Hydricity of an Fe-H species and catalytic CO2 hydrogenation

Article abstract of DOI:10.1021/ic502508p

Despite renewed interest in carbon dioxide (CO₂) reduction chemistry, examples of homogeneous iron catalysts that hydrogenate CO₂ are limited compared to their noble-metal counterparts. Knowledge of the thermodynamic properties of iron hydride complexes, including M-H hydricities (G_H^-), could aid in the development of new iron-based catalysts. Here we present the experimentally determined hydricity of an iron hydride complex: (SiP^iPr₃)Fe(H₂)(H), G_H^- = 54.3 ± 0.9 kcal/mol [SiP^iPr₃ = [Si(o-C₆H₄PiPr₂)₃]^-]. We also explore the CO₂ hydrogenation chemistry of a series of triphosphinoiron complexes, each with a distinct apical unit on the ligand chelate (Si^-, C^-, PhB^-, N, B). The silyliron (SiP^R₃)Fe (R = iPr and Ph) and boratoiron (PhBP^iPr₃)Fe (PhBP^iPr₃ = [PhB(CH₂PiPr₂)₃]^-) systems, as well as the recently reported (CP^iPr₃)Fe (CP^iPr₃ = [C(o-C₆H₄PiPr₂)₃]^-), are also catalysts for CO₂ hydrogenation in methanol and in the presence of triethylamine, generating methylformate and triethylammonium formate at up to 200 TON using (SiP^Ph₃)FeCl as the precatalyst. Under stoichiometric conditions, the iron hydride complexes of this series react with CO₂ to give formate complexes. Finally, the proposed mechanism of the (SiP^iPr₃)-Fe system proceeds through a monohydride intermediate (SiP^iPr₃)Fe(H₂)(H), in contrast to that of the known and highly active tetraphosphinoiron, (tetraphos)Fe (tetraphos = P(o-C₆H₄PPh₂)₃), CO₂ hydrogenation catalyst.

Full text of DOI:10.1021/ic502508p

Forum Article
pubs.acs.org/IC
Hydricity of an Fe−H Species and Catalytic CO2 Hydrogenation
Henry Fong and Jonas C. Peters*
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125, United States
S
* Supporting Information
ABSTRACT: Despite renewed interest in carbon dioxide
(CO2) reduction chemistry, examples of homogeneous iron
catalysts that hydrogenate CO2 are limited compared to their
noble-metal counterparts. Knowledge of the thermodynamic
properties of iron hydride complexes, including M−H
−
hydricities (ΔGH ), could aid in the development of new
iron-based catalysts. Here we present the experimentally
determined hydricity of an iron hydride complex: (SiPiPr3)Fe-
0.9 kcal/mol [SiPiPr = [Si(o-
3
−
(H2)(H), ΔGH = 54.3
C6H4PiPr2)3]−]. We also explore the CO2 hydrogenation
chemistry of a series of triphosphinoiron complexes, each with
a distinct apical unit on the ligand chelate (Si−, C−, PhB−, N,
B). The silyliron (SiPR )Fe (R = iPr and Ph) and boratoiron
3
(PhBPiPr3)Fe (PhBPiPr = [PhB(CH2PiPr2)3]−) systems, as well as the recently reported (CPiPr3)Fe (CPiPr = [C(o-
3
3
C6H4PiPr2)3]−), are also catalysts for CO2 hydrogenation in methanol and in the presence of triethylamine, generating
methylformate and triethylammonium formate at up to 200 TON using (SiPPh3)FeCl as the precatalyst. Under stoichiometric
conditions, the iron hydride complexes of this series react with CO2 to give formate complexes. Finally, the proposed mechanism
of the (SiPiPr3)-Fe system proceeds through a monohydride intermediate (SiPiPr3)Fe(H2)(H), in contrast to that of the known
and highly active tetraphosphinoiron, (tetraphos)Fe (tetraphos = P(o-C6H4PPh2)3), CO2 hydrogenation catalyst.
by the recent work of Linehan and co-workers on a cobalt
hydride catalyst,26,27 in which the design of this efficient CO2-
to-formate hydrogenation system was achieved, in part, by
using a cobalt hydride that was more hydridic (i.e., <43 kcal/
mol) than the formate36 (eq 2).
INTRODUCTION
■
The reduction of carbon dioxide (CO2) into value-added
chemicals and liquid fuels has received considerable attention
recently because of increasing interest in the development of
carbon neutral energy sources.1 The production of liquid fuels
such as methanol2 or formic acid3 from CO2 and H2 (or its
formal equivalents) is particularly attractive. However, selective
production of these products using heterogeneous catalysts
remains challenging.4−6 One interesting approach toward CO2
reduction is to use molecular catalysis, where product selectivity
may be better controlled than heterogeneous systems.7 The
catalytically active species in molecular systems can often be
probed either directly or indirectly, thereby offering oppor-
tunities to understand the catalytic mechanism and synthetically
tune systems in a well-defined manner.8
[M−H]n+ → Mn+1 + H−
OCHO− → CO2 + H−
ΔGH
−
(1)
(2)
−
ΔGH = 43 kcal/mol
As part of our exploratory research of phosphine-supported
iron complexes in small-molecule activation reactions,37−42 we
were interested in studying the catalytic CO2 hydrogenation
chemistry of a series of triphosphinoiron species (Chart 1):
(SiPR )Fe(L)(H) (L = H2 or N2; SiPR3 = [Si(o-C6H4PR2)3]−, R
3
= iPr or Ph),37,43,44 (PhBPiPr3)Fe(H)3(PMe3) (PhBPiPr
PhB(CH2PiPr2)3),45 [(NPiPr3)Fe(N2)(H)](PF6) (NPiPr
=
=
3
One of the simplest CO2 reduction reactions is its
hydrogenation to formic acid.3 While a number of noble-
metal catalysts for the hydrogenation of CO2 to formic acid
exist,9−17 there are only a handful of examples using first-row
transition metals such as iron18−24 and cobalt,25−28 and
information about their thermodynamic properties and
elementary reaction steps is needed.29−34 For example, the
3
N(CH2CH2PiPr2)3),46 (TPB)(μ-H)Fe(L)(H) (L = N2 or H2;
TPB = B(o-C6H4PiPr3)3),44 (CPiPr3)Fe (CPiPr = [C(o-
3
C6H4PiPr2)3]−),42 and (CSiPPh3)Fe (CSiPPh = [C(Si-
3
(CH3)2CH2PPh2)3]−)47 (Chart 1). These systems are structur-
ally related to two tetraphosphinoiron hydride CO2 hydro-
genation catalysts ([(PP3)Fe(H2)(H)](BF4)19,48 and
−
hydricity (ΔGH ), which is the heterolytic dissociation energy
of [M−H]n+ into Mn+ and H− (eq 1), has only been
experimentally determined for one iron hydride complex
(FpH)35 despite recent reports of iron-catalyzed CO2 hydro-
genation.18−24 Knowledge of the hydricities of hydrogenation
catalysts can aid the design of new catalysts. This is highlighted
Special Issue: Small Molecule Activation: From Biological Principles
to Energy Applications
Received: October 18, 2014
© XXXX American Chemical Society
A
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
conjugate acid of the metal hydride of interest. We previously
reported H2 chemistry of the (SiPiPr3)Fe system,44 including
deprotonation of the cationic iron dihydrogen complex
[(SiPiPr3)Fe(H2)](BArF ) (BArF = [(3,5-(CF3)2−C6H3)4B]−)
Chart 1. Select Phosphine−Metal Complexes of Relevance
to Catalytic CO2 Hydrogenation
4
4
by Hunig’s base under H (1 atm) to afford (SiPiPr3)Fe(H2)-
̈
2
(H).43 This motivated us to use this deprotonation reaction to
experimentally determine the hydricity of (SiPiPr3)Fe(H2)(H)
using the series of equations in Scheme 1. The equilibrium in
Scheme 1. Reactions Relevant to Determination of the pKa
a
and Hydricity of (SiPiPr3)Fe (B = Base)
a
See Chart 1 for a full detailed ligand representation.
(tetraphos)Fe(H2)(H)](BF4),20 where PP3 = P-
(CH2CH2PPh2)3 and tetraphos = P(o-C6H4PPh2)3), studied
in a similar context by the groups of Beller and Laurenczy
(Chart 1). A distinguishing feature of the present series of
triphosphinoiron complexes is that each of the present ligand
scaffolds possesses a different apical unit. These include an X-
1
eq 3 was followed by H NMR spectroscopy independently
with 1,8-bis(dimethylamino)naphthalene (proton sponge; Keq
= 4.3), 2,6-lutidine (Keq = 3.3 × 10−5), and 2,4,6-
trimethylpyridine (Keq = 5.1 × 10−5) in deuterated tetrahy-
drofuran (THF-d8). The reverse protonation of (SiPiPr3)Fe-
type silyl in SiPR , an X-type alkyl in (CPiPr3)Fe and
3
(CSiPPh3)Fe, a noncoordinating borate in PhBPiPr3, an L-type
amine in NPiPr3, and a Z-type borane in TPB. Each of these
apical units can confer different (i) geometries at iron, (ii)
formal oxidation states at iron, and (iii) reactivity patterns for
otherwise structurally similar species, as we have studied
previously with respect to N2 activation chemistry.37−42
(H2)(H) with the BArF salt of 1,8-bis(dimethylammonium)-
4
naphthalene (Keq = 2.6) was also followed by 1H NMR
spectroscopy in THF-d8.49
We note that pKa of the H2 ligand in [(SiPiPr3)Fe(H2)]-
(BArF ) can be estimated using eqs 3 and 4. The experimentally
4
determined pKa in THF using this method is pKaTHF = 10.8
0.6 for [(SiPiPr3)Fe(H2)](BArF ). Notably, the pKaTHF agrees
In the present work, we experimentally determined the pKa
and hydricity for the (SiPiPr3)Fe system and studied the
catalytic and stoichiometric hydrogenation of CO2 in this and
the related triphosphinoiron species shown in Chart 1. Under
elevated temperatures and pressures of CO2 and H2 and with
triethylamine as the base, the (SiPiPr3)Fe, (SiPPh3)Fe,
(PhBPiPr3)Fe, and (CPiPr3)Fe systems catalytically hydro-
genated CO2 to triethylammonium formate and methylformate,
while (NPiPr3)Fe, (TPB)Fe, and (CSiPPh3)Fe did not catalyze
CO2 hydrogenation. We also show that despite a low hydricity
4
very well with the predicted value of 10.2 obtained from the
ligand acidity constant method recently developed by
Morris.50,51 We caution that this is only a rough estimate of
the pKa of [(SiPiPr3)Fe(H2)](BArF ) because the pKa of
4
[(SiPiPr3)Fe(H2)](BArF ) is a measure of the removal of a
4
proton to afford “(SiPiPr3)Fe(H)”, whereas in the observed
deprotonation reaction, H2 coordinates to this species to afford
(SiPiPr3)Fe(H2)(H) and contributes to the equilibrium depicted
in eq 3.
