Synthetic Metallodithiolato Ligands as Pendant Bases in [FeIFeI], [FeI[Fe(NO)]II], and [(μ-H)FeIIFeII] Complexes

Article abstract of DOI:10.1021/acs.inorgchem.9b03409

The development of ligands with specific stereo- and electrochemical requirements that are necessary for catalyst design challenges synthetic chemists in academia and industry. The crucial aza-dithiolate linker in the active site of [FeFe]-H₂ase has inspired the development of synthetic analogues that utilize ligands which serve as conventional σ donors with pendant base features for H⁺ binding and delivery. Several MN₂S₂ complexes (M = Ni²⁺, [Fe(NO)]²⁺, [Co(NO)]²⁺, etc.) utilize these cis-dithiolates to bind low valent metals and also demonstrate the useful property of hemilability, i.e., alternate between bi- and monodentate ligation. Herein, synthetic efforts have led to the isolation and characterization of three heterotrimetallics that employ metallodithiolato ligand binding to di-iron scaffolds in three redox levels, (μ-pdt)[Fe(CO)₃]₂, (μ-pdt)[Fe(CO)₃][(Fe(NO))^II(IMe)(CO)]⁺, and (μ-pdt)(μ-H)[Fe^II(CO)₂(PMe₃)]₂⁺ to generate (μ-pdt)[(Fe^I(CO)₃][Fe^I(CO)₂·NiN₂S₂] (1), (μ-pdt)[Fe^I(CO)₃][(Fe(NO))^II(IMe)(CO)]⁺ (2), and (μ-pdt)(μ-H)[Fe^II(CO)₂(PMe₃)][Fe^II(CO)(PMe₃)·NiN₂S₂]⁺ (3) complexes (pdt = 1,3-propanedithiolate, IMe = 1,3-dimethylimidazole-2-ylidene, NiN₂S₂ = [N,N′-bis(2-mercaptidoethyl)-1,4-diazacycloheptane] nickel(II)). These complexes display efficient metallodithiolato binding to the di-iron scaffold with one thiolate-S, which allows the free unbound thiolate to potentially serve as a built-in pendant base to direct proton binding, promoting a possible Fe-H^-···⁺H-S coupling mechanism for the electrocatalytic hydrogen evolution reaction (HER) in the presence of acids. Ligand substitution studies on 1 indicate an associative/dissociative type reaction mechanism for the replacement of the NiN₂S₂ ligand, providing insight into the Fe-S bond strength.

Full text of DOI:10.1021/acs.inorgchem.9b03409

pubs.acs.org/IC
Article
Synthetic Metallodithiolato Ligands as Pendant Bases in [FeIFeI],
[FeI[Fe(NO)]II], and [(μ-H)FeIIFeII] Complexes
Kavindu Dilshan Kariyawasam Pathirana, Pokhraj Ghosh, Chung-H. Hsieh, Lindy Chase Elrod,
Nattamai Bhuvanesh, Donald J. Darensbourg,* and Marcetta Y. Darensbourg*
Cite This: https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: The development of ligands with specific stereo- and
electrochemical requirements that are necessary for catalyst design
challenges synthetic chemists in academia and industry. The crucial aza-
dithiolate linker in the active site of [FeFe]-H2ase has inspired the
development of synthetic analogues that utilize ligands which serve as
conventional σ donors with pendant base features for H+ binding and
delivery. Several MN2S2 complexes (M = Ni2+, [Fe(NO)]2+, [Co(NO)]2+,
etc.) utilize these cis-dithiolates to bind low valent metals and also
demonstrate the useful property of hemilability, i.e., alternate between bi-
and monodentate ligation. Herein, synthetic efforts have led to the isolation
and characterization of three heterotrimetallics that employ metal-
lodithiolato ligand binding to di-iron scaffolds in three redox levels, (μ-
pdt)[Fe(CO)3]2, (μ-pdt)[Fe(CO)3][(Fe(NO))II(IMe)(CO)]+, and (μ-
pdt)(μ-H)[FeII(CO)2(PMe3)]2 to generate (μ-pdt)[(FeI(CO)3][FeI(CO)2·NiN2S2] (1), (μ-pdt)[FeI(CO)3][(Fe(NO))II(IMe)-
(CO)]+ (2), and (μ-pdt)(μ-H)[FeII(CO)2(PMe3)][FeII(CO)(PMe3)·NiN2S2]+ (3) complexes (pdt = 1,3-propanedithiolate, IMe =
1,3-dimethylimidazole-2-ylidene, NiN2S2 = [N,N′-bis(2-mercaptidoethyl)-1,4-diazacycloheptane] nickel(II)). These complexes
display efficient metallodithiolato binding to the di-iron scaffold with one thiolate-S, which allows the free unbound thiolate to
potentially serve as a built-in pendant base to direct proton binding, promoting a possible Fe−H−···+H−S coupling mechanism for
the electrocatalytic hydrogen evolution reaction (HER) in the presence of acids. Ligand substitution studies on 1 indicate an
associative/dissociative type reaction mechanism for the replacement of the NiN2S2 ligand, providing insight into the Fe−S bond
strength.
+
We have long had interest in metallodithiolato ligands,
particularly MN2S2.10 As cis-dithiolates, they are generally
found as bidentate ligands creating M(μ-S)2M″ butterfly type
structures.10,11 These ligands are tunable based on the
hydrocarbon linkers between S and N and between N and
N, as well as the characteristics of the central metal, M.10 The
most common of these are square planar MN2S2 with M = NiII;
however square pyramidal systems (M = (VO)2+, [(NO)-
Fe]II or [(NO)Co]II) are also known to act as sulfur donor
based metallo-ligands to various receivers, particularly low
valent metals with additional carbonyl or nitrosyl ligands.12
Several bimetallic complexes have been developed in various
laboratories as electrocatalysts for the hydrogen evolution
reaction, HER, despite the absence of a well-positioned
pendant base as a proton shuttle.13−16 Nevertheless, computa-
INTRODUCTION
■
The search for innovative ligands to meet specific stereo- and
electrochemical requirements in catalyst development is of
imminent interest to synthetic chemists. Biomimetics of the
hydrogenase active sites have provided particular challenges
and opportunities to explore fundamental properties of abiotic
ligands that might provide function to the small molecules
similar to that of the metalloenzymes that exist with many
layers of control via the protein matrix.1−5 The two most
prominent hydrogenases, [FeFe]- and [NiFe]-H2ase, are
thiolate-S bridged bimetallics, with diatomics attached to
iron, CN−, and CO, in both.1,6,7 Hence, for biomimetic
purposes, organometallic chemists have used familiar ligands
such as phosphines and N-heterocyclic carbenes, of similar
donor ability to CN−, but having the advantage that they avoid
the problem of ligand-based, i.e., cyanide-nitrogen, protonation
that does not direct the proton toward iron, an ultimate
requirement for H+/H− coupling.5−8 Such proton shuttling
capability is readily arranged by the proteins, via an
azadithiolate in the [FeFe]-H2ase; a terminal cysteinyl sulfur
operates in this function for the [NiFe]-H2ase.9
Received: November 20, 2019
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
© XXXX American Chemical Society
Inorg. Chem. XXXX, XXX, XXX−XXX
A

Inorganic Chemistry
pubs.acs.org/IC
Article
tional studies aimed at exploring the mechanism of proton/
electron coupling according to the electrochemical events
displayed on the acid concentration-dependent cyclic
voltammograms indicated that one of the guiding principles
was the actuation of hemilability, Figure 1.17 As shown in
binuclear oxidative addition reaction, yielding the bridging
+
hydride, (μ-pdt)(μ-H)[FeII(CO)2PR3]2 , with oxidation state
conversions of FeI to FeII.21 While these CO/L substitution
reactions are on alternate iron centers, in another study it was
found that the nitrosonium ligand, NO+, isoelectronic with CO
but greater engendering electron deficiency on the metal to
which it adds, promotes a second CO/L exchange on the same
iron.22 That is, a completely dissymmetric di-iron complex is
produced: (μ-pdt)[Fe(CO)3][Fe(L′)(NO)(L)]+ where L =
N-heterocyclic carbene or phosphine. Notably, the NO+ ligand
effectively oxidizes one iron in the di-iron complex, giving the
opportunity to contrast di-iron complexes in three redox levels,
FeIFeI, FeIFeII, and FeIIFeII. We have attempted using the
NiN2S2 metalloligand as a surrogate for phosphine or NHC in
di-iron complexes. We are particularly interested in assessing
whether the metalloligand binds in bidentate or monodentate
fashion, and if the latter, how does the unbound thiolate orient
itself when twisted away from the first coordination sphere.
