Generating Potent C-H PCET Donors: Ligand-Induced Fe-to-Ring Proton Migration from a Cp*FeIII-H Complex Demonstrates a Promising Strategy

Article abstract of DOI:10.1021/jacs.0c09363

Highly reactive organometallic species that mediate reductive proton-coupled electron transfer (PCET) reactions are an exciting area for development in catalysis, where a key objective focuses on tuning the reactivity of such species. This work pursues ligand-induced activation of a stable organometallic complex toward PCET reactivity. This is studied via the conversion of a prototypical Cp*FeIII-H species, [FeIII(η5-Cp*)(dppe)H]+ (Cp? = C5Me5-, dppe = 1,2-bis(diphenylphosphino)ethane), to a highly reactive, S = 1/2 ring-protonated endo-Cp*H-Fe relative, triggered by the addition of CO. Our assignment of the latter ring-protonated species contrasts with its previous reported formulation, which instead assigned it as a hypervalent 19-electron hydride, [FeIII(η5-Cp*)(dppe)(CO)H]+. Herein, pulse EPR spectroscopy (1,2H HYSCORE, ENDOR) and X-ray crystallography, with corresponding DFT studies, cement its assignment as the ring-protonated isomer, [FeI(endo-η4-Cp*H)(dppe)(CO)]+. A less sterically shielded and hence more reactive exo-isomer can be generated through oxidation of a stable Fe0(exo-η4-Cp*H)(dppe)(CO) precursor. Both endo- and exo-ring-protonated isomers are calculated to have an exceptionally low bond dissociation free energy (BDFEC-H ≈ 29 kcal mol-1 and 25 kcal mol-1, respectively) cf. BDFEFe-H of 56 kcal mol-1 for [FeIII(η5-Cp*)(dppe)H]+. These weak C-H bonds are shown to undergo proton-coupled electron transfer (PCET) to azobenzene to generate diphenylhydrazine and the corresponding closed-shell [FeII(η5-Cp*)(dppe)CO]+ byproduct.

Full text of DOI:10.1021/jacs.0c09363

pubs.acs.org/JACS
Article
Generating Potent C−H PCET Donors: Ligand-Induced Fe-to-Ring
Proton Migration from a Cp*FeIII−H Complex Demonstrates a
Promising Strategy
Dirk J. Schild,† Marcus W. Drover,† Paul H. Oyala, and Jonas C. Peters*
Cite This: https://dx.doi.org/10.1021/jacs.0c09363
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Highly reactive organometallic species that mediate
reductive proton-coupled electron transfer (PCET) reactions are
an exciting area for development in catalysis, where a key objective
focuses on tuning the reactivity of such species. This work pursues
ligand-induced activation of a stable organometallic complex
toward PCET reactivity. This is studied via the conversion of a
prototypical Cp*FeIII−H species, [FeIII(η5-Cp*)(dppe)H]+ (Cp*
−
= C5Me5 , dppe = 1,2-bis(diphenylphosphino)ethane), to a highly
reactive, S = 1/2 ring-protonated endo-Cp*H−Fe relative,
triggered by the addition of CO. Our assignment of the latter
ring-protonated species contrasts with its previous reported
formulation, which instead assigned it as a hypervalent 19-electron
hydride, [FeIII(η5-Cp*)(dppe)(CO)H]+. Herein, pulse EPR spectroscopy (1,2H HYSCORE, ENDOR) and X-ray crystallography,
with corresponding DFT studies, cement its assignment as the ring-protonated isomer, [FeI(endo-η4-Cp*H)(dppe)(CO)]+. A less
sterically shielded and hence more reactive exo-isomer can be generated through oxidation of a stable Fe0(exo-η4-
Cp*H)(dppe)(CO) precursor. Both endo- and exo-ring-protonated isomers are calculated to have an exceptionally low bond
dissociation free energy (BDFEC−H ≈ 29 kcal mol−1 and 25 kcal mol−1, respectively) cf. BDFEFe−H of 56 kcal mol−1 for [FeIII(η5-
Cp*)(dppe)H]+. These weak C−H bonds are shown to undergo proton-coupled electron transfer (PCET) to azobenzene to
generate diphenylhydrazine and the corresponding closed-shell [FeII(η5-Cp*)(dppe)CO]+ byproduct.
ined an Fe(III) indenide hydride [FeIII(η5-Ind)(depe)H]+ (Ind
INTRODUCTION
■
= C8H7-, depe = 1,2-bis(diethylphosphino)ethane) and its low
Proton-coupled electron transfer (PCET) reactions have
temperature isomerism to give a η6-indene complex, [Fe(η6-
emerged as powerful strategies for mediating sundry reductive
IndH)(depe)]+, (BDFEX‑H = 50 kcal mol−1; X = Fe or C).9
transformations in organic synthesis, with clever new
The highly reactive nature of these ring-protonated species
approaches being discovered to generate in situ highly reactive
has frustrated independent studies of their reactivity profiles
H atom surrogates as intermediates.1−3 Reactive fragments
toward exogenous substrates. We hence targeted a related
with weak element-hydrogen (E-H) bonds have thus been
system wherein a reactive ring−C-H PCET donor might be
targeted to facilitate hydrogen atom delivery, revealing
generated in solution, via association of a ligand to the metal to
desirable substrate reductions that are fascinating in
cause concomitant isomerization of a metal-bound hydride to a
scope.1,2,4 As a salient example, dissolution of SmI2 in H2O/
Cp*-ring position. Such a strategy would, in principle, enable
THF mixtures confers coordination-induced bond-weakening
two unreactive partners to reside in solution, with a net PCET
in the resulting [Sm(H2O)n]2+ complex to provide weak and
reaction being triggered by addition of a donor ligand L (see
hence highly reactive O−H bonds (BDFEO−H ≈ 26 kcal
Chart 1).