(i.e., large ΔGH value) for the complex (SiPiPr3)Fe(H2)(H)
−
With the equilibrium of eq 3 in hand, the hydricity of the
conjugate base (SiPiPr3)Fe(H2)(H) can be determined by the
summation of eqs 3−5 to give eq 6.52,53 Most hydricity values
have been reported in acetonitrile in part because of the known
heterolytic dissociation energy of H2 in acetonitrile (eq 5).
However, irreversible coordination of acetonitrile to [(SiPiPr3)-
−
(ΔGH = 54.3
0.9 kcal/mol) coordination of the formate
product to the iron center following hydride transfer to CO2
provides enough driving force to make the reaction thermally
accessible.
Fe(H2)](BArF ) precluded the use of this solvent. An empirical
RESULTS AND DISCUSSION
■
4
pKa and Hydricity for (SiPiPr3)Fe. Because Fe−H species
have been invoked as intermediates for CO2 hydrogenation, we
were curious if (SiPiPr3)Fe(H2)(H) was sufficiently hydridic to
react with CO2. One method for determining hydricities is to
use a thermochemical cycle that involves deprotonating the
relationship relates the pKaTHF of a metal complex to the pKa in
acetonitrile (pKaMeCN).54 Using this relationship, the pKaMeCN of
[(SiPiPr3)Fe(H2)](BArF ) is 15.9 0.7. When the pKaMeCN of
[(SiPiPr3)Fe(H2)](BArF4) is combined with eq 5, the hydricity
4
of (SiPiPr3)Fe(H2)(H) in MeCN is 54.3 0.9 kcal/mol. To the
B
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
best of our knowledge, this is only the second experimentally
estimated hydricity value of an iron hydride complex.35
Formal hydride transfer from phosphine-ligated iron hydride
in C6D6 at RT). Similarly, (SiPiPr3)Fe(N2)(H) reacted with
CO2 to afford (SiPiPr3)Fe(OCHO). The ATR-IR spectrum
showed an asymmetric O−C−O stretch at 1623 cm−1 (13CO2:
1583 cm−1). While the symmetric O−C−O stretch could not
be reliably discerned (Table 1), its S = 1 spin state (2.8 μB in
C6D6 at room temperature (RT)) and yellow color indicate a
five-coordinate (SiPiPr3)Fe(OCHO) complex.
complexes to CO2 to give formate is well-known.29−34,55
A
comparison of the hydricity of (SiPiPr3)Fe(H2)(H) to that of
formate (eq 2) indicates that the reaction for hydride transfer
from (SiPiPr3)Fe(H2)(H) to CO2 to afford formate is
endergonic by over 10 kcal/mol. Yet, as will be shown below,
(SiPiPr3)Fe(H2)(H) can still react with CO2 both stoichio-
metrically and catalytically to afford formate.
Similarly, a THF solution of (PhBPiPr3)Fe(H)3(PMe3)
reacted with CO2 (1 atm) at RT to afford the κ2-bound
formate adduct (PhBPiPr3)Fe(OCHO) (Scheme 2b). The
ATR-IR spectrum of this S = 2 species (5.0 μB, C6D6 at RT)
exhibited features of a formate ligand at 1595 and 1362 cm−1
(13CO2: 1546 and 1355 cm−1) with a Δν(O−C−O) = 233
cm−1 that is consistent with the κ2-bound formate assignment.56
The formate coordination mode is in contrast to the κ1-bound
Stoichiometric Reactivity of Fe−H Species with CO2.
In addition to (SiPiPr3)Fe(L)(H)43,44 (where L = N2 or H2), we
have previously reported the synthesis and characterization of
three other related triphosphinoiron hydride complexes,
(PhBPiPr3)Fe(H)3(PMe3),45 [(NPiPr3)Fe(N2)(H)](PF6),46 and
(TPB)(μ-H)Fe(N2)(H)44 (Chart 1), and demonstrated that
the two former complexes, (PhBPiPr3)Fe(H)3(PMe3) and
(TPB)(μ-H)Fe(N2)(H), are olefin hydrogenation catalysts.
The iron hydride species of the tris(diphenylphosphino)silyl
ligand, (SiPPh3)Fe(N2)(H), had not previously been reported,
but it has now been synthesized in a manner analogous to that
of the preparation of the isopropyl analogue (SiPiPr3)Fe(N2)-
(H) (vide infra). The reactivity of these iron hydrides to CO2
was probed.
formate ligands in (SiPR )Fe(OCHO), [(NPiPr3)Fe(OCHO)]-
3
(PF6), and (TPB)Fe(OCHO). We presume this arises because
of the lower coordination number in (PhBPiPr3)Fe.
A solution of [(NPiPr3)Fe(N2)(H)](PF6) also reacted with
CO2 (1 atm) at RT to afford the formate adduct [(NPiPr3)-
Fe(OCHO)](PF6) (Scheme 2c). [(NPiPr3)Fe(OCHO)](PF6)
is S = 2 (5.1 μB, C6D6 at RT), analogous to [(NPiPr3)FeCl]-
(PF6),46 with a diagnostic νasym(O−C−O) vibrational feature at
1613 cm−1 (13CO2: 1579 cm−1) in the ATR-IR spectrum.
However, the accompanying lower-energy νsym(O−C−O)
vibrational feature could not be reliably assigned because of
overlapping ligand vibrational modes in the 1200−1300 cm−1
region. The obscured νsym(O−C−O) feature prevented
assignment of the formate binding mode, but νasym(O−C−O)
most closely matches κ1-bound formate ligands (Table 1).56
Mixing a benzene solution of (TPB)(μ-H)Fe(N2)(H) with
CO2 (1 atm) afforded the κ1-formate complex (TPB)Fe-
Synthesis of Iron Formate Species. A solution of
(SiPPh3)Fe(N2)(H) reacted with CO2 (1 atm) at 50 °C to
afford the yellow iron formate species (SiPPh3)Fe(OCHO)
(Scheme 2a). Consistent with the κ1-bound formate ligand,56
attenuated total reflectance infrared (ATR-IR) spectroscopy
showed two signature vibrational features at 1618 and 1316
cm−1 (13CO2: 1587 and 1254 cm−1) with a Δν(O−C−O) of
302 cm−1 (Table 1). As expected for a five-coordinate
1
(SiPR )FeII complex,57 (SiPPh3)Fe(OCHO) is S = 1 (2.7 μB
(OCHO) (Scheme 2d) as a yellow solution. The color, H
3
NMR spectrum, and solution magnetic moment (4.2 μB, S = 3/2
in C6D6 at RT) are consistent with the formulation of
(TPB)Fe(OCHO) as a {Fe−B}7 species,58,59 and vibrational
modes in the IR spectrum at 1627 and 1291 cm−1 (13CO2:
1588 and 1269 cm−1) with a Δν(O−C−O) value of 336 cm−1
(Table 1) are diagnostic for a κ1-formate ligand.56 The IR
spectrum of (TPB)Fe(OCHO) lacks any feature that is
diagnostic for a B−H unit.60 For comparison, in the related S
= 2 (TPBH)Fe(CCAr) (Ar = phenyl or tolyl) complex, where a
terminal B−H is present, the IR spectra exhibit diagnostic B−H
vibrations at 2490 cm−1 for Ar = phenyl and 2500 cm−1 for Ar
= tolyl.44
Scheme 2. Stoichiometric CO2 Hydrogenation to
Triethylammonium Formate
a
The formation of (TPB)Fe(OCHO) from the reaction of
(TPB)(μ-H)Fe(N2)(H) with CO2 (1 atm) is notable in that
there is a formal loss of an hydrogen atom (Schemes 2d and 3).
The loss of 0.5 equiv of H2 (relative to the starting iron
complex) was confirmed by gas chromatography with a thermal
conductivity detector (GC-TCD; 0.44 equiv of H2 quantified).
The reaction between the previously reported (TPB)Fe(N2)58
with formic acid also formed (TPB)Fe(OCHO), with 0.42
equiv of H2 detected by GC-TCD as a product (Scheme 3).
Reactivity of Fe(OCHO) Species. The formate ligands in
all five of the aforementioned iron formate complexes are
substitutionally labile. The addition of triethylammonium
chloride (10 equiv) into either benzene or methanol solutions
of these complexes resulted in the formation of the respective
iron chloride complexes and triethylammonium formate
(Scheme 2). Furthermore, the iron chloride products from
these metathesis reactions are synthons for the respective iron
hydride complexes.
a
See Chart 1 for detailed representations of the ligands indicated.
C
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
Table 1. IR Stretching Frequencies and Solution Magnetic Moments for Iron Formate Complexes
b
b
νasym(O−C−O) (cm−1
)
νsym(O−C−O) (cm−1
)
a
c
μeff (μB)
CO2
13CO2
CO2
1316
13CO2
Δνasym (O−C−O) (cm−1
)
(SiPPh3)Fe(OCHO)
(SiPiPr3)Fe(OCHO)
(PhBPiPr3)Fe(OCHO)
[(NPiPr3)Fe(OCHO)][PF6]
(TPB)Fe(OCHO)
2.7
2.8
5.0
5.1
4.2
1618
1623
1595
1613
1627
1587
1583
1546
1579
1254
302
1362
1291
1355
1269
233
336
1588
a
b
c
Solution magnetic moments at RT. ATR-IR data of solution thin films. Difference between νasym(O−C−O) and νsym(O−C−O).
triethylamine in the presence of H2 and D2 (ca. 1 atm:1 atm)
gives HD (Scheme 4h). Related, the cationic H2 adduct
Scheme 3. Reactivity of (TPB)Fe Complexes with CO2 and
Formic Acid
a
[(SiPiPr3)Fe(H2)][BArF ], a model for [(SiPiPr3)Fe(H2)]+,
4
scrambles a mixture of H2 and D2 (ca. 1 atm:1 atm) to HD
in a CD3OD/THF-d8 solution (10:1). (SiPiPr3)FeCl is the sole
1
observed iron-containing species by H NMR spectroscopy in
the former experiment, indicating that the equilibrium with the
putative [(SiPiPr3)Fe(H2)]+ responsible for scrambling H2/D2
heavily favors (SiPiPr3)FeCl.
a
See Chart 1 for a full detailed ligand representation.
Catalytic Hydrogenation. Having realized a synthetic
cycle for CO2 hydrogenation to formate, we explored whether
the process could be made catalytic. Following literature
precedent, the triphosphinoiron chloride complexes were tested
in an initial screen for catalysis,9−16,18−26,28 and triethylamine
was added to serve as a base.61 1H NMR spectroscopy with
N,N-dimethylformamide (DMF) added as an integration
standard was used to quantify triethylammonium formate
yields. Other known products of CO2 hydrogenation are
formate esters such as methylformate that are obtained from
the esterification of formate with methanol.19,20 Because of the
volatility and low yields of MeOCHO, GC-FID (GC with flame
ionization detection) was used to quantify this product.
