The Role of Proton Shuttles in the Reversible Activation of Hydrogen via Metal-Ligand Cooperation

Article abstract of DOI:10.1021/jacs.9b09062

The reversible activation of H₂ via a pathway involving metal-ligand cooperation (MLC) is proposed to be important in many transition metal catalyzed hydrogenation and dehydrogenation reactions. Nevertheless, there is a paucity of experimental information probing the mechanism of this transformation. Here, we present an in-depth kinetic study of the 1,2-addition of H₂ via an MLC pathway to the widely used iron catalyst [(^iPrPNP)FeH(CO)] (1) (^iPrPNP = N(CH₂CH₂PⁱPr₂)₂ ^-). We report one of the first experimental demonstrations of an enhancement in rate for the activation of H₂ using protic additives, which operate as "proton shuttles". Our results indicate that proton shuttles need to be able to both simultaneously donate and accept a proton, and the best shuttles are molecules that are strong hydrogen bond donors but sufficiently weak acids to avoid deleterious protonation of the transition metal complex. Additionally, comparison of the rate of H₂ activation via an MLC pathway between 1 and two widely used ruthenium catalysts enables more general conclusions about the role of the metal, ancillary ligand, and proton shuttles in H₂ activation. The results of this study provide guidance about the design of catalysts and additives to promote H₂ activation via an MLC pathway.

Full text of DOI:10.1021/jacs.9b09062

Article
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
pubs.acs.org/JACS
The Role of Proton Shuttles in the Reversible Activation of
Hydrogen via Metal−Ligand Cooperation
Nicholas E. Smith,† Wesley H. Bernskoetter, and Nilay Hazari*
‡
,†
†
The Department of Chemistry, Yale University, P.O. Box 208107, New Haven, Connecticut 06520, United States
The Department of Chemistry, The University of Missouri, Columbia, Missouri 65211, United States
‡
*
S Supporting Information
ABSTRACT: The reversible activation of H via a pathway
2
involving metal−ligand cooperation (MLC) is proposed to be
important in many transition metal catalyzed hydrogenation and
dehydrogenation reactions. Nevertheless, there is a paucity of
experimental information probing the mechanism of this
transformation. Here, we present an in-depth kinetic study of
the 1,2-addition of H via an MLC pathway to the widely used
2
iPr
iPr
i
−
iron catalyst [( PNP)FeH(CO)] (1) ( PNP = N(CH CH P Pr ) ). We report one of the first experimental demonstrations
2
2
2 2
of an enhancement in rate for the activation of H using protic additives, which operate as “proton shuttles”. Our results indicate
2
that proton shuttles need to be able to both simultaneously donate and accept a proton, and the best shuttles are molecules that
are strong hydrogen bond donors but sufficiently weak acids to avoid deleterious protonation of the transition metal complex.
Additionally, comparison of the rate of H activation via an MLC pathway between 1 and two widely used ruthenium catalysts
2
enables more general conclusions about the role of the metal, ancillary ligand, and proton shuttles in H activation. The results
2
of this study provide guidance about the design of catalysts and additives to promote H activation via an MLC pathway.
2
INTRODUCTION
cleave H through an alternative mechanism involving metal−
2
■
4
ligand cooperation (MLC). Specifically, H is split to form a
2
The transition metal mediated hydrogenation of unsaturated
molecules is one of the most widely used catalytic trans-
formations in both academia and industry. In particular, the
metal hydride and a protonated ligand (Figure 1b). Addition-
1
ally, the microscopic reverse reaction in which H is eliminated
2
from a metal hydride and a protonated ligand is proposed to be
a key step in the dehydrogenation of a number of polar
use of homogeneous catalysts for these reactions has enabled
in-depth mechanistic understanding that has led to important
advances in catalyst performance and selectivity. Historically,
many catalysts for the hydrogenation of nonpolar double or
triple bonds proceed through a mechanism in which the
cleavage of the H−H bond occurs through an elementary
oxidative addition step (Figure 1a). In this well understood
reaction the ancillary ligand(s) have no direct role in H2
activation. In contrast, although some transition metal catalysts
5
substrates such as alcohols and amines. Although 1,2-
addition/elimination of H via MLC is a commonly proposed
2
step, there have been a paucity of experimental studies
5a,6
exploring the kinetics of this reaction, and it is not clear how
2
catalysts can be altered to promote or inhibit H addition/
2
elimination. Further, predominantly on the basis of computa-
tional studies, it has been suggested that molecules with protic
functional groups, such as alcohols and water (which are often
present in the reaction mixture as substrates or products), can
for the hydrogenation of polar double or triple bonds, such as
ketones and imines, activate H via oxidative addition, many
catalysts for these and other polar substrates are proposed to
3
2
assist in H addition/elimination by acting as proton shuttles
2
4
a,c,d,7
(
Figure 2).
However, there are no experimental studies
comparing the rates of H activation with and without a shuttle
2
or establishing which features of a shuttle should be modified
8
,9
to increase their effectiveness.
The first (de)hydrogenation catalysts proposed to activate
H via a pathway involving MLC were based on precious
2
4a,10
metals such as ruthenium.
However, a number of catalysts
incorporating first-row transition metals are also proposed to
4d,e,11
operate through a MLC pathway.
In particular, recently
developed cobalt, nickel, manganese, and iron catalysts
i
iPr
H
featuring the tridentate ligand HN(CH CH P Pr ) ( PN P)
2
2
2 2
Figure 1. Mechanisms for H activation by a transition metal complex
2
involving (a) oxidative addition and (b) metal−ligand cooperation
Received: August 21, 2019
(
MLC).
©
XXXX American Chemical Society
A
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
iPr
iPr
H
Figure 2. Activation of H by [( PNP)Fe(H)(CO)] (1) to form [( PN P)Fe(H) (CO)] (2) in toluene, (a) in the absence of protic additives
2
2
t
and (b) in the presence of a protic additive such as BuOH. Line plots of kinetic traces for the conversion of 1 and H to an equilibrium mixture
with 2 in the presence of varying equivalents of (c) BuOH and methyl tert-butyl ether (MTBE) and (d) p-X-2-phenyl-2-propanol (X = H, Me, or
2
t
Br). See SI for additional information about reaction conditions.