mol−1);1h such species can be employed toward substrate
With this goal in mind, we turned to a known iron(III)
reductions such as ketones to alcohols and anthracene to
hydride half-sandwich complex, [FeIII(η5-Cp*)(dppe)H]+[1]+,
dihydroanthracene5 and also toward catalytic N2 fixation to
NH3 in the presence of a Mo catalyst.6
Received: August 31, 2020
As an outgrowth of our own mechanistic studies of Fe-
mediated N2 fixation, we recently reported the rigorous
characterization of a highly reactive [Cp*(endo/exo-η4-Cp*H)-
−
Co]+ (Cp* = C5Me5 ) species (BDFEC−H < 29 kcal mol−1 for
the exo-analogue), generated in situ via ring protonation of
decamethylcobaltocene, Cp*2Co.7,8 Relatedly, we have exam-
https://dx.doi.org/10.1021/jacs.0c09363
© XXXX American Chemical Society
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
A

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
cobaltocenes,4 we posited its structure to be [FeI(endo-η4-
Cp*H/D)(dppe)(CO)]+ (endo-isomer confirmed by XRD
vide infra), endo-[3-H/D]+, rather than the terminal hydride
[2]+ (Scheme 1).
Chart 1. (A) Reported Generation of a 19-Electron FeIII−
H,11 Reassigned Herein; (B) Generalized Donor-Induced
a
PCET; (C) Donor-Induced PCET Described in This Study
Scheme 1. Reported Preparation and (Incorrect)
Assignment of a 19-Electron Fe(III)-H Complex11 and
Synthesis of endo-[3]+/[4]+ by Addition of CO or CNXyl
a
with Accompanying H2 Release to Give [5]+ and [6]+
a
BDFE values (calculated by DFT) are in kcal mol−1.
−
(Cp* = C5Me5 , dppe = 1,2-bis(diphenylphosphino)ethane),
first prepared by Lapinte and co-workers in 1992.10 Complex
[1]+ was reported to undergo associative binding of CO to
provide the 19-electron hydride [FeIII(η5-Cp*)(dppe)(CO)-
H]+ ([2]+), the latter having been characterized by X-band
EPR spectroscopy. The reaction was deemed reversible with
warming, returning [1]+.11
a
BDFE values (calculated by DFT) are in kcal mol−1.
Consumption of [1]+ is easily identified by EPR spectros-
copy based on the significant shift in the value of gmax (Figure
1). Besides coupling to two distinct 31P (I = 1/2) nuclei with
Based on our findings with [Cp*(endo-η4-Cp*H)Co]+ and
[Fe(η6-IndH)(depe)]+, we wondered whether [2]+ would be
best described as [FeI(η4-Cp*H)(dppe)(CO)]+, which would
occur via proton migration to the ring upon CO binding.
Herein, we use a range of methods, including crystallography
and various EPR techniques, with corresponding DFT analysis,
to establish this is indeed the case. We then demonstrate that
an intermolecular PCET reaction is possible, with azobenzene
as a model substrate, via this ligand-induced trigger.
1
similar magnitude, significant additional coupling to H is
1
2
evident when the H and H isotopologues are compared,
consistent with the presence of a relatively strongly coupled 1H
nucleus (Figure 2). Collection of a series of X-band Davies
ENDOR spectra acquired across the EPR envelope of endo-[3-
D]+ (see the Supporting Information) provides additional data
consistent with large couplings to two nonequivalent
phosphines (A(31P1) = [72, 59, 58] MHz and A(31P2) =
[49, 42, 51] MHz).
X-band HYSCORE spectroscopy of the 1H and 2H
isotopologues was used to determine the hyperfine parameters
for the hydride-derived hydrogen nucleus (Figure 2).
Simulation of field-dependent HYSCORE spectra of endo-[3-
H]+ reveals a relatively isotropic proton hyperfine tensor,
A(1H) = [24, 20, 34.5] MHz (aiso(1H) = 26.2 MHz), with a
small Euler rotation of the hyperfine tensor relative to the g-
tensor of (α,β,γ) = (0, 30, 0)°. HYSCORE spectra of endo-[3-
D]+ exhibit intense features arising from deuterium which are
RESULTS AND DISCUSSION
■
We began our study with (re)examining the solution
spectroscopy of [FeIII(η5-Cp*)(dppe)H]+[1]+ in the presence
of CO. Addition of CO to the 1H and 2H isotopologues (Fe−
H/D) of [1]+ results in a change in the EPR spectrum,
indicating the formation of a single S = 1/2 species exhibiting
rhombic symmetry (g = [2.085, 2.039, 2.004]), parameters that
are similar to those previously reported by Lapinte (g =
[2.0777, 2.0367, 2.0019], A(31P) = [48, 50, 50] MHz, and
A(1H) = [14, 17, 34] MHz (aiso(1H)= 22 MHz)).11 Based
on preliminary DFT calculations (vide infra) and the similarity
of the proton hyperfine coupling to ring protonated
1
well simulated by scaling the H hyperfine tensor determined
from 1H HYSCORE by the proportion of 1H/2H gyromagnetic
ratios (γ1H/γ2H = 6.514), with A(2H) = [3.7, 3.1, 5.3] MHz.