Under the standardized reaction conditions of 29 atm of CO2
and 29 atm of H2 in a methanol solvent with triethylamine,
(SiPiPr3)FeCl, (SiPPh3)FeCl, and (PhBPiPr3)FeCl are precata-
lysts for hydrogenation of CO2 to triethylammonium formate
and methylformate (Table 2, entries 1−3). (SiPPh3)FeCl is the
most active, having an average turnover number of 200. These
three systems are also more selective for (Et3NH)(OCHO)
than MeOCHO, with (PhBPiPr3)FeCl being the most selective
of the three with a 10:1 (Et3NH)(OCHO) to MeOCHO
product ratio. It is also worth noting that the primary
coordination sphere of the zwitterionic (PhBPiPr3)Fe system
is structurally similar to a known cationic ruthenium system
(triphos)Ru (triphos = CH3C(CH2PPh2)3) that hydrogenates
CO2 to methanol62 and also dehydrogenates formic acid63
(Chart 1). We have also reported the reduction of CO2 to
oxalate by (PhBPiPr3)Fe.64
With this metathesis reaction and known reaction chemistry
for the (SiPiPr3)Fe scaffold, we can construct a synthetic cycle
for CO2 hydrogenation, which may inform the catalytic CO2
hydrogenation reaction (vide infra). Starting from the Fe−Cl
species, (SiPiPr3)FeCl reacts with MeMgCl (1 equiv) to afford
the iron methyl complex (SiPiPr3)FeMe (this work, Scheme 4a).
Scheme 4. Synthetic Cycle for CO2 Hydrogenation to
a
Formate by (SiPiPr3)Fe
a
See Chart 1 for a full detailed ligand representation. Conditions: (a)
MeMgCl, THF; (b) H2, THF (plus N2 workup for L = N2); (c)
(HBArF )(Et2O)2, C6H6; (d) H2 (forward), N2 (reverse), THF; (e)
4
Et3N, THF (plus N2 workup for L = N2); (f) CO2, MeOH, THF, or
C6H6; (g) (Et3NH)Cl, C6H6 or MeOH; (h) 1 atm:1 atm H2/D2, Et3N,
10:1 CD3OD/THF-d8 (HD is produced); (i) for L = N2, Et3NHCl,
10:1 CD3OD/THF-d8.
Under the standard conditions, [(NPiPr3)FeCl](PF6) and
(TPB)FeCl are not precatalysts for the reaction (Table 2,
entries 4 and 5). The recently reported (CPiPr3)FeCl complex
(Chart 1),42 where the silicon atom in (SiPiPr3)FeCl is
substituted by a carbon atom, is also catalytically competent
(Table 2, entry 6) but is significantly less active than
(SiPPh3)FeCl. Another carbon variant of the triphosphinoiron
series of complexes, (CSiPPh3)FeCl47 (Chart 1), is not
catalytically competent (Table 2, entry 7).
For a direct comparison with known iron CO2 hydro-
genation catalysts and as a benchmark of the method employed,
we subjected the (PP3)Fe19 and (tetraphos)Fe20 systems to our
standardized conditions. Beller and Laurency reported that a
mixture of the PP3 ligand with Fe(BF4)2 is one of the more
Subsequent reaction with H2 affords (SiPiPr3)Fe(N2)(H)
(Scheme 4b).43 Alternatively, the iron methyl complex
(SiPiPr3)FeMe can be converted to the cationic H2 complex
[(SiPiPr3)Fe(H2)](BArF )43 (Scheme 4c,d). The H2 ligand in
4
the latter complex can be deprotonated by triethylamine (this
work) to generate (SiPiPr3)Fe(L)(H) (where L = N2 or H2;
Scheme 4e). As shown above, (SiPiPr3)Fe(N2)(H) reacts with
CO2 to afford (SiPiPr3)Fe(OCHO) (Scheme 4f), which can
undergo metathesis with (Et3NH)Cl to afford the starting iron
chloride complex (Scheme 4g).
Reaction of (SiPiPr3)FeCl with H2 is also possible. A
CD3OD/THF-d8 (10:1) solution of (SiPiPr3)FeCl with excess
D
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
a
Table 2. Triphosphinoiron-Catalyzed CO2 Hydrogenation
Table 3. (SiPiPr3)FeCl-Catalyzed CO2 Hydrogenation under
Varied Conditions
deviation from standard
(Et3NH)
a
b
f
(Et3NH)(OCHO):MeOCHO
entry
conditions
TON
(OCHO):MeOCHO ratio
b
f
entry
precatalyst
TON
ratio
0
1
none
53
0
3:1
0
1
(SiPiPr3)FeCl
53
200
27
0
3:1
2:1
10:1
0
0 atm of H2
CD3OD
150 °C
20 °C
2
(SiPPh3)FeCl
c
2
32
40
0
2:1
2:1
0
3
(PhBPiPr3)FeCl
[(NPiPr3)FeCl](PF6)
(TPB)FeCl
(CPiPr3)FeCl
(CSiPPh3)FeCl
3
4
4
5
0
0
5
2 h
16
41
93
69
45
33
26
57
1:0
5:1
6:1
2:1
8:1
12:1
9:1
21:1
6
27
0
6:1
0
d
6
0.5 equiv of (Et3NH)Cl
7
d
7
0.5 equiv of NaBF4
0.5 equiv of NaBArF
c,d
8
PP3/Fe(BF4)2
486
3:1
1:1
0
d
8
4
c,e
9
[(tetraphos)FeF]BF4
FeCl2
1661
d
9
0.5 equiv of NaF
10
11
12
0
0
0
d,e
10
11
12
0.5 equiv of TBAF
FeCl2/4PPh3
no iron
0
d
0.5 equiv of CsF
0
d
0.5 equiv of K2CO3
a
Conditions: 0.1 mol % (0.7 mM) iron precatalyst (relative to Et3N),
a
Standard conditions: 0.1 mol % (0.7 mM) iron precatalyst (relative to
methanol, 651 mM Et3N, 29 atm of CO2 (RT), 29 atm of H2 (RT),
Et3N), methanol, 651 mM Et3N, 29 atm of CO2 (RT), 29 atm of H2
b
100 °C, 20 h. Turnover number: combined yield (moles) of
b
(RT), 100 °C, 20 h. Turnover number: combined yield (moles) of
(Et3NH)(OCHO) and MeOCHO divided by moles of precatalyst.
(Et3NH)(OCHO) and MeOCHO divided by moles of precatalyst.
c
d
Previously studied under slightly different conditions. See ref 19.
c
(Et3NH)(OCHO) was detected by 1H NMR spectroscopy, but
e
f
See ref 20. Ratio of the amount of (Et3NH)(OCHO) product to the
neither (Et3ND)(OCDO), (Et3NH)(OCDO), nor (Et3ND)(OCHO)
amount of MeOCHO product.
d
was detected by 2H NMR spectroscopy. Relative to moles of
e
f
(SiPiPr3)FeCl. TBAF = tetrabutylammonium fluoride. Ratio of the
amount of (Et3NH)(OCHO) product to the amount of MeOCHO
product.
active conditions for CO2 hydrogenation in the PP3 system.
Under the standard conditions of this study, a 1:1 mixture of
PP3 and Fe(BF4)2 hydrogenates CO2 to triethylammonium
formate and methylformate, as well, at a total TON of 486
(Table 2, entry 8). The (tetraphos)Fe(F)(BF4)2 complex also
catalyzes CO2 hydrogenation to the same products at a total
TON of 1661 (Table 2, entry 9). These values are in near
agreement with the respective literature reports. The
(tetraphos)Fe(F)(BF4)2 complex is the least selective of the
series in Table 2 for formate production.
examples.3,9−16,19−26,28 Milstein and co-workers have reported
CO2 hydrogenation to formate at a relatively low 10 total atm
of H2 and CO2.18 Also critical is methanol because the catalytic
activity does not occur in neat THF under any pressures of
CO2 and H2 studied here (see the SI), highlighting the
importance of polar, protic solvents in phosphinoiron CO2
hydrogenation catalysis.61
A series of control experiments were performed to probe the
homogeneity of the reaction. The catalytic reaction is
uninhibited by the addition of elemental mercury (see the
Supporting Information, SI). Also, CO2 hydrogenation does
not occur with the iron salt FeCl2 (Table 2, entry 10) or with a
1:4 mixture of FeCl2 and triphenylphosphine (Table 2, entry
11), nor does it proceed in the absence of an iron source (Table
2, entry 12). These experiments do not preclude a role for
heterogeneous species but provide evidence consistent with a
homogeneous process.
To gain insight into the reaction, we chose to study the
hydrogenation catalysis by the (SiPiPr3)Fe system further
because it is more active than (PhBPiPr3)FeCl and because its
coordination chemistry has been studied in greater detail than
that of its phenyl analogue (SiPPh3)Fe.43,44
It was determined that 100 °C and 20 h are optimal for the
reaction under the conditions studied here. Running the
reaction at 150 °C slightly reduces the turnover relative to the
standard conditions (Table 3, entry 3), which is likely a result
of catalyst decomposition (vide infra). At 20 °C, no reaction
occurs (Table 3, entry 4), and the starting precatalyst
(SiPiPr3)FeCl is the only iron-containing species at the end of
the reaction. Reducing the reaction time to 2 h at 100 °C
reduces the TON by a factor of 3 (Table 3, entry 5) compared
to the standard conditions.
Using the stoichiometric reactions as a guide, we also probed
the effects of additives and precatalysts on the catalysis. The
stoichiometric metathesis reaction for the transformation of
(SiPiPr3)Fe(OCHO) to (SiPiPr3)FeCl suggests that chloride
substitution for formate may be a route for formate release.
However, the addition of 0.5 equiv of (Et3NH)Cl (relative to
iron) into the reaction reduces the TON, although the
selectivity for (Et3NH)(OCHO) over MeOCHO slightly
increases to 5:1 (Table 3, entry 6). It appears that while
chloride may substitute for formate, excess chloride may also
slow H2 substitution at iron (vide infra) and reduces the overall
TON. The addition of a noncoordinating anion in the form of
NaBF4 to the catalytic mixture is beneficial, yielding a TON of
93 and 6:1 selectivity for (Et3NH)(OCHO) (Table 3, entry 7),
Under standard conditions but in the absence of H2,
triethylammonium formate and methylformate were not
detected (Table 3, entry 1). Furthermore, when the reaction
was run in CD3OD instead of CH3OH, (Et3NH)(OCHO) was
1
detected by H NMR spectroscopy at the conclusion of the
reaction, while (Et3ND)(OCDO) was not detected by 2H
NMR spectroscopy (Table 3, entry 2). The data collectively
indicate that H2 is the source of the hydrogen atom equivalents.
High pressures of CO2 and H2 are critical because the
reaction does not proceed at or near atmospheric pressures of
H2 and CO2 (see the SI) in agreement with most literature
while the addition of Na(BArF ) only modestly increases the
4
TON to 69 and without significantly affecting the selectivity
E
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
(Table 3, entry 8). The origin of the effect from the Na+ and/or
borate anion is not understood, but one effect may be that Na+
facilitates removal of the inner-sphere chloride as NaCl.
Additionally, alkali metals are known to facilitate CO2
coordination to cobalt centers.65 It is also noteworthy that
BF4− is the counteranion of the highly active tetraphosphinoir-
on (PP3)Fe19 and (tetraphos)Fe20 systems and is also beneficial
for iron-catalyzed formic acid dehydrogenation.66 It is unlikely
less catalytically competent than the synthesized iron complex
(SiPiPr3)FeCl (Table 4, entry 4). All four of these precatalysts
are more selective than (SiPiPr3)FeCl for (Et3NH)(OCHO).