4
e
have shown remarkable activity. For example, both our group
in this reaction and extend our findings to ruthenium systems
iPr
H
and others have reported that [( PN P)Fe(H)(BH )(CO)]
that activate H via a pathway involving MLC. Our results
4
2
iPr
and related precatalysts, which access [( PNP)Fe(H)(CO)]
provide general information about how proton shuttles can be
used to assist (de)hydrogenation reactions that use catalysts
iPr
i
−
iPr
H
(
(
1) ( PNP = N(CH CH P Pr ) ) and [( PN P)Fe-
2 2 2 2
H) (CO)] (2) in situ (Figure 2), are active for the
hydrogenation of esters, ketones, nitriles, amides,
nitrogen-containing heterocycles, olefins, and CO2. The
same precatalysts also give high activity for the dehydrogen-
ation of formic acid, primary and secondary alcohols,
polyols such as glycerol, nitrogen-containing heterocycles,
which add or release H using MLC.
2
2
1
2
7b
13
14
5
e
15
16
RESULTS AND DISCUSSION
■1
,2-Addition of H to 1 in the Presence and Absence
2
17
7b
5e
t
of BuOH. Given its widespread use as a catalyst for
5i
(
de)hydrogenation reactions we selected 1 as our initial
5f
and ammonia borane and mediate the dehydrogenative
model complex to study H activation via an MLC
2
5h
5h
5k
5
d−f,h−l,7b,12a,13−17
synthesis of lactones, lactams, and amides. In almost all
of these reactions, a key step is proposed to be either the 1,2-
pathway.
1
At room temperature (25 °C) and
atm of H , 1 and 2 exist in equilibrium (Table 1). The
2
addition of H to 1 to form 2 or H elimination from 2 to form
(Figure 2a) with the latter often proposed as the turnover-
2
2
conversion of 1 and H to an equilibrium mixture of 1, H , and
2
2
1
2
in toluene was monitored using UV−visible spectroscopy.
limiting step for dehydrogenation reactions. Previous computa-
tional work suggests that methanol and water act as proton
shuttles for the interconversion of 1 and 2, but this has not
The addition of H is first-order in [1] and [H ], and there is
2
2
no evidence for autocatalytic behavior from the N−H moiety
of 2, which could in principle act as a proton shuttle.
Evaluating the role of water and methanol as proton shuttles is
complicated by the fact that the O−H bond of both of these
species undergo reversible 1,2-addition across the Fe−N bond
5
g,7b
been verified experimentally.
In this work, we describe a
detailed mechanistic study of the interconversion between 1,
H , and 2. We conclusively establish that proton shuttles assist
2
B
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Table 1. Thermodynamic Effects on the Equilibrium between 1, H , and 2 in the Presence of Additives
2
thermodynamic parameters
a
entry
additive
Keq at 298 K
ΔH (kcal/mol)
ΔS (cal/mol K)
ΔG298 K (kcal/mol)
1
2
3
4
5
6
7
none
BuOH
3700 ± 200
3600 ± 200
16000 ± 800
3600 ± 200
11900 ± 600
9800 ± 500
7600 ± 400
−7.8 ± 0.2
−9.5 ± 0.3
−14.8 ± 0.4
−16.1 ± 0.5
−11.6 ± 0.3
−8.8 ± 0.3
−8.4 ± 0.3
−9.7 ± 0.3
−15.7 ± 0.5
−30.3 ± 0.9
−37.8 ± 1.1
−20.3 ± 0.6
−11.3 ± 0.3
−10.3 ± 0.3
−4.9 ± 0.2
−4.8 ± 0.2
−5.8 ± 0.3
−4.8 ± 0.2
−5.5 ± 0.3
−5.4 ± 0.3
−4.5 ± 0.2
t
2-phenyl-2-propanol
p-bromo-2-phenyl-2-propanol
p-methyl-2-phenyl-2-propanol
2,6-lutidine
b
MTBE
a
Equilibrium constants were recorded using 1 atm of H in the presence of 10 equiv of the additive over a temperature range spanning ca. 50 °C.
2
For a typical experiment, [1] before H addition was 20 mM. The concentration of H was determined from literature values for the solubility of H
in toluene at the given temperature and 1 atm. See Supporting Information for van’t Hoff plots and further details. MTBE = methyl tert-butyl
2
2
2
b
ether.
5g,k
of 1 (see Supporting Information (SI) for more details).
Additionally, methanol is dehydrogenated to form form-
entropically favorable (ΔΔS = −6 cal/mol K). This
observation is consistent with a hydrogen bonding complex
5
g
t
t
aldehyde and 2. Instead, BuOH was selected as a model
alcohol for exploring proton shuttling since it cannot be
dehydrogenated and does not add across the Fe−N bond of 1
at room temperature (when 50 equiv are present relative to 1).
between BuOH and 2 being present in solution (along with 2
t
and free BuOH). Further evidence for hydrogen bonding
t
between 2 and BuOH was obtained using solution state IR
spectroscopy, which showed noticeable changes in the
spectrum of 2 in the presence of excess BuOH (see SI for
t
When the reaction between 1 and H is performed in the
2
t
presence of 10 equiv of BuOH, there is a significant
details).