These parameters provide excellent agreement with simu-
B
https://dx.doi.org/10.1021/jacs.0c09363
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
The more stable, neutral endo/exo-species (endo/exo-3) were
also prepared. Treatment of [FeII(η5-Cp*)(dppe)(CO)]-
[BArF ] ([5]BArF ) with LiBEt3H provides exo-[Fe0(η4-
4
4
Cp*H)(dppe)(CO)] (exo-3) in good yield (Scheme 1). The
alternative isomer, endo-3, was prepared via reduction of endo-
[3]+ with Cp2Co at −78 °C.11 The room temperature stability
allowed for isolation and growth of crystalline material suitable
for single crystal X-ray diffraction. The structures of exo/endo-3
are presented in Figure 3. Exo- and endo-3 feature an η4-diene
unit bound to a zerovalent iron center. Notably, the C(1)−
C(6) bond distance in endo-3 (1.538(3) Å) is slightly longer
than that in exo-3 (1.514(2) Å), possibly suggestive of Fe →
C−C (σ*) donation in the former. Such an interaction is
predicted based on the singularly occupied molecular orbitals
(SOMOs) of endo/exo-[3]+ (Figure 4). Two distinct C−H
stretches are observed for exo-3-H at 2711 and 2612 cm−1 that
shift to 2009 and 1955 cm−1 for exo-3-D; C−H stretches for
endo-3-H are not discernible from the bulk C−H stretching
region i.e., >2711 cm−1 (see the Supporting Information).14
Similar observations have been made for the pairs: [Cp*(exo-
η4-Cp*H)CoII]/[Cp*(exo-η4-Cp*D)CoII]8 and [Cp(η4-
C5H6)Co]/[Cp(exo-η4-C5H5D)Co].15
Despite its exceptionally weak C−H bond, the assignment of
endo-[3]+ could be corroborated by X-ray crystallography
(Figure 3) via growth of suitable crystals at low temperature.
Gratifyingly, the solid-state structure, which was unobtainable
for [Cp*(endo-η4-Cp*H)Co]+, confirms the formation of an
endo-Cp*(C−H) bond, with metrics associated with a
dearomatized five-membered Cp*H ring that differs from
that of a Cp* anion with elongation along the C(1)−C(5) and
C(1)−C(2) vectors [1.51(1)/1.53(1) Å] from ca. 1.42 Å
(average C−C bond distances in [7]+, vide infra), signifying
new C−C single bonds. The low temperature stability of endo-
[3]+ likely derives from a high degree of steric shrouding of the
reactive ring-bound H atom via phenyl rings from the dppe
ligand.
Figure 1. X-band CW-EPR spectra of [1]+, endo-[3-H]+, exo-[3-H]+,
and endo-[4-H]+ in 2-MeTHF at 77 K.
lations of the X-band CW-EPR spectra of these two
isotopologues.
In contrast to terminal metal-hydrides,12 the anisotropic
1
component (T) of the H hyperfine of endo-[3]+, T(1H)=
Next, we sought to generate the alternative, and presumably
more reactive, exo-isomer wherein steric shrouding by dppe is
[-2.2, −6.2, 8.3] MHz), is considerably smaller, indicating
greater distance between the nucleus and Fe-centered spin
density. Similarly, low anisotropy is also observed for the endo-
Cp*(H) protonation product of Cp*2Co. The low anisotropy
suggests ring-protonation (endo-[3]+), and not a terminal
hydride (i.e., [2]+) species. Hyperfine coupling constants and
the predicted anisotropy as calculated by DFT further support
the assignment and are independent of the method used (see
the Supporting Information).
far less prominent. Oxidation of neutral exo-3 with [Fc]BArF
4
at −78 °C provides green solutions of reactive exo-[3]+ that
can be analyzed by EPR spectroscopy. From the CW data, we
assign two conformational isomers of exo-[3]+ (A and B), both
with rhombic symmetry (Figure 5): A with g = [2.116, 2.073,
1.997] and another, B with g = [2.093, 2.045, 2.013] that
presumably differ by rotation of the Cp*H ligand. The
potential presence of multiple conformers is supported by
DFT, as four distinct minima are found within 2 kcal mol−1
upon Cp*H rotation. The exact nature of the conformers
present in solution could not be determined, due to the small
differences in energies and predicted 1H hyperfine tensors (see
the Supporting Information). Least-squares optimization of
simulations of the X-band CW spectra converged at relative
populations of 0.6:0.4 for conformers A:B, indicating that the
two conformations are indeed of similar relative energies.
We also explored the addition of an isocyanide donor,
CNXyl (Xyl = 2,6-dimethylphenyl). Treatment of [1]+ with
CNXyl at −78 °C provides [FeI(endo-η4-Cp*H)(dppe)-
(CNXyl)]+ (endo-[4]+) as evidenced by CW X-band EPR
spectroscopy at 77 K; g = [2.132, 2.042, 2.004], A(31P1) =
[75, 35, 54] MHz, and A(31P2) = [76, 64, 64] MHz.