We also probed the fate of the iron precatalyst (SiPiPr3)FeCl
under the reaction conditions. At the end of the reaction under
standard conditions, the 31P NMR spectrum showed a mixture
of phosphorus-containing material, including significant quanti-
ties of free ligand (HSiPiPr3). If the reaction was run at RT, only
the starting precatalyst (SiPiPr3)FeCl was observed by 1H NMR
spectroscopy. These observations indicate that while the
catalysis requires heating, elevated temperatures lead to
eventual catalyst decomposition.
−
that fluoride, which may be a decomposition product of BF4 , is
the source of the positive response because fluoride salts
decrease the TON but increase the selectivity for (Et3NH)-
(OCHO) (Table 3, entries 9−11). Finally, the addition of
K2CO3, which has been reported to enhance CO2 hydro-
genation catalysis for some noble- and non-noble-metal
systems,67 has no effect on the TON but significantly increases
the selectivity for (Et3NH)(OCHO) compared to the other
additives (Table 3, entry 12). The additives containing
coordinating anions are more selective for (Et3NH)(OCHO)
over MeOCHO, presumably a result of anion coordination
inhibiting iron-catalyzed esterification of formate to methyl-
formate.68 However, we caution that these results are
qualitative. A systematic study of the effects of these and
other additives on catalysis would be warranted to draw
quantitative conclusions.
A possible catalytic cycle based in part on the observed
stoichiometric reactions discussed in Scheme 4 is proposed in
Scheme 5 for the (SiPiPr3)Fe system. Starting from precatalyst
Scheme 5. Proposed Catalytic Cycle for (SiPiPr3)Fe in
a
MeOH
Other important factors known to affect catalysis are the base
identity26 and base concentrations.61 A careful study of the
effect of different bases and concentrations on catalysis in the
present series is beyond the scope of this report, but we point
out that the pKa of triethylamine is suitably matched to the pKa
of [(SiPiPr3)Fe(H2)]+ (vide supra), an intermediate in the
catalytic cycle of the (SiPiPr3)Fe system (Scheme 5; vide infra).
Other (SiPiPr3)Fe species are also competent precatalysts.
The iron formate (SiPiPr3)Fe(OCHO) and iron hydride
complex (SiPiPr3)Fe(N2)(H) are each catalytically competent
precatalysts (Table 4, entries 1 and 2), with TONs comparable
to that of (SiPiPr)FeCl. The cationic N2 complex [(SiPiPr3)Fe-
a
See Chart 1 for a full detailed ligand representation.
(SiPiPr3)FeCl in Scheme 5, H2 substitution forms the cationic
H2 adduct [(SiPiPr3)Fe(H2)]+. The viability of this H2 for Cl−
substitution step is demonstrated by H/D scrambling experi-
ments discussed above. However, we cannot rule out a process
wherein methanol is involved in the displacement of Cl−
through, for example, methanol for Cl− substitution at iron
followed by H2 substitution of methanol. The cationic H2
adduct [(SiPiPr3)Fe(H2)]+ in the catalytic cycle can be
deprotonated by triethylamine to give (SiPiPr3)Fe(H2)(H), as
we have observed in the stoichiometric reaction (Scheme 4e).
The reverse of this reaction is also possible: a 1:1 mixture of
(Et3NH)Cl and (SiPiPr3)Fe(H2)(H) reacts to afford (SiPiPr3)-
FeCl (Scheme 4i).
(N2)](BArF ), which is a synthon for (SiPiPr3)Fe(N2)(H) in the
4
presence of H2 and triethylamine (Scheme 4), is also a
catalytically competent precatalyst (Table 4, entry 3). Finally, a
1:1 mixture of the free ligand HSiPiPr3 and FeCl2 is significantly
Table 4. CO2 Hydrogenation Catalyzed by Various (SiPiPr
Fe Species
)
3
a
(Et3NH)(OCHO):MeOCHO
b
d
The iron hydride intermediate (SiPiPr3)Fe(H2)(H) can then
react with CO2 to form the iron formate complex (SiPiPr3)Fe-
(OCHO), which can subsequently react with (Et3NH)Cl,
reform (SiPiPr3)FeCl, and release (Et3NH)(OCHO). The
direct conversion of (SiPiPr3)Fe(OCHO) to [(SiPiPr3)Fe(H2)]+
may also be a viable pathway because chloride-free [(SiPiPr3)-
entry
precatalyst
(SiPiPr3)FeCl
TON
ratio
0
1
2
3
53
52
47
18
3:1
15:1
3:1
8:1
(SiPiPr3)Fe(OCHO)
(SiPiPr3)Fe(N2)(H)
[(SiPiPr3)Fe(N2)]
(BArF )
4
c
HSiPiPr3/FeCl2 (1:1)
12
4:1
Fe(N2)](BArF ), (SiPiPr3)Fe(N2)(H), and (SiPiPr3)Fe(OCHO)
4
4
are catalytically competent.
a
Conditions: 0.1 mol % (0.7 mM) iron precatalyst (relative to Et3N),
An alternative mechanism involving an iron dihydride species
cannot be ruled out but is unlikely for the (SiPiPr3)Fe system
(Scheme 6a). A similar mechanism was proposed by Beller et
al. for the cationic (tetraphos)Fe catalyst based on in situ NMR
data, where the intermediate [(tetraphos)Fe(H2)(H)]+ was
deprotonated by Et3N to give (tetraphos)Fe(H)2 (Scheme
methanol, 651 mM Et3N, 29 atm of CO2 (RT), 29 atm of H2 (RT),
100 °C, 20 h. Turnover number: combined yield (moles) of
b
(Et3NH)(OCHO) and MeOCHO divided by moles of precatalyst.
c
1:1 mixture of HSiPiPr3:FeCl2 (0.7 mM) was used as the precatalyst in
d
place of (SiPiPr3)FeCl. Ratio of the amount of (Et3NH)(OCHO)
product to the amount of MeOCHO product.
F
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
An additional factor preventing catalysis in the (TPB)Fe system
is the unproductive loss of 0.5 equiv of H2 following the
reaction of (TPB)(μ-H)Fe(N2)(H) with CO2, which generates
the catalytically incompetent (TPB)Fe(OCHO) (Scheme 3).
Influence of the Hydricity on the Reaction with CO2.
On the basis of only the hydricity (54.3 0.9 kcal/mol), the
reaction of (SiPiPr3)Fe(H2)(H) with CO2 to afford formate
Scheme 6. Dihydride Pathways for Catalytic CO2
Hydrogenation
a
−
(ΔGH = 43 kcal/mol) is endergonic by over 10 kcal/mol.
However, comparisons of only the hydricities of the iron
hydride and formate neglect to take into account the observed
formate coordination to iron (Scheme 2). To estimate the free
energy afforded by formate coordination to iron, we
determined the formate binding constant by UV−vis titration
for the reaction of [(SiPiPr3)Fe(N2)](BArF ) and Li(OCHO) to
4
(SiPiPr3)Fe(OCHO) (Scheme 7, eq 7). The titration in THF
Scheme 7. Gibbs Free Energies for the Reactions of
a
Relevance to CO2 Hydrogenation by (SiPiPr3)Fe(H2)(H)
a
See Chart 1 for a full detailed representation of the ligands indicated.
6b).20 This iron dihydride intermediate was suggested to react
with CO2 to give the iron hydride formate intermediate
(tetraphos)Fe(H)(OCHO). However, we note that a dihydride
intermediate in the (SiPiPr3)Fe system would be unlikely
because the analogous deprotonaton of (SiPiPr3)Fe(H2)(H)
would form an anionic iron dihydride species “[(SiPiPr3)Fe-
(H)2]−”, which is likely to be thermodynamically inaccessible.
For example, a solution of (SiPiPr3)Fe(H2)(H) with excess
triethylamine is stable for hours at 90 °C. While this does not
rule out the possibility of an equilibrium mixture of
(SiPiPr3)Fe(H2)(H) and [(SiPiPr3)Fe(H)2]−, heavily favoring
the neutral monohydride species, we also note that the
estimated pKa of the H2 and H− ligands in (SiPiPr3)Fe(H2)(H)
is greater than 45 in THF,51 vastly higher than that for
triethylamine ([Et3NH]+; pKa = 12.5 in THF).69
a
See Chart 1 for a full detailed ligand representation.
It is of interest to compare the (SiPR )Fe system to the
3
catalytically incompetent (TPB)Fe system because (TPB)(μ-
H)Fe(N2)(H) is an olefin hydrogenation catalyst.44 A key step
that may be required for catalysis is the substitution of Cl− by
H2 in (SiPiPr3)FeCl to give the cationic H2 adduct [(SiPiPr3)-
Fe(H2)]+. Deprotonation of the H2 ligand in a C6D6/THF-d8
mixture by triethylamine leads to the CO2-reactive iron hydride
complex (SiPiPr3)Fe(H2)(H). The initial H2 substitution step,
therefore, is critical towards forming (SiPiPr3)Fe(H2)(H).
However, the {Fe−B}7 complexes (TPB)Fe(OCHO) and
(TPB)FeCl do not react with H2 (4 atm). Related, the
indicates that the binding constant of formate to the iron
complex is on the order of 106 M−1. This is equivalent to ΔG <
−8 kcal/mol for formate binding. Thus, the added driving force
from formate coordination brings the free-energy change for
the reaction of (SiPiPr3)Fe(H2)(H) and CO2 to form
(SiPiPr3)Fe(OCHO) to about 3 kcal/mol (Scheme 7, eq 10;
from the sum of eqs 7−9). This is thermally accessible at the
elevated temperatures at which the stoichiometric and catalytic
reactions are run.
previously reported {Fe−B}7 [(TPB)Fe](BArF ) complex,41
We caution that (SiPiPr3)Fe(H2)(H) may not be the actual
iron hydride intermediate that reacts with CO2; i.e., an
intermediate elementary step may occur prior to CO2 reacting
with the iron complex. The hydricity of such a species is likely
different from that of (SiPiPr3)Fe(H2)(H) owing to the trans-
influencing Si−. We also note that these hydricity values are for
acetonitrile, while the catalytic reactions were run in methanol.
The magnitude of the difference in hydricities between formate
and metal hydrides is known to decrease upon a change from
acetonitrile to water.70 A similar phenomenon may be
occurring in methanol, where the difference in the hydricity
between (SiPiPr3)Fe(H2)(H) and formate may not be as large
as the values in acetonitrile. This, combined with formate
4
which has a vacant fifth coordination site, does not react with
H2 in the presence of excess triethylamine under 1 atm of H2 at
90 °C for 12 h. Furthermore, [(TPB)Fe](BArF ) does not
4
hydrogenate CO2 under the catalytic conditions (see the SI).
[(NPiPr3)FeCl](PF6) is not a hydrogenation precatalyst for
possibly the same reason. Qualitatively, it appears that the
inability of these latter systems to coordinate H2, presumably a
reflection of their weaker ligand-field strengths by comparison
to the SiPR3 system and hence their tendency to populate high-
spin configurations ([(NPiPr3)FeCl](PF6), S = 2;46 (TPB)FeCl,
S = /258), limits their efficacy toward CO2 hydrogenation by
3
comparison with the (SiPR )Fe system [(SiPiPr3)FeCl; S = 157].