Thus, the trimetallic complexes were synthesized and
characterized by XRD: (μ-pdt)[FeI(CO)3][FeI(CO)2·
NiN2S2] (1), (μ-pdt)[FeI(CO)3][{Fe(NO)}II(IMe)·
−
NiN2S2]+BF4 (2), and (μ-pdt)(μ-H)[FeII(CO)2(PMe3)]-
[FeII(CO)(PMe3)·NiN2S2]+PF6 (3). Complex 1 was exam-
−
ined for the response of the cyclovoltammogram to added
equivalents of acid, i.e., for indications of electrocatalytic
proton reduction. Furthermore, a kinetic analysis of NiN2S2
ligand displacement in an attempt to obtain a measure of the
Fe−S bond strength of the neutral complex 1 was also
performed. Throughout this manuscript, NiN2S2 refers to the
[N,N′-bis(2-mercaptidoethyl)-1,4-diazacycloheptane] nickel-
(II) complex.
Figure 1. Hemilability in MN2S2 bridging thiolates.17
Figure 1, dissociation of one dative RS → M bond opens up
two reactive sites: a base on thiolate sulfur and a hole on the M
available for the addition of an electron. Specific examples are
presented in Figure 1; pathway A indicates the addition of an
electron, first to the open space on the 16-electron NiII side of
the bimetallic. A fast electron transfer to the more electrophilic
iron reduces FeII to FeI and opens the Fe−S bond. The B route
indicates that the first added electron, buffered by the electron-
delocalizing NO ligand, is insufficient to induce hemilability,
but rather both an electron and proton are needed prior to S−
Fe bond cleavage.12 In this way, the MN2S2 ligands may create
a temporary but necessary pendant base, avoiding the synthetic
requirement of building a permanently available strategically
located Lewis base.17
RESULTS AND DISCUSSION
■
Synthesis. Complex 1. Scheme 1 describes the synthetic
protocol employed to obtain the three complexes of this study.
Direct combination of the (μ-pdt)[FeI(CO)3]2 parent
compound with NiN2S2 over the course of 1 h showed
negligible reaction even on heating. However, with addition of
the decarbonylating agent Me3NO, the solution color
immediately changed from red to dark brown with a release
of CO2, giving the desired complex 1 (Scheme 1). Infrared
spectroscopy showed a bathochromic shift of ν(CO) stretching
frequencies to 2032(s), 1962(s), and 1900(w) upon binding of
the σ-donating NiN2S2 ligand (ν(CO) of (μ-pdt)[FeI(CO)3]2:
2072(m), 2037(s), 1990(s)). The pattern of ν(CO) stretching
frequencies matched that of the (μ-pdt)[FeI(CO)3]-
[FeI(CO)2PR3] species, indicative of monodentate binding of
the NiN2S2 ligand to the di-iron core (Table 1). The slight
lowering of ν(CO) values of complex 1 vs (μ-pdt)[FeI(CO)3]-
[FeI(CO)2PR3] indicates that the NiN2S2 metallodithiolato
ligand is a better donor than phosphines.23,24 Nevertheless, the
addition of PMe3 to 1 results in displacement of the NiN2S2
ligand, forming (μ-pdt)[FeI(CO)3][FeI(CO)2PMe3], 1-PMe3.
It is interesting to note that, unlike NiN2S2, PMe3 reacts
directly with the parent compound (μ-pdt)[FeI(CO)3]2
without the need for oxidative assistance in the elimination
of CO. As the NiN2S2 is a better electron-donating ligand than
PMe3, we postulate that this discrepancy is due to the larger
steric hindrance from NiN2S2 in the associative substitution
process.
Synthetic analogues of the [2Fe2S] active site of [FeFe]-
H2ase provide a unique opportunity to explore the influences
of specific changes in binuclear complexes. Previous studies on
the ubiquitous (μ-pdt)[Fe(CO)3]2 concluded that rates of
CO/CN− ligand substitution processes correlate with fluxion-
ality in the Fe(CO)3 rotors, that is, the intramolecular COap/
COba site exchange on individual metals.18−20 Such rotation,
likely promoted by an incoming ligand, coupled with shifting
the electron density in the Fe−Fe bond, leads to an open site
readily available to said donor, with concomitant release of a
CO (SI Figure S12).18 In this way, an associative mechanism
for ligand substitution on a pseudo 18-electron metal center is
made possible, on individual and alternate iron centers. The
monosubstituted (μ-pdt)[Fe(CO)3][Fe(CO)2CN]− is fol-
lowed by the second process and yields the much appreciated
20
2−
disubstituted dianion (μ-pdt)[Fe(CO)2CN]2
.
Phosphine
Complex 2. On mixing the parent compound (μ-pdt)[Fe-
−
mimics of this process lead to the analogous (μ-pdt)-
(CO)3][(Fe(NO))II(IMe)(CO)]+BF4 with NiN2S2, a clean
[FeI(CO)2PR3]2, which is capable of taking up a proton in a
conversion led to the paramagnetic complex 2 (Scheme 1) that
B
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Scheme 1. Synthesis of Complexes 1, 2, and 3
Figure 2. (A) Overlay of IR spectra recorded in CH2Cl2 solution of 2
−
(red) and the (μ-pdt)[FeI(CO)3][{Fe(NO)}II(IMe)(CO)]+BF4
(blue) starting material. (B) CH2Cl2 solution EPR spectra of 2 at
298 K The g value reported is from a simulation shown in the red
trace.
seems unlikely that the NiN2S2 would react easily with the iron
center that already contains a bulky IMe ligand. N-heterocyclic
carbene (IMes or IMe) stabilized (μ-pdt)[FeI(CO)3][{Fe-
(NO)}II(IMe)L]+ complexes have been previously studied
with respect to CO exchange (within the [(Fe(NO))II(CO)L]
moiety) using PMe3 and CN−, which seem to react similarly to
NiN2S2. The redox noninnocent ligand NO of the parent
compound creates a {FeNO}7 unit in the [{Fe(NO)}II(IMe)-
(CO)] center. Together with the electron withdrawing
capability of NO, the [{Fe(NO)}II(IMe)(CO)] center
presents a better electrophile compared to the [Fe(CO)3]
center to the incoming NiN2S2 ligand, which leads to the
formation of trisubstituted iron.