Simulations of HYSCORE spectra of endo-[4-H/D]+ give
A(1H) = [17, 22, 32.5] MHz, aiso(1H) = 23.8 MHz, and
T(1H) = [−6.8, −1.8, 8.7], which are nearly identical to
those discussed above for the endo-Cp*(C−H) Fe(CO)
adduct endo-[3]+. Additional features present in the (-,+)
quadrant of HYSCORE of both endo-[4-H]+ and endo-[4-D]+
can be assigned to hyperfine coupling to 14N in the CNXyl
2
Simulations of H−1H difference ENDOR spectra of this this
2
mixture of exo-[3]+ products provided constraints on the H
1
(and by proxy H) hyperfine couplings, with higher frequency
ENDOR providing the same for 31P hyperfine couplings for
the two conformers in the above CW simulations, with A(1H)
ligand, with A(14N) = [7.4, 7.4, 9] MHz, a relatively small 14
N
[85, 84, 83] MHz, aiso(1H) = 84 MHz, T(1H) = [1, 0,
nuclear quadrupole coupling constant e2qQ/h(14N) = 1.0
MHz, and a negligible electric field gradient rhombicity η(14N)
≈ 0, as expected for a triply bonded CN moiety with axial
symmetry.13
=
−1], A(31P1) = [96, 88, 47] MHz, and A(31P2) = [78, 75,
63] MHz for conformer A and A(1H) = [76, 74, 70] MHz,
aiso(1H) = 73 MHz, T(1H) = [3, 1, −3], A(31P1) = [46,
C
https://dx.doi.org/10.1021/jacs.0c09363
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 2. X-band CW-EPR spectra and corresponding X-band HYSCORE spectra of freeze-quenched samples in 2-MeTHF. (A) endo-[3-H]BArF
4
and endo-[3-D]BArF ; (B) endo-[4-H]BArF4 and endo-[4-D]BArF . HYSCORE simulations of features from 1H, 2H (red), and 14N (green, CNXyl
4
4
only) hyperfine couplings are overlaid over the data, which is plotted in gray. ** = traces of [1]+.
44, 15] MHz, and A(31P2) = [70, 64, 64] MHz for conformer
B. The aiso(1H) values for these isomers are much larger than
that observed for the endo adduct, endo-[3]+ (aiso(1H) = 26.2
MHz), while the magnitude of T(1H) tensors are similar to
that observed for endo-[3]+ and are consistent with a ligand
C−H, rather than M-H unit. This trend, aiso(1H) (exo) >
aiso(1H) (endo) is also been observed for the protonated
[Cp*(endo/exo-η4-Cp*H)CoII]+ derivative (Figure 4) and
correlates with greater predicted spin density on the exo
(versus the endo) hydrogen atom for the staggered16 isomers
(0.06 versus 0.02 e−).
CNXyl (endo-[4]+) adducts, IV were calculated to be −3 and
+3 kcal mol−1 in energy relative to [1]+. The process involves
adduct formation and then reductive C−H elimination (or vice
versa). H2 formation from IV to give complexes VI ([5]+ and
[6]+) is calculated to be highly favorable (CO and CNXyl:
BDFEC−H = 29 kcal mol−1 vs BDFEH−H = 102.3 kcal mol−1 in
CH3CN)1, correlating with a decrease in BDFEX‑H of almost
30 kcal mol−1 (cf. 56 kcal mol−1 for [1]+). The BDFEC−H of
exo-[3]+ is calculated to be weaker (25 kcal mol−1), which in
combination with the reduced steric crowding, manifests
experimentally via facile H2 evolution at temperatures for
which endo-[3]+ is stable enough to isolate.
Annealing frozen solutions of the 17-electron η4-Cp*H
complexes endo-[3]+, exo-[3]+, or endo-[4]+ (Scheme 1)
provides ca. 0.5 equiv H2 and their corresponding 18-electron
To experimentally benchmark the H• transfer propensity of
[1]+, we aimed to determine an upper limit for its BDFEFe−H
.
S = 0 stable adducts, [FeII(η5-Cp*)(dppe)(CY)][BArF ]
Given the reduced FeII−H congener undergoes hydride
transfer to 1-benzyl-3-acetamidopyridinium [BNAP]OTf
(ΔGH‑ ≈ 59 kcal mol−1)18 and its known FeII/FeIII oxidation
potential (E1/2 = −0.71 V vs Fc/Fc+),19 the FeIII−H bond of
[1]+ is estimated to have an upper bound of BDFEFe−H < 50
kcal mol−1;20 this can be compared with a DFT-predicted
value of 56 kcal mol−1. In spite of its weak BDFEFe−H, complex
[1]+ is stable at room temperature in ethereal solvents (THF,
Et2O; <5% yield of H2 and [Fe(η5-Cp*)(dppe)(N2)]+ [7]+ at
+80 °C, 1 week; Scheme 3); details pertaining to the
characterization of [7]+ are presented in the Supporting
Information. Although [7]+ is thermodynamically poised to
release H2, a substantial kinetic barrier presumably attenuates
the rate of H2 loss (BDFEH−H = 102.3 kcal mol−1).
4
([5]BArF (Y = O), [6]BArF (Y = NXyl)). The FeII−CO
4
4
adduct [5]+ has been described previously;17 full character-
ization data for [6]BArF is in the Supporting Information. By
4
contrast to endo-[3]+, which can be isolated and characterized
at temperatures below −70 °C, H2 evolution and formation of
[5]+ from exo-[3]+ is observed at temperatures as low as −100
°C. As previously noted, the difference in stability between
endo-[3]+ and exo-[3]+ is likely due to a dramatic difference in
steric hindrance between the endo- and exo-H positions.