3
G
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
NMR spectra were recorded on Varian 300, 400, and 500 MHz
spectrometers. 1H and 13C chemical shifts are reported in ppm relative
to residual solvent as internal standards. 31P and 11B chemical shifts are
reported in ppm relative to 85% aqueous H3PO4 and BF3·Et2O,
respectively. Multiplicities are indicated by br (broad), s (singlet), d
(doublet), t (triplet), quart (quart), quin (quintet), multiplet (m), d−d
(doublet of doublets), and t−d (triplet of doublets).
The ATR-IR measurements were performed in a glovebox on a thin
film of the complex obtained from evaporation of a drop of the
solution on the surface of a Bruker APLHA ATR-IR spectrometer
probe (Platinum Sampling Module, diamond, OPUS software
package) at 2 cm−1 resolution. IR intensities indicated by s (strong),
m (medium), and w (weak).
coordination to iron (Scheme 7, eq 7), may, in fact, make this
formal CO2 insertion step exergonic in methanol.
CONCLUSIONS
■
In summary, we studied a series of triphosphinoiron hydride
complexes, including (SiPR )Fe(L)(H), (PhBPiPr3)Fe-
3
(H)3(PMe3), [(NPiPr3)Fe(N2)(H)](PF6), and (TPB)(μ-H)Fe-
(N2)(H) in the context of CO2 hydrogenation. These iron
hydride complexes react with CO2 to afford iron formate
complexes, which can undergo metathesis with triethylammo-
nium chloride to release triethylammonium formate and well-
defined iron chloride complexes, which are themselves
synthons for the CO2-reactive iron hydride complexes (Scheme
2). Subjecting these iron complexes to catalytic conditions
under elevated pressures of H2 and CO2, we found that
(SiPiPr3)FeCl, (SiPPh3)FeCl, and (PhBPiPr3)FeCl are precata-
lysts for catalytic CO2 hydrogenation to formate and
methylformate (Table 2). (CPiPr3)FeCl, in which carbon
replaces the silicon atom in (SiPiPr3)FeCl, was also a competent
catalyst. The catalytic reactions proceeded in methanol but not
in THF, highlighting the importance of solvent in the catalytic
reaction.61
UV−vis spectra were collected on a Cary 60 UV−vis spectropho-
tometer. The titration experiments were performed in a glovebox using
an Ocean Optics HR4000CG spectrometer.
H2 Quantification by GC-TCD. H2 was quantified on an Agilent
7890A gas chromotograph (HP-PLOT U, 30 m, 0.32 mm i.d.; 30 °C
isothermal; 1 mL/min flow rate; helium carrier gas) using a thermal
conductivity detector. The total amount of H2 produced was
determined as the sum of H2 in the headspace plus dissolved H2 in
the solution calculated by Henry’s law with a constant of 328 MPa.71
Methylformate Quantification by GC-FID. Methylformate
quantification was performed on a 1.2 mL aliquot of the crude
reaction mixture by GC-FID against a methylformate calibration curve.
GC-FID instrument: Hewlett-Packard 5890 with a 57 m Restek RTX-
VRX column (0.32 mm i.d., 1.8 μm films). Method parameters: helium
carrier gas, 1 μL injection volume, 200 °C inlet temperature, 250 °C
detector temperature, 7:1 split ratio, 2.9 mL/min flow rate, 20 psi
pressure, 35 cm/s velocity. Ramp rate: 35 °C initial temperature held
for 8 min, followed by 10 °C/min steps up to 100 °C, then
immediately followed by 25 °C/min steps up to 230 °C, which was
held for 4 min.
As depicted in Scheme 5, we believe that H2 substitution into
(SiPiPr3)FeCl or (SiPiPr3)Fe(OCHO) to form [(SiPiPr3)Fe-
(H2)]+ followed by deprotonation to form the CO2-reactive
(SiPiPr3)Fe(H2)(H) are key steps in the catalytic cycle and
determine catalytic competency. The proposed mechanism for
(SiPiPr3)Fe also differs from the mechanism for the highly active
(tetraphos)Fe system, which proceeds through a dihydride
intermediate.
Synthetic Protocols. Synthesis of (SiPiPr3)FeMe from (SiPiPr3)FeCl.
A yellow solution of (SiPiPr3)FeCl (44.4 mg, 73 μmol) in THF (10
mL) was cooled to −78 °C. A solution of MeMgCl (24 μL of a 3 M
THF solution, 73 μmol) was diluted with THF (1 mL) and then
added dropwise to the stirring reaction, causing a gradual change to a
red solution. The stirring solution was allowed to warm to RT
overnight. The crude mixture was filtered through a glass frit to
remove black precipitate, and the volatiles were removed in vacuo to
reveal a red solid. The material was taken up in a minimal amount of
pentane and allowed to sit at −35 °C overnight, revealing red crystals
of (SiPiPr3)FeMe (11.3 mg, 22%). The 1H and 31P NMR spectra of this
material were identical with the reported spectra.57
Finally, the hydricity value of an iron hydride species has also
been experimentally determined. The hydricity of (SiPiPr3)Fe-
(H2)(H) is 54.3
estimated pKaMeCN of the related conjugate acid [(SiPiPr3)Fe-
(H2)](BArF ) is 15.9
0.7. Despite the low hydricity,
0.9 kcal/mol in acetonitrile, and the
4
(SiPiPr3)Fe(H2)(H) hydrogenates CO2 to formate, and part
of the driving force for the reaction is coordination of formate
to the iron center. Thus, the free-energy change for the reaction
between (SiPiPr3)Fe(H2)(H) and CO2 to (SiPiPr3)Fe(OCHO)
is only slightly uphill at 3 kcal/mol and accessible under the
reactions conditions. It will be of interest to measure the
hydricities of other iron hydrides, including within the present
series of complexes, in the context of CO2 hydrogenation to
better understand the factors that may lead to improved
catalytic activity.
Synthesis of (SiPiPr3)Fe(OCHO). A yellow THF solution (10 mL) of
(SiPiPr3)Fe(N2)(H) (50 mg, 72 μmol) was degassed by freeze−
pump−thaw cycles (3×). Subsequently, CO2 (1 atm) was introduced
to the thawed solution. The reaction was sealed and then heated for 1
h at 50 °C to give a yellow solution. The volatiles were removed in
vacuo to give a yellow solid. The material was extracted with C6H6 and
lyophilized to give (SiPiPr3)Fe(OCHO) as a yellow solid (46 mg,
90%). Analytically pure material was obtained by layering a
concentrated solution of (SiPiPr3)Fe(OCHO) in THF (1 mL) under
HMDSO (5 mL) and allowing the solution to sit at −35 °C for 3 days.
1H NMR (C6D6, 400 MHz): δ 12.2, 12.1, 4.7, 4.6, 1.3, 0.2, −2.0, −2.2,
−4.9. μeff (C6D6, method of Evans, 20 °C): 2.8 μB (S = 1). IR (thin
film, cm−1): 1623 (m, νasym(O−C−O)). UV−vis (THF, nm {M−1
cm−1}): 357 {shoulder, 3247}, 426 {2243}, 478 {253}, 963 {br abs
starting at 884 nm, 451}. Anal. Calcd for C37H55FeO2P3Si: C, 62.71;
H, 7.82. Found: C, 61.81; H, 7.24.
EXPERIMENTAL SECTION
■
General Considerations. All manipulations were carried out using
standard glovebox or Schlenk techniques under an N2 atmosphere.
Unless otherwise noted, solvents were deoxygenated and dried by
thorough sparging with N2 gas followed by passage through an
activated alumina column in the solvent purification system by SG
Water, USA LLC, Nashua, NH. Deuterated solvents and 13CO2 gas
were purchased from Cambridge Isotope Laboratories, Inc., Tewks-
bury, MA. The deuterated solvents were degassed and dried over
activated 3 Å sieves prior to use. Unless otherwise noted, all
compounds were purchased commercially and used without further
purification. (SiPiPr3)Fe(N2)(H),43 (SiPiPr3)FeCl,57 [(SiPiPr3)Fe(N2)]-
(BArF4),37 (SiPPh3)FeCl,57 (SiPPh3)FeMe,57 (PhBPiPr3)Fe-
(H)3(PMe3),45 (PhBPiPr3)FeCl,38 [(NPiPr3)Fe(N2)(H)](PF6),46
[(NPiPr3)FeCl](PF6),46 (TPB)(μ-H)Fe(N2)(H),44 (CPiPr3)FeCl,42
and (CSiPPh3)FeCl47 were synthesized by literature procedures.
Elemental analyses were performed by Robertson Microlit Labo-
ratories, Ledgewood, NJ.
Synthesis of (SiPiPr3)Fe(O13CHO). The procedures used to
synthesize (SiPiPr3)Fe(OCHO) were used here, except that 13CO2
1
was used in place of CO2. The H NMR spectrum was identical with
that of (SiPiPr3)Fe(OCHO). IR (thin film, cm−1): 1583 (m,
νasym(O−13C−O)).
Synthesis of (SiPPh3)Fe(N2)(H). A procedure nearly identical with
that used to synthesize (SiPiPr3)Fe(N2)(H) was used to synthesize
(SiPPh3)Fe(N2)(H). In a 100 mL Schlenk tube, a red solution of
(SiPPh3)FeMe (26.3 mg, 30 μmol) in THF (20 mL) was degassed by
H
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
freeze−pump−thaw cycles (3×). H2 gas (1 atm) was charged into the
thawed solution. The reaction was then sealed and heated to 60 °C for
over 1 week. The reaction was then filtered through Celite, and
volatiles were removed in vacuo to give a light-yellow powder. The
solid was collected on a glass frit and washed with pentane (3 mL ×
2). The resulting product (SiPPh3)Fe(N2)(H) (24.4 mg, 91%) was
obtained as a light-yellow powder after drying under vacuum. Layering
a THF solution of (SiPPh3)Fe(N2)(H) under Et2O and letting the
solution stand for 2 days yielded an analytically pure powder of
(SiPPh3)Fe(N2)(H). 1H NMR (C6D6, 300 MHz): δ 8.55 (2H, d, 2JH−H
(PhBPiPr3)Fe(O13CHO) was identical with that of (PhBPiPr3)Fe-
(OCHO). IR (thin film, cm−1): 1546 (m, νasym(O−13C−O)), 1355
(m, νsym(O−13C−O)).
Synthesis of [(NPiPr3)Fe(OCHO)](PF6). A yellow THF solution (10
mL) of [(NPiPr3)Fe(N2)(H)](PF6) (29 mg, 40 μmol) was degassed by
freeze−pump−thaw cycles (3×). Subsequently, CO2 (1 atm) was
introduced. The reaction was then sealed and stirred for 3 h at RT to
give a colorless solution. The solvent was removed in vacuo to give a
colorless solid. The material was triturated with pentane, and the
solvent was removed in vacuo. The solid was washed with diethyl ether
(3 × 1 mL) to give [(NPiPr3)Fe(OCHO)](PF6) (29 mg, 97%) as a
white solid. Analytically pure material was obtained by layering Et2O
on top of a THF solution of [(NPiPr3)Fe(OCHO)](PF6) and allowing
it to stand overnight at −35 °C. 1H NMR (3:2 mixture of C6D6/THF-
d8, 300 MHz): δ 27.6, 9.4, 8.9, 6.4, 2.1, 1.9, 0.4, −8.3. 31P NMR (3:2
2
= 6 Hz, Ar−H), 8.32 (2H, d, JH−H = 3 Hz, Ar−H), 7.62 (3H, br s,
2
Ar−H), 7.45 (2H, br s, Ar−H), 7.34 (4H, d, JH−H = 6 Hz, Ar−H),
6.85 (2H, t, 2JH−H = 6 Hz, Ar−H), 6.69 (3H, q, 2JH−H = 3 Hz, Ar−H),
2
2
6.52 (2H, q, JH−H = 3 Hz, Ar−H), −11.88 (1H, t-d, JP
= 54 Hz,
cis−H
2JP −H = 12 Hz, Fe−H). 31P NMR (C6D6, 121 MHz): δ 85.3 (2P, s),
trans
1
mixture of C6D6/THF-d8, 121 MHz): δ −144.4 (h, JP−F = 708 Hz,
78.7 (1P, s). 13C NMR (THF with 1 drop of C6D6, 125 MHz): δ 156.3
(d, JC−P = 38 Hz, CAr), 155.8 (d, JC−P = 35 Hz, CAr), 150.7 (d, JC−P = 5
Hz, CAr), 150.4 (s, CAr), 150.2 (s, CAr), 143.0 (s, CAr), 141.6 (d, J = 23
Hz, CAr), 141.0 (s, CAr), 139.6 (d, J = 28 Hz, CAr), 138.6 (d, J = 10 Hz,
CAr), 133.8 (s, CAr), 132.6 (s, CAr), 132.3 (s, CAr), 129.5 (s, CAr), 128.4
(s, CAr), 128.2 (s, CAr), 128.1 (s, CAr), 127.5 (d, J = 5 Hz, CAr). IR (thin
film, cm−1): 2073 (s, ν(N−N)), 1889 (w, ν(Fe−H)). UV−vis (THF,
nm {M−1 cm−1}): 335 {shoulder, 8125}, 437 {shoulder, 4500}. Anal.
Calcd for C54H43FeN2P3Si: C, 72.32; H, 4.83; N, 3.12. Found: C,
72.94; H, 5.22; N, 2.83.
PF6). 19F NMR (3:2 mixture of C6D6/THF-d8, 282 MHz): −73.4 (d,
1JP−F = 710 Hz, PF6). μeff (C6D6, method of Evans, 20 °C): 5.1 μB (S =
2). IR (thin film, cm−1): 1613 (m, νasym(O−C−O)). UV−vis (THF,
nm {M−1 cm−1}): 311 {shoulder, 660}, 379 {shoulder, 249}. Anal.
Calcd for C25H55F6FeNO2P4: C, 43.18; H, 7.97; N, 2.01. Found: C,
44.10; H, 8.25; N, 1.86.
Synthesis of [(NPiPr3)Fe(O13CHO)](PF6). The same procedures as
those used to synthesize [(NPiPr3)Fe(OCHO)](PF6) were used here,
1
except that 13CO2 was used in place of CO2. The H NMR spectrum
of [(NPiPr3)Fe(O13CHO)](PF6) was identical with that of [(NPiPr3)-
Fe(OCHO)](PF6). IR (thin film, cm−1): 1579 (m, νasym(O−13C−O)).
Synthesis of (TPB)FeCl. The procedures used to synthesize
(TPB)FeBr58 were used to synthesize (TPB)FeCl, except that FeCl2
was used in place of FeBr2. A Schlenk tube was charged with TPB (117
mg, 172 μmol), FeCl2 (26 mg, 200 μmol), iron powder (113 mg, 2000
μmol), and THF (20 mL). The reaction was heated to 90 °C for 3
days with vigorous stirring, resulting in a color change of the liquid
phase from light yellow to dark green-brown. The remaining iron
powder was removed by filtration, and the solvent was removed in
vacuo. The residue was taken up in toluene (5 mL), and the solvent
was removed in vacuo. Pentane (200 mL) was added, and the mixture
was stirred for 3 h and filtered. Removal of the solvent in vacuo yielded
a yellow-brown powder of (TPB)FeCl (123 mg, 91%). 1H NMR
(C6D6, 300 MHz): δ 97.6, 35.1, 23.6, 9.6, 5.8, 3.4, 1.9, −0.2, −2.3,
−22.5. μeff (C6D6, method of Evans, 20 °C): 4.1 μB (S = 2). UV−vis
(THF, nm {M−1 cm−1}): 275 {14086}, 317 {10385}, 556 {sh, 80},
774 {66}, 897 {91}. Anal. Calcd for C36H54BClFeP3: C, 63.41; H, 7.98.
Found: C, 64.06; H, 8.89.
Synthesis of (SiPPh3)Fe(OCHO). A yellow THF solution (10 mL) of
(SiPPh)Fe(N2)(H) (51 mg, 57 μmol) was degassed by freeze−pump−
thaw cycles (3×). CO2 (1 atm) was introduced to the thawed solution.
The reaction was sealed and then heated for 1 h at 50 °C to give a
yellow solution. The volatiles were removed in vacuo to give a yellow
solid. The material was redissolved in C6H6 and filtered through a
pipet filter to remove a small amount of black material. The filtrate was
lyophilized in vacuo to give (SiPPh3)Fe(OCHO) as a yellow solid (41
mg, 79%). Analytically pure material was obtained by layering a
concentrated THF solution of (SiPPh3)Fe(OCHO) (3 mL) under
1
pentane (5 mL) and allowing it to stand for 2 days at RT. H NMR
(3:2 mixture of C6D6/THF-d8, 300 MHz): δ 12.2, 6.5, 5.7, 4.8, −2.1,
−4.7. μeff (THF-d8, method of Evans, 20 °C): 2.7 μB (S = 1). IR (thin
film, cm−1): 1618 (m, νasym(O−C−O)), 1316 (m, νsym(O−C−O)).
UV−vis (THF, nm {M−1 cm−1}): 325 {shoulder, 4775}, 415 {4100},
474 {3700}, 995 {br abs starting at 900 nm, 263}. Anal. Calcd for
C55H43FeO2P3Si: C, 72.37; H, 4.75. Found: C, 73.21; H, 5.48.
Synthesis of (SiPPh3)Fe(O13CHO). The same procedures as those
used to synthesize (SiPPh3)Fe(OCHO) were used here, except that
13CO2 was used in place of CO2. The 1H NMR spectrum of
(SiPPh3)Fe(O13CHO) was identical with that of (SiPPh3)Fe(OCHO).
IR (thin film, cm−1): 1587 (m, νasym(O−13C−O)), 1254 (m,
νsym(O−13C−O)).
Synthesis of (TPB)Fe(OCHO). A yellow benzene solution (6 mL) of
(TPB)(μ-H)Fe(N2)(H) (20.7 mg, 31 μmol) was degassed by freeze−
pump−thaw cycles (3×). Subsequently, CO2 (1 atm) was introduced.
The reaction was then sealed, and the yellow solution was mixed for 1
h at RT. The solvent was lyophilized in vacuo to give (TPB)Fe-
(OCHO) as a dark-yellow solid (21.0 mg, 99%). Analytically pure
material was obtained by cooling a concentrated pentane solution of
(TPB)Fe(OCHO) to −35 °C overnight. 1H NMR (C6D6, 300 MHz):
δ 86.1, 66.3, 38.5, 26.3, 15.5, 4.3, 2.7, 1.4, 1.0, −0.7, −2.6, −3.5, −24.1.
μeff (C6D6, method of Evans, 20 °C): 4.2 μB (S = 2). IR (thin film,
cm−1): 1627 (m, νasym(O−C−O)), 1291 (m, νsym(O−C−O)). UV−vis
(THF, nm {M−1 cm−1}): 278 {16400}, 317 {12800}, 773 {br abs, 98},
958 {127}. Anal. Calcd for C37H55BFeO2P3: C, 64.27; H, 8.02. Found:
C, 63.16; H, 7.75.
Synthesis of (PhBPiPr3)Fe(OCHO). A yellow THF solution (1 mL)
of (PhBPiPr3)Fe(H)3(PMe3) (6.7 mg, 12 μmol) was degassed by
freeze−pump−thaw cycles (3×). Subsequently, CO2 (1 atm) was
introduced to the thawed solution. The reaction was sealed and then
stirred for 12 h at RT to give a light-yellow solution. The volatiles were
removed to give an light-yellow solid. The material was triterated with
pentane, and the solvent was removed in vacuo. The material was then
redissolved in C6H6 (3 mL) and filtered through a glass frit to remove
a black solid. Removal of the solvent in vacuo gave (PhBPiPr3)Fe-
(OCHO) (4.7 mg, 70%) as a light-yellow solid. Analytically pure
material was obtained by layering HDMSO on top of a THF solution
of (PhBPiPr3)Fe(OCHO) and allowing it to stand overnight. 1H NMR
(C6D6, 300 MHz): δ 41.1, 19.9, 18.6, 13.5, 9.2, 4.5, 3.6, 1.6, −1.2,
−11.2, −12.1, −32.6, −37.7. μeff (C6D6, method of Evans, 20 °C): 5.0
μB (S = 2). IR (thin film, cm−1): 1595 (m, νasym(O−C−O)), 1362 (m,
νsym(O−C−O)). UV−vis (THF, nm {M−1 cm−1}): 298 {1173}, 410
{274}. Anal. Calcd for C28H54FeO2P3: C, 57.75; H, 9.35. Found: C,
58.12; H, 9.67.
Synthesis of (TPB)Fe(O13CHO). The same procedures as those used
to synthesize (TPB)Fe(OCHO) were used, except that 13CO2 was
1
used in place of CO2. The H NMR spectrum was identical with that
of (TPB)Fe(OCHO). IR (thin film, cm−1): 1588 (m, νasym(O−13C−
O)), 1269 (m, νsym(O−13C−O)).
Reaction of (TPB)Fe(N2) with Formic Acid. (TPB)Fe(N2) (8.2 mg,
12.1 μmol) in 2 mL of THF was charged into a round-bottomed flask,
and the flask was sealed with a rubber septum. Formic acid (3 μL, 80.3
μmol) was added by syringe through the septum, immediately
resulting in effervescence of H2 and a yellow-brown solution. The
solution was allowed to stir for a few minutes before the volatiles were
removed in vacuo to reveal a brown solid. The material was
Synthesis of (PhBPiPr3)Fe(O13CHO). The same procedures as those
used to synthesize (PhBPiPr3)Fe(OCHO) were used here, except that
13CO2 was used in place of CO2. The 1H NMR spectrum of
I
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
redissolved in benzene and filtered. Removal of the volatiles in vacuo
revealed a brown powder of (TPB)Fe(OCHO) (8.1 mg, 96%). NMR
and IR spectral data for this material were identical with those of
(TPB)Fe(OCHO).
and the reactor was then heated at 100 °C for 20 h. Changes to these
conditions were made as described in Tables 2−4 and S1 in the SI. At
the conclusion of the reaction, the reactor was cooled to 10 °C with an
ice bath over ca. 1.5 h, and the pressure was slowly released through a
needle valve. An aliquot of the crude solution was immediately taken
for methylformate quantification by GC-FID. The aliquot was then
recombined with the crude solution, DMF was added (1 mmol), and
25 μL of this solution was taken into 0.5 mL of CD2Cl2 for
Deprotonation of [(SiPiPr3)Fe(H2)](BArF ) with Et3N. [(SiPiPr3)Fe-
4
(N2)](BArF ) (14.6 mg, 9.4 μmol) and triethylamine (1.7 μL, 9.7
4
μmol) were charged into an NMR tube with a J-young valve with C6D6
and THF-d8 (ca. 0.4 and 0.1 mL, respectively), yielding a green
solution. The solution was degassed by freeze−pump−thaw cycles
(3×), revealing an orange solution consistent with [(SiPiPr3)Fe-
1
triethylammonium formate quantification by H NMR spectroscopy.
Similar procedures were followed for the low-pressure reactions (1
atm of CO2 and 1−4 atm of H2), where the reactions were run in a 15
mL Schlenk tube that has a Teflon valve. The solution was degassed by
freeze−pump−thaw cycles (3×), and CO2 (1 atm) was introduced
into the vessel at RT. For the reactions requiring 4 atm of H2, the
entire body of the Schlenk tube was then cooled in a liquid-N2 bath
and 1 atm (RT) of H2 was introduced. For reactions requiring 1 atm of
H2, the Schlenk tube was cooled with liquid N2 up to the solution level
only and 1 atm (RT) of H2 was introduced.
(THF)](BArF ). H2 (1 atm) was charged into the reaction mixture,
4
yielding a transient gray solution (consistent with [(SiPiPr3)Fe(H2)]-
(BArF ) that immediately changed to orange-yellow upon mixing. The
4
reaction was mixed overnight. The NMR data of the iron species in
this reaction mixture were identical with those of (SiPiPr3)Fe(H2)-
(H).43 The volatiles were removed in vacuo, and the resulting yellow
solid was extracted with pentane. The pentane was removed in vacuo
1
to yield a yellow solid of (SiPiPr3)Fe(N2)(H) (5.1 mg, 88%). The H
and 31P NMR spectra of this material are identical with those of
(SiPiPr3)Fe(N2)(H).
Analysis of the Iron Content Postcatalytic Reaction. The
reaction was worked up similarly to the procedures described above
for the hydrogenation runs. After the reactor was depressurized, it was
brought into the glovebox for workup. The crude solution was
transferred to a scintillation vial, and the volatiles were removed in
vacuo. The resulting light-yellow solid was dissolved in C6D6 and
Quantifying H2 Loss from the Reaction of (TPB)(μ-H)Fe(N2)-
(H) with CO2. Procedures similar to those of the synthesis of
(TPB)Fe(OCHO) were followed. (TPB)(μ-H)Fe(N2)(H) (20.0 mg,
31 μmol) was dissolved in 6 mL of benzene and charged into a
calibrated 200 mL Schlenk tube having a Teflon valve and a 24/40 side
joint. The solution was degassed by freeze−pump−thaw cycles (3×),
opened to CO2 (1 atm) and agitated for ca. 5 s to ensure adequate
dissolution of CO2. The reaction was then sealed at the Teflon valve
joint and also with a rubber septum at the 24/40 joint. The reaction
was stirred vigorously for 30 min, the Teflon valve was opened, and
the headspace was sampled through the rubber septum with a 10 mL
gastight syringe, being careful to ensure adequate mixing of the gases
from the reaction headspace into the 24/40 joint’s headspace by
repeated extraction and reinjection (3×) of the headspace gas with the
gastight syringe before a final aliquot was taken for analysis by GC-
TCD. A total of 0.44 equiv of H2 [relative to (TPB)(μ-H)Fe(N2)(H)]
was found.
1
analyzed by H and 31P NMR spectroscopy.
Hydricity Determination. The hydricity was experimentally
determined using the method presented by DuBois et al.52,53 The
equilibrium of eq 3 (Scheme 1) with a given base (proton sponge, 2,6-
lutidine, or 2,4,6-trimethylpyridine) was measured in THF-d8. With
proton sponge as the base, [(SiPiPr3)Fe(N2)](BArF ) (8.0 mg, 5.1
4
μmol) was mixed with proton sponge (1.1 mg, 5.1 μmol) and the
integration standard 1,3,5-trimethoxybenzene (1.2 mg, 7.1 μmol) in
THF-d8 (0.5 mL). With 2,6-lutidine as the base, [(SiPiPr3)Fe(N2)]-
(BArF ) (8.7 mg, 5.6 μmol) was mixed with 2,6-lutidine (34 μL, 292
4
μmol) and the integration standard 1,3,5-trimethoxybenzene (1.1 mg,
6.5 μmol) in THF-d8 (0.5 mL). With 2,4,6-trimethylpyridine as the
base, [(SiPiPr3)Fe(N2)](BArF ) (8.3 mg, 5.3 μmol) was mixed with
4
2,4,6-trimethylpyridine (1.6 μL, 19.8 μmol) and the integration
standard 1,3,5-trimethoxybenzene (1.3 mg, 7.7 μmol) with THF-d8
(0.5 mL). The solutions were degassed by freeze−pump−thaw cycles
(3×), and H2 (1 atm) was introduced. The solutions were mixed using
a mechanical rotator at a rate of ca. 12 min−1, and the reaction was
monitored by 1H NMR spectroscopy until equilibration: proton
sponge, 6 days; 2,6-lutidine, 5 days; 2,4,6-trimethylpyridine, 5 days.
Quantifying H2 Loss from the Reaction of (TPB)Fe(N2) with
Formic Acid. Procedures similar to those of the synthesis of
(TPB)Fe(OCHO) from formic acid and (TPB)Fe(N2) were followed.
(TPB)Fe(N2) (24.5 mg, 36 μmol) in benzene (3 mL) was charged
into a calibrated 100 mL round-bottomed flask and sealed with a
rubber septum. Formic acid (1.3 μL, 36 μmol) was added by syringe.
Effervescence was immediately visible. The reaction was allowed to stir
for a few minutes before the headspace was sampled through the
rubber septum with a 10 mL gastight syringe for analysis by GC-TCD.
A total of 0.42 equiv of H2 [relative to (TPB)Fe(N2)] was found.
Metathesis Reactions of Iron Formate Complexes with
(Et3NH)Cl. (Et3NH)Cl (10 equiv) was charged into a methanol or
benzene solution of (SiPiPr3)Fe(OCHO), (SiPPh3)Fe(OCHO),
(PhBPiPr3)Fe(OCHO), [(NPiPr3)Fe(OCHO)](PF6), or (TPB)Fe-
(OCHO). The resulting suspension was stirred overnight and then
filtered through a glass frit. The filtrate was concentrated in vacuo into
a solid and then extracted. For (SiPiPr3)Fe(OCHO), (SiPPh3)Fe-
(OCHO), (PhBPiPr3)Fe(OCHO), and (TPB)Fe(OCHO), pentane (3
× 1 mL) was used for extraction, while a 4:1 C6H6/THF mixture (3 ×
−
Equation 3−5 were used to calculate ΔGH . The equilibrium between
[(SiPiPr3)Fe(H2)](BArF ) and its THF adduct [(SiPiPr3)Fe(H2)]-
4
(BArF ) was also taken into account in the calculations (see the SI).
4
UV−Vis Titration. The UV−vis titration experiments of [(SiPiPr3)-
Fe(N2)](BArF ) (4.8 mM) with Li(OCHO) (48 mM) in THF was
4
performed by adding aliquots of the formate solution to a solution of
the iron complex. The decay of [(SiPiPr3)Fe(N2)](BArF ) was
4
monitored, with the absorbance at 752 nm used for fitting to a
quadratic equation against Keq. After the addition of 1 equiv of
Li(OCHO), a spectrum corresponding to (SiPiPr3)Fe(OCHO) was
observed.
1
1 mL) was used for [(NPiPr3)Fe(OCHO)](PF6)). The respective H
ASSOCIATED CONTENT
* Supporting Information
■
NMR and IR spectra of the extract showed conversion to
(SiPiPr3)FeCl, (SiPPh3)FeCl, (PhBPiPr3)FeCl, [(NPiPr3)FeCl](PF6), or
(TPB)FeCl.
S
NMR, IR, and UV−vis spectra, along with detailed catalytic
hydrogenation results and a detailed description of the hydricity
calculations. This material is available free of charge via the
Internet at http://pubs.acs.org.
CO2 Hydrogenation Catalysis Protocols. High-pressure hydro-
genation reactions were run in a Parr Instruments Company 4590
Micro Bench Top Reactor, with a 20 mL reaction vessel, controlled by
a Parr Instruments Company 4834 controller. In a typical catalytic run,
an iron precatalyst in 0.1 mL of THF (to solubilize iron precatalyst),
10 mL of methanol, and 1 mL of triethylamine were charged into the
Parr reactor. The reactor was sealed, stirred with the attached
mechanical stirrer (100 rpm), and charged with CO2 until the desired
pressure at equilibrium was achieved (ca. 10 min). H2 was
subsequently added into the reactor. The gas inlet port was closed,
AUTHOR INFORMATION
Corresponding Author
*E-mail: jpeters@caltech.edu.
■
Notes
The authors declare no competing financial interest.
J
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
(23) Inoue, Y.; Sasaki, Y.; Hashimoto, H. J. Chem. Soc., Chem.
Commun. 1975, 718−719.
ACKNOWLEDGMENTS
■
This material is based upon work performed by the Joint
Center of Artificial Photosynthesis, a DOE Energy Innovation
Hub, supported through the Office of Science of the U.S.
Department of Energy under Award DE-SC0004993. The
Bercaw Group at the California Institute of Technology is
acknowledged for their assistance with the high-pressure
reactions. GC-FID instrumentation at the Environmental
Analysis Center (EAC) at the California Institute of
Technology was used in this work. We acknowledge Dr.
Nathan Dalleska of the EAC for his assistance with the GC-FID
measurements.
(24) Haynes, P.; Slaugh, L. H.; Kohnle, J. F. Tetrahedron Lett. 1970,
11, 365−368.
(25) Federsel, C.; Ziebart, C.; Jackstell, R.; Baumann, W.; Beller, M.
Chem.Eur. J. 2012, 18, 72−75.
(26) Jeletic, M. S.; Mock, M. T.; Appel, A. M.; Linehan, J. C. J. Am.
Chem. Soc. 2013, 135, 11533−11536.
(27) Kumar, N.; Camaioni, D.; Dupuis, M.; Raugei, S.; Appel, A. M.
Dalton Trans. 2014, 43, 11803−11106.
(28) Badiei, Y. M.; Wang, W.-H.; Hull, J. F.; Szalda, D. J.;
Muckerman, J. T.; Himeda, Y.; Fujita, E. Inorg. Chem. 2013, 52,
12576−12586.
(29) Field, L. D.; Lawrenz, E. T.; Shaw, W. J.; Turner, P. Inorg. Chem.
2000, 39, 5632−5638.
REFERENCES
(30) Allen, O. R.; Dalgarno, S. J.; Field, L. D. Organometallics 2008,
27, 3328−3330.
■
(1) Carbon Dioxide as Chemical Feedstock; Wiley-VCH Verlag GmbH
& Co. KGaA: Weinheim, Germany, 2010.
(2) Olah, G. A. Angew. Chem., Int. Ed. 2013, 52, 104−107.
(3) Enthaler, S.; von Langermann, J.; Schmidt, T. Energy Environ. Sci.
2010, 3, 1207−1217.
(4) Arakawa, H.; Aresta, M.; Armor, J. N.; Barteau, M. A.; Beckman,
E. J.; Bell, A. T.; Bercaw, J. E.; Creutz, C.; Dinjus, E.; Dixon, D. A.;
Domen, K.; DuBois, D. L.; Eckert, J.; Fujita, E.; Gibson, D. H.;
Goddard, W. A.; Goodman, D. W.; Keller, J.; Kubas, G. J.; Kung, H.
H.; Lyons, J. E.; Manzer, L. E.; Marks, T. J.; Morokuma, K.; Nicholas,
K. M.; Periana, R.; Que, L.; Rostrup-Nielson, J.; Sachtler, W. M. H.;
Schmidt, L. D.; Sen, A.; Somorjai, G. A.; Stair, P. C.; Stults, B. R.;
Tumas, W. Chem. Rev. 2001, 101, 953−996.
(5) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois,
D. L.; Dupuis, M.; Ferry, J. G.; Fujita, E.; Hille, R.; Kenis, P. J. A.;
Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F.; Portis, A. R.; Ragsdale,
S. W.; Rauchfuss, T. B.; Reek, J. N. H.; Seefeldt, L. C.; Thauer, R. K.;
Waldrop, G. L. Chem. Rev. 2013, 113, 6621−6658.
(6) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M. Chem.
Soc. Rev. 2009, 38, 89−99.
(7) Cokoja, M.; Bruckmeier, C.; Rieger, B.; Herrmann, W. A.; Kuhn,
(31) Field, L. D.; Shaw, W. J.; Turner, P. Chem. Commun. 2002, 46−
47.
(32) Hills, A.; Hughes, D. L.; Jimenez-Tenorio, M.; Leigh, G. J. J.
Organomet. Chem. 1990, 391, C41−C44.
(33) Bianco, V. D.; Doronzo, S.; Gallo, N. Inorg. Nucl. Chem. Lett.
1980, 16, 97−99.
(34) Bianco, V. D.; Doronzo, S.; Rossi, M. J. Organomet. Chem. 1972,
35, 337−339.
(35) Estes, D. P.; Vannucci, A. K.; Hall, A. R.; Lichtenberger, D. L.;
Norton, J. R. Organometallics 2011, 30, 3444−3447.
(36) DuBois, D. L.; Berning, D. E. Appl. Organomet. Chem. 2000, 14,
860−862.
(37) Lee, Y.; Mankad, N. P.; Peters, J. C. Nat. Chem. 2010, 2, 558−
565.
(38) Betley, T. A.; Peters, J. C. J. Am. Chem. Soc. 2003, 125, 10782−
10783.
(39) Betley, T. A.; Peters, J. C. J. Am. Chem. Soc. 2004, 126, 6252−
6254.
(40) Anderson, J. S.; Rittle, J.; Peters, J. C. Nature 2013, 501, 84−87.
(41) Anderson, J. S.; Moret, M.-E.; Peters, J. C. J. Am. Chem. Soc.
2012, 135, 534−537.
(42) Creutz, S. E.; Peters, J. C. J. Am. Chem. Soc. 2013, 136, 1105−
1115.
̈
F. E. Angew. Chem., Int. Ed. 2011, 50, 8510−8537.
(8) The Handbook of Homogeneous Hydrogenation; Wiley-VCH Verlag
GmbH: Weinheim, Germany, 2008.
(9) Jessop, P. G.; Ikariya, T.; Noyori, R. Nature 1994, 368, 231−233.
(10) Tanaka, R.; Yamashita, M.; Nozaki, K. J. Am. Chem. Soc. 2009,
131, 14168−14169.
(43) Lee, Y.; Kinney, R. A.; Hoffman, B. M.; Peters, J. C. J. Am. Chem.
Soc. 2011, 133, 16366−16369.
(44) Fong, H.; Moret, M.-E.; Lee, Y.; Peters, J. C. Organometallics
2013, 32, 3053−3062.
(11) Graf, E.; Leitner, W. J. Chem. Soc., Chem. Commun. 1992, 623−
624.
(45) Daida, E. J.; Peters, J. C. Inorg. Chem. 2004, 43, 7474−7485.
(46) MacBeth, C. E.; Harkins, S. B.; Peters, J. C. Can. J. Chem. 2005,
83, 332−340.
(12) Lau, C. P.; Chen, Y. Z. J. Mol. Catal. A: Chem. 1995, 101, 33−
36.
(13) Erlandsson, M.; Landaeta, V. R.; Gonsalvi, L.; Peruzzini, M.;
Phillips, A. D.; Dyson, P. J.; Laurenczy, G. Eur. J. Inorg. Chem. 2008,
2008, 620−627.
(14) Huff, C. A.; Sanford, M. S. J. Am. Chem. Soc. 2011, 133, 18122−
18125.
(47) Rittle, J.; Peters, J. C. Proc. Natl. Acad. Sci. U. S. A. 2013, 110,
15898−15903.
(48) Heinekey, D. M.; van Roon, M. J. Am. Chem. Soc. 1996, 118,
12134−12140.
(49) THF is known to bind competitively with H2 on [(SiPiPr
)
3
(15) Huff, C. A.; Sanford, M. S. ACS Catal. 2013, 2412−2416.
(16) Schaub, T.; Paciello, R. A. Angew. Chem., Int. Ed. 2011, 50,
7278−7282.
(17) Wang, W.-H.; Muckerman, J. T.; Fujita, E.; Himeda, Y. ACS
Catal. 2013, 3, 856−860.
(18) Langer, R.; Diskin-Posner, Y.; Leitus, G.; Shimon, L. J. W.; Ben-
David, Y.; Milstein, D. Angew. Chem., Int. Ed. 2011, 50, 9948−9952.
(19) Federsel, C.; Boddien, A.; Jackstell, R.; Jennerjahn, R.; Dyson, P.
J.; Scopelliti, R.; Laurenczy, G.; Beller, M. Angew. Chem., Int. Ed. 2010,
49, 9777−9780.
(20) Ziebart, C.; Federsel, C.; Anbarasan, P.; Jackstell, R.; Baumann,
W.; Spannenberg, A.; Beller, M. J. Am. Chem. Soc. 2012, 134, 20701−
20704.
Fe(L)](BArF ] (where L = THF or H2), and the THF/H2 binding
4
equilibrium was taken into account in the calculations; see the SI.
(50) The pKa of [(SiPiPr3)Fe(H2)](BArF ) was estimated using the
4
Morris method. These calculations rely on the ligand acidity constants
for each of the ligands of the conjugate base metal complex, which in
this case is the deprotonation product (SiPiPr3)Fe(H2)(H). We note
that the ligand acidity constants for H2 and the formally Si− ligands of
the conjugate base complex are not known. We therefore used the
reported ligand acidity constants of C2H4 as a model for the H2 ligand
−
and of CH3 /H− as a model for the Si− ligand unit for these
calculations. See ref 51.
(51) Morris, R. H. J. Am. Chem. Soc. 2014, 136, 1948−1959.
(52) Ciancanelli, R.; Noll, B. C.; DuBois, D. L.; DuBois, M. R. J. Am.
Chem. Soc. 2002, 124, 2984−2992.
(53) Price, A. J.; Ciancanelli, R.; Noll, B. C.; Curtis, C. J.; DuBois, D.
L.; DuBois, M. R. Organometallics 2002, 21, 4833−4839.
(21) Evans, G. O.; Newell, C. J. Inorg. Chim. Acta 1978, 31, L387−
L389.
(22) Tai, C.-C.; Chang, T.; Roller, B.; Jessop, P. G. Inorg. Chem.
2003, 42, 7340−7341.
K
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
(54) Abdur-Rashid, K.; Fong, T. P.; Greaves, B.; Gusev, D. G.;
Hinman, J. G.; Landau, S. E.; Lough, A. J.; Morris, R. H. J. Am. Chem.
Soc. 2000, 122, 9155−9171.
(55) Chen, J. Z.; Szalda, D. J.; Fujita, E.; Creutz, C. Inorg. Chem.
2010, 49, 9380−9391.
(56) Gibson, D. H. Coord. Chem. Rev. 1999, 185−186, 335−355.
(57) Whited, M. T.; Mankad, N. P.; Lee, Y.; Oblad, P. F.; Peters, J. C.
Inorg. Chem. 2009, 48, 2507−2517.
(58) Moret, M.-E.; Peters, J. C. Angew. Chem., Int. Ed. 2011, 50,
2063−2067.
(59) The {M−B}n notation refers to the number of valence electrons
that are formally assigned to the metal (e.g., iron) and shared with the
boron atom. For a discussion of M−B bonding, see: (a) Hill, A. F.
Organometallics 2006, 25, 4741−4743. (b) Parkin, G. Organometallics
2006, 25, 4744−4747.
(60) Marks, T. J.; Kolb, J. R. Chem. Rev. 1977, 77, 263−293.
(61) Munshi, P.; Main, A. D.; Linehan, J. C.; Tai, C.-C.; Jessop, P. G.
J. Am. Chem. Soc. 2002, 124, 7963−7971.
(62) Wesselbaum, S.; vom Stein, T.; Klankermayer, J.; Leitner, W.
Angew. Chem., Int. Ed. 2012, 51, 7499−7502.
(63) Mellone, I.; Peruzzini, M.; Rosi, L.; Mellmann, D.; Junge, H.;
Beller, M.; Gonsalvi, L. Dalton Trans. 2013, 42, 2495−2501.
(64) Saouma, C. T.; Lu, C. C.; Day, M. W.; Peters, J. C. Chem. Sci.
2013, 4, 4042−4051.
(65) Gambarotta, S.; Arena, F.; Floriani, C.; Zanazzi, P. F. J. Am.
Chem. Soc. 1982, 104, 5082−5092.
(66) Bielinski, E. A.; Lagaditis, P. O.; Zhang, Y.; Mercado, B. Q.;
Wurtele, C.; Bernskoetter, W. H.; Hazari, N.; Schneider, S. J. Am.
̈
Chem. Soc. 2014, 136, 10234−10237.
(67) Drake, J. L.; Manna, C. M.; Byers, J. A. Organometallics 2013, 32,
6891−6894.
(68) Otera, J.; Nishikido, J. Esterification; Wiley-VCH Verlag GmbH
& Co. KGaA: Weinheim, Germany, 2010.
(69) Kaljurand, I.; Kutt, A.; Soovali, L.; Rodima, T.; Maemets, V.;
̈
̈
̈
Leito, I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019−1028.
(70) Matsubara, Y.; Fujita, E.; Doherty, M. D.; Muckerman, J. T.;
Creutz, C. J. Am. Chem. Soc. 2012, 134, 15743−15757.
(71) Brunner, E. J. Chem. Eng. Data 1985, 30, 269−273.
L
DOI: 10.1021/ic502508p
Inorg. Chem. XXXX, XXX, XXX−XXX