Table 1. IR Stretching Frequencies for Complexes 1, 2, and
3 and Starting Materials and Their Corresponding
Monosubstituted Phosphine Derivatives
diatomic vibrational
a
complexes
frequencies (cm−1
)
[FeIFeI]
(μ-pdt)[FeI(CO)3]2
Complex 3. Photolysis by sunlamp irradiation of (μ-pdt)(μ-
2072(m), 2037(s),
b 25
,
H)[FeII(CO)2(PMe3)]2 PF6 in the presence of NiN2S2
formed complex 3, over a period of 30 min in CH2Cl2
(Scheme 1). The change in pattern and position of ν(CO)
values, from 2032 and 1990 to 2023, 1968, and 1929 cm−1
reflects the increase in electron density in the di-iron core,
upon binding of NiN2S2 (Table 1). Thus, CO substitution
results in a tricarbonyl species making use of a metal-
lodithiolato ligand, NiN2S2, as has been established using
neutral or anionic σ-donor ligands. It is important to note that
the visible light irradiation is required to drive the reaction
forward. Complexes similar to (μ-pdt)(μ-H)-
+
−
1990(s)
2044(s), 1984(s),
(μ-pdt)[FeI(CO)3][FeI(CO)2(PPh3)]
b
26
,
1931(m)
(μ-pdt)[FeI(CO)3][FeI(CO)2(NiN2S2)] (1)
2032(s), 1962(s),
c d
,
1900(w)
[FeI(Fe(NO))II]
(μ-pdt)[FeI(CO)3][(Fe(NO))II(IMe)(CO)]+
2085(s), 2058(s), 2018(s),
c
22
,
1809(s)
2061(s), 1996(s),
(μ-pdt)[FeI(CO)3][(Fe(NO))II(IMe)
c
22
,
(PMe3)]+
1759(m)
(μ-pdt)[FeI(CO)3][(Fe(NO))II(IMe)
2056(s), 1990(s),
cd
,
(NiN2S2)]+ (2)
1740(m)
(μ-H)[FeIIFeII]
[FeII(CO)2(PMe3)]2 PF6 are known to eject a CO ligand
upon photolysis, producing an open site for the incoming
ligand. During the formation of complex 2, NiN2S2 readily
reacts with the [{Fe(NO)}II(IMe)(CO)] center without any
chemical/physical assistance to remove one of its CO’s to
generate a fully substituted iron center. However, formally
being in the FeII oxidation state, parent compound (μ-pdt)(μ-
+
−
c
21
,
(μ-H) (μ-pdt)[FeII(CO)2(PMe3)]2
2032(s), 1990(s)
2023(s), 1968(m),
+
(μ-H) (μ-pdt)[FeII(CO)2(PMe3)][FeII(CO)
cd
,
(PMe3)(NiN2S2)]+ (3)
1929(m)
a
b
Values underlined are for NO. Spectra measured in CH3CN.
c
d
Spectra measured in CH2Cl2. This work.
−
contains a fully substituted iron. Consistent with the better
donor ability of NiN2S2, similar bathochromic shifts in ν(CO)
and ν(NO) stretching frequencies were observed. The X-band
EPR spectra of complex 2, collected in perpendicular mode
field polarization in CH2Cl2 solutions at 298 K, showed the
presence of an S = 1/2 species with a g⊥ value of 2.09 (Figure
2B). It is important to note that the NiN2S2 ligand substitutes
the CO on the [{Fe(NO)}II(IMe)(CO)] center without any
external assisance from a reagent such as Me3NO, unlike in the
case of complex 1. Furthermore, in terms of steric hindrance, it
H)[FeII(CO)2(PMe3)2]+PF6 requires photolysis in order to
create an open site for the incoming NiN2S2.
X-ray Diffraction Analysis. X-ray quality crystals of
complexes 1, 2, and 3 were obtained by layering concentrated
CH2Cl2 solutions with hexane at −28 °C under nitrogen. All
three complexes show monodentate binding of the NiN2S2
ligand to the di-iron unit. The distance between the bound
sulfur and iron is around 2.3 Å, and the distance between
nonbound sulfur and the nearest iron within the trimetallic is
around 4 Å in all complexes (Figure 3). The NiII in the
C
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
−
Figure 3. Crystal structures of 1, 2 as its BF4 salt, and 3 as its BArF− salt.
metalloligand is maintained within a square planar geometry,
and the Ni···Feproximal distances are around 3.6 Å. (Table 2)
hydride distance to 3.06 Å (Table 2). This feature serves as a
potential structural platform to the “free thiolate as a pendant
base” thesis, such that the unbound thiolate can hold a proton
for a convenient proton/hydride coupling in dihydrogen
formation.
Electrochemistry. Complex 1 is the most well-behaved,
and its stability permits further electrochemical analysis.
Complex 2 is quite unstable, especially in the presence of
acid, and the opposite is the case with complex 3 and its very
stable bridging hydride. Two irreversible reductions (Epc) at
−1.82 V and −2.1 V are assigned for FeIFeI/FeIFe0 and
FeIFe0/Fe0Fe0 couples, respectively27 (Figure 4). The starting
Table 2. Selected Metric Parameters of 1, 2, and 3
1
2
3
Fe−Fe/Å
Fe−Sb*/Å
2.551 (2)
2.308 (4)
3.843 (5)
3.197 (5)
0.020
2.592 (9)
2.369 (9)
3.916 (1)
3.206 (1)
0.013
2.579 (8)
2.326 (1)
3.995 (2)
3.206 (2)
0.057
a
b
Fe−Snb*/Å
c
S*−S*/Å
Nidisp/Å
d
S*−Ni−S*/deg
93.6 (1)
95.03 (4)
94.9 (1)
a
b
Fe-bonding sulfur distance. Fe-nonbonding sulfur distance.
Sulfur−sulfur distance in NiN2S2. Displacement of Ni from N2S2
best plane.
c
d
Complex 1 shows that NiN2S2 binds to the iron center from
a basal position with an Fe···Sb distance of 2.30 Å. The
FeS2(CCO)2SNiN S moiety of 1 is square pyramidal (τ = 0.03)
2
2
The best plane of NiN2S2 is moderately coplanar with the
FeSpFe (where Sp is the S of pdt, proximal to NiN2S2) plane
for complex 1.
Complex 2 shows that the NiN2S2 binds to the iron center
from the basal position of the pseudosquare pyramidal Fe. The
Fe−Fe distance is slightly elongated (0.03 Å) compared to the
analogous CN− derivative. Due to the steric hindrance on the
substituted iron side, both the flap of the pdt unit and the Fe−
N−O angle are bent toward the unsubstituted iron. The
torsion angles, defined by Nap−Fe−Fe−Cap, are ∼7° for
complex 2 and the reported CN− derivative, while it is much
lower, ∼1°, for the reported PMe3 derivative. Conversely, the τ
value for the substituted Fe shows a discernible square
pyramidal geometric distortion for complex 2 and the reported
PMe3 derivative (0.22 and 0.26, respectively), while the
reported CN− substituted complex maintains a moderately
regular square pyramid (τ = 0.1). This can be attributed to the
steric interference from the PMe3 and the metallodithiolato
ligand as compared to the linear diatomic cyanide. The NiII,
present in a perfect square plane, is 3.4 Å away from the
proximal Fe, negating any bonding interaction (Table 2).
In complex 3, substitution of CO by NiN2S2 in (μ-pdt)(μ-
Figure 4. Scan rate dependence of complex 1 as a 2.0 mM CH3CN
solution containing 0.1 M [tBu4N][PF6]. (Scans were initiated in
negative direction.)
material (μ-pdt)[Fe(CO)3]2 shows two irreversible reductions
similarly at −1.34 V and −1.95 V,25 and it is reasonable to
assume that in complex 1 the shift in reduction potentials
toward a more negative value is due to the electron-donating
capability of the appended NiN2S2 ligand, which increases the
electron density at the di-iron core similarly to the (μ-
pdt)[Fe(CO)3][Fe(CO)2PMe3].27 Similar trends have been
observed in complexes such as (μ-pdt)[Fe(CO)2(PMe3)]2,
where a single reduction event corresponding to the FeIFeI/
FeIFe0 couple occurs at −1.85 V.25 A quasi-reversible
reduction event is formed at −2.41 V, which corresponds to
NiII/I reduction (SI Figure S9). Bimetallic complexes that are
composed of a metallodithiolato ligand and a Lewis acidic
metal center usually show a positive shift in the reduction
potential of the metal center in the metallodithiolato ligand
due to the decrease in electron density around it.10 However,
the reduction potential of Ni in complex 1 shows no significant
shift compared to free NiN2S2.28 One possible reason for this
might be the initial reduction of the di-iron core prior to the Ni
reduction. The effect of the reduced iron on the reduction
+
H)[FeII(CO)2(PMe3)]2 repositions the PMe3 ligands from
trans-basal to apical-basal, minimizing the steric interactions
between the NiN2S2 and PMe3. The flap created by the pdt is
oriented toward the disubstituted iron. Unlike other
complexes, NiN2S2 positions itself in a manner such that the
N2S2 plane is oriented toward the di-iron scaffold in complex 3,
thereby shortening the free-thiolate to potential bridging-
D
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
potential of the nickel is thus minimized, indeed almost
indistinguishable from that of free NiN2S2.29 Interestingly,
there is an enhancement in the current intensity at the NiII/I
reduction event which would support the hypothesis of NiN2S2
ligand dissociation and the buildup of free NiN2S2 at the
diffusion layer of the electrode. However (contrary to this
hypothesis), such an enhancement of intensity has been
observed in several examples of heterobimetallics containing
the NiN2S2 metaloligand, which are highly unlikely to be
dissociated.12 We conjecture that some type of kinetic control
(efficient electron transfer from electrode to the Ni center)
governs this process in the intact bimetallic complexes.