To gauge the relative energies of the species discussed
herein, we turned to DFT calculations to estimate relevant
BDFEs (Scheme 2). The starting hydride complex [1]+ was
calculated to have a BDFEFe−H of 56 kcal mol−1, providing the
vacant S = 1 cation, [Fe(η5-Cp*)(dppe)]+ (III).
Subsequently, we aimed to determine an upper limit for the
BDFEC−H of endo-[3]+ (Scheme 4). Reaction of an MeCN
solution of endo-3 with CO2 results in hydride transfer and
Associative binding of an L-type donor (prior to H• loss)
was then considered. In this way, the CO (endo-[3]+) and
D
https://dx.doi.org/10.1021/jacs.0c09363
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 3. Solid-state structure of exo-3, endo-3, and endo-[3]+ with
ellipsoids shown at 50% probability. Counteranion omitted for endo-
[3]+.
Figure 5. (A) X-Band CW-EPR spectra of freeze-quenched samples
of exo-[3-H/D]+ in 2-MeTHF at 77 K for with simulations of
conformers A and B. (B) X-band 2H−1H Difference Davies ENDOR
of exo-[3-H]+ with data (black), total 2H simulation (red), conformer
A (blue), and conformer B (green).
Figure 4. Frontier molecular orbitals (SOMOs) of endo/exo-[3]+
7,8
(staggered conformers) and [Cp*(endo/exo-η4-Cp*H)CoII]+ opti-
BDFE (ΔG([3]+)H•) of <36 kcal mol−1 (DFT prediction: 29
kcal mol−1), generating [5]+ and H• as products. On the basis
of thermodynamics alone, H2 elimination from endo-[3]+
should be facile. The experimentally predicted upper bound
for the hydricity of endo-[3]+ results in a pKa of 23 in
acetonitrile (see the Supporting Information for square
scheme). In line with the predicted lower BDFE of exo-[3]+,
its Fe0/I redox potential is more positive (E1/2 = −0.70 V vs
Fc/Fc+) compared to endo-[3]+. This ∼110 mV difference
between the endo and exo isomers corresponds to a ∼2.5 kcal
mol−1 difference in BDFE, suggesting that an additional
contributing factor to the difference in C−H BDFE’s between
these isomers is due to their different acidities.
mized using TPSS; def2tzvp (Fe), def2svp (all other atoms). Inset
shows aiso(1H) as obtained by 1,2H HYSCORE and/or ENDOR.
formation of [Fe(η5-Cp*)(dppe)(CO)]+ [5]+. The hydricity of
endo-3 in MeCN must therefore be less than the hydricity of
HCO2 (ΔGH− ≈ 43 kcal mol−1). In other words, endo-3 is
−
21
−
more hydridic that HCO2 . Utilizing thermodynamic
relationships that relate H• and H− transfer in MeCN,22 the
upper bound for the free energy of H− transfer from endo-3
and H• transfer from endo-[3]+ can then be related. The endo-
Cp*(C−H) bond of endo-[3]+ (Fe0/I, E1/2 = −0.81 V vs Fc/
Fc+) is thus conservatively estimated to have an upper bound
E
https://dx.doi.org/10.1021/jacs.0c09363
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Scheme 2. Free Energy Change (kcal mol−1) for PCET
Involving endo-[3]+/[4]+ [TPSS; def2tzvp (Fe), def2svp (All
Other Atoms)]
Scheme 4. (A) Experimental Determination of BDFEC−H for
endo-[3]+; (B) Relevant Thermodynamic Equations
Relating H− and H• Transfer for the endo-Variant Only
drazine and vice versa has been studied over the years with
various PCET donors and acceptors,23 with some donors in a
similar BDFE range.3
Complex [1]+ does not react with a 20-fold excess of
azobenzene (PhN=NPh) over a period of days (for an average
BDFEN−H = 65 kcal mol−1 over two transfers), but upon
addition of CO, which generates endo-[3]+ in situ at −78 °C, a
reaction is triggered to generate 1,2-diphenylhydrazine (25%
yield, Scheme 5), H2, and [5]+. Isotopic labeling studies using
the deuterated analogue, endo-[3-D]+, result in the formation
of the labeled isotopologue, PhDN-NDPh. In situ generation
of exo-[3]+ through oxidation of exo-3 and reaction with
PhNNPh produces 1,2-diphenylhydrazine in 78% yield
a
Scheme 3. (A) Synthesis of [7]+ by H2 Evolution (<5%)
and (B) Non-Productive H• Transfer to Azobenzene Using
b
[1]+
Scheme 5. PCET Transfers to Azobenzene Using (A) endo-
[3]+, Generating 25% PhNHNHPh, and (B) exo-[3]+,
a
Giving 78% PhNHNHPh
a
Inset shows solid-state structure of [7]+ with ellipsoids shown at 50%
b
probability. BDFE values (kcal mol−1) calculated by DFT using
TPSS; def2tzvp (Fe), def2svp (all other atoms).
We next turned our focus to exploring whether a productive
H• transfer to an exogenous substrate from either endo-[3]+ or
exo-[3]+ could be accomplished, as opposed to undesired H2
evolution. Azobenzene proved an interesting choice of
substrate for this purpose as its conversion to diphenylhy-
a
BDFE values (kcal mol−1) calculated by DFT using TPSS; def2tzvp
(Fe), def2svp (all other atoms).
F
https://dx.doi.org/10.1021/jacs.0c09363
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(Scheme 5). In line with endo-[3]+, only H2 and [5]+ are
observed as by products, and labeling with deuterium results in
the formation of PhDN-NDPh. No other organic or iron
containing products can be detected by NMR spectroscopy. A
proton transfer followed by electron transfer or vice versa is
unlikely as endo and exo-[3]+ are weak acids (vide supra), and
the protonation step is calculated to be 27 kcal mol−1 uphill.
AUTHOR INFORMATION
■
Corresponding Author
Jonas C. Peters − Division of Chemistry and Chemical
Engineering, California Institute of Technology, Pasadena,
California 91125, United States; orcid.org/0000-0002-
6610-4414; Email: jpeters@caltech.edu
Furthermore, endo and exo-[3]+ are weak reductants (E1/2
>
Authors
−0.81 V vs Fc/Fc+)24 and not capable of reducing azobenzene
to the radical anion (E1/2 = −1.73 V vs Fc/Fc+).25 It is worth
emphasizing that the increase in reaction rate upon addition of
CO to [1]+ is orders of magnitude larger, as no H2 formation
or reduction of azobenzene is observed at room temperature in
the absence of added CO, while instantaneous formation of
[5]+, the terminal product of H• transfer, can be observed even
at −78 °C. This remarkable difference in reactivity via addition
of a simple L-type ligand highlights the potential for in situ
generation of a powerful PCET reagent.
Dirk J. Schild − Division of Chemistry and Chemical
Engineering, California Institute of Technology, Pasadena,
California 91125, United States
Marcus W. Drover − Division of Chemistry and Chemical
Engineering, California Institute of Technology, Pasadena,
California 91125, United States; orcid.org/0000-0002-
2186-1040
Paul H. Oyala − Division of Chemistry and Chemical
Engineering, California Institute of Technology, Pasadena,
California 91125, United States; orcid.org/0000-0002-
8761-4667
CONCLUSION
■
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.0c09363
In closing, building on recent work describing protonated
metallocene and related species with very weak and hence
reactive C−H bonds, we have herein described the character-
ization of endo-[FeI(η4-Cp*H)(dppe)(L)]+ (L = CO, CNXyl)
and exo-[FeI(η4-Cp*H)(dppe)(L)]+ (L = CO). Notably, one
of these systems, endo-[FeI(η4-Cp*H)(dppe)(CO)], was
previously studied but had been instead assigned as the 19-
electron hydride species [FeIII(η5-Cp*)(dppe)(CO)(H)]+.
Use of pulse EPR techniques (1,2H HYSCORE, ENDOR) as
well as the solid-state crystal structure of endo-[FeI(η4-
Cp*H)(dppe)(CO)]+, obtained at low temperature, cements
this reassignment.
Author Contributions
†D.J.S. and M.W.D. contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors are grateful to the Department of Energy Basic
Energy Sciences for support via Grant No. DOE-0235032. The
Caltech EPR facility was previously supported by the National
Science Foundation via Grant No. NSF MRI-153194 and is
currently supported by the Dow Next Generation Educator
Fund. The Beckman Institute is thanked for X-ray support.
M.W.D. acknowledges NSERC (Banting PDF award), and
M.W.D. and D.J.S. thank the Resnick Sustainability Institute at
Caltech for fellowships.
Our interest in these particular endo-[FeI(η4-Cp*H)(dppe)-
(L)]+ systems is that they can be generated in situ via L
addition to a far more stable iron hydride precursor, [FeIII(η5-
Cp*)(dppe)H]+. This affords the opportunity to transfer H• to
a substrate, demonstrated for L = CO using the reduction of
azobenzene to diphenylhydrazine as a model, where the H•
transfer reaction is triggered at low temperature via the
addition of CO to a mixture of [FeIII(η5-Cp*)(dppe)H]+ and
PhNNPh. The latter two partners do not otherwise react
under the same conditions; CO triggers the isomerization that
leads to the weak and hence reactive C−H bond. This
approach demonstrates an attractive strategy for designing a
powerful, in situ generated PCET reagent (BDFEC−H ≈ 29
kcal mol−1 for endo-[FeI(η4-Cp*H)(dppe)(CO)]+), with
tunability of the BDFE via the choice of L donor.
REFERENCES
■
(1) (a) Warren, J. J.; Tronic, T. A.; Mayer, J. M. Thermochemistry of
Proton-Coupled Electron Transfer Reagents and its Implications.
Chem. Rev. 2010, 110, 6961. (b) Weinberg, D. R.; Gagliardi, C. J.;
Hull, J. F.; Murphy, C. F.; Kent, C. A.; Westlake, B. C.; Paul, A.; Ess,
D. H.; McCafferty, D. G.; Meyer, T. J. Proton-Coupled Electron
Transfer. Chem. Rev. 2012, 112, 4016. (c) Hammes-Schiffer, S.
Theory of Proton-Coupled Electron Transfer in Energy Conversion
Processes. Acc. Chem. Res. 2009, 42, 1881. (d) Gentry, E. C.; Knowles,
R. R. Synthetic Applications of Proton-Coupled Electron Transfer.
Acc. Chem. Res. 2016, 49, 1546. (e) Matos, J. L. M.; Green, S. A.;
Shenvi, R. A. Markovnikov Functionalization by Hydrogen Atom
Transfer. Organic Reactions 2019, 100, 383. (f) Tsui, E.; Metrano, A.
J.; Tsuchiya, Y.; Knowles, R. R. Catalytic Hydroetherification of
Unactivated Alkenes Enabled by Proton-Coupled Electron Transfer.
Angew. Chem., Int. Ed. 2020, 59, 11845. (g) Roos, C. B.; Demaerel, J.;
Graff, D. E.; Knowles, R. R. Enantioselective Hydroamination of
Alkenes with Sulfonamides Enabled by Proton-Coupled Electron
Transfer. J. Am. Chem. Soc. 2020, 142, 5974. (h) Kolmar, S. S.; Mayer,
J. M. SmI2(H2O)n Reduction of Electron Rich Enamines by Proton-
Coupled Electron Transfer. J. Am. Chem. Soc. 2017, 139, 10687.
(2) Crossley, S. W. M.; Obradors, C.; Martinez, R. M.; Shenvi, R. A.
Mn-, Fe-, and Co-Catalyzed Radical Hydrofunctionalizations of
Olefins. Chem. Rev. 2016, 116, 8912.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.0c09363.
Additional experimental details including figures, tables,
and references as discussed in the text (PDF)
Crystallographic data for exo/endo-3, [3]BArF , [6]-
4
BArF , and [7]BArF . CCDC 1943147−1943149 and
4
4
2021150-2021151 (CIF)
1H, 13C{1H}, and 31P{1H} NMR spectra for complexes;
̈
Mossbauer, UV−vis, IR, and EPR spectroscopy; and
computational models (MOL)
G
https://dx.doi.org/10.1021/jacs.0c09363
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(3) Bezdek, M. J.; Guo, S.; Chirik, P. J. Coordination-induced
weakening of ammonia, water, and hydrazine X−H bonds in a
molybdenum complex. Science 2016, 354, 730.
(4) Chalkley, M. J.; Garrido-Barros, P.; Peters, J. C. A molecular
mediator for reductive concerted proton-electron transfers via
electrocatalysis. Science 2020, 369, 850.
(5) (a) Chciuk, T. V.; Anderson, W. R.; Flowers, R. A., II. High-
Affinity Proton Donors Promote Proton-Coupled Electron Transfer
by Samarium Diiodide. Angew. Chem., Int. Ed. 2016, 55, 6033.
(b) Chciuk, T. V.; Anderson, W. R.; Flowers, R. A., II. Proton-
Coupled Electron Transfer in the Reduction of Carbonyls by
Samarium Diiodide-Water Complexes. J. Am. Chem. Soc. 2016, 138,
8738. (c) Chciuk, T. V.; Anderson, W. R., jr.; Flowers, R. A., II.
Interplay between Substrate and Proton Donor Coordination in
Reductions of Carbonyls by SmI2-Water Through Proton-Coupled-
Electron Transfer. J. Am. Chem. Soc. 2018, 140, 15342.
(6) Ashida, Y.; Arashiba, K.; Nakajima, K.; Nishibayashi, Y.
Molybdenum-catalyzed ammonia production with samarium diiodide
and alcohols or water. Nature 2019, 568, 536.
(7) (a) Chalkley, M.; Del Castillo, T.; Matson, B.; Roddy, J.; Peters,
J. C. Catalytic N2-to-NH3 Conversion by Fe at Lower Driving Force:
A proposed Role for Metallocene-Mediated PCET. ACS Cent. Sci.
2017, 3, 217. (b) Chalkley, M. J.; Del Castillo, T. J.; Matson, B.;
Peters, J. C. Fe-Mediated Nitrogen Fixation with a Metallocene
Mediator: Exploring pKa Effects and Demonstrating Electrocatalysis.
J. Am. Chem. Soc. 2018, 140, 6122.
experimentally determined BDFE (50 kcal mol‑1) to be lower by as
much as the binding strength of MeCN to the vacant Fe(II) complex,
[Fe(η5-Cp*)(dppe)]+. By DFT, we estimate this value to be 12 kcal
mol−1, bringing the BDFE to 62 kcal mol−1. See: Zhang, F.; Jia, J.;
Dong, S.; Wang, W.; Tung, C.-H. Hydride Transfer from Iron(II)
Hydride Compounds to NAD(P)+ Analogues. Organometallics 2016,
35, 1151.
(19) Tilset, M.; Fjeldahl, I.; Hamon, J.-R.; Hamon, P.; Toupet, L.;
Saillard, J.-Y.; Costuas, K.; Haynes, A. Theoretical, Thermodynamic,
Spectroscopic, and Structural Studies of the Consequences of One-
Electron Oxidation on the Fe-X Bonds in 17- and 18-Electron
Cp*Fe(dppe)X Complexes (X = F, Cl, Br, I, H, CH3). J. Am. Chem.
Soc. 2001, 123, 9984.
(20) This value does not take into account the free energy associated
with acetonitrile binding.
(21) (a) Ilic, S.; Pandey Kadel, U.; Basdogan, Y.; Keith, J. A.; Glusac,
K. D. Thermodynamic Hydricities of Biomimetic Organic Hydride
Donors. J. Am. Chem. Soc. 2018, 140, 4569. (b) Ellis, W. W.; Raebiger,
J. W.; Curtis, C. J.; Bruno, J. W.; DuBois, D. L. Hydricities of
BzNADH, C5H5Mo(PMe3)(CO)2H, and C5H5Mo(PMe3)(CO)2H in
Acetonitrile. J. Am. Chem. Soc. 2004, 126, 2738.
(22) (a) Wiedner, E. S.; Chambers, M. B.; Pitman, C. L.; Bullock, R.
M.; Miller, A. J. M.; Appel, A. M. Thermodynamic hydricity of
Transition Metal Hydrides. Chem. Rev. 2016, 116, 8655. (b) Pearson,
R. G. The Transition-Metal-Hydrogen Bond. Chem. Rev. 1985, 85, 41.
́
(23) (a) Saveant, J.-M.; Tard, C. Proton-Coupled Electron Transfer
in Azobenzene/Hydrazobenzene Couples with pendant Acid-Base
Functions. Hydrogen-Bonding and Structural Effects. J. Am. Chem.
Soc. 2014, 136, 8907. (b) Cattaneo, M.; Ryken, S. A.; Mayer, J. M.
Outer-Sphere 2e−/2H+ Transfer Reactions of Ruthenium(II)-Amine
and Ruthenium(IV)-Amido Complexes. Angew. Chem., Int. Ed. 2017,
56, 3675.
(8) Chalkley, M. J.; Oyala, P. H.; Peters, J. C. Cp* Noninnocence
Leads to a Remarkably Weak C-H bond via Metallocene protonation.
J. Am. Chem. Soc. 2019, 141, 4721.
(9) Drover, M. W.; Schild, D. J.; Oyala, P. H.; Peters, J. C. Snapshots
of a Migrating H-Atom: Characterization of a Reactive Iron(III)
Indenide Hydride and its Nearly Isoenergetic Ring-Protonated
Iron(I) Isomer. Angew. Chem., Int. Ed. 2019, 58, 15504.
(10) Hamon, P.; Toupet, L.; Hamon, J. R.; Lapinte, C. Novel
diamagnetic and paramagnetic iron(II), iron(III), and iron(IV)
classical and nonclassical hydrides. X-ray crystal structure of
[Fe(C5Me5)(dppe)D]PF6. Organometallics 1992, 11, 1429.
(11) Hamon, P.; Hamon, J. R.; Lapinte, C. Isolation and
characterization of a cationic 19-electron iron(III) hydride complex;
electron transfer induced hydride migration by carbon monoxide at an
iron(III) center. J. Chem. Soc., Chem. Commun. 1992, 1602.
(12) For example: (a) Chiang, K. P.; Scarborough, C. C.; Horitani,
M.; Lees, N. S.; Ding, K.; Dugan, T. R.; Brennessel, W. W.; Bill, E.;
Hoffman, B. M.; Holland, P. L. Characterization of the Fe-H Bond in
a Three-Coordinate Terminal Hydride Complex of Iron(I). Angew.
Chem., Int. Ed. 2012, 51, 3658. (b) Gu, N. X.; Oyala, P. H.; Peters, J.
C. An S = 1/2 Iron Complex Featuring N2, Thiolate, and Hydride
Ligands: Reductive Elimination of H2 and Relevant Thermochemical
Fe-H Parameters. J. Am. Chem. Soc. 2018, 140, 6374.
(24) The FeII/I redox potential of endo and exo-[3]+ could not be
determined experimentally but the FeI/0 couples give conservative
lower limits of, respectively, −0.81 and −0.70 V vs Fc/Fc+.
(25) Goulet-Hanssens, A.; Utecht, M.; Mutruc, D.; Titov, E.;
Schwarz, J.; Grubert, L.; Bleger, D.; Saalfrank, P.; Hecht, S.
Electrocatalytic Z − > E Isomerization of Azobenzenes. J. Am.
Chem. Soc. 2017, 139, 335−341.
(13) Lucken, E. A. C. Nuclear Quadrupole Coupling Constants;
Academic Press: London, 1969; pp 227−131.
(14) For the IR spectrum of free Cp*H, see: Threlkel, R. S.; Bercaw,
J. E.; Seidler, P. F.; Stryker, J. M.; Bergman, R. G. 1,2,3,4,5-
pentamethylcyclopentadiene. Org. Synth. 1987, 65, 42.
(15) Green, M. L. H.; Pratt, L.; Wilkinson, G. 760. A new type of
transition metal− cyclopentadiene compound. J. Chem. Soc. 1959, 0,
3753−3767.
(16) Here, staggered refers to the puckered CH(CH3) group as
being on the opposite side of the [Fe]-coordinated CO ligand.
(17) (a) Paul, F.; Toupet, L.; Roisnel, T.; Hamon, P.; Lapinte, C.
Solid-state characterisation of the [(η2-dppe)(η5-C5Me5)FeCO]+
cation: an unexpected ‘oxidation’ product of the [(η2-dppe)(η5-
C5Me5)FeC = C(C6H4)NMe2]+ radical cation. C. R. Chim. 2005, 8,
1174. (b) Catheline, D.; Astruc, D. Synthesis and characterization of
+
−
−
C5(CH3)5Fe(CO)3 PF6 and C5(CH3)5Fe(CO)2 K+. J. Organomet.
Chem. 1982, 226, No. c52.
(18) In the reaction between [Fe(η5-Cp*)(dppe)(H)] and
[BNAP]+, the solvent adduct, [Fe(η5-Cp*)(dppe)(NCMe)]+, and
BNAPH are produced, in this case, acetonitrile binding will cause the
H
https://dx.doi.org/10.1021/jacs.0c09363
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX