigt, A.; Abram, U.; Strauch, P.; Kirmse, R. An Electron
Paramagnetic Resonance Study of Tetra-n-butylammonium
Tetrabromonitridorhenate(VI), [(n-C4H9)4N][ReNBr4]. Inorg.
Chim. Acta 1998, 271, 199−202.
(53) Abram, U.; Braun, M.; Abram, S.; Kirmse, R.; Voigt, A.
[NBu4][ReNCl4]: Facile Synthesis, Structure, Electron Paramagnetic
Resonance Spectroscopy and Reactions. J. Chem. Soc., Dalton Trans.
1998, 231−238.
(54) Badger, R. M. Relation Between the Internuclear Distances and
Force Constants of Molecules and Its Application to Polyatomic
Molecules. J. Chem. Phys. 1935, 3, 710−14.
(55) Laihia, K.; Kolehmainen, E.; Kauppinen, R.; Lorenc, J.; Puszko,
A. Multinuclear 1H, 13C and 15N NMR Study of Some Substituted 2-
amino-4-nitropyridines and Their N-oxides. Spectrochim. Acta, Part A
2002, 58A, 1425−1435.
(56) Wrackmeyer, B. Calculation of 15N NMR Parameters of Azides
and Some Related Compounds. Revisiting the Methylation of Nitrous
Oxide N2O. Z. Naturforsch., B: J. Chem. Sci. 2011, 66, 1079−1082.
(57) Perinu, C.; Saramakoon, G.; Arstad, B.; Jens, K.-J. Application
of 15N-NMR Spectroscopy to Analysis of Amine Based CO2 Capture
Solvents. Energy Procedia 2014, 63, 1144−1150.
(58) Galajov, M.; Garcia, C.; Gomez, M. Tri-chlorido, 2-methylallyl
and 2-butenyl Tert-butylimido Niobium and Tantalum Complexes:
Synthesis, Multinuclear NMR Spectroscopy and Reactivity. Dalton
Trans 2011, 40, 413−420.
(59) Galajov, M.; Garcia, C.; Gomez, M.; Gomez-Sal, P.; Temprado,
M. Synthesis and DFT, Multinuclear Magnetic Resonance, and X-ray
Structural Studies of Iminoacyl Imido Hydridotris(3,5-
dimethylpyrazolyl)borate Niobium and Tantalum(V) Complexes.
Organometallics 2014, 33, 2277−2286.
(60) Scepaniak, J. J.; Margarit, C. G.; Harvey, J. N.; Smith, J. M.
Nitrogen Atom Transfer from Iron(IV) Nitrido Complexes: A Dual-
Nature Transition State for Atom Transfer. Inorg. Chem. 2011, 50,
9508−9517.
(61) Johnson, N. P.; Lock, C. J. L.; Wilkinson, G. Amine, Phosphine,
Arsine, and Stilbine Complexes of Rhenium(III), -(IV), and -(V). J.
Chem. Soc. 1964, 1054−66.
(62) Rouschias, G.; Wilkinson, G. Preparation and Reactions of
Trihalo(alkanonitrile)bis(triphenylphosphine) Rhenium(III) Com-
plexes. J. Chem. Soc. A 1967, 993−1000.
(63) Schulz, A.; Villinger, A. Binary Polyazides of Cadmium and
Mercury. Chem. - Eur. J. 2015, 21, 3649−3663.
(64) Demadis, K. D.; Meyer, T. J.; White, P. S. Reactivity of
Osmium(VI) Nitrides with the Azide Ion. A New Synthetic Route to
Osmium(II) Polypyridyl Complexes. Inorg. Chem. 1998, 37, 3610−
3619.
(65) Rhile, I. J.; Markle, T. F.; Nagao, H.; DiPasquale, A. G.; Lam,
O. P.; Lockwood, M. A.; Rotter, K.; Mayer, J. M. Concerted Proton−
Electron Transfer in the Oxidation of Hydrogen-Bonded Phenols. J.
Am. Chem. Soc. 2006, 128, 6075−6088.
(66) Fulmer, G. R.; Miller, A. J. M.; Sherden, N. H.; Gottlieb, H. E.;
Nudelman, A.; Stoltz, B. M.; Bercaw, J. E.; Goldberg, K. I. NMR
Chemical Shifts of Trace Impurities: Common Laboratory Solvents,
K
DOI: 10.1021/acs.inorgchem.9b01075
Inorg. Chem. XXXX, XXX, XXX−XXX