To test the ability of complex 1 to respond electrochemically
to added protons, cyclic voltammograms were recorded in the
presence of trifluoroacetic acid (TFA) in CH3CN. In the
presence of 10 equiv of acid, the ν(CO) IR spectrum shows
little change in pattern, indicating that complex 1 maintains its
structural integrity (Figure 5). However, minor shifts (∼7
Figure 7. Cyclic voltammograms of 1 in 2 mM CH3CN solution with
added equivalents of TFA. The black trace shows 16 equiv of free
TFA in the absence of a catalyst.
electrocatalytic waves vs acid concentration indicates steeper
slopes for the event at −1.48 V compared to the all-CO
starting material as well as (μ-pdt)[FeI(CO)3][FeI(CO)2PPh3]
(Figure 8), where the increase of sensitivity is on the order of 1
Figure 5. Overlay of IR spectra of 1 and 1H+ in CH2Cl2 solution.
*Likely decomposition product.
cm−1) toward more positive values are consistent with the
protonated dangling thiolate (Figure 5). Computational DFT
studies find that the HOMO of complex 1 is centered around
the unbound thiolate sulfur, further indicating its potential to
undergo protonation (Figure 6). The cyclic voltammogram of
Figure 8. Dependence of current heights of electrocatalytic waves on
acid concentration. (a) 1 (−1.48 V), (b) (μ-pdt)[Fe(CO)3][Fe-
(CO)2PPh3] (−1.73 V), (c) (μ-pdt)[Fe(CO)3]2 (−1.95 V).
> (μ-pdt)[FeI(CO)3][FeI(CO)2PPh3] > (μ-pdt)[[Fe(CO)3]2.
Many reported mimics of [FeFe]-H2ase show catalytic
properties following reduction of the Fe center which makes
it amenable to oxidatively add a proton converting it to a
hydride.21 Previous studies have shown that the all-CO
complex is likely to have an EECC type electrocatalytic
mechanism for proton reduction, especially at higher scan
rates; at lower rates, there is a decomposition.25,30 The mono-
substituted (μ-pdt)[FeI(CO)3][FeI(CO)3PPh3] displays a
similar electrochemistry to that of complex 1 and shows a
higher sensitivity toward the added acid compared to the all-
CO complex.25 The greater sensitivity of complex 1 to added
acid compared to both the monophosphine substituted and the
all-CO complexes is in agreement with the postulate that the
metallodithiolato ligand might act as a pendant base. As a good
donor, the NiN2S2 metallodithiolato ligand has the capability
of increasing the electron density in the di-iron core and in
addition will stabilize the oxidized center following the
oxidative addition of a proton to the iron. In the monodentate
Figure 6. HOMO of complex 1. See SI Table S1 for details.
complex 1 upon sequential addition of aliquots of acid showed
an increase in the cathodic current at the first reduction event
at −1.82 V, which is presumed to be the FeIFeI/FeIFe0
reduction (Figure 7). Furthermore, a new, steadily increasing
current event at −1.48 V can also be observed when increasing
the concentration of acid in the medium (Figure 7). The
attribution of the reduction-induced cathodic current enhance-
ment upon addition of TFA to electrocatalytic H2 production
was verified by gas chromatography and quantified by bulk
electrolysis experiments performed at −1.61 V for 30 min, in a
CH3CN solution containing 2 mM of catalyst and 50 equiv of
TFA (TOF, 17.5 s−1; TON, 0.22; FE, 78.4%; SI Figures S10−
S16, Tables S1−S3). The plot of current heights of
E
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
form, it harbors a nearby site for a proton to promote efficient
proton/hydride coupling. It is interesting to note that complex
1 shows an increase of catalytic current at −1.48 V at lower pH
values prior to the reduction in di-iron core. We postulate that
at lower pH values protonation at unbound thiolate leads to
the protonated complex 1H which is capable of accepting an
electron at a more positive reduction potential; this is arguably
the first step toward the catalytic proton reduction, i.e., the “C”
in C[ECEC]. Previous studies have shown that an electron-
rich di-iron core may be protonated to yield an FeII(μ-H)FeII
bridging hydride.21 Similarly, in complex 1, the substituted iron
has an increased electron density which we expect will take up
a proton to yield a hydride-bound di-iron core at lower pH
values probably via a concerted electron/proton transfer
pathway; however, a pH dependent study is required to
confirm this postulate. Computational studies have shown that
complex 1H has the capability to generate an asymmetric
hydride localized on the iron with the bound NiN2S2 (1HH)
where the H+···H− distance of 2.03 Å is favorable for coupling
(Figure 9).
Figure 10. (A) Disappearance of absorbance at 393 nm for complex
1, (B) 522 nm for complex 2, (C) 442 nm for complex 3 in DCM.
A simplistic relationship of the qualitative results of NiN2S2
ligand displacement is with the expected electrostatic
contribution to the Fe−S bonds in the metallodithiolato
ligand: FeIFeI−S2N2Ni < FeIFeII−S2N2Ni < < FeIIFeII−
S2N2Ni. This correlation is assuming positive character of
the NO-substituted Fe in complex 2 following the addition of
NO+, thus generating the Enemark−Feltham31 {Fe(NO)}7
unit as the NO+ withdraws sufficient electron density from FeI,
effectively generating FeII(NO)•. However, the steric encum-
brance from the NHC ligand in complex 2 is also a factor as
expressed in the longest Fe−S bond distance found, while the
all-CO derivative, complex 1, generates the shortest distance.
The conundrum of steric vs electronic control thus encouraged
further study.
The PMe3/NiN2S2 exchange in complex 1 proceeds at a rate
that is readily monitored by conventional techniques. We
designed such studies in order to establish the mechanistic
path and, if possible, to determine a kinetic measure of the
FeFe−S2N2Ni bond strength.
Rates of NiN2S2/PMe3 Ligand Exchange in Complex 1.
Kinetic parameters for the reactions of 1 with PMe3 were
monitored by IR and UV−vis spectroscopies. For IR
measurements, disappearance of the ν(CO) band at 1962
cm−1 with time was monitored and for UV−vis measurements;
disappearance of the absorbance at 393 nm which corresponds
to 1 was used, Figure 11.
In situ IR spectroscopy was initially used to identify the
reactants and products. Due to the smaller requirement of
sample concentration and faster spectral acquisition rates,
UV−vis spectroscopy was used for the determination of rates
as a function of temperature. The plots of ln(A0/At) versus
time showed good linearity as seen in Figure 12B. The derived
values of kobs at various PMe3 concentrations and reaction
temperatures, as determined by UV−vis spectroscopy, are
given in SI Table S4, entries 1−12. Each kobs was obtained
from a single kinetic run, with at least ∼100 spectral
accumulations per run. The plot of kobs versus [PMe3], Figure
12A, shows linearity. However, a nonzero intercept indicates
the reaction does not follow a strict second-order rate
expression, but rather both associative (A) and dissociative
(D) substitution processes are assumed to be involved, eq 4.
Hence, the plot of kobs versus [PMe3] at 20 °C, Figure 12A
and eq 5, yields the dissociative substitution rate constant, k1 =
2.29 × 10−5 s−1, derived from the intercept, and the rate
constant for the associative path is obtained from the slope, k2
= 4.90 × 10−3 M−1 s−1.