t
enhancement (1.7x) in the rate at which the reaction
approaches equilibrium compared to a reaction with no
additive (Figure 2c). The shuttled reaction is first-order in [1],
To obtain more information about the effect of BuOH on
the kinetics of H addition/elimination across the Fe−N bond,
2
the concentration profiles of both the shuttled and unshuttled
t
t
18
[
H ], and [ BuOH] (although we note that BuOH is not
reactions were modeled using the COPASI program. This
2
consumed in the reaction), and as a result even greater
enabled elementary rate constants for the forward and reverse
steps of both the shuttled and unshuttled reactions to be
enhancements can be obtained by adding more equivalents of
t
BuOH. A control reaction with 10 equiv of methyl tert-butyl
determined (Table 2, entries 1 and 2). The observed first-order
t
ether (MTBE) gave no rate enhancement, demonstrating the
dependence on [1], [H
2
], and [ BuOH] suggests that if the
t
shuttled reaction to form 2 is a multistep process, rather than a
importance of the proton in BuOH. These preliminary results
concerted trimolecular reaction, the individual steps are
represent one of the first experimental observations of a proton
19
kinetically indistinguishable. As a consequence, reaction
shuttling effect in the activation of H via a pathway involving
2
profiles in the absence of additives were modeled using a two-
MLC.
term rate law of the form Rate = k [1][H ] − k [2] (Figure
In reactions containing methanol, previous DFT calculations
suggest that a hydrogen bonded complex between 2 and
methanol is in equilibrium with 2 and free methanol, and is an
1
2
−1
2
a). Reaction profiles in the presence of additives, such as
t
BuOH, were modeled using a four-term rate law of the form
t
intermediate in the pathway for H activation involving proton
Rate = k
1
[1][H
2
] − k−1[2] + k [1][H
and k are the forward unshuttled
2
2
2
][ BuOH] − k−2[2]-
2
5
g,7b
t
shuttling.
This interaction between 2 and methanol was
[ BuOH]. The values of k
1
calculated to be exothermic (−12.1 kcal/mol relative to 2 and
free methanol) but endergonic at catalytic temperatures (+1.1
kcal/mol relative to 2 and free methanol, calculated at 393
and shuttled rate constants, respectively, and k−1 and k−2 are
the reverse unshuttled and shuttled rate constants, respectively
(Figure 2b).
7
b
1
K). Analysis using both H NMR and UV−vis spectroscopy
Activation parameters determined from an Eyring plot
(Table 3) show that the barriers for both the forward and
t
indicates that the presence of BuOH causes no significant
−
1
change in the position of the equilibrium between 1 and H2
and 2 at room temperature (Table 1, entries 1 and 2). When
the temperature is changed, however, there are differences in
the position of the equilibrium (see SI). At room temperature,
the enthalpy and entropy changes (derived from a van’t Hoff
reverse reactions are approximately 2.2 kcal mol lower at
t
room temperature in the presence of BuOH. This is in
t
agreement with our observation that the addition of BuOH
does not change the position of the equilibrium between 1, H ,
2
and 2 at room temperature (vide supra). In the absence of
t
t
plot) of the reaction are offset, as in the presence of BuOH,
BuOH, the predominant barrier to the formation of 2 is
and the conversion of 1 and H2 to 2 becomes more
enthalpically favorable (ΔΔH = −1.7 kcal/mol), but less
entropic, with only a small enthalpic contribution to the free
energy barrier (entry 1). In the presence of BuOH (entry 2),
t
C
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
t
Table 2. Rate Constants for Reactions between 1, H , and
2
freedom when BuOH is also incorporated into the transition
state structure, this is counteracted by gains in rotational and
vibrational entropy upon moving from a four-membered to a
six-membered ring in the transition state. For the reverse
2
a
reaction, 1,2-elimination of H to form 1, there is no change
2
(
within error) in the entropic barrier upon the addition of
t
t
BuOH (entries 3 and 4). This is consistent with BuOH
hydrogen bonding to 2 prior to H loss and forming 2- BuOH
t
2
as the ground state structure, which avoids a significant
entropic penalty in the transition state.
a
b
entry
1
additive
kforward
k1 = 4.3 ± 0.4 M− s−1
kreverse
1
= 1.2(1) × 10−3
Kinetic Isotope Effects on the 1,2-Addition of H to 1
none
k
2
−1
−
1
t
s
in the Absence and Presence of BuOH. To further probe
−
1
−7
(
0.6 ± 0.06 M s−1 at (6.2(6) × 10 s−1 at
the mechanism of H addition/elimination, several kinetic
2
−
60 °C)
−60 °C)
isotope effects (KIEs) were measured for reactions in the
tBuOH
k2 = 160 ± 16 M s−1 k−2 = 4.5(5) × 10
−2
−2
2
3
4
5
t
absence and presence of BuOH (Table 4). In order to avoid
−
1
−1
M
s
complications due to chemical exchange between protonated
and deuterated sites, KIEs were measured by comparing initial
rates at low conversion. A KIE of 3.2 ± 0.4 was measured for
−
2
2-phenyl-
-propanol
p-bromo-2-phenyl-
-propanol
p-methyl-2-phenyl-
-propanol
k = 1200 ± 120
k−2 = 7.5(8) × 10
−1 −1
2
M
−
2
−1
2
s
M
s
k = 1600 ± 160
k−2 = 0.44 ± 0.04
2
M
−
2
−1
−1 −1
t
2
s
M
s
the 1,2-addition of H or D to 1 in the absence of BuOH
2
2
k2 = 600 ± 60 M s−1 k−2 = 5.0(5) × 10
−2
−2
(entry 1). This primary KIE is consistent with H being
2
−
1
−1
2
M
s
cleaved in the rate-determining transition state and is similar to
values that have been reported for σ-bond metathesis and 1,2-
addition reactions for early transition metals, which are
proposed to proceed through a related four-membered
a
For entry 1 this is the elementary rate constant (k ) for the
unshuttled conversion of 1 and H to 2 (Figure 2a). For all other
entries these are the elementary rate constants (k ) for the shuttled
conversion of 1 and H to 2 (Figure 2b). For entry 1 this is the
elementary rate constant (k ) for the unshuttled conversion of 2 to 1
and H (Figure 2a). For all other entries these are the elementary rate
constants (k ) for the shuttled conversion of 2 to 1 and H (Figure
b). See SI for more details on how elementary rate constants were
1
2
2
b
2
21
transition state. Replacing the hydride ligand in 1 with a
−
1
iPr
deuteride, [( PNP)Fe(D)(CO)] (1-d ), resulted in a KIE of
1
2
3
.8 ± 0.4 when the rate of addition of H to 1 is compared
2
−
2
2
with the rate of addition of D to 1-d (entry 2), which is
2
2
1
determined.
indistinguishable (within error) from that measured in entry 1.
This comparison suggests that the spectator hydride ligand in 1
does not have an effect on the rate of 1,2-addition.
Table 3. Activation Parameters for Reactions
Interconverting 1 and H and 2 in the Presence and
Absence of BuOH
We next analyzed the impact of deuteration on the shuttled
2
t
reaction by measuring KIEs using different combinations of d -
1
t
t
tert-butanol ( BuOD), BuOH, D , and H . Specifically, we
determined KIEs using BuOD with D (entry 3), BuOD with
H (entry 4), and BuOH with D (entry 5), all relative to a
reaction utilizing BuOH and H . Our initial experiment with
2
2
‡
ΔG at
t
t
‡
‡
2
activation
parameters
ΔH
ΔS
298 K
t
entry
1
(kcal/mol) (cal/(mol K)) (kcal/mol)
2
2
t
formation of 2,
unshuttled, k1
2.9 ± 0.3
−0.9 ± 1
10.7 ± 0.6
8.6 ± 0.5
−46 ± 2
−51 ± 5
−36 ± 2
−36 ± 2
16.6 ± 0.8
14.3 ± 2.4
21.4 ± 1.1
19.3 ± 1.0
2
t
BuOD and D gave a KIE of 1.7 ± 0.2. This result is consistent
2
2
3
4
formation of 2,
shuttled, k2
with the shuttle changing the nature of the transition state, as
the value is significantly smaller than the KIE for the
unshuttled reaction (entry 1). At this stage, the exact reasons
why the KIE is smaller are unclear. It could suggest that there
is less bond breaking in the shuttled transition state compared
to the unshuttled case, which would be consistent with the
proposed pathways for the reaction (Figure 2). However,
interpreting isotope effects for H2 addition reactions to
transition metal complexes is complicated, as equilibrium
effects can play a large role, and this possibility cannot be
formation of 1 and
H , unshuttled, k
2
−1
formation of 1 and
H , shuttled, k
2
−2
a
Activation parameters determined using Eyring plots (see SI) with
rate constants measured from 258 to 308 K in 10 K increments. See
SI for reaction conditions under which activation parameters were
determined.
22
the contribution of enthalpy to the barrier for the formation of
excluded in this case.
t
2
is less than or close to zero presumably due to hydrogen
The KIE for the reaction using BuOD and H (entry 4), 5.0
2
t
bonding type interactions in the transition state (vide infra).
For the unshuttled formation of 2 the entropic barrier is −46 ±
2
± 0.6, is much larger than the KIE for the reaction with BuOD
and D (entry 3). In contrast, the reaction with BuOH and D
t
2
2
cal/mol K (entry 1). This large entropy of activation is
(entry 5) gives a similar KIE, 1.8 ± 0.2, to entry 3. Given that
these three reactions proceed through the same transition state
(Figure 2b), and have the same restrictions on rotational and
vibrational modes, this difference likely results from enthalpic
effects rather than entropic effects. We suggest that it is
possible to explain the difference by considering the bonds that
are broken and formed in the transition states (see SI for a
more complete discussion). In all cases, an Fe−N π-bond is
broken, and therefore this bond is not relevant to the
similar to those reported for σ-bond metathesis reactions with
0
0
d /f metal centers that invoke a “highly ordered” four-
20
membered transition structure. The transition state for the
unshuttled 1,2-addition of H is also proposed to proceed
through a four-membered transition state (Figure 2),
2
5
l,7b,c
where there is likely a significant loss of both vibrational and
rotational freedom. Interestingly, the shuttled formation of 2
t
with BuOH has an entropic barrier that is only marginally
t
higher at −51 ± 5 cal/mol K (entry 2). We suggest that even
though there is a significant loss in translational degrees of
differences in KIEs. In entry 4, when BuOD is used as the
shuttle with H , O−D, and H−H bonds are broken and O−H,
2
D
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
t
a
Table 4. Kinetic Isotope Effect Measurements on the Addition of H to 1 in the Absence and Presence of BuOH
2
a
KIEs measured via the method of initial rates to avoid complications due to isotopic scrambling. The KIE for the unshuttled reaction was
measured first, to determine the elementary rate constant for 1,2-addition of D . After this, rate constants for the reactions in the presence of a
2
shuttle were determined by taking the observed rate and subtracting the background rate of the unshuttled reaction under the given conditions,
with H or D . The remaining rate was used to determine the elementary rate constant for the shuttled reaction and the KIE calculated relative to
2
2
t
b
the shuttled rate constant with BuOH and H . See SI for more details. Initial isotopic substitution drawn for the product. Chemical exchange
2
produces a mixture over time.
N−D, and Fe−H bonds are formed. The O−D bond is
stronger than the N−D bond, and consequently, it is more
preferable for the D atom to be bound to the O atom and it is
enthalpically harder to reach the transition state relative to the
explain the observed KIEs. Overall, our results strongly suggest
that there is a difference in transition state between the
shuttled and unshuttled reaction and that deuteration of the
shuttle impacts the reaction.
t
comparative reaction when BuOH is used as the shuttle. This
Effects of Other Additives on the 1,2-Addition of H2
t
23
t
to 1. Finding other additives apart from BuOH that could act
results in the large KIE. In entry 3, when BuOD is used as
as proton shuttles for the addition of H2 to 1 proved
challenging (see SI for a list of all additives tested). Strongly
acidic additives such as thiols, phenols, and urea underwent
the shuttle with D , O−D, and D−D bonds are broken and
2
O−D, N−D, and Fe−D bonds are formed. Here, there is little
relative difference in the strengths of the bonds formed and
1
,2-addition across the Fe−N bond of 1, which precluded their
broken relative to the comparison reaction with H and
2
t
use as shuttles due to this deleterious side reaction. In small
BuOH, so only a small KIE is observed. In entry 5, when D
and BuOH are used, O−H and D−D bonds are broken and
2
amounts (0.75−2.0 equiv), water accelerates the conversion of
t
1
and H to 2, but the presence of a pre-equilibrium of 1,2-
2
O−D, N−H, and Fe−D bonds are formed. This situation is
comparable to entry 3, and as a result the KIE is also small.
Even though the different KIEs observed in entries 3−5 can be
explained by bond enthalpies, caution needs to be taken with
this interpretation due to the possibility that equilibrium
isotope effects are also contributing. Additionally, the
uncertainty in the exact speciation of 1 and 2 when proton
addition of water across the Fe−N bond of 1 prevents the
determination of rate constants and when a higher number of
equivalents of water are used the reaction is inhibited (see SI
for kinetic traces). Our results with water are consistent with
previous proposals that adventitious water can act as a proton
7a
shuttle. Interestingly, the addition of a variety of amines to
the reaction of 1 with H did not result in any measurable rate
2
t
shuttles such as BuOH (which can hydrogen bond to the
enhancement (see below for further discussion). This included
substrates that are typically used in N-heterocycle dehydrogen-
complexes) are present further complicates our ability to
E
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. (a) Dependence of the shuttled rate constant, k , on the enthalpy of the formation of 2 from 1 and H , which is an indicator of hydrogen
2
2
+
bond strength. (b) Schematic showing the two interactions that a shuttle must be able to engage in, to accept and donate H in the transition state,
for the shuttle-mediated process to occur. See SI for reaction conditions under which rate constants were determined.
5l
ation, such as piperidine and 1,2,3,4-tetrahydroquinoline.
from 1 and H in the presence of Br-Ar at room temperature is
2
However, 2-phenyl-2-propanol (H-Ar) and the related para-
substituted derivatives p-bromo-2-phenyl-2-propanol (Br-Ar)
and p-methyl-2-phenyl-2-propanol (Me-Ar) could be used as
the same as for the unshuttled reaction.
Analysis of the types of compounds that can act as proton
shuttles allows us to make some general statements about what
is required to facilitate shuttling in the 1,2-addition of H to 1:
(i) The shuttle requires an acidic functional group in order to
proton shuttles (Table 2, entries 3−5 and Figure 2d). All three
2
t
compounds gave larger values of k than BuOH, indicating
2
t
that they are more efficient shuttles. When comparing the 2-
phenyl-2-propanol series, the presence of an electron-donating
group slows the reaction, while the electron-poor shuttle
enhances the rate, suggesting that the acidity of the shuttle is a
key factor in proton shuttle activity. In fact, reactions were
performed with only 5 equiv of Br-Ar because the reaction
proceeds too quickly to obtain reliable rate information in the
be capable of donating a proton. For example, BuOH can act
as a proton shuttle, but MTBE and 2,6-lutidine cannot (Figure
3). (ii) The shuttle cannot be too strong of an acid, or it will
react directly with the transition metal complex and inhibit H2
activation. For example, phenol and tert-butylthiol cannot be
used as shuttles with 1. (iii) In addition to being able to donate
a proton, the shuttle needs to be able to accept a proton which
is facilitated by a hydrogen bonding interaction. In fact, the
presence of 10 equiv. The value of k , the shuttled rate
2
constant, is almost three times larger for Br-Ar (1600 ± 160
enthalpy of the formation of 2 from 1 and H in the presence
2
−
2
−1
−2 −1
M
s ) than for Me-Ar (600 ± 60 M s ) (Table 2, entries
of different proton shuttles (a measure of the strength of the
hydrogen bonding) demonstrates a negative correlation with
the elementary rate constant for the shuttle-mediated process
(Figure 3). For a compound to be an effective shuttle,
requirements (i)−(iii) must all be met. MTBE and 2,6-lutidine
can participate in hydrogen bonding with 2 (Table 1, entries 6
and 7), but the lack of an acidic functional group means they
cannot shuttle hydrogen (Figure 3). In contrast, secondary
amines such as piperidine have a basic group, but cannot easily
donate a proton and do not engage in strong hydrogen
4
and 5).
When 2-phenyl-2-propanol and related species are used as
proton shuttles they not only affect the kinetics of the reaction
but also change the thermodynamics (Table 1, entries 3−5).
For example, the addition of H-Ar makes the reaction between
1
and H to form 2 approximately 1 kcal/mol more exergonic
2
than the unshuttled reaction (entry 3). Specifically, H-Ar
changes the thermodynamics by making the reaction more
enthalpically favorable, but less entropically favorable. This is
consistent with our model that a proton shuttle can hydrogen
bond to 2, and it has commonly been observed in the literature
that hydrogen bonding in the products makes reactions more
24
bonding relative to alcohols, so have no real effect as shuttles.
The relative performance of the different proton shuttles can
be further rationalized by considering the nature of the 1,2-
24
enthalpically, but less entropically, favorable. When the
proton shuttle is more electron-rich, it would be expected to be
a worse hydrogen bond donor and form weaker hydrogen
addition of H in the presence of a proton shuttle on a
2
molecular level (Figure 4). The proton shuttling reaction can
be described as consisting of two concomitant proton transfers:
(i) a proton transfer from the shuttle to the amido moiety of
the ligand and (ii) a proton transfer from a coordinated
dihydrogen to the proton shuttle. These transfers could be
either concerted or asynchronous, in which case either proton
transfer could occur first. At one extreme, the fully
asynchronous transfer of each proton would result in a
stepwise mechanism (corresponding to the outside edges of
Figure 4a). In nonpolar solvents like toluene, the high degree
of charge separation necessary for fully asynchronous proton
transfers will be disfavored. Therefore, it is probable that the
2
5
bonds with 2. In agreement with this hypothesis, Me-Ar
forms weaker hydrogen bonds with 2 than H-Ar as evidenced
by the fact that the enthalpy for the conversion of 1 and H to
2
2
is not as favorable and the entropy is more favorable. Overall,
it is less favorable to form 2 in the presence of Me-Ar than H-
Ar. In the case of the less electron-rich proton shuttle Br-Ar, it
is enthalpically very favorable to form 2, but entropically highly
disfavored, presumably because of extremely strong hydrogen
bonding. However, at room temperature the entropic term
dominates and the equilibrium constant for the formation of 2
F
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
proceeds in the intermediate asynchronous fashion depicted in
path II in Figure 4.
The results of our kinetic studies on proton shuttling
confirm computational results suggesting that proton shuttles
can lower the barrier for H addition to 1 or elimination from
2
5g,l,7b,c,26
2
.
They also show that in catalytic reactions that
generate alcohols, such as methanol as products, or use
alcohols as the substrate, there is no need to use an additive to
facilitate H addition or elimination, as there are native shuttles
2
in the system. In contrast, in reactions that do not use or
5
e
generate alcohols, such as N-heterocycle (de)hydrogenation
15
or olefin hydrogenation, the addition of a small quantity of a
proton shuttle may aid the reaction by facilitating steps
involving H addition or elimination. We propose the use of
2
t
tertiary alcohols such as BuOH or substituted 2-phenyl-2-
proponols as additives, as they are less likely to undergo side
reactions with 1, which lead to catalyst deactivation or
consumption of the proton shuttle. Alternatively, our results
suggest that the main barrier to 1,2-addition of H in 1 relates
2
to the entropic barrier associated with forming the highly
ordered four-membered transition state and if ligands can be
designed with similar basicity but which are preorganized for
Figure 4. (a) Square scheme showing the different pathways by which
the two proton transfers involved in the 1,2-addition of H to 1 in the
2
H addition, it may be possible to significantly reduce the
2
presence of a proton shuttle could occur. Vertical arrows correspond
to proton transfer from the shuttle to the amido nitrogen, while
horizontal arrows correspond to proton transfer from coordinated H2
to the shuttle. The dashed arrows correspond to intermediate
asynchronous proton transfers, which could involve initial deproto-
barrier for the reaction without using a proton shuttle.
Exploring Addition of H via an MLC Pathway in
2
Ruthenium Complexes. We were interested in evaluating
the applicability of our findings on proton shuttling to other
transition metal complexes that engage in the activation of H2
nation of coordinated H (I) or protonation of the amido ligand (II).
2
The bold diagonal arrow corresponds to concerted transfer of both
protons. (b) Generic representations of proposed transition state
structures for asynchronous intermediate pathway I or asynchronous
intermediate pathway II. The proposed transition state structure for
the concerted pathway is shown in Figure 2b.
via MLC. Given that the ruthenium congener of 1,
iPr
[
( PNP)Ru(H)(CO)] (3), is also proposed to be an
important catalytic intermediate in many (de)hydrogenation
5c,m,7c,27
reactions,
it was selected for kinetic studies in both the
presence and absence of a proton shuttle. Complex 3
undergoes a rapid, irreversible reaction with H at room
2
effects of proton shuttles in nonpolar solvents are much more
iPr
H
temperature to form [( PN P)Ru(H) (CO)] (4) (Figure 5).
2
important than in polar solvents where stepwise mechanisms
The rate constant for the conversion of 3 to 4 without any
7d
for the activation of H via MLC have been proposed. If the
2
4
additive present is 10 times larger than for the transformation
two proton transfers are truly concerted, there should be
minimal charge buildup on the proton shuttle, and the acidity
or basicity of the proton shuttle would be expected to have
little effect on its performance. A slight buildup of positive
charge on the shuttle, however, would be expected if proton
transfer from a coordinated dihydrogen molecule to the proton
shuttle dominates the beginning of an intermediate asynchro-
nous reaction coordinate (path I in Figure 4). In this case,
stronger bases, such as amines, should be superior proton
shuttles. Instead, our results indicate that stronger acids, such
as alcohols, in particular electron poor alcohols, are superior
proton shuttles. This suggests that the proton transfer to the
amido moiety (initial deprotonation of the shuttle) of the
ligand likely begins prior to deprotonation of the bound
dihydrogen molecule. On this basis, we propose the reaction
of 1 to 2. This large rate enhancement for the heavier
transition metal element is similar to observations made for
1
,2-addition reactions of C−H bonds across early transition
metal systems with metal ligand multiple bonds, which are also
proposed to react via four-centered transition states.
21c,28
While the unshuttled addition of H to 3 can be monitored at
2
room temperature, the shuttled reaction proceeds too quickly
and must be monitored at low temperatures (−60 °C) to allow
for collection of full kinetic profiles for the reaction. Addition
t
of only 1 equiv of BuOH increases the rate of addition of H
2
to 3 by a factor of 2 (see SI). The reaction is first-order in [3]
t
and BuOH, suggesting that the mechanism of proton shuttling
is comparable to that described for 1. The larger effect of
t
BuOH in the case of 3 is likely due to the ruthenium complex
iPr
t
Figure 5. Rate constants for 1,2-addition of H to [( PNP)Ru(H)(CO)](3) in the presence and absence of BuOH. See SI for reaction conditions
2
under which rate constants were determined.
G
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
tBu
Figure 6. Effect of proton shuttles on the 1,3-addition of H to[( PN P)Ru(H)(CO)](5). (a) Rate constants for 1,3-addition of H enhanced by
2
py
2
t
p-bromo-2-phenyl-2-propanol (Br−Ar). (b) Line plots for the conversion of 5 and H to 6 with varying equivalents of BuOH. See SI for reaction
2
conditions under which rate constants were determined.
t
being a better hydrogen bonding partner for BuOH as both a
hydrogen bond donor and acceptor.
alcohol we could inhibit coordination to the metal center by
the shuttle. Indeed, when 2 equiv of p-bromo-2-phenyl-2-
propanol (Br-Ar) are used as an additive, a significant
We were interested in investigating whether or not proton
shuttling was a general phenomenon applicable to other MLC
platforms besides the aliphatic PNP ligand. One of the most
common platforms for MLC is the use of pincer ligands that
incorporate pyridyl groups that can undergo dearomatization
upon deprotonation of a distal C−H bond. The ruthenium
complex [( PN P*)Ru(H)(CO)] (5,
NC H (CH P Bu )(CHP Bu ) ) can activate H by under-
going a 1,3-addition reaction to form [( PN P)Ru-
enhancement (1.5x) in the rate of 1,3-addition of H is
2
observed. The enhancement becomes more pronounced as the
number of equivalents of Br-Ar are increased, and when 10
equiv of the additive are utilized, the reaction proceeds too
quickly for a rate to be measured. This suggests that proton
shuttling is relevant to multiple ligand scaffolds that activate H2
via a pathway involving MLC.
tBu
py
tBu
py
PN P* =
t
t
−
5
3
2
2
2
2
tBu
py
The results in this section provide experimental confirma-
tion that proton shuttling is general to complexes beyond 1.
However, they also indicate that an additive which facilitates
proton shuttling with one transition metal catalyst may not be
suitable for a different system. Instead, the identity of a shuttle
needs to be matched with a catalyst to ensure that background
reactions between the shuttle and the catalyst are minimized.
At this stage, the matching of a catalyst with the best shuttle
will likely have to be performed through empirical
optimization. In the future, more detailed computational
studies which explore the transition state structure in the
presence of different shuttles may assist in the rational design
of proton shuttles for a variety of catalysts, with our results
enabling computational models to be benchmarked. Our data
also suggest that, in general, ruthenium systems are
thermodynamically more likely to undergo 1,2-addition of H2
than iron systems, suggesting that ruthenium systems may be
better for hydrogenation reactions, whereas iron catalysts may
be better suited for dehydrogenation. Kinetically, 1,2-addition
is more facile for ruthenium systems, suggesting that proton
shuttles are likely to be more important with iron catalysts than
with ruthenium systems and that more basic sites in the
ancillary ligand are required for iron to display comparable
reactivity to ruthenium.
tBu
py
t
(
H) (CO)] (6, PN P = NC H (CH P Bu ) ) analogous
2 5 3 2 2 2
to the 1,2-addition of H for 1 and 3, and is a widely proposed
intermediate in (de)hydrogenative catalysis (Figure 6a).
By monitoring the addition of H2 to 5 using UV−vis
spectroscopy, we determined that the reaction is first-order
2
4c,29
in [5] but, in contrast to H activation by 3, proceeds with a
2
rate constant only ca. 5 times larger than that for the
conversion of 1 to 2 at 25 °C (Figures 5 and 6a). We propose
that the slower rate of addition is because the carbon center
which accepts a proton from H in 5 is less basic than the
nitrogen center which receives the proton from H in 3, and
consequently there is less stabilization of the TS. We note,
however, that this difference may also arise simply due to
structural differences in the transition state because the
addition is a 1,3 rather than 1,2 process. Interestingly, addition
of BuOH to the reaction between 5 and H leads to an
inhibition in the rate of H activation (Figure 6b). Although
BuOH does not undergo a 1,3 addition to 5, a variable-
temperature experiment using NMR spectroscopy suggests a
fluxional process on the NMR time scale, which we propose is
related to transient binding of the alcohol directly to the metal
center (see SI). We postulated that by using a less electron-rich
2
2
t
2
2
t
H
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
CONCLUSIONS
REFERENCES
■
■
Our work demonstrates that there are mechanistic similarities
between the activation of H via an MLC pathway and both σ-
bond metathesis reactions and 1,2-addition reactions across
metal−ligand multiple bonds. We propose that this is because
all of these reactions involve similar four-membered transition
(1) (a) The Handbook of Homogeneous Hydrogenation; de Vries, J. G.,
Elsevier, C. J., Eds.: Wiley-VCH: Weinheim, 2008. (b) Morris, R. H.
2
Mechanisms of the H -Hydrogenation and Transfer Hydrogenation of
2
Polar Bonds Catalyzed by Iron Group Hydrides. Dalton Trans 2018,
7, 10809−10826. (c) Filonenko, G. A.; van Putten, R.; Hensen, E. J.
4
M.; Pidko, E. A. Catalytic (De)Hydrogenation Promoted by Non-
Precious Metals − Co, Fe and Mn: Recent Advances in an Emerging
Field. Chem. Soc. Rev. 2018, 47, 1459−1483. (d) Mukherjee, A.;
Milstein, D. Homogeneous Catalysis by Cobalt and Manganese
Pincer Complexes. ACS Catal. 2018, 8, 11435−11469. (e) Sa, J.;
Medlin, J. W. On-The-Fly Catalyst Modification: Strategy to Improve
Catalytic Processes Selectivity and Understanding. ChemCatChem
states. To improve the rates of H addition or elimination
2
reactions via an MLC pathway, systems need to be designed
that limit the entropic penalty currently associated with these
reactions. Additionally, we show that proton shuttles can
enhance H addition or elimination in systems that operate via
2
MLC. To act as a shuttle, a compound needs to be able to
simultaneously donate and accept a proton and should not
undergo background reactions with the transition metal
catalyst. Compounds which meet the above criteria and are
good hydrogen bond donors are likely to be better shuttles and
2
019, 11, 3355−3365. (f) Mironenko, R. M.; Belskaya, O. B.;
Likholobov, V. A. Approaches to the Synthesis of Pd/C Catalysts with
Controllable Activity and Selectivity in Hydrogenation Reactions.
Catal. Today 2019, in press; DOI: 10.1016/j.cattod.2019.03.023.
(
g) Junge, K.; Papa, V.; Beller, M. Cobalt-Pincer Complexes in
Catalysis. Chem. - Eur. J. 2019, 25, 122−143.
2) (a) Halpern, J. Homogeneous Catalytic Hydrogenation: A
Retrospective Account. J. Organomet. Chem. 1980, 200, 133−144.
(b) Crabtree, R. H. The Organometallic Chemistry of the Transition
Metals, 5th ed.; Wiley: New Jersey, USA, 2009; pp 224−258.
(c) Kubas, G. J. Activation of Dihydrogen and Coordination of
cause a larger increase in the rate of H addition or elimination
2
compared with compounds that are poorer hydrogen bond
donors. More specifically, our data suggest that common
alcohols are good shuttles, whereas common amines are
ineffective. In many catalytic reactions, reagents that can
operate as proton shuttles are likely already present in the
reaction, as they are substrates or products. However, in cases
where native proton shuttles are not present, the addition of a
small amount of an exogenous shuttle could enhance catalytic
rates. The selection of a shuttle needs to be matched with the
transition metal catalyst, as a compound that is an effective
shuttle with one catalyst may undergo deleterious side
reactions with a different catalyst. Overall, our results provide
fundamental information on how to modulate the rates of H2
addition and elimination reactions that proceed via an MLC
pathway, and future work in our groups will aim to apply these
principles to improving catalytic reactions.
(
Molecular H on Transition Metals. J. Organomet. Chem. 2014, 751,
2
3
2
(
3−49. (d) Crabtree, R. H. Dihydrogen Complexation. Chem. Rev.
016, 116, 8750−8769.
3) Blaser, H. U.; Malan, C.; Pugin, B.; Spindler, F.; Steiner, H.;
Studer, M. Selective Hydrogenation for Fine Chemicals: Recent
Trends and New Developments. Adv. Synth. Catal. 2003, 345, 103−
151.
(
4) (a) Clapham, S. E.; Hadzovic, A.; Morris, R. H. Mechanisms of
the H -Hydrogenation and Transfer Hydrogenation of Polar Bonds
2
Catalyzed by Ruthenium Hydride Complexes. Coord. Chem. Rev.
̈
2004, 248, 2201−2237. (b) Grutzmacher, H. Cooperating Ligands in
Catalysis. Angew. Chem., Int. Ed. 2008, 47, 1814−1818. (c) Khusnut-
dinova, J. R.; Milstein, D. Metal−Ligand Cooperation. Angew. Chem.,
Int. Ed. 2015, 54, 12236−12273. (d) Morris, R. H. Exploiting Metal−
Ligand Bifunctional Reactions in the Design of Iron Asymmetric
Hydrogenation Catalysts. Acc. Chem. Res. 2015, 48, 1494−1502.
e) Alig, L.; Fritz, M.; Schneider, S. First-Row Transition Metal
De)Hydrogenation Catalysis Based On Functional Pincer Ligands.
Chem. Rev. 2019, 119, 2681−2751.
5) (a) Casey, C. P.; Johnson, J. B.; Singer, S. W.; Cui, Q. Hydrogen
ASSOCIATED CONTENT
■
*
S
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.9b09062.
(
(
(
Additional information about selected experiments,
NMR spectra, and other details (PDF)
Elimination from a Hydroxycyclopentadienyl Ruthenium(II) Hy-
dride: Study of Hydrogen Activation in a Ligand-Metal Bifunctional
Hydrogenation Catalyst. J. Am. Chem. Soc. 2005, 127, 3100−3109.
(
b) Casey, C. P.; Guan, H. Cyclopentadienone Iron Alcohol
AUTHOR INFORMATION
■
*
Complexes: Synthesis, Reactivity, and Implications for the Mechanism
of Iron-Catalyzed Hydrogenation of Aldehydes. J. Am. Chem. Soc.
009, 131, 2499−2507. (c) Bertoli, M.; Choualeb, A.; Lough, A. J.;
Moore, B.; Spasyuk, D.; Gusev, D. G. Osmium and Ruthenium
Corresponding Author
nilay.hazari@yale.edu
2
ORCID
Catalysts for Dehydrogenation of Alcohols. Organometallics 2011, 30,
3479−3482. (d) Alberico, E.; Sponholz, P.; Cordes, C.; Nielsen, M.;
Nicholas E. Smith: 0000-0003-2504-0191
Wesley H. Bernskoetter: 0000-0003-0738-5946
Nilay Hazari: 0000-0001-8337-198X
Drexler, H.-J.; Baumann, W.; Junge, H.; Beller, M. Selective Hydrogen
Production from Methanol with a Defined Iron Pincer Catalyst under
Mild Conditions. Angew. Chem. 2013, 125, 14412−14416. (e) Chak-
raborty, S.; Brennessel, W. W.; Jones, W. D. A Molecular Iron Catalyst
for the Acceptorless Dehydrogenation and Hydrogenation of N-
Notes
The authors declare no competing financial interest.
Heterocycles. J. Am. Chem. Soc. 2014, 136, 8564−8567. (f) Glu
̈
er, A.;
Forster, M.; Celinski, V. R.; Schmedt auf der Gunne, J.; Holthausen,
̈
̈
ACKNOWLEDGMENTS
M. C.; Schneider, S. Highly Active Iron Catalyst for Ammonia Borane
■
Dehydrocoupling at Room Temperature. ACS Catal. 2015, 5, 7214−
This work was supported by the U.S. Department of Energy,
Office of Science, Basic Energy Sciences, Catalysis Science
Program, under Award DE-SC0018222. We thank Professor
James Mayer for the use of UV−vis instrumentation and Julia
Curley, Jessica Heimann, and Tanya Townsend for assistance
in editing this manuscript.
7
217. (g) Bielinski, E. A.; Forster, M.; Zhang, Y.; Bernskoetter, W. H.;
̈
Hazari, N.; Holthausen, M. C. Base-Free Methanol Dehydrogