Bioinspired CNP Iron(II) Pincers Relevant to [Fe]-Hydrogenase (Hmd): Effect of Dicarbonyl versus Monocarbonyl Motifs in H2 Activation and Transfer Hydrogenation

Article abstract of DOI:10.1021/acs.inorgchem.9b03476

A set of bioinspired carbamoyl CNP pincer complexes are reported that are relevant to [Fe]-hydrogenase (Hmd). The dicarbonyl species [(C^NHN^NHP^R2)Fe(CO)₂I] [R = Ph, 1; R = ⁱPr, 2] undergoes ligand deprotonation, resulting in the dearomatized complexes of formulas [(C^NHN^N=P^R2)Fe(CO)₂] (5 and 6). The crystal structure and ¹H{³¹P} NMR spectroscopy of the iodide-bound dearomatized species [Na(18-crown-6)][(C^NHN^N=P^Ph2)Fe(CO)₂I] (7) showed that the deprotonated moiety was the phosphoramine N(H) linkage. Separately, the monocarbonyl complexes [(C^NHN^NHP^R2)Fe(CO)(MeCN)₂](BF₄) (8 and 9) synthesized, as well as deprotonated and dearomatized in similar fashion. Reactivity studies revealed that the parent dicarbonyl complexes require more forceful conditions for H₂ activation, compared with the monocarbonyl complexes. The ligand backbone was not found to participate in H₂ activation and H₂ → hydride transfer to an organic substrate was not observed in either case. Density functional theory calculations revealed that the higher reactivity of the monocarbonyl complex in H₂ splitting could be attributed to its higher affinity for H₂. This behavior is attributed to two key points related to the requisite d_I(Fe) → σ*(H₂) back-bonding interaction in a conventional M-H₂ Kubas interaction: (i) generally, the weaker πdonor capacity of the dicarbonyls, and (ii) specifically, the detrimental effect of a strongly πacidic CO ligand (versus weakly πacidic MeCN ligand) trans to the H₂ activation site. The higher reactivity of the monocarbonyl complex is also evidenced by the catalytic transfer hydrogenation by monocarbonyl 8, whereas dicarbonyl 1 was ineffective. Overall, the results suggest that Nature uses the dicarbonyl motif in [Fe]-hydrogenase to diminish the interaction between the Fe center and dihydrogen, thereby preventing premature H₂ activation prior to substrate (H₄MPT⁺) binding and any resulting nonspecific hydride transfer reactivity.

Full text of DOI:10.1021/acs.inorgchem.9b03476

pubs.acs.org/IC
Article
Bioinspired CNP Iron(II) Pincers Relevant to [Fe]-Hydrogenase
(Hmd): Effect of Dicarbonyl versus Monocarbonyl Motifs in H2
Activation and Transfer Hydrogenation
Zhu-Lin Xie, Wenrui Chai, Spencer A. Kerns, Graeme A. Henkelman, and Michael J. Rose*
Cite This: https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: A set of bioinspired carbamoyl CNP pincer
complexes are reported that are relevant to [Fe]-hydrogenase
(Hmd). The dicarbonyl species [(CNHNNHPR2)Fe(CO)2I] [R = Ph,
i
1; R = Pr, 2] undergoes ligand deprotonation, resulting in the
dearomatized complexes of formulas [(CNHNN=PR2)Fe(CO)2] (5
1
and 6). The crystal structure and H{31P} NMR spectroscopy of
the iodide-bound dearomatized species [Na(18-crown-6)]-
[(CNHNN=PPh2)Fe(CO)2I] (7) showed that the deprotonated
moiety was the phosphoramine N(H) linkage. Separately, the
monocarbonyl complexes [(CNHNNHPR2)Fe(CO)(MeCN)2](BF4)
(8 and 9) synthesized, as well as deprotonated and dearomatized in
similar fashion. Reactivity studies revealed that the parent dicarbonyl
complexes require more forceful conditions for H2 activation,
compared with the monocarbonyl complexes. The ligand backbone was not found to participate in H2 activation and H2 → hydride
transfer to an organic substrate was not observed in either case. Density functional theory calculations revealed that the higher
reactivity of the monocarbonyl complex in H2 splitting could be attributed to its higher affinity for H2. This behavior is attributed to
two key points related to the requisite dπ(Fe) → σ*(H2) back-bonding interaction in a conventional M−H2 Kubas interaction: (i)
generally, the weaker π donor capacity of the dicarbonyls, and (ii) specifically, the detrimental effect of a strongly π acidic CO ligand
(versus weakly π acidic MeCN ligand) trans to the H2 activation site. The higher reactivity of the monocarbonyl complex is also
evidenced by the catalytic transfer hydrogenation by monocarbonyl 8, whereas dicarbonyl 1 was ineffective. Overall, the results suggest
that Nature uses the dicarbonyl motif in [Fe]-hydrogenase to diminish the interaction between the Fe center and dihydrogen,
thereby preventing premature H2 activation prior to substrate (H4MPT+) binding and any resulting nonspecific hydride transfer
reactivity.
genase.7 In contrast to the thorough understanding of the
bimetallic hydrogenases, studies of [Fe]-hydrogenase remain
relatively nascent. The enzyme [Fe]-hydrogenase catalyzes a
INTRODUCTION
■
As the availability and environmental cost of fossil fuels make
their extraction increasingly difficult, the world’s energy sources
are slowly shifting from traditional energies to renewable
non-redox hydride transfer from H2 to the substrate methenyl-
tetrahydromethanopterin (H4MPT+), which serves as a C1
energies.1−3 One promising alternative renewable energy source
carrier in methanogenic carbon dioxide (CO2) reduction (see
is dihydrogen (H2), which has many advantages: zero carbon
Scheme 1).7−9 This metalloenzyme (also called Hmd: H2-
emission, high mass energy density (120 MJ/kg), and high
energy efficiency in combustion.4 Despite the benefits of
dihydrogen, its generation limits its applications, because 95%
forming H4methylene-PT dehydrogenase) plays an obligate role
in “nickel-free” CO2 → CH4 metabolism in the absence of
bioavailable nickel (and thereby, [NiFe] hydrogenase).7−9 The
of H2 production is derived from carbon-intensive methods,
such as steam reforming and water−gas shift.5 Although
active site of [Fe]-hydrogenase (FeGP cofactor) exhibits a
unique array of nonproteinaceous ligands (except for Cys176),
including a cis-dicarbonyl motif, a bidentate pyridone-acyl unit
methods such as biomass fermentation and direct water-splitting
have drawn substantial attention from researchers, the efficiency
of H2 generation remains an issue.6 This motivates the
development of more-efficient catalysts for both H2 generation
and utilization derived from Earth-abundant elements.
Received: December 4, 2019
In Nature, H2 is both generated and utilized by hydrogenases.
To date, three types of hydrogenases have been discovered:
[NiFe]-hydrogenase, [FeFe]-hydrogenase, and [Fe]-hydro-
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
© XXXX American Chemical Society
Inorg. Chem. XXXX, XXX, XXX−XXX
A

Inorganic Chemistry
pubs.acs.org/IC
Article
Scheme 1. Reaction of Methenyl-H4MPT+ to Methylene-H4MPT by [Fe]-Hydrogenase and the Structure of FeGP Cofactor
that presents a unique (to biology) organometallic Fe−C bond,
and a substrate binding site occupied by a weakly bound H2O in
the resting state.10,11
2). Later, Hu and Shima26 reported semisynthetic [Fe]-
hydrogenases comprised of apoenzyme reconstituted with
pyridine- and pyridinol-containing iron complexes. The
construct containing 2-hydroxypyridine (conjugate base:
pyridone) exhibited detectable turnover frequencies in the
forward and reverse reactions of [Fe]-hydrogenase (2 s−1 and 1
s−1, respectively). In contrast, the methoxy analogue did not
exhibit any catalytic activity above the detection limit. These
results revealed the importance of the 2-hydroxy group
presumably as a pendant base that accepts H+ during H2
cleavage. The above reports also supported the second
mechanism13 (proposed by QM/MM) in which the protein
scaffold is imperative for the catalytic activity.
Computational studies of [Fe]-hydrogenase have evaluated
several plausible mechanisms of H2 splitting. Hall and co-
workers12 performed density functional theory (DFT) calcu-
lations on a simplified active site. This study suggested that the
active site binds and activates H2 in a stepwise fashion without
the participation of substrate (H4MPT+). The coordination site
trans to the acyl unit is ideal for binding and cleavage of H2, as it
requires the minimum structural change on the ligand backbone
during the catalytic circle. The main pathway of H2 heterolytic
splitting utilized metal−ligand bifunction between iron(II) and
the (deprotonated) pyridone-oxygen. The resulting iron-
hydride intermediate was calculated along the reaction
trajectory and provides hydride transfer to H4MPT+ with an
accessible activation barrier (15.2 kcal/mol). In 2014, Reiher
and co-workers13 reported a theoretical study based on the full
protein structure using a multiscale modeling method (quantum
mechanics/molecular mechanics (QM/MM)). In contrast to
the previously proposed mechanism, Reiher suggested that the
protein scaffold that harbors the FeGP cofactor provides the
necessary conformation change for the substrate to approach the
active site. As such, the coordinated H2 can be cleaved by the
orbital push−pull effect between the pyridone oxygen and the
cationic carbon on the substrate imidazolium ring (ΔEa = 1.0
kcal/mol). As a result, no formal iron-hydride intermediate was
implicated.
Our own foray into this area commenced with the
development of carbamoyl-pyridine-thioether (CNS) iron
carbonyl complexes (Scheme 2).27 The dicarbonyl bromide
species (j) bears a tridentate ligand backbone with C, N, S
ligating in meridional mode. With this pincer-type mimic, we
observed the first Fe−H species relevant to [Fe]-hydrogenase by
treating the bromide salt with KHBEt3. However, the thermal
instability of the hydride species resulted in the isolation of
desulfurized complex (k) with a [Fe2SMe2] core derived from C−
S bond cleavage. Similarly, treating the bromide complex with a
strong base, KOtBu, led to C−S bond cleavage, forming a CNC
pentacoordinate Fe complex (l). It is likely that KOtBu
generated a putative deprotonated carbamoyl intermediate
[(O=CN=NSMe)Fe(CO)2], which triggered the desulfurization
process. We have also reported model systems based on an
“anthracene scaffold”.28,29 The anthracene scaffold tethers the
C,N,S donors to provide the facial motif found in the active site.
Reactivity studies showed that the complexes [(Anth·
In a nonbiomimetic synthetic system (Scheme 2), Kirchner14
reported PNHNNHP iron(II) carbonyl complexes, in which the
dicarbonyl complex [(PNHN=NP)Fe(CO)2]+ cleaves H2 via a
metal−ligand bifunctional mechanism, leading to the hydride
complex [(PNHNNHP)Fe(CO)2(H)] (a; L = CO). Later, the
same group15 reported that the monocarbonyl complexes
[(PNHNNHP)Fe(CO)(L)(H)] (a), with labile ligands (L =
C
NHNSMe)Fe(CO)2(THF)]+ (m) and [(Anth·CNHNS)Fe-
(CO)2(PhCN)] (n) are functional for H2 activation.
Br−, MeCN, BH4 ) trans to the hydride, are efficient catalysts for
−
In this work, we report a bioinspired “hybrid” system that
contains elements of both the biological motif (iron-acyl
bonding) and proven catalytic platforms (phosphine pincer),
while maintaining a pendant base at the pyridine ortho position.
We aspired to “connect” the reactivity patterns of purely
bioinspired complexes (fac-CNS chelates) with purely synthetic
systems (mer-PNP chelates), as well as to clearly understand the
discrepancy between the dicarbonyl motif found in [Fe]-
hydrogenase and the monocarbonyl motif found in synthetic
phosphine catalysts. Herein, we report a bioinspired CNP pincer
complex with a biomimetic pendant base and carbamoyl group.
The experimental and computational studies were performed
regarding the synthesis, dearomatization, H2 activation, and
transfer hydrogenation within this model system.
hydrogenation of ketones and aldehydes to alcohols. In parallel,
methylene-spaced PNP iron monocarbonyl complexes (b)
developed by Milstein have been applied to hydrogenation of
CO2,16 ketones17,18 and trifluoroacetic esters19 with high
efficiency. Relatedly, Hu synthesized a third type of PNP iron
complex with an oxygen spacer (c) that performed methanol-
assisted H/D scrambling and hydrogenation and transfer
hydrogenation of aldehydes.20
Returning to [Fe]-hydrogenase, the mechanisms proposed in
DFT calculations warrant experimental investigation by
synthesizing model complexes that faithfully mimic the active
site. In this vein, Hu21−23 and (separately) Pickett24,25 reported
a series structural models containing acyl-methylpyridine (d, e,
and f) and “carbamoyl”-pyridine ligands (g, h, and i) (Scheme
B
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Scheme 2. Non-biomimetic and Biomimetic Iron Carbonyls for H2 Reactivity and Hydride Transfer: (i) PNP Pincer Iron
Carbonyls,14−20 (ii) Acyl-Methylpyridine Mimics,21−23 (iii) Carbamoyl-Pyridine Mimics,24,25 (iv) Carbamoyl-Pyridine-
Thioether Mimics,27 (v) Scaffold-Based Functional Mimics,28,29 and (vi) Structures Obtained in This Work
C
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Scheme 3. Synthesis of Monosubstituted Phosphine Ligands (LPh and LiPr) and Dicarbonyl Iron(II) Complexes (1, 2, 3, and 4)
Table 1. Selected Bond Lengths and Bond Angles
complex
3
4-H2O
7
cis-8
trans-8
1.955(4)
2.2930(11)
1.963(3)
1.366(6)
1.696(4)
1.793(4)
cis-9
Bond Lengths (Å)
1.959(4)
Fe−(C)carbamoyl
Fe−P
1.969(5)
2.3037(14)
1.947(4)
1.372(6)
1.690(4)
1.792(5)
1.751(7)
1.149(6)
1.146(7)
1.962(10)
2.288(3)
1.940(8)
1.377(14)
1.711(9)
1.809(12)
1.722(13)
1.109(14)
1.150(17)
1.972(7)
2.2729(18)
1.957(5)
1.343(9)
1.658(6)
1.778(7)
1.963(2)
2.2769(10)
1.937(3)
1.372(5)
1.701(4)
2.2976(5)
1.9456(16)
1.362(3)
Fe−NPy
N(−P)−CPy
N−P
1.7012(17)
Fe−C(O)in‑plane
Fe−C(O)out‑of‑plane
COin‑plane
COout‑of‑plane
Fe−I
a
a
1.831(8), 1.793(7)
1.089(9)
1.118(8), 1.198(7)
1.770(5)
1.143(5)
1.765(2)
1.139(5)
1.147(3)
a
2.4984(15), 2.4929(14)
b
Fe−NCMe
Fe−OH2
1.974(4), 1.932(4)
1.916(3), 1.908(4)
1.9717(17), 1.9291(17)
2.001(9)
Fe−OTf
2.064(5)
Bond Angles (deg)
164.94(14)
(O)C−Fe−P
NPy−Fe−P
MeCN−Fe−NCMe
163.60(16)
81.86(12)
165.0(3)
82.4(2)
161.3(2)
79.74(17)
164.01(13)
81.81(10)
175.76
164.29(6)
82.24(5)
87.15(7)
82.23(9)
86.11
a
b
Substitutional disorder of CO and I. trans to CO.
isopropyl substituents. In the 31P{H} NMR spectra, complex 1
displays a singlet at 90.8 ppm, whereas complex 2 gives rise to a
singlet at 121.2 ppm. The IR spectrum of 1 shows two carbonyl
stretches at 2032 and 1976 cm−1, and for 2 the carbonyl peaks
are red-shifted to 2019 and 1967 cm−1, because of the stronger
(iPr)2P donor. The CO stretching frequencies are slightly blue-
shifted, relative to the native enzyme data (Hmd: 1996, 1928
cm−1; FeGP cofactor: 2004, 1934 cm−1) and comparable to the
anthracene-based model complexes ([(Anth·CNHNSMe)Fe-
(CO)2 I]: 2031, 1981 cm− 1 ; [(Anth·CN H NS)Fe-
(CO)2(PhCN)]: 2016, 1956 cm−1).
RESULTS AND DISCUSSION
■
Synthesis, Spectroscopy, and X-ray Structure of
Dicarbonyl Complexes. The CNP pincer complexes herein
are derived from monosubstituted phosphine ligands (LPh and
LiPr; see Scheme 3), which were synthesized by a nucleophilic
substitution reaction of 2,6-diaminopyridine and chlorodiphe-
nylphosphine or chlorodiisopropylphosphine (1 equiv) by
modification of a reported procedure.30−32 The metalation of
LPh and LiPr proceeded at room temperature with
[Fe(CO)4(I)2] in DCM, which generated the dicarbonyl
complexes (1 and 2) with a bioinspired carbamoyl unit related
to Hmd. Complexes 1 and 2 were characterized by nuclear
magnetic resonance (1H NMR, 31P{1H} NMR) and infrared
(IR) spectroscopy. The 1H NMR spectrum of 1 exhibits features
at 9.64 and 8.74 ppm that correspond to carbamoyl-NH and
phosphino-NH, respectively. For 2, the resonances of NH
protons are shifted upfield (8.70 ppm for carbamoyl-NH, 6.93
ppm for phosphino-NH), indicating the increased electronic
density at the Fe center in 2 resulting from the electron-donating
To provide structural characterization and provide a more
labile coordinate site for reactivity studies, the thermodynami-
cally formed dicarbonyl complexes were derivatized by halide
abstraction with silver triflate, forming the triflate-bound
complex and AgI precipitate. Treatment of a 1,2-dicholoro-
ethane (DCE) solution of complex 1 or 2 with Ag(OTf)
resulting in the formation of either the adduct [(CNHNNHPPh2)-
Fe(CO)2(OTf)] (3) or the triflate salt [(CNHNNHPiPr2)Fe-
i
(CO)2(OTf)] (4), depending on the R group (R = Ph or Pr,
D
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 1. ORTEP diagrams for (a) [(CNHNNHPPh2)Fe(CO)2(OTf)] (3) (30% thermal ellipsoids) and (b) [(CNHNNHPiPr2)Fe(CO)2(H2O)](OTf)
(4-H2O) (50% thermal ellipsoids). The H atoms except for the NH protons are omitted for clarity.
Scheme 4. Synthesis of Dearomatized Complexes (5, 6)
respectively). The IR spectra exhibited the two carbonyl peaks at
higher wavenumber (2049, 1993 cm−1 for 3; 2045, 1994 cm−1
for 4) and the 31P resonance of 4 was observed as a singlet at
122.8 ppm. However, the 31P NMR of 3 showed no resonance
due to paramagnetic impurity. The crystal structures of the
triflate salts were obtained upon recrystallization of the products
(vapor diffusion of cyclohexane into the THF solution of
complexes). Surprisingly, complex 4 underwent ligand exchange
of OTf− by ambient H2O during crystallization (the
crystallization was set up outside the N2 box), generating
dicarbonyl complexes; the in-plane CO is further from the Fe
center than the axial CO ligand (1.792(5) Å vs 1.751(7) Å in 3;
1.809(12) Å vs 1.722(13) Å in 4), which can be ascribed to the
stronger trans influence of pyridine versus halide/H2O. This is
comparable to the Fe−CO bond length in cis-[Fe(PNP-iPr)-
(CO)2(Br)]+, where the CO in the plane of the ligand backbone
has a longer distance to the Fe center, compared with the CO
perpendicular to the plane (1.772(1) Å vs 1.758(2) Å).39 Both
complexes display an intermolecular hydrogen bonding motif on
the amide proton linkers in the solid state (Figures S37 and S38
in the Supporting Information). The NH−P proton interacts
with carbamoyl oxygen with bond metrics of O···H = 1.93 Å,
O···N = 2.781(7) Å, bond angle = 163° for 3, and O···H = 1.98
Å, O···N = 2.854(10) Å, bond angle = 174° for 4. In 4, the outer
sphere triflate anion displays hydrogen bonding with the NH−
carbamoyl proton with O···H = 2.07 Å, O···N = 2.932(7) Å,
bond angle = 168° for 1, and O···H = 2.10 Å, O···N = 2.913(15)
Å, bond angle = 154°.
Activation via Deprotonation/Dearomatization of
Dicarbonyl Complexes. Treatment of dicarbonyl complex 1
in THF with 1 equiv of the bulky phenolate, sodium 2,6-di-tert-
butyl-4-methoxyphenolate (NaDBHA), yielded an orange
solution containing the deprotonated/dearomatized product
formulated as [(CNHNN=PPh2)Fe(CO)2] (5) (Scheme 4). The
ν(CO) peaks were shifted to lower energy at 2006 and 1952
cm−1, indicating deprotonation of the ligand backbone. While
the precursor 1 possesses two acidic NH protons on the side
arms, 1H NMR after deprotonation provides no information on
determining the deprotonation site, as the spectrum shows the
disappearance of both NH resonances. The 31P{1H} NMR
spectrum of 5 exhibits a singlet at 90.6 ppm, which is a trivial
shift, compared to that of 1 (90.8 ppm). These results may be
H2O-bound complex [(CNHNNH iPr2)Fe(CO)2(H2O)](OTf)
P
(4-H2O). Selected bond parameters and crystal information are
tabulated in the Table 1 and Table S1 in the Supporting
Information. Complex 3 crystallized as a neutral complex
(Figure 1), wherein the Fe center is coordinated with a
carbamoyl-C, pyridine-N, and diphenyl-P in the conventional
“pincer” equatorial fashion. The CO ligands are arranged in cis
fashion, and the sixth coordination site is occupied by triflate
ligand. The bond distances are unremarkable, except that the
Fe−P bond distance (2.3037(14) Å) is longer than regular range
of bond length for iron(II) carbonyl phosphine complexes
(2.16−2.25 Å).14,15,17,18,33−36 This is attributed to the trans
influence of the strongly σ-donating carbamoyl unit.21 Complex
4 crystallized as a cationic complex, wherein the Fe center is
coordinated to the CNP ligand, cis CO ligands, and a water
moleculewith the charge balance provided by the outer-
sphere triflate anion. Presumably, the stronger σ-donating effect
of isopropyl-substituted phosphine ligand than diphenylphos-
phine37,38 gives rise to (i) the more labile triflate anion due to
increased electronic density at the Fe center, and (ii) the Fe−P
bond of 4 is slightly shortened (2.288(3) Å), compared with 3
(2.3037(14) Å). Two types of Fe−CO bond are presented in the
E
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 2. CO region in the IR spectra (drop-cast) of (a) 1 only, (b) reaction of 1 + NaDBHA producing 5, (c) reaction of 1 + [NEt4][DBHA] +
TlBArF producing 5, (d) reaction of 1 + NaDBHA + 18-crown-6 ether producing 7, and (e) reaction of 1 + [NEt4][DBHA] producing 7.
ascribed to a fast proton exchange process the NH proton with
protic species. Similarly, Kirchner et al. reported a fast proton
exchange process between the two N-sites of deprotonated
[Fe(PNP-BIPOL)(CH3CN)3]BF4 in MeCN, giving rise to a
singlet in 31P{1H} NMR.40 Since NMR results were ambiguous
in determining the exact deprotonation site of 1, DFT
calculations of P−NH-deprotonated 5 and carbamoyl NH-
deprotonated 5′ were performed at PBE0 theory level. The P−
NH-deprotonated 5DFT is shown to be 12.3 kcal/mol lower than
5′DFT in energy (see Table S2 in the Supporting Information),
indicating that the NH on the phosphine side arm is
deprotonated. The reversibility of the deprotonation was
demonstrated by treating 5 with pyridinium bromide in THF,
as 1H NMR and IR spectroscopies revealed the regeneration of
complex 1Br (Br-bound). For complex 2, similar deprotonation
was observed upon the treatment of NaDBHA and the IR
spectrum of the deprotonated complex 6 demonstrated
analogously red-shifted CO bands at 2001 and 1942 cm−1,
again consistent with deprotonation of the ligand backbone. The
31P{1H} NMR spectrum exhibits a singlet at 116.3 ppm.
To verify the dissociation of iodide from the Fe center in
complex 5 (lost as NaI), we substituted NaDBHA with other
bases and compared the changes of CO stretches in the solid-
state IR spectra recorded via the drop-cast method (see the
Supporting Information for details regarding the physical
measurements). As shown in Figure 2, treatment of 1 with the
corresponding ammonium salt, tetraethylammonium 2,6-di-tert-
butyl-4-methoxyphenolate ([NEt4][DBHA]) and a halide
abstracting reagent, thallium tetrakis(3,5-bis(trifluoromethyl)-
phenyl)borate (TlBArF), affords an IR spectrum that displays
two CO bands at 2004 and 1949 cm−1similar to those of 5.
This observation means that, in the reaction of NaDBHA and 1,
F
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
iodide dissociates from the Fe center, forming a pentacoordi-
nate/deprotonated species 5. Interestingly, the addition of a
reagent, [(18-crown-6)Na][DBHA] (Figure 2d) or [NEt4]-
[DBHA] (Figure 2e) to complex 1 gave rise to two CO peaks
red-shifted much further in the solid IR (1992, 1936 cm−1 and
1991, 1934 cm−1, respectively). The difference in the CO region
can be explained by the formation of an iodide-bound
deprotonated species 7, as evidenced by the crystal structure
(vide infra). It is noteworthy that the IR spectra in Figure 2 were
obtained by drop casting the sample solution that was taken
directly from the reaction. These results indicate that the iodide
is so tightly bound to the Fe center that, in solid state, without
Tl+ or Na+ cation, it ligates the iron, regardless of the
deprotonation of the ligand backbone.
While attempts to obtain the X-ray structure of the
dearomatized species 5 or 6 were not fruitful, the reaction of 1
with NaDBHA and 1 equiv of 18-crown-6 (Figure 2d) did afford
[(18-crown-6)Na][(CNHNN=PPh2)Fe(CO)2] (7) as orange
crystals. The structure of complex 7 exhibits an Fe center
coordinated to the CNP ligand, two cis-COs and I− anion,
displaying a pseudo-octohedral geometry (Figure 3). Interest-
Figure 4. Comparison of 1H NMR spectra of 7 and 5 in THF-d8. Peaks
corresponding to residual solvent impurity and phenol are labeled with
an asterisk (*).
suggest that, in solution, complex 7 loses iodide and converts to
the pentacoordinate complex 5. Similar dehydrohalogenation
has been observed or proposed in PNP iron complexes; the
deprotonation/dearomatization of the ligand backbone pro-
motes halide dissociation from the Fe center.14,17 Importantly,
such dearomatized and dehalogenated intermediate has been
proposed to be the active species to cleave H2 by the fashion of
metal−ligand cooperation.41 It is promising that the CNP pincer
complex could also function as a H2 activation complex;
however, with an anionic carbamoyl unit, the reactivity might be
different from the neutral PNP complexes. Since the iodide
anion is still dissolved in the solution, upon crystallization, it
1
binds with the Fe center, forming 7. While the H NMR
spectrum of complex 7 matches that of 5, the 1H NMR spectrum
of 7 exhibits the NH resonance at 8.90 ppm, which was not seen
in the original spectrum of 5. This can be attributed to the lack of
protic solvent to facilitate proton exchange of NH, whereas in 5,
phenol is present, leading to the extinction of the NH peak.
Synthesis of Monocarbonyl Complexes. To study the
effect of the number of CO ligands on tuning hydrogen
activation reactivity, we synthesized the corresponding mono-
carbonyl CNP complexes. Direct addition of stoichiometric
amounts of the decarbonylating reagent trimethylammonium N-
oxide (Me3NO) to 1 resulted in no reaction, and no change in
the CO bands was observed in the IR spectroscopy. Meanwhile,
treatment of 1 with excessive Me3NO led only to complete
decarbonylation of both CO ligands. The high electron density
on the Fe center produces strong backbonding from Fe to CO,
which stabilizes CO ligands on Fe. Previously, we reported that
decarbonylation can be facilitated by replacing the anionic
halide with neutral L-type ligand.42 Therefore, to attain the
monocarbonyl complex 8, we utilized a two-step synthesis
shown in Scheme 5. Starting with 1, the halide abstracting
reagent AgBF4 was added to the acetonitrile solution of 1,
generating the acetonitrile-bound dicarbonyl species. Then,
Me3NO was used to selectively eliminate one CO ligand,
affording the monocarbonyl species with two acetonitrile
molecules bound in trans or cis fashion, trans-8 and cis-8. The
IR spectrum of 8 (dropcast) exhibits a single CO band at 1987
Figure 3. ORTEP diagram (30% thermal ellipsoids) for [Na(18-crown-
6)][(CNHNN=PPh2)Fe(CO)2I] (7). H atoms except for the NH proton
are omitted for clarity.
ingly, the iodide anion shows substitutional disorder, because of
its coordination from both sides of the CNP plane (see Figure
S39 in the Supporting Information). The occupations of the two
isomers are ∼50% for each. The 18-crown-6-encapsulated Na+
cation binds with carbamoyl-O, at a distance of 2.309(5) Å.
Notably, the NH on the phosphine side arm was verifiably
deprotonated. As a result, the N(−P)−CPy and N−P distances
are shortened to 1.343(9) Å and 1.658(6) Å, respectively (∼1.37
Å and ∼1.70 Å for 3 and 4, respectively). The Fe−P bond length
also decreases (2.2729(18) Å), with respect to 3 (2.3037(14) Å)
and 4 (2.288(3) Å). The bond angle of P−N−C in 7 is more
acute (114°) than those of 3 and 4 (120° and 119°,
respectively). In addition, hydrogen bonding is observed
between carbamoyl-NH and crown-O (O···H = 2.24 Å, O···N
= 3.028(9) Å, bond angle = 153°). Importantly, no sign of
hydrogen bonding is observed on the phosphine side arm, which is
consistent with the deprotonation state of the phosphine-NH
linker.
Surprisingly, the dissolution of 7 in d8-THF gives rise to a 1H
NMR spectrum equivalent to that of 5 (Figure 4), and the
31P{1H} NMR spectrum of 7 displays a singlet peak at 91.4 ppm,
almost the same to the31P resonance of 5 (90.6 ppm). These
G
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Scheme 5. Synthesis of Monocarbonyl Complexes (trans-8 and cis-8)
1
cm−1. In H NMR, the cis and trans isomers can be readily
were recorded along with the reaction. As shown in Figure 6, the
1H NMR resonances derived from trans-8 gradually decreased,
as those of cis-8 increased; the same trend was observed in the
31P NMR spectrum. These results indicate that trans-8 is the
kinetic product and cis-8 is the thermodynamic product.
The energy difference between cis-8 and trans-8 was obtained
through DFT calculation by geometry optimization at PBE0/
opt (see the Supporting Information for details of the basis set)
level of theory. The converged structures (see Figures S44 and
S45 in the Supporting Information) precisely simulated the
structural characteristics of 8, indicating the accuracy of PBE0 in
structural calculation with transition-metal complexes. Impor-
tantly, the electronic energy of cis-8 is lower than trans-8 by
∼1.5 kcal/mol (see Table S2 in the Supporting Information).
This result is consistent with the NMR study of conversion of
the trans isomer to the cis isomer.
The isopropyl analogue complex 9 was synthesized by the
similar method, in which we only isolated the cis-9 in 75% yield.
A small amount (∼8%) of trans-9 complex could be observed
from the crude NMR (see Figures S33 and S34 in the
Supporting Information). The 1H NMR displays the NH
protons at 8.75 and 6.86 ppm, which are assignable to carbamoyl
NH and phosphine NH, respectively. The signals of isopropyl
protons exhibit in an asymmetric pattern, because the cis
acetonitrile ligands make the two isopropyl groups chemical
unequivalent (methine: 2.82, 2.69 ppm; methyl: 1.46−1.27
ppm). The 31P NMR shows a singlet at 110.6 ppm, lower than
cis-8. The CO stretching frequency of cis-9 appears at 1969
cm−1, red-shifted relative to 8, indicating a higher electron
density on the Fe of cis-9. The crystal structure of cis-9 is shown
in Figure 7, and the selected bond parameters are tabulated in
Table 1. The Fe atom coordinates with the CNP ligand, one CO
ligand, and two cis-acetonitrile molecules in a pseudo-octahedral
geometry. The Fe−C(O) bond distance is comparable with 4
(1.963(2) Å for cis-9 and 1.962(10) Å for 4). Note that, in cis-9,
the Fe−P bond length (2.2976(5) Å) is the longest among all
the structures.
identified at a ratio of ∼3:5. The assignments for each feature
were based on the 1H−31P HMBC NMR on the isolated product
in MeCN-d3 (see Figure S32 in the Supoorting Information).
The cis isomer exhibits two singlets at 8.85 and 7.88 ppm, which
correspond to the carbamoyl-NH and phosphino-NH, respec-
tively. The trans isomer displays the specific NH signals at 8.80
and 8.05 ppm. The protons at 5- and 3-positions on pyridine ring
give rise to two doublets at 6.47, 6.34 ppm for cis-8 and 6.72,
6.52 ppm for trans-8. In the aliphatic region, two singlet
resonances at 2.31 and 1.74 ppm are attributed to the methyl
protons of the coordinated acetonitrile of cis-8, and the singlet
observed at 1.56 ppm corresponds to the methyl protons of
trans-8. These methyl proton resonances persisted for about 1
day and disappeared upon ligand exchange with the deuterated
acetonitrile solvent. In 31P{1H} NMR, a singlet is observed at
89.4 ppm for cis-8, and trans-8 gives rise to a singlet at 100.5
ppm.
The crystal structure of 8 was obtained upon diffusion of
diethyl ether to the acetonitrile solution, and, interestingly,
trans-8 and cis-8 co-crystallized in the crystal lattice. The
structure is shown in Figure 5, and the selected bond parameters
Figure 5. Asymmetric part of the unit cell for 8: ORTEP diagrams (30%
thermal ellipsoids) for cis-[(CNHNNHPPh2)Fe(CO)(MeCN)2](BF4)
(cis-8) and trans-[(CNHNNHPPh2)Fe(CO)(MeCN)2](BF4) (trans-8).
Both complexes were found co-crystallized in the same unit cell. H
atoms except for the NH protons are omitted.
H2 Reactivity. Having proven that the phosphoramide NH of
CNP complexes can be deprotonated, the possibility of the CNP
pincers performing H2 cleavage was investigated. This would be
consistent with numerous literature examples of organometallic
iron pincer complexes. Kirchner et al. reported that the
dearomatized PNHNNHP−Fe pincer complex cleaves H2, the
proton of which protonates the ligand backbone and the hydride
binds to the Fe center.14 Milstein et al., synthesized a series of
methylene-derived PCH2NCH2P−Fe complexes, which perform
hydrogenation of aldehydes, ketones, and trifluoroacetic esters
with high efficiency.17,19,46 The active species of methylene-
deprotonated intermediate is also proven spectroscopically and
computationally to be critical for hydrogen splitting and hydride
transfer. Note that the organometallic pincer complexes are
mostly constructed via neutral phosphine tridentate ligands. It is
likely that, in our CNP complexes, the increase of the negative
charge of the ligand backbone from neutral to −1 would increase
are tabulated in Table 1. Both Fe atoms exhibit pseudo-
octahedral geometry and coordinate with the CNP ligand, one
CO ligand, and two acetonitrile molecules. The charge of the
−
complex is balanced by BF4 anion. The Fe−C(O) bond
distances are shortened (1.959(4) Å for cis-8 and 1.955(4) Å for
trans-8), compared with dicarbonyl complexes. It is noteworthy
that, in cis-8, the orientation of acetonitrile applies the weakest
trans-influence on pyridine, leading to the shortest Fe−Npy bond
length (1.937(3) Å) among all the structures. The Fe−NCMe
bond distances are among comparable range of iron(II)
acetonitrile bond distance,40,43−45 except for the longest
Fe(2)−N(9) bond (1.974(4) Å).
While room-temperature synthesis of the monocarbonyl
complexes afforded trans-8 as the major product, conversion of
trans-8 to cis-8 was observed at elevated temperature. The
solution of 8 in MeCN-d3 was heated at 55 °C and NMR spectra
H
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 6. 1H NMR and 31P{1H} NMR spectra showing interconversion of trans-[(CNHNNHPPh2)Fe(CO)(MeCN)2](BF4) (trans-8) to cis-
[(CNHNNHPPh2)Fe(CO)(MeCN)2](BF4) (cis-8) in MeCN-d3 solution incubated at 55 °C.
Figure 7. ORTEP diagrams (30% thermal ellipsoids) for cis-
[(CNHNNHPiPr2)Fe(CO)(MeCN)2](BF4) (cis-9). H atoms except for
the NH protons are omitted.
the reactivity in H2 activation. Moreover, in relation to the
bioinspiration for Hmd, we hypothesized that the non-
biomimetic monocarbonyl motif would provide a reactivity
advantage over the dicarbonyl motif found in the enzyme,
because of the higher electron density on the Fe center.
Figure 8. 2H NMR spectra of D2 activations with dicarbonyl complex 1
or monocarbonyl complex 8. Reaction conditions: (a) 1 + D2 (7 atm) +
KOtBu (1 equiv), THF, RT, 1 d; (b) 1 + D2 (7 atm) + KOtBu (2 equiv),
THF, RT, 1 d; (c) 8+ D2 (4 atm) + NaDBHA (1 equiv), THF, RT, 2 d;
(d) 8 + D2 (4 atm) + NaDBHA (1 equiv) + EtOH, THF, RT, 2 d. The
identity of the asterisk (*) feature in spectrum (c) remains unknown.
To test these hypotheses, we performed the H2 activation
studies using D2 and 2H NMR spectroscopy. To a high-pressure
NMR tube were charged the THF solution of complexes (1 or 8,
dicarbonyl and monocarbonyl complexes, respectively), base
(bulky phenolate or KOtBu), and D2 gas (4−7 atm); after a
specified time period, 2H NMR spectra were obtained to identify
the formation of deuterated products. As shown in Figure 8a,
treatment of dicarbonyl complex 1 with D2 and 1 equiv of either
NaDBHA (weaker phenolate base) or KOtBu (stronger alkoxide
base) did not result in any product peak, as evidenced by the 2H
NMR spectrum. However, inclusion of 2 equiv of KOtBu
resulted in a new feature at 3.16 ppm assigned to the tert-butanol
OD resonance (Figure 8b), indicating heterolytic D2 cleavage.
By comparison, complex 8 was found to readily activate D2 gas
(4 atm) in the presence of only 1 equiv of the weaker base
2
NaDBHA (Figure 8c). The H NMR spectrum exhibited two
product peaks at 5.56 and 2.56 ppm, assigned to deuterated 2,6-
di-tert-butyl-4-methoxyphenol (PhOD) and D2O, respectively,
as well as a feature at 3.27 ppm, which was derived from an as-
yet-unidentifiable alcohol −OD resonance, while the D2
splitting was not observed in the control experiments without
complex (same concentration of NaDBHA + 7 atm D2 in THF).
However, attempts to detect the putative iron deuteride species
with NMR were also not fruitful, even at lower temperature
(−40 °C). It is likely that the low concentration, as well as the
transient nature of the deuteride species, prohibited its
detection.
t
The BuOD formation suggests that D2 is cleaved by the
exogenous tert-butanoxide base. It is evident that the
deprotonated ligand backbone does not participate in the
2
reaction, because of the absence of ND signal in H NMR. In
Previously, it has been proposed that ethanol can promote H2
addition, the putative Fe−D intermediate was not detected.
cleavage catalyzed by a PNHNNHP iron complex by forming a
I
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
proton shuttle between the ligand-backbone pendant base and
the iron-bound dihydrogen molecule, thus decreasing the
reaction barrier for H2 splitting.15,47 To examine this possibility,
20 μL of ethanol (∼65 equiv) was added to the reaction
containing 8 (Figure 8d). Surprisingly, the reactivity of 8
a base-assisted mechanism by calculating the enthalpy change
(ΔH) of two subsequent steps involving dearomatized ligand
backbone and external bases: (i) H2 binding to the
pentacoordinate dearomatized complexes (A), generating H2
adducts (B); and (ii) H2 cleavage by external bases, forming
hydride intermediates (C). The ΔH via the reaction trajectory is
illustrated in Figure 9 and the enthalpies of base molecules have
2
decreased, as the H NMR spectrum only showed a negligible
peak attributed to EtOD signal and no features assignable to
PhOD or D2O was observed. The same result was also obtained
when NaDBHA was substituted by KOtBu. The disappearance
of PhOD and D2O peaks indicates that the smaller-sized
ethoxide, generated by NaDBHA or KOtBu, was involved in D2
cleavage, instead of the bulkier phenolate or tert-butoxide,
because it was easier for ethoxide to approach the Fe center.
Furthermore, the decreased reactivity of 8 is likely due to the
coordination of ethanol molecule to the Fe center, leading to a
lower amount of the active species cleaving D2. Overall, the
adverse effect of ethanol toward D2 activation suggests that the
ligand−metal cooperation pathway in which D2 is cleaved by Fe
center and the deprotonated NH moiety is not feasible. It is
external bases, e.g., tert-butoxide, bulky phenolate, or ethoxide,
that actually facilitate the reaction.
The aforementioned reactions were also investigated in
MeCN as solvent in lieu of THF. The 2H NMR spectra display
no resonances for deuterated species, except for those derived
from solvent and D2. This result indicates that D2 is activated
upon binding to the Fe center; with MeCN as a solvent, the
vacant site of Fe center is occupied by MeCN, which precludes
any reactivity with D2.
Next, we investigated the D2/deuteride transfer reactivity of 1
and 8 to assess the possibility of CNP complexes to perform the
complete Hmd-relevant process. Complexes 1 or 8 were treated
with D2, base and a deuteride acceptorsuch as an imidazolium
cation (1,3-bis(2,6-difluorophenyl)-2-(p-tolyl)imidazolium,
Im+), or the more strongly deuteride-accepting acridinium
cation (9-phenyl-10-methylacridinium, Ac+).48 However, nei-
ther 2H NMR spectroscopy nor mass spectrometry could detect
the corresponding deuteride-transferred product, Im−D or Ac−
D. The lack of deuteride transfer could be attributed to the high
Lewis acidity of the Fe center. Bound with only one anionic
donor, carbamoyl unit, as well as strong π-accepting CO ligand,
the Fe center functions as a deuteride acceptor, rather than a
deuteride donor.
DFT Study of H2 Activation. Based on the above reactivity
study, it is evident that the monocarbonyl complex exhibits
higher reactivity in H2 activation than the dicarbonyl complex.
To obtain theoretical insight, DFT was used to investigate H2
binding and activation. Three putative pentacoordinate active
species were considered (Scheme 6): the dicarbonyl complex
(AdiCO), the monocarbonyl complex with MeCN trans to the
vacant site (AMeCN), and the monocarbonyl complex with CO
trans to the vacant site (ACO). According to the experimental
evidence, the basic amide linker on the ligand backbone does not
participate in the reaction (vide infra). We therefore considered
Figure 9. Calculated relative enthalpies in the H2 binding and cleavage
reactions with external bases of ACO + NaDBHA (red), AdiCO
+
NaDBHA (solid green), AMeCN + NaDBHA (blue) and AdiCO + KOtBu
(dashed green) proceeding from the corrected iso-energetic state A
through B and forming C.
been taken into account in the total energy calculation, as shown
in the equation in Figure 9. In addition, we must point out that
the ΔG values of the reactions are dominated by ΔH as the
entropy change is negligible (TΔSA→B = −4.96 kcal/mol), and,
therefore, it is convenient to only consider the enthalpy change
in these reactions.
First, the H2 binding processes are all exothermic. AMeCN
shows the highest affinity for H2 (ΔHA→B = −22.7 kcal/mol)
and ACO has the lowest affinity for H2 (ΔHA→B = −13.6 kcal/
mol) with AdiCO exhibiting a moderate affinity (ΔHA→B = −15.2
kcal/mol) (Figure 9). While all three Fe complexes exhibit some
affinity for H2, the difference in reactivities does not solely lie on
the number of CO ligandsi.e., the electron density at the Fe
center. The lowest ΔHA→B for MeCN and the similar ΔHA→B
values between diCO and CO suggest that the ligand trans to the
H2 binding site determines the H2 affinity. According to Kubas
principles,49−51 the Fe−H2 interaction resembles the classic
Dewar−Chatt−Duncanson model, in which the filled σ(H2)
2
2
donates electron density to the empty dx −y (Fe) orbital, forming
a σ-bond, and the filled dxy(Fe) back-donates electron density to
σ*(H2), forming a π-interaction. By examining the MO diagram
of B (see Figure S57 in the Supporting Information), we
identified the σ bonding orbital of Fe−H2 for the complexes in
low energy orbitals (HOMO−40 for BCO; HOMO−27 for
B
diCO; MOHO−45 for BMeCN) and the π-interaction emerging
from back-donation are closer to the frontier orbitals: HOMO−
5 for BCO; HOMO−8 for BdiCO; MOHO−4 for BMeCN. The
weak π-accepting ligand MeCN trans to the vacant site is less
competitive (versus trans CO) with H2 for back-donation from
dxz(Fe)which is an orbital that the two trans ligands (H2 and
MeCN/CO) must share.52 Therefore, the trans MeCN case
provides the strongest Fe−H2 bonding interaction in the set.
Notably, the identity of the external base does not have an
apparent influence on ΔH of the H2 binding process. For the
dicarbonyl complex AdiCO, changing the base from NaDBHA
Scheme 6. Dearomatized Pentacoordinate Complexes
Considered in the DFT Calculations
J
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
a
Table 2. Reaction Optimization for Transfer Hydrogenation of iPrOH to Benzylaldehyde
c
Yield (%)
catalyst
time, t
base
temperature,
conversion
turnover number,
d
entry
(10 mol %)
(h)
base
equiv
T (°C)
solvent
iPrOH
iPrOH
iPrOH
iPrOH
iPrOH
iPrOH
iPrOH
iPrOH
iPrOH
iPrOH
iPrOH+H2O
iPrOH+H2O
iPrOH+H2O
iPrOH
iPrOH
iPrOH
iPrOH
(%)
a
b
c
TON
1
2
8
8
8
8
8
8
8
8
8
8
8
8
8
8
1
−
9
20
20
20
20
20
20
20
20
20
20
20
20
20
1
NaDBHA
NaDBHA
NaDBHA
NaDBHA
NaDBHA
DBU
1
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
60
60
60
40
RT
40
40
40
60
80
40
40
40
80
80
80
80
0
3
0
0
0
0
1
22
21
2
0
6
3
87
33
4
53
3
2
2
4
7
5
0
2
6
0
0
0
0
7
KOtBu
38
34
53
84
3
17
18
35
31
3
11
7
9
2
2
3
3
8
KOH
8
9
KOH
7
11
2
10
11
12
13
14
15
16
17
KOH
51
0
b
NaDBHA
KOtBu
KOH
0
26
2
8
11
0
6
2
0
KOH
83
40
37
58
29
18
13
21
47
22
23
32
7
3
2
1
2
20
20
20
KOH
1
KOH
1
KOH
5
a
b
General reaction conditions: 0.0157 mmol of base, 0.0523 mmol of benzylaldehyde, 10 mol % of catalyst, 1 mL of solvent. 0.9 mL iPrOH + 0.1
c
d
mL H2O. Yields calculated from the integration of GC-MS data. TON = (mmol of benzyl alcohol)/(mmol of catalyst).
(solid green line in Figure 9) to KOtBu (dashed green line)
slightly increases the enthalpy of the system (from −15.2 kcal/
mol to −14.7 kcal/mol). This is reasonable in that the base is not
directly participating in this step.
three. Importantly, although the number of CO ligands does
affect the electron density of the Fe center, it seems that the
identity of the trans ligand to the H2 binding site is more pivotal
in our pincer system. The trans MeCN ligand features weak σ-
donating and weak π-accepting ability, which promotes the
bonding of H2 to Fe center. This strategy has also been utilized
by [Fe]-hydrogenase, as the acyl moiety is a strong σ-donor and
weak π-acceptor and unquestionably enables the enzyme to
exhibit a stronger affinity for H2. Therefore, we believe this is a
crucial way to modulate trans influence to tune the activity of
catalysts for H2 activation.
Transfer Hydrogenation. The affinity of CNP complexes
to bind alcohols (vide supra) prompted us to test their function
as transfer hydrogenation (TH) catalysts to demonstrate the
functional reactivity of the CNP system. The selected reaction
was transfer hydrogenation of benzylaldehyde using iPrOH as a
hydrogen source. A complete screening of the base, temperature,
solvent, and catalysts was performed (see Table 2). Starting with
the monocarbonyl complex 8 and NaDBHA as a base at 60 °C,
increasing to the amount of NaDBHA to 3 equiv led to higher
reactivity with 87% of conversion of benzylaldehyde (see Table
2, entry 3). However, GC-MS results revealed the major product
to be the dienone b (53%) with the minor product as enone c
(6%). The desired hydrogen-transferred product benzyl alcohol
a formed with yield of 22%. The side products b and c were
produced from the Claisen−Schmidt condensation in which
acetone reacted with benzylaldehyde under basic conditions.53
Decreasing the temperature provided smaller amounts of side
products (see Table 2, entries 4 and 5), but the yield of benzyl
alcohol also decreased. Screening of different bases revealed that
the neutral nitrogen-based base, 1,8-diazabicyclo[5.4.0]undec-
7-ene (DBU), is not suitable to promote the reaction, most likely
Next, we considered the energy change of the H2 cleavage step
via base-assisted mechanism, in which H2 on B is heterolytically
cleaved, producing hydride intermediate C and conjugate acid.
When NaDBHA is chosen as a base (Figure 9, solid lines), all
three complexes give rise to exothermic results in which CMeCN
shows the lowest energy (−31.9 kcal/mol), CdiCO in the middle
(−26.4 kcal/mol), and CCO is the highest in energy (−18.2 kcal/
mol). Although the ΔHB→C values are comparable for all three
complexes, the total ΔHA→C for all three reactions shows clear
trend in that the monocarbonyl complex AMeCN is the most
reactive species in H2 activation and, significantly, ACO is the
least active species among them, indicating that ACO is not
responsible for H2 activation. This is consistent with the
experimental evidence that monocarbonyl complex 8 is more
easily activated (1 equiv of NaDBHA is sufficient for H2
heterolysis). Interestingly, substitution of NaDBHA for KOtBu
led to a drastic decrease in enthalpy for diCO (−52.1 kcal/mol),
making the H2 cleavage with AdiCO complex more feasible. This
is in agreement with the experimental results showing that
dicarbonyl complex 1 was able to activate H2 in the presence of 2
equiv KOtBu. These results demonstrate that the propensity of
H2 cleavage is highly dependent on the basicity of the external
base.
The calculated energy profile demonstrates the additive
effects of the trans influence in the primary coordination sphere
and the basicity of external bases in H2 activation. Both
theoretically and experimentally, we have showed that the
monocarbonyl complex AMeCN is the most reactive among the
K
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
because of its weaker basicity (Table 2, entry 6). Instead, using
the strong inorganic base KOH (Table 2, entry 8) afforded a
higher yield of benzyl alcohol and lower formation of side
products, compared with other organic bases (NaDBHA, DBU,
and KOtBu). The optimal temperature with KOH was
determined to be 60 °C, with the highest yield of a at 35%
(Table 2, entries 9 and 10). Water has also been reported to
benefit the TH reaction with PNP complexes, because of its
formation of a proton bridge, which facilitates the proton
transfer between the pendant base and the substrate.47 However,
complexes. Conditions for reactions (1 equiv of PhONa for 8
and 2 equiv of KOtBu for 1) show that monocarbonyl complex 8
exhibits higher reactivity for H2 activation than dicarbonyl
complex 1.
(4) DFT calculations show that the efficiency of H2 reactivity
follows the trend of CO < diCO < MeCN. This reactivity
difference is largely derived from the step of H2 binding, in which
electron density at the Fe center (minor effect) and influence of
ligand trans to the H2 binding site (major effect) jointly
modulate the affinity of H2 of the pentacoordinate species (A).
Subsequently, the cooperation of the Fe center and external
bases leads to H2 cleavage. The efficiency should be related to
the pKa of the base and the extent of H−H bond weakening,
which are consistent with the experimental results.
(5) The catalytic activity of the monocarbonyl CNP complex
8 toward transfer hydrogenation of benzylaldehyde is observed,
whereas the dicarbonyl complex 1 and −PiPr2-substituted
complex 9 exhibit lower to no reactivity. This could be ascribed
to the latter species’ more-stable hydride species, leading to low
hydricity and hydride transfer.
i
the addition of ∼10% water (VH O:VPrOH = 1:9) to the reaction
2
mixture suppressed the reaction, regardless of the base identity
(Table 2, entries 11−13). Shortening the reaction time from 20
h to 1 h (Table 2, entry 14) resulted in only a slightly lower yield
of a (29%), indicating that benzylaldehyde was hydrogenated
within the first hour of reaction. Lastly, the effects of changing
complexes on TH were investigated. Exchanging the mono-
carbonyl complex 8 for the dicarbonyl complex 1 led to decrease
of the yield of a to 18% (Table 2, entry 15), indicating lower
catalytic activity of the dicarbonyl system. It is noteworthy that,
in the absence of Fe complex, the TH reaction still occurred,
with benzyl alcohol formed at 13% yield (Table 2, entry 16).
Indeed, it has been reported that quantitative reduction of
acetophenone to phenylethanol was observed in a concentrated
NaOH solution.54 Surprisingly, substituting the isopropyl
monocarbonyl analogue (with −PiPr2) complex 9 for 8 (with
−PPh2) again provided a lower yield (21%; see Table 2, entry
17). This is comparable to the reaction with 1, indicating that the
isopropyl substitution does not promote the TH reactivity in
this series of CNP complexes.
(6) These results highlight three key items regarding the [Fe]-
hydrogenase enzyme:
(i) The counterintuitive use of the less H2-reactive
dicarbonyl motif may restrain the reactivity of [Fe]-
hydrogenase, thus providing more selectivity for hydride
transfer to the imidazolium unit of the H4MPT+.
(ii) The methylene-acyl carbanion present in the active site
occurs trans to the site of H2 activation; this strong σ
donor and weak π acceptor likely drives significant H−H
elongation in the putative Kubas intermediate.
Although some catalytic activity was observed with our CNP
complexes, the overall performance of the catalysts does not
compare favorably with other well-established catalysts.55 Such
low TH activity could be due to the low hydricity of the hydride
species (thermodynamic hydride donating ability),56 consider-
ing that the active species in TH is usually a metal-hydride
intermediate.55 For example, Hu reported the [Fe(H)(Br)-
(iii) The pyridone/pyridonate-O atom directly adjacent to the
H2 binding site in the enzyme likely provides both the
optimum basicity and orientation that promotes H2
heterolysis (ΔEa ≈ 1 kcal/mol, as estimated via the
QM/MM method).13
(7) Future bioinspired catalysts (with or without phosphine)
should aspire to achieve such a second-coordination sphere
feature. In addition, new catalysts and complexes with strong σ-
donating and weak π-accepting ligand trans to the hydride (e.g.,
H−, phosphine, and cyclopentadiene) would increase reactivities
toward both H2 splitting and hydride transfer.
i
(iPr2PONOP Pr2)] catalyzes the TH reaction to aldehydes
using formate as the hydride source, in which the dihydride
i
species [Fe(H)2(iPr2PONOP Pr2)] was proposed to be the
active species.20 Although we have not yet observed the hydride
intermediate in the present set of CNP complexes, it is
reasonable to hypothesize that the catalytic cycle involves the
formation of hydride species. In the previous section, we have
established the energy profile of H2 cleavage with the CNP
complexes. The DFT calculations suggest high stability of the
hydride species (C), which likely results in weak hydricity and,
thus, TH activity.
ASSOCIATED CONTENT
* Supporting Information
■
sı
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.inorgchem.9b03476.
Details of synthetic procedures, NMR spectra, IR spectra,
X-ray crystallography, and DFT calculations (PDF)
CONCLUSION
■
Accession Codes
The main conclusions of this work can be summarized as
follows:
CCDC 1907564−1907568 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/data_request/cif, or by email-
ing data_request@ccdc.cam.ac.uk, or by contacting The Cam-
bridge Crystallographic Data Centre, 12 Union Road, Cam-
bridge CB2 1EZ, UK; fax: +44 1223 336033.
(1) Bioinspired iron(II) carbonyl pincers are synthetically
accessible via metalation of the apo-ligand with a ferrous
carbonyl salt, which forms the Fe−C(carbamoyl) bond in situ.
(2) Base addition results in deprotonation of the phosphor-
amide unit (analogous to the Kirchner system), rather than the
carbamoyl unit.
AUTHOR INFORMATION
Corresponding Author
(3) Incubation of the pentacoordinate dearomatized/
deprotonated complexes in D2 gas leads to the appearance of
■
t
the PhOD or BuOD signals in 2H NMR spectroscopy,
Michael J. Rose − Department of Chemistry, The University of
Texas at Austin, Austin, Texas 78712, United States;
indicating the heterolytic cleavage of D2 with the CNP
L
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
orcid.org/0000-0002-6960-6639; Email: mrose@
pubs.acs.org/IC
Article
Hδ+-O, Bond and Methenyl-H4MPT+ Triggered Hydride Transfer. J.
Am. Chem. Soc. 2009, 131 (31), 10901−10908.
cm.utexas.edu
(13) Finkelmann, A. R.; Senn, H. M.; Reiher, M. Hydrogen-Activation
Mechanism of [Fe] Hydrogenase Revealed by Multi-Scale Modeling.
Chem. Sci. 2014, 5 (11), 4474−4482.
Authors
Zhu-Lin Xie − Department of Chemistry, The University of Texas
at Austin, Austin, Texas 78712, United States; orcid.org/
0000-0002-1969-2831
Wenrui Chai − Department of Chemistry, The University of Texas
at Austin, Austin, Texas 78712, United States; orcid.org/
0000-0002-5801-0135
Spencer A. Kerns − Department of Chemistry, The University of
Texas at Austin, Austin, Texas 78712, United States;
orcid.org/0000-0001-6936-8098
Graeme A. Henkelman − Department of Chemistry, The
University of Texas at Austin, Austin, Texas 78712, United
States; orcid.org/0000-0002-0336-7153
̈
(14) Bichler, B.; Holzhacker, C.; Stoger, B.; Puchberger, M.; Veiros, L.
F.; Kirchner, K. Heterolytic Cleavage of Dihydrogen by an Iron(II)
PNP Pincer Complex via Metal-Ligand Cooperation. Organometallics
2013, 32 (15), 4114−4121.
̈
(15) Gorgas, N.; Stoger, B.; Veiros, L. F.; Pittenauer, E.; Allmaier, G.;
Kirchner, K. Efficient Hydrogenation of Ketones and Aldehydes
Catalyzed by Well-Defined Iron(II) PNP Pincer Complexes: Evidence
for an Insertion Mechanism. Organometallics 2014, 33 (23), 6905−
6914.
(16) Langer, R.; Diskin-Posner, Y.; Leitus, G.; Shimon, L. J. W.; Ben-
David, Y.; Milstein, D. Low-Pressure Hydrogenation of Carbon Dioxide
Catalyzed by an Iron Pincer Complex Exhibiting Noble Metal Activity.
Angew. Chem., Int. Ed. 2011, 50 (42), 9948−9952.
(17) Langer, R.; Leitus, G.; Ben-David, Y.; Milstein, D. Efficient
Hydrogenation of Ketones Catalyzed by an Iron Pincer Complex.
Angew. Chem., Int. Ed. 2011, 50 (9), 2120−2124.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.inorgchem.9b03476
Notes
(18) Langer, R.; Iron, M. A.; Konstantinovski, L.; Diskin-Posner, Y.;
Leitus, G.; Ben-David, Y.; Milstein, D. Iron Borohydride Pincer
Complexes for the Efficient Hydrogenation of Ketones under Mild,
Base-Free Conditions: Synthesis and Mechanistic Insight. Chem. - Eur.
J. 2012, 18 (23), 7196−7209.
(19) Zell, T.; Ben-David, Y.; Milstein, D. Unprecedented Iron-
Catalyzed Ester Hydrogenation. Mild, Selective, and Efficient Hydro-
genation of Trifluoroacetic Esters to Alcohols Catalyzed by an Iron
Pincer Complex. Angew. Chem., Int. Ed. 2014, 53 (18), 4685−4689.
(20) Mazza, S.; Scopelliti, R.; Hu, X. Chemoselective Hydrogenation
and Transfer Hydrogenation of Aldehydes Catalyzed by Iron(II)
PONOP Pincer Complexes. Organometallics 2015, 34 (8), 1538−1545.
(21) Chen, D.; Scopelliti, R.; Hu, X. A Five-Coordinate Iron Center in
the Active Site of [Fe]-Hydrogenase: Hints from a Model Study. Angew.
Chem., Int. Ed. 2011, 50 (25), 5671−5673.
(22) Hu, B.; Chen, D.; Hu, X. Synthesis and Reactivity of
Mononuclear Iron Models of [Fe]-Hydrogenase That Contain an
Acylmethylpyridinol Ligand. Chem. - Eur. J. 2014, 20 (6), 1677−1682.
(23) Xu, T.; Yin, C.-J. M.; Wodrich, M. D.; Mazza, S.; Schultz, K. M.;
Scopelliti, R.; Hu, X. A Functional Model of [Fe]-Hydrogenase. J. Am.
Chem. Soc. 2016, 138 (10), 3270−3273.
(24) Turrell, P. J.; Hill, A. D.; Ibrahim, S. K.; Wright, J. A.; Pickett, C. J.
Ferracyclic Carbamoyl Complexes Related to the Active Site of [Fe]-
Hydrogenase. Dalton Trans. 2013, 42 (22), 8140−8146.
(25) Turrell, P. J.; Wright, J. A.; Peck, J. N. T.; Oganesyan, V. S.;
Pickett, C. J. The Third Hydrogenase: A Ferracyclic Carbamoyl with
Close Structural Analogy to the Active Site of Hmd. Angew. Chem., Int.
Ed. 2010, 49 (41), 7508−7511.
(26) Shima, S.; Chen, D.; Xu, T.; Wodrich, M. D.; Fujishiro, T.;
Schultz, K. M.; Kahnt, J.; Ataka, K.; Hu, X. Reconstitution of [Fe]-
Hydrogenase Using Model Complexes. Nat. Chem. 2015, 7 (12), 995−
1002.
(27) Durgaprasad, G.; Xie, Z.-L.; Rose, M. J. Iron Hydride Detection
and Intramolecular Hydride Transfer in a Synthetic Model of Mono-
Iron Hydrogenase with a CNS Chelate. Inorg. Chem. 2016, 55 (2),
386−389.
(28) Seo, J.; Manes, T. A.; Rose, M. J. Structural and Functional
Synthetic Model of Mono-Iron Hydrogenase Featuring an Anthracene
Scaffold. Nat. Chem. 2017, 9 (6), 552−557.
(29) Kerns, S. A.; Magtaan, A.-C.; Vong, P. R.; Rose, M. J. Functional
Hydride Transfer by a Thiolate-Containing Model of Mono-Iron
Hydrogenase Featuring an Anthracene Scaffold. Angew. Chem., Int. Ed.
2018, 57 (11), 2855−2858.
(30) Schroeder-Holzhacker, C.; Stoeger, B.; Pittenauer, E.; Allmaier,
G. G.; Veiros, L. F.; Kirchner, K.; et al. High-Spin Iron(II) Complexes
with Mono-Phosphorylated 2,6-Diaminopyridine Ligands. Monatsh.
Chem. 2016, 147 (9), 1539−1545.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Z.X., S.A.K., and M.J.R. acknowledge funding from the National
Science Foundation (NSF) (No. CHE-1808311) and the Welch
Foundation (No. F-1822), and are grateful to Dr. Vince Lynch
for assistance in solving X-ray structures. W.C. and G.H.
acknowledge funding from the Welch Foundation (No. F-1841)
and computational resources from the Texas Advanced
Computing Center.
REFERENCES
■
(1) Santhanam, K. S. V.; Press, R. J.; Miri, M. J.; Bailey, A. V.; Takacs,
G. A. Introduction to Hydrogen Technology, 2nd Edition; Wiley:
Hoboken, NJ, 2017.
(2) Momirlan, M.; Veziroglu, T. Current Status of Hydrogen Energy.
Renewable Sustainable Energy Rev. 2002, 6 (1−2), 141−179.
(3) Zhang, F.; Zhao, P.; Niu, M.; Maddy, J. The Survey of Key
Technologies in Hydrogen Energy Storage. Int. J. Hydrogen Energy
2016, 41 (33), 14535−14552.
(4) Møller, K. T.; Jensen, T. R.; Akiba, E.; Li, H. Hydrogen - A
Sustainable Energy Carrier. Prog. Nat. Sci. 2017, 27 (1), 34−40.
(5) Hydrogen Production Technical Team Roadmap. Technical Report;
U.S. Department of Energy Office of Scientific and Technical
Information, 2017 ( DOI: 10.2172/1220219).
(6) Holladay, J. D.; Hu, J.; King, D. L.; Wang, Y. An Overview of
Hydrogen Production Technologies. Catal. Today 2009, 139 (4), 244−
260.
(7) Lubitz, W.; Ogata, H.; Rudiger, O.; Reijerse, E. Hydrogenases.
̈
Chem. Rev. 2014, 114 (8), 4081−4148.
(8) Schilter, D.; Camara, J. M.; Huynh, M. T.; Hammes-Schiffer, S.;
Rauchfuss, T. B. Hydrogenase Enzymes and Their Synthetic Models:
The Role of Metal Hydrides. Chem. Rev. 2016, 116 (15), 8693−8749.
(9) Evans, D. J.; Pickett, C. J. Chemistry and the Hydrogenases. Chem.
Soc. Rev. 2003, 32 (5), 268−275.
(10) Shima, S.; Pilak, O.; Vogt, S.; Schick, M.; Stagni, M. S.; Meyer-
Klaucke, W.; Warkentin, E.; Thauer, R. K.; Ermler, U. The Crystal
Structure of [Fe]-Hydrogenase Reveals the Geometry of the Active
Site. Science 2008, 321 (5888), 572−575.
(11) Hiromoto, T.; Ataka, K.; Pilak, O.; Vogt, S.; Stagni, M. S.; Meyer-
Klaucke, W.; Warkentin, E.; Thauer, R. K.; Shima, S.; Ermler, U. The
Crystal Structure of C176A Mutated [Fe]-Hydrogenase Suggests an
Acyl-Iron Ligation in the Active Site Iron Complex. FEBS Lett. 2009,
583 (3), 585−590.
(12) Yang, X.; Hall, M. B. Monoiron Hydrogenase Catalysis:
Hydrogen Activation with the Formation of a Dihydrogen, Fe-Hδ−···
M
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
(31) Holzhacker, C.; Stoeger, B.; Carvalho, M. D.; Ferreira, L. P.;
Pittenauer, E.; Allmaier, G.; Veiros, L. F.; Realista, S.; Gil, A.; Calhorda,
M. J.; et al. Synthesis and Reactivity of TADDOL-Based Chiral Fe(II)
PNP Pincer Complexes-Solution Equilibria between K2P, N- and
K3P,N,P-Bound PNP Pincer Ligands. Dalton Trans. 2015, 44 (29),
13071−13086.
(48) Kalz, K. F.; Brinkmeier, A.; Dechert, S.; Mata, R. A.; Meyer, F.
Functional Model for the [Fe] Hydrogenase Inspired by the Frustrated
Lewis Pair Concept. J. Am. Chem. Soc. 2014, 136 (47), 16626−16634.
(49) Kubas, G. J. Hydrogen Activation on Organometallic Complexes
and H2 Production, Utilization, and Storage for Future Energy. J.
Organomet. Chem. 2009, 694 (17), 2648−2653.
(50) Kubas, G. J. Molecular Hydrogen Complexes: Coordination of a.
Sigma. Bond to Transition Metals. Acc. Chem. Res. 1988, 21 (3), 120−
128.
(32) Lackner-Warton, W.; Tanaka, S.; Standfest-Hauser, C. M.;
̈
̈
Oztopcu, O.; Hsieh, J.-C.; Mereiter, K.; Kirchner, K. Synthesis and
Characterization of Ruthenium P-Cymene Complexes Bearing
Bidentate P-N and E-N Ligands (E = S, Se) Based on 2-Aminopyridine.
Polyhedron 2010, 29 (16), 3097−3102.
(51) Kubas, G. J. Fundamentals of H2 Binding and Reactivity on
Transition Metals Underlying Hydrogenase Function and H2
Production and Storage. Chem. Rev. 2007, 107 (10), 4152−5205.
(52) Coe, B. J.; Glenwright, S. J. Trans-Effects in Octahedral
Transition Metal Complexes. Coord. Chem. Rev. 2000, 203 (1), 5−80.
(53) Satrawala, N.; Sharma, K. N.; Matsinha, L. C.; Maqeda, L.;
Siangwata, S.; Smith, G. S.; Joshi, R. K. Base-Catalyzed Cross Coupling
of Secondary Alcohols and Aryl-Aldehydes with Concomitant
Oxidation of Alcohols to Ketones: An Alternative Route for Synthesis
of the Claisen-Schmidt Condensation Products. Tetrahedron Lett.
2017, 58 (28), 2761−2764.
̈
(33) Gorgas, N.; Stoger, B.; Veiros, L. F.; Kirchner, K. Highly Efficient
and Selective Hydrogenation of Aldehydes: A Well-Defined Fe(II)
Catalyst Exhibits Noble-Metal Activity. ACS Catal. 2016, 6 (4), 2664−
2672.
̈
(34) Mastalir, M.; Glatz, M.; Gorgas, N.; Stoger, B.; Pittenauer, E.;
Allmaier, G.; Veiros, L. F.; Kirchner, K. Divergent Coupling of Alcohols
I
II
and Amines Catalyzed by Isoelectronic Hydride Mn and Fe PNP
Pincer Complexes. Chem. - Eur. J. 2016, 22 (35), 12316−12320.
̈
(35) Bertini, F.; Gorgas, N.; Stoger, B.; Peruzzini, M.; Veiros, L. F.;
(54) Ouali, A.; Majoral, J.-P.; Caminade, A.-M.; Taillefer, M. NaOH-
Promoted Hydrogen Transfer: Does NaOH or Traces of Transition
Metals Catalyze the Reaction? ChemCatChem 2009, 1 (4), 504−509.
(55) Clapham, S. E.; Hadzovic, A.; Morris, R. H. Mechanisms of the
H2-Hydrogenation and Transfer Hydrogenation of Polar Bonds
Catalyzed by Ruthenium Hydride Complexes. Coord. Chem. Rev.
2004, 248 (21−24), 2201−2237.
(56) Wiedner, E. S.; Chambers, M. B.; Pitman, C. L.; Bullock, R. M.;
Miller, A. J.; Appel, A. M. Thermodynamic Hydricity of Transition
Metal Hydrides. Chem. Rev. 2016, 116 (15), 8655−8692.
Kirchner, K.; Gonsalvi, L. Efficient and Mild Carbon Dioxide
Hydrogenation to Formate Catalyzed by Fe(II) Hydrido Carbonyl
Complexes Bearing 2,6-(Diaminopyridyl)Diphosphine Pincer Ligands.
ACS Catal. 2016, 6 (5), 2889−2893.
(36) Huber, R.; Passera, A.; Mezzetti, A. Iron(II)-Catalyzed
Hydrogenation of Acetophenone with a Chiral, Pyridine-Based PNP
Pincer Ligand: Support for an Outer-Sphere Mechanism. Organo-
metallics 2018, 37 (3), 396−405.
(37) Tolman, C. A. Steric Effects of Phosphorus Ligands in
Organometallic Chemistry and Homogeneous Catalysis. Chem. Rev.
1977, 77 (3), 313−348.
(38) Crabtree, R. H. Carbonyls, Phosphine Complexes, and Ligand
Substitution Reactions. In The Organometallic Chemistry of the
Transition Metals; John Wiley & Sons, Inc.: Hoboken, NJ, 2005; pp
87−124 ( DOI: 10.1002/0471718769.ch4).
̈
(39) Gorgas, N.; Alves, L. G.; Stoger, B.; Martins, A. M.; Veiros, L. F.;
Kirchner, K. Stable, Yet Highly Reactive Nonclassical Iron(II)
Polyhydride Pincer Complexes: Z-Selective Dimerization and Hydro-
boration of Terminal Alkynes. J. Am. Chem. Soc. 2017, 139 (24), 8130−
8133.
(40) Benito-Garagorri, D.; Becker, E.; Wiedermann, J.; Lackner, W.;
Pollak, M.; Mereiter, K.; Kisala, J.; Kirchner, K. Achiral and Chiral
Transition Metal Complexes with Modularly Designed Tridentate PNP
Pincer-Type Ligands Based on N-Heterocyclic Diamines. Organo-
metallics 2006, 25 (8), 1900−1913.
(41) Hale, L. V. A.; Szymczak, N. K. Hydrogen Transfer Catalysis
beyond the Primary Coordination Sphere. ACS Catal. 2018, 8 (7),
6446−6461.
(42) Xie, Z.-L.; Durgaprasad, G.; Ali, A. K.; Rose, M. J. Substitution
Reactions of Iron(II) Carbamoyl-Thioether Complexes Related to
Mono-Iron Hydrogenase. Dalton Trans. 2017, 46 (33), 10814−10829.
(43) Nakahara, Y.; Toda, T.; Kuwata, S. Iron and Ruthenium
Complexes Having a Pincer-Type Ligand with Two Protic Amidepyr-
azole Arms: Structures and Catalytic Application. Polyhedron 2018,
143, 105−110.
́
(44) Aloisi, A.; Berthet, J.-C.; Genre, C.; Thuery, P.; Cantat, T.
Complexes of the Tripodal Phosphine Ligands PhSi(XPPh2)3 (X =
CH2, O): Synthesis, Structure and Catalytic Activity in the Hydro-
boration of CO 2. Dalton Trans. 2016, 45 (37), 14774−14788.
(45) Sui-Seng, C.; Freutel, F.; Lough, A. J.; Morris, R. H. Highly
Efficient Catalyst Systems Using Iron Complexes with a Tetradentate
PNNP Ligand for the Asymmetric Hydrogenation of Polar Bonds.
Angew. Chem., Int. Ed. 2008, 47 (5), 940−943.
(46) Zell, T.; Ben-David, Y.; Milstein, D. Highly Efficient, General
Hydrogenation of Aldehydes Catalyzed by PNP Iron Pincer
Complexes. Catal. Sci. Technol. 2015, 5 (2), 822−826.
(47) Li, H.; Zheng, B.; Huang, K.-W. A New Class of PN3-Pincer
Ligands for Metal-Ligand Cooperative Catalysis. Coord. Chem. Rev.
2015, 293−294, 116−138.
N
https://dx.doi.org/10.1021/acs.inorgchem.9b03476
Inorg. Chem. XXXX, XXX, XXX−XXX