Figure 9. Modeled structure for complex 1HH. See SI Table S2 for
details.
Reactivity Studies Directed toward Evaluating FeFe−
S2N2Ni Bond Strengths. In order to rank the Fe−S bond
lability in the three FeFe complexes in different oxidation
states of iron and bearing NiN2S2 metalloligands, complexes 1,
2, and 3 in CH2Cl2 solvent and at 25 °C were exposed to a 20-
fold excess of PMe3 (eqs 1, 2, and 3). Under these conditions,
displacement of the NiN2S2 metalloligand in complex 1
yielding (μ-pdt)[Fe(CO)3][Fe(CO)2PMe3] (1-PMe3) was
complete within 15 min. Complex 2 more slowly converts to
the PMe3 analogue, requiring ≤2 h, and complex 3 shows no
reaction whatsoever over the course of 3 h.
complex 1 + PMe3
→ (μ‐pdt)[(FeI(CO)3][FeI(CO)2PMe3]
(1)
complex 2 + PMe3
→ (μ‐pdt)[FeI(CO)3][{Fe(NO)} (IMe)(PMe3)]+BF−4
II
(2)
complex 3 + PMe3 −×→(μ‐pdt)(μ‐H)[FeII(CO)2
(PME3)][FeII(CO)(PMe3)2]+PF−6
(3)
The disappearance of respective absorbances of complexes 1,
2, and 3 in the UV−vis spectra were used to monitor the
reactions (Figure 10); mass spectrometry was used to identify
both initial and final products (SI Figures S17−S19).
F
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 11. (a) Infrared spectra of 5 mM solution of 1 in DCM with 20-fold excess of PMe3, in the 1800−2100 cm−1 region, as NiN2S2/PMe3
exchange takes place to produce the complex 1-PMe3. (b) UV−vis spectra of converting 0.3 mM solution of complex 1 (initial: blue trace) to 1-
PMe3 (final: green trace) in DCM.
Figure 12. (a) Plots of kobs vs [PMe3] for the formation of 1-PMe3 from 1 measured at 20/25/30/35 °C (conditions: SI Table S4, entries 1−12).
(b) Example plot of ln(A0/At) vs time over four half-lives (conditions: SI Table S4, entry 1).
rate = k1[1] + k2[1][PMe3]
Table 3. Temperature Dependence of Reaction of PMe3
with 1
(4)
(5)
rate = kobs[1]; kobs = k1 + k2[PMe3]
T,
b
rxn path
°C
k
activation parameters
a
1 → 1-PMe3
10−3k2, M−1 s−1
The same reaction was monitored using in situ IR spectros-
copy at −40 °C. The plot of kobs vs [PMe3] obtained from this
data also resulted in a linear plot with nonzero intercept (SI
Table S3, entries 1−3, Figure S14), confirming that the rate
expression contains both associative and dissociative compo-
nents. It should be noted that the IR and UV−vis traces
represent a stoichiometric reaction, and reproducibility is on
the order of 95%.
The temperature dependence of rate constants for both A
and D pathways for the formation of complex 1-PMe3 from 1
obtained by UV−vis are given in Table 3 and presented
graphically as Eyring plots, Figure 13. Extrapolation of Eyring
plots led to a calculated value of k2 = 7.88 × 10−3 M−1 s−1 and
k1 = 1.31 × 10−7 s−1 at −40 °C. These calculated values match
the experimentally determined values for k2 and k1 at −40 °C
using the in situ IR spectroscopic monitor (SI Table S3) within
an acceptable margin, further confirming the accuracy of the
method.
(A path)
20
25
30
35
4.90
5.56
6.42
6.91
Ea = 17 kJ/mol
ΔH⧧ = 15 ( 1) kJ/mol
ΔS⧧ = −237 ( 5) J/mol.K
ΔG⧧ = 86 kJ/mol (298 K)
10−4k1, s−1
0.23
a
1 → 1-PMe3
(D path)
20
25
30
35
Ea = 101 kJ/mol
0.51
1.01
1.70
ΔH⧧ = 98 ( 5) kJ/mol
ΔS⧧ = 3 ( 17) J/mol.K
ΔG⧧ = 97 kJ/mol (298 K)
a
b
[a] = 3.1 × 10−4 M; [PMe3] = 0.0627−0.1254 M. Errors in
activation parameters reported at 95% confidence level.
(μ‐pdt)[FeI(CO)3][FeI(CO)2PMe3] + PMe3
Eact
⎯⎯→ (μ‐pdt)[Fe(CO)2(PMe3)]2
(6)
This reaction was previously reported to follow a similar
associative pathway;20 however, the reported activation energy
barrier for the CO displacement reaction is higher compared to
the PMe3 substitution reaction of NiN2S2 in 1.20 This is
possibly due to the fact that the NiN2S2 ligand, according to
the ν(CO) IR data, is a better electron donor compared to
PMe3 (Table 1). The proposed mechanism for such reactions
(Scheme 2) specifies that Eact(overall) is composed of two
contributory factors: Eact(attack) and Eact(rot’n).18 Accepting
Calculated values for the enthalpy and entropy of activation
are listed in Table 3. The small values of ΔH⧧ (15 ( 1) kJ/
mol) and large negative values for ΔS⧧ (−237 ( 5) J/mol·K)
are consistent with the associative pathway. Equation 6
describes the PMe3/CO substitution reaction of a mono-
substituted phosphine complex, affording a disubstituted
phosphine complex.
G
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
active site. Interestingly, in no case were we successful in
generating the bidentate form of complex 1·NiN2S2. Electro-
chemical studies of complex 1 in the presence of added
equivalents of trifluoraacetic acid showed a catalytic current
response with the formation of a new event at −1.48 V, which
is prior to the reduction of the di-iron core. This, we posit, is
due to the protonation at free thiolate, which makes the
trimetallic complex more susceptible to reduction. Further-
more, complex 1 shows higher sensitivity toward the added
acid compared to (μ-pdt)[[Fe(CO)3]2 and (μ-pdt)-
[FeI(CO)3][FeI(CO)3PPh3]. This result is taken as consistent
with unbound thiolate acting as a pendant base that assists in
guiding the proton toward the di-iron core. Ligand substitution
studies indicate the ability of metallodithiolato ligand to
undergo substitution reactions in the presence of the PMe3
ligand according to a dual, associative + dissociative pathway;
the dissociative path provided activation parameters giving a
kinetic measure of the Fe−S bond strength of ca. 98 kJ/mol.
Thus, for synthetic purposes, such metallodithiolato ligands
can add to the organometallic ligand tool box with features32
that give flexibility in positioning a second nucleophilic site.
Figure 13. Eyring plots for the formation of complex 1-PMe3 from
the associative ( ) and dissociative ( ) paths.
■
●
this analysis, we conclude that during the associative pathway,
PMe3 attacks the more electrophilic [Fe(CO)3] center, leading
to shift one of its CO’s underneath the Fe−Fe vector (rotation
of the Fe(CO)3 unit) where it is stabilized by the electron-rich
side to create the transition state (Scheme 2). Due to the
assistance from the [Fe(CO)2·NiN2S2] center to stabilize the
rotated transition state, the contribution from Eact(rot’n)
becomes significantly smaller, thereby lowering the
Eact(overall) compared to the reaction described by eq 6.20
Large values of ΔH⧧ and positive values for ΔS⧧ are
hallmarks for dissociative pathways. The dissociative path for
the reaction between 1 and PMe3 yields a ΔH⧧ value of 98 kJ/
mol, possibly due to the detachment of the NiN2S2 ligand from
the di-iron core, which is preparatory to the transition state,
which permits the attachment of the incoming PMe3 ligand
(Scheme 3). Thus, we conclude that for complex 1 the Fe−S
bond energy is ca. 98 kJ/mol.
EXPERIMENTAL SECTION
■
Methods and Materials. An MBraun Manual Solvent Purifica-
tion System with Alcoa F200 activated alumina desiccant was used to
purify dichloromethane (CH2Cl2), acetonitrile (CH3CN), tetrahy-
drofuran (THF), hexane, and diethyl ether. All reactions and
operations were carried out using standard Schlenk conditions
under a N2 atmosphere. All regents were used as received from
standard vendors unless otherwise stated. The known complexes (μ-
−
pdt)[Fe(CO)3]2, (μ-pdt)[Fe(CO)3][(Fe(NO))II(IMe)(CO)]+BF4 ,
+
−
(μ-pdt)(μ-H)[FeII(CO)2(PMe3)]2 PF6 , and Ni(bme-dach) (bme-
dach = [N,N′-bis(2-mercaptidoethyl)-1,4-diazacycloheptane] nickel-
(II) were prepared according to the literature procedures.22,33
Physical Measurements. The Bruker Tensor 37 Fourier
transform IR (FTIR) spectrometer was used to record solution
infrared spectra, using a CaF2 cell with a 0.2 mm path length. Kinetic
measurements using solution infrared spectroscopy were carried out
and recorded on a Bruker Tensor 27 FTIR spectrometer. UV−vis
measurements were obtained using a Shimadzu UV-2450 UV−vis
spectrophotometer. Eyela PSL-1400 and VWR polyscience 1162
instruments were used as a controlled temperature baths in kinetic
experiments. The Texas A&M Laboratory for Biological Mass
Spectrometry performed all mass spectrometry. The Atlantic Microlab
Inc. located in Norcross, Georgia, performed all elemental analyses.
1H/31P NMR were recorded on the Bruker 400 MHz super-
conducting NMR instrument.
CONCLUSIONS
■
The metallodithiolato ligand (NiN2S2) efficiently replaces
phosphines or even π-acid ligands such as CN− or CO in
binding to the propanedithiolate bridged di-iron scaffold, a
famous precursor to numerous biomimetics of the [FeFe]-
H2ase active site. The [FeFe]·NiN2S2 trimetallics were
characterized in three redox states, [FeIFeI] (1), [FeI[Fe-
(NO)]II] (2), and [(μ-H)FeIIFeII] (3), using IR spectroscopy,
ESI-MS, and X-ray crystallography. In all, the cis-dithiolate,
NiN2S2, displays monodentate binding using one sulfur, while
the unbound thiolate has the potential to act as a pendant base,
modeling the bridge-head amine of the [FeFe] enzyme active
site or the terminal cysteinyl thiolate in the [NiFe]-H2ase
X-ray Diffraction Analysis. The crystal data for 1, 2, and 3 were
determined using a Bruker Apex 2 X-ray (three-circle) diffractometer
with a Mo sealed X-ray tube (Kα = 0.70173 Å) at 110 K. Weighted
least-squares refinement on F0 was used for refining the structures. At
idealized positions, hydrogen atoms were placed, and fixed isotropic
Scheme 2. Proposed Mechanism for the Associative Path
H
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Scheme 3. Dissociative Path for PMe3/NiN2S2 Exchange
displacement parameters were used to refine them. Anisotropic
displacement parameters were employed for all non-hydrogen atoms.
The following programs were used: data collection, APEX3; data
reduction, SAINT; absorption correction, SADABS; and cell refine-
ment and structure solutions, SHELXS/XT. The final data
presentation and structure plots were generated in Olex2.30.
Crystallographic data for the complexes 1, 2, and 3 are deposited in
the Cambridge Crystallographic Data Centre; their numbers are
CCDC 1965616 (1), CCDC 1965614 (2), and CCDC 1965615 (3).
Crystal data and structure refinement parameters are summarized in
SI Tables S5, S6, and S7.
Electrochemistry. A custom-made three-necked conical shaped
flask with an outlet port/gas inlet equipped with a three-electrode
setup was used to obtain cyclic voltammograms (CVs) using a
CH1600E electrochemical analyzer (HCH instruments, Inc.). The
working electrode was a 0.071 cm2 glassy carbon dish electrode. The
counter electrode was a platinum wire, and the reference electrode
was Ag/Ag+ using a AgNO3 1 mM solution in CH3CN in a Vycor-
tipped glass tube. Diamond paste (3 μm) was used to polish the
working electrode, and a blanket of Ar was maintained over the
solution throughout the experiment following an initial Ar purge for 5
min. All CVs were obtained in CH3CN at room temperature using 0.1
M [n-Bu4N][PF6] as the electrolyte and 2.0 mM in analyte. All
potentials were shown relative to the Fc/Fc+ couple at 0.00 V. Bulk
electrolysis experiments were conducted using a 3 mm glassy carbon
working electrode, a Ni−Cr coiled wire counter electrode, and a 1
mM Ag/AgNO3(CH3CN) reference electrode. The electrochemical
cell was charged with 10 mL of 0.1 M [n-Bu4N][PF6] in CH3CN and
purged with Ar to deplete O2. To the cell, 2 × 10−5 mol of the catalyst
and 50 equiv of trifluoracetic acid (TFA) were added. CV was
recorded in order to ensure the experiment was under catalytic
conditions, and the bulk electrolysis was performed at −1.61 V vs Fc/
Fc+ for 1 h. The overhead gas was analyzed using gas chromatography
(GC) to detect H2 (SI Figures S10, S11).
energies, Ea, were obtained from Arrhenius plots (SI Figures S12−
S19, Tables S3 and S4).
Synthesis. Three new complexes containing a di-iron unit and
NiN2S2 were prepared, isolated, and characterized.
(μ-pdt)[(FeI(CO)3][FeI(CO)2·NiN2S2] (1). A solution of (μ-pdt)[Fe-
(CO)3]2 (386 mg, 1.00 mmol), NiN2S2 (277 mg, 1.00 mmol), and
trimethylamine N-oxide (TMANO; 75 mg, 1.00 mmol) in CH2Cl2
(20 mL) was stirred for 30 min at ca. 22 °C. The resulting solution
was dried under a vacuum and washed with 20 mL (×3) of hexane.
The mixture was redissolved in CH2Cl2 and filtered through Celite to
remove any unreacted starting material, and X-ray quality crystals for
complex 1 were obtained by layering moderately concentrated
solutions in CH2Cl2 with hexanes at −28 °C under a N2 atmosphere.
IR (CH2Cl2, cm−1): ν(CO) 2032, 1962, 1900. ESI-MS+: m/z
633.85 (calcd for [M]+, 633.86). Elem anal. calcd (found) for
C17H24Fe2N2NiO5S4 (MW = 635.01 g mol−1): C, 32.15 (32.21); H,
1
3.81 (3.90); N, 4.41 (4.52). H NMR (400 MHz, CD2Cl2): δ 2.00−
4.50 (18H), 1.56 (4H), 1.27 (2H) (SI Figures S1, S4, S7).
−
(μ-pdt)[FeI(CO)3][(Fe(NO))II(IMe)(CO)]+ BF4 (2). A solution of (μ-
−
pdt)[Fe(CO)3][(Fe(NO))II(IMe)(CO)]+ BF4 (543 mg, 1.00
mmol) and NiN2S2 (277 mg, 1.00 mmol) in CH2Cl2 (20 mL) was
stirred for 3 h at room temperature. The resulting solution was dried
under a vacuum and washed with 20 mL (×3) of diethyl ether. The
mixture was redissolved in CH2Cl2 and filtered through Celite to
remove any unreacted starting material, and X-ray quality crystals for
complex 2 were obtained by layering moderately concentrated
solutions in CH2Cl2 with hexane and ether (3:1) at −28 °C under
N2. IR (CH2Cl2, cm−1): ν(CO) 2056, 1990; ν(NO) 1740. ESI-MS+:
m/z 703.93 (calcd for [M]+, 703.94). Elem anal. calcd (found) for
C20H32Fe2N5NiO4S4BF4 (MW = 791.93 g mol−1): C, 30.33 (30.95);
H, 4.07 (3.92); N, 8.84 (8.17) (SI Figures S2, S5)
−
(μ-pdt)(μ-H)[FeII(CO)2(PMe3)][FeII(CO)(PMe3+)·NiN2S2]+BF4 (3). A
−
solution of (μ-pdt)(μ-H)[FeII(CO)2(PMe3)]2 PF6 (570 mg, 1.00
Kinetic Measurements. Pseudo-first-order reaction conditions
were employed for determination of the order of the reaction.
Reactions were monitored using infrared spectroscopy and UV−vis.
For IR measurements, 2 mL of DCM solution containing 4.7 mmolar
di-iron complex 1 in a two-neck flask was placed in a temperature-
controlled bath. Following temperature equilibration, a solution of
PMe3 at the same temperature was injected into the two-neck flask.
While being magnetically stirred, aliquots for IR spectra were
removed, and the reaction was monitored through four half-lives.
Rates of reactions were measured by following the decrease in
absorption of the reactant’s most intense nonoverlapping ν(CO)
vibration, which was centered at ca. 1966 cm−1. The IR data were
reproduced using UV−vis spectroscopy where the reactions were
monitored in a cuvette. Three milliliters of 0.3 mmolar di-iron
complex in DCM was directly injected with the PMe3 solution, and a
decrease of the peak at 393 nm, which corresponds to the starting
material, was monitored. Rate constants were calculated from plots of
ln(A0/At) versus time typically using >30 data points over four to five
half-lives, giving a correlation constant of 0.99. Reproducibility of a
run under identical conditions was 95% or better. Activation
parameters ΔH⧧ and ΔS⧧ were obtained from Eyring plots; activation
mmol) and NiN2S2 (277 mg, 1.00 mmol) in CH2Cl2 (20 mL) was
stirred for 30 min at room temperature while irradiated by a sunlamp.
The resulting solution was dried under a vacuum and washed with 20
mL (×3) of hexane. The mixture was redissolved in CH2Cl2 and
filtered through Celite to remove any decomposed products. An
equimolar CH2Cl2 solution of sodium tetrakis[3,5-bis-
(trifluoromethyl)phenyl]borate (Sodium BArF−) (886 mg, 1.00
mmol) was added to the above solution and stirred for 1 h. The
−
final solution was filtered over Celite to remove Na PF6 . X-ray
quality crystals for complex 3 as its BArF− salt were obtained by
layering moderately concentrated solutions in CH2Cl2 with hexanes at
−28 °C under N2.
IR (CH2Cl2, cm−1): ν(CO) 2023, 1968, 1929. ESI-MS+: m/z
730.95 (calcd for [M]+, 730.97). Elem anal. calcd (found) for
C21H43Fe2N2NiO3P2S4BF4 (MW = 818.96 g mol−1): C, 30.80
(31.00); H, 5.29 (5.40); N, 3.42 (3.76). 31P NMR (400 MHz,
dimethyl sulfoxide-d6): δ 21.13 (PMe3−, [Fe(CO)2PMe3]), −8.69
(PMe3, [Fe(CO)(PMe3)(NiN2S2)]; PF6 , −144.24; SI Figures S3,
S6, S8).
I
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.inorgchem.9b03409.
ABBREVIATIONS
■
■
sı
A path, Associative path; D path, Dissociative path; TOF,
turnover frequency; TON, turnover number; FE, Faradaic
efficiency
FTIR spectra of 1, 2, and 3; ESI+ mass spectra of 1, 2,
and 3; 1H NMR spectrum of 1; 31P NMR spectrum of 3;
cyclic voltammetry of 1; bulk electrolysis of 1; gas
chromatogram of the bulk electrolysis product of 1;
absorbance vs time plot for complex 1; IR character-
ization of products after UV−vis kinetics for complex 1;
rate data for the reaction of PMe3 with 1 at −40 °C in
DCM using IR; plot of kobs vs [PMe3] for the formation
of 1-PMe3 at −40 °C; rate data for the reaction of PMe3
with 1 at 20/25/30/35 °C in DCM using UV−vis; ESI+
mass spectrum for the reaction between 2 and PMe3
after 1 h; ESI+ mass spectrum for the reaction between
2 and PMe3 after 2 h; ESI+ mass spectrum for the
reaction between 3 and PMe3 after 3 h; thermal ellipsoid
at 50% probability for 1, 2, and 3; crystal data and
structure refinement for 1, 2, and 3 (PDF)
REFERENCES
■
(1) Lubitz, W.; Ogata, H.; Ru
Chem. Rev. 2014, 114 (8), 4081−4148.
̈
diger, O.; Reijerse, E. Hydrogenases.
(2) 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.
(3) Darensbourg, M. Y. Hydrogenase Active Sites: a New Paradigm
for Natural Product-Inspired Synthesis Based on Organometallic
Chemistry. Comments Inorg. Chem. 2010, 31 (3−4), 144−152.
(4) Liu, T.; Darensbourg, M. Y. A Mixed-Valent, Fe(II)Fe(I), Diiron
Complex Reproduces the Unique Rotated State of the [FeFe]-
Hydrogenase Active Site. J. Am. Chem. Soc. 2007, 129 (22), 7008−
7009.
(5) Li, Y.; Rauchfuss, T. B. Synthesis of Diiron(I) Dithiolato
Carbonyl Complexes. Chem. Rev. 2016, 116 (12), 7043−7077.
(6) Peters, J. W.; Lanzilotta, W. N.; Lemon, B. J.; Seefeldt, L. C. X-
ray Crystal Structure of the Fe-Only Hydrogenase (CpI) from
Clostridium pasteurianum to 1.8 Angstrom Resolution. Science 1998,
282 (5395), 1853−1858.
(7) Volbeda, A.; Charon, M.-H.; Piras, C.; Hatchikian, E. C.; Frey,
M.; Fontecilla-Camps, J. C. Crystal structure of the nickel−iron
hydrogenase from Desulfovibrio gigas. Nature 1995, 373 (6515),
580−587.
(8) Henry, R. M.; Shoemaker, R. K.; DuBois, D. L.; DuBois, M. R.
Pendant Bases as Proton Relays in Iron Hydride and Dihydrogen
Complexes. J. Am. Chem. Soc. 2006, 128 (9), 3002−3010.
(9) Chiang, C.-Y.; Miller, M. L.; Reibenspies, J. H.; Darensbourg, M.
Y. Bismercaptoethanediazacyclooctane as a N2S2 Chelating Agent
and Cys−X−Cys Mimic for Fe(NO) and Fe(NO)2. J. Am. Chem. Soc.
2004, 126 (35), 10867−10874.
(10) Denny, J. A.; Darensbourg, M. Y. Metallodithiolates as Ligands
in Coordination, Bioinorganic, and Organometallic Chemistry. Chem.
Rev. 2015, 115 (11), 5248−5273.
(11) Rao, P. V.; Bhaduri, S.; Jiang, J.; Holm, R. H. Sulfur Bridging
Interactions of Cis-Planar NiII−S2N2 Coordination Units with
Nickel(II), Copper(I,II), Zinc(II), and Mercury(II): A Library of
Bridging Modes, Including NiII(μ2-SR)2MI,II Rhombs. Inorg. Chem.
2004, 43 (19), 5833−5849.
(12) Ghosh, P.; Ding, S.; Chupik, R. B.; Quiroz, M.; Hsieh, C.-H.;
Bhuvanesh, N.; Hall, M. B.; Darensbourg, M. Y. A matrix of
heterobimetallic complexes for interrogation of hydrogen evolution
reaction electrocatalysts. Chemical Science 2017, 8 (12), 8291−8300.
(13) Brazzolotto, D.; Gennari, M.; Queyriaux, N.; Simmons, T. R.;
Accession Codes
CCDC 1965614−1965616 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
Corresponding Authors
■
Donald J. Darensbourg − Department of Chemistry, Texas
A&M University, College Station, Texas 77843, United States;
orcid.org/0000-0001-9285-4895; Email: djdarens@
chem.tamu.edu
Marcetta Y. Darensbourg − Department of Chemistry, Texas
A&M University, College Station, Texas 77843, United States;
orcid.org/0000-0002-0070-2075; Email: marcetta@
chem.tamu.edu
Authors
Kavindu Dilshan Kariyawasam Pathirana − Department of
Chemistry, Texas A&M University, College Station, Texas
77843, United States
Pokhraj Ghosh − Department of Chemistry, Texas A&M
University, College Station, Texas 77843, United States
Chung-H. Hsieh − Department of Chemistry, Tamkang
University, New Taipei City, Taiwan
Lindy Chase Elrod − Department of Chemistry, Texas A&M
University, College Station, Texas 77843, United States;
orcid.org/0000-0002-5583-2731
́
Pecaut, J.; Demeshko, S.; Meyer, F.; Orio, M.; Artero, V.; Duboc, C.
Nickel-centred proton reduction catalysis in a model of [NiFe]
hydrogenase. Nat. Chem. 2016, 8, 1054.
(14) Nguyen, N. T.; Mori, Y.; Matsumoto, T.; Yatabe, T.; Kabe, R.;
Nakai, H.; Yoon, K.-S.; Ogo, S. A [NiFe]hydrogenase model that
catalyses the release of hydrogen from formic acid. Chem. Commun.
2014, 50 (87), 13385−13387.
Nattamai Bhuvanesh − Department of Chemistry, Texas A&M
University, College Station, Texas 77843, United States
(15) Brazzolotto, D.; Wang, L.; Tang, H.; Gennari, M.; Queyriaux,
N.; Philouze, C.; Demeshko, S.; Meyer, F.; Orio, M.; Artero, V.; Hall,
M. B.; Duboc, C. Tuning Reactivity of Bioinspired [NiFe]-
Hydrogenase Models by Ligand Design and Modeling the CO
Inhibition Process. ACS Catal. 2018, 8 (11), 10658−10667.
(16) Ghosh, P.; Quiroz, M.; Wang, N.; Bhuvanesh, N.; Darensbourg,
M. Y. Complexes of MN2S2·Fe(η5-C5R5)(CO) as platform for
exploring cooperative heterobimetallic effects in HER electrocatalysis.
Dalton Transactions 2017, 46 (17), 5617−5624.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.inorgchem.9b03409
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was financially supported by the National Science
Foundation (CHE-1665258) and the Robert A. Welch
Foundation (A-0924 to M.Y.D. and A-0923 to D.J.D.)
(17) Ding, S.; Ghosh, P.; Darensbourg, M. Y.; Hall, M. B. Interplay
of hemilability and redox activity in models of hydrogenase active
sites. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (46), E9775−E9782.
J
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
(18) Darensbourg, M. Y.; Lyon, E. J.; Zhao, X.; Georgakaki, I. P. The
organometallic active site of [Fe]hydrogenase: Models and entatic
states. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (7), 3683.
(19) Lyon, E. J.; Georgakaki, I. P.; Reibenspies, J. H.; Darensbourg,
M. Y. Coordination Sphere Flexibility of Active-Site Models for Fe-
Only Hydrogenase: Studies in Intra- and Intermolecular Diatomic
Ligand Exchange. J. Am. Chem. Soc. 2001, 123 (14), 3268−3278.
(20) Li, B.; Liu, T.; Singleton, M. L.; Darensbourg, M. Y. Influence
of Sulf-Oxygenation on CO/L Substitution and Fe(CO)3 Rotation in
Thiolate-Bridged Diiron Complexes. Inorg. Chem. 2009, 48 (17),
8393−8403.
(21) Zhao, X.; Georgakaki, I. P.; Miller, M. L.; Mejia-Rodriguez, R.;
Chiang, C.-Y.; Darensbourg, M. Y. Catalysis of H2/D2 Scrambling
and Other H/D Exchange Processes by [Fe]-Hydrogenase Model
Complexes. Inorg. Chem. 2002, 41 (15), 3917−3928.
(22) Bethel, R. D.; Crouthers, D. J.; Hsieh, C.-H.; Denny, J. A.; Hall,
M. B.; Darensbourg, M. Y. Regioselectivity in Ligand Substitution
Reactions on Diiron Complexes Governed by Nucleophilic and
Electrophilic Ligand Properties. Inorg. Chem. 2015, 54 (7), 3523−
3535.
(23) Rampersad, M. V.; Jeffery, S. P.; Golden, M. L.; Lee, J.;
Reibenspies, J. H.; Darensbourg, D. J.; Darensbourg, M. Y.
Characterization of Steric and Electronic Properties of NiN2S2
Complexes as S-Donor Metallodithiolate Ligands. J. Am. Chem. Soc.
2005, 127 (49), 17323−17334.
(24) Rampersad, M. V.; Jeffery, S. P.; Reibenspies, J. H.; Ortiz, C.
G.; Darensbourg, D. J.; Darensbourg, M. Y. N2S2Ni Metallothiolates
as a Class of Ligands that Support Organometallic and Bioorgano-
metallic Reactivity. Angew. Chem. 2005, 117 (8), 1243−1246.
(25) Chong, D.; Georgakaki, I. P.; Mejia-Rodriguez, R.; Sanabria-
Chinchilla, J.; Soriaga, M. P.; Darensbourg, M. Y. Electrocatalysis of
hydrogen production by active site analogues of the iron hydrogenase
enzyme: structure/function relationships. Dalton Transactions 2003,
No. 21, 4158−4163.
(26) Liu, T.; Li, B.; Singleton, M. L.; Hall, M. B.; Darensbourg, M. Y.
Sulfur Oxygenates of Biomimetics of the Diiron Subsite of the [FeFe]-
Hydrogenase Active Site: Properties and Oxygen Damage Repair
Possibilities. J. Am. Chem. Soc. 2009, 131 (23), 8296−8307.
(27) Li, P.; Wang, M.; He, C.; Liu, X.; Jin, K.; Sun, L. Phosphane
and Phosphite Unsymmetrically Disubstituted Diiron Complexes
Related to the Fe-Only Hydrogenase Active Site. Eur. J. Inorg. Chem.
2007, 2007 (23), 3718−3727.
(28) Farmer, P. J.; Reibenspies, J. H.; Lindahl, P. A.; Darensbourg,
M. Y. Effects of sulfur site modification on the redox potentials of
derivatives of [N,N′-bis(2-mercaptoethyl)-1,5-diazacyclooctanato]-
nickel(II). J. Am. Chem. Soc. 1993, 115 (11), 4665−4674.
(29) Zhao, T.; Ghosh, P.; Martinez, Z.; Liu, X.; Meng, X.;
Darensbourg, M. Y. Discrete Air-Stable Nickel(II)−Palladium(II)
Complexes as Catalysts for Suzuki−Miyaura Reactions. Organo-
metallics 2017, 36 (9), 1822−1827.
̈
(30) Trautwein, R.; Almazahreh, L. R.; Gorls, H.; Weigand, W.
Steric effect of the dithiolato linker on the reduction mechanism of
[Fe2(CO)6{μ-(XCH2)2CRR′}] hydrogenase models (X = S, Se).
Dalton Transactions 2015, 44 (43), 18780−18794.
(31) Enemark, J. H.; Feltham, R. D. Principles of structure, bonding,
and reactivity for metal nitrosyl complexes. Coord. Chem. Rev. 1974,
13 (4), 339−406.
(32) Denny, J. A.; Darensbourg, M. Y. Approaches to quantifying the
electronic and steric properties of metallodithiolates as ligands in
coordination chemistry. Coord. Chem. Rev. 2016, 324, 82−89.
(33) Smee, J. J.; Miller, M. L.; Grapperhaus, C. A.; Reibenspies, J.
H.; Darensbourg, M. Y. Subtle Bite-Angle Influences on N2S2Ni
Complexes. Inorg. Chem. 2001, 40 (14), 3601−3605.
K
https://dx.doi.org/10.1021/acs.inorgchem.9b03409
Inorg. Chem. XXXX, XXX, XXX−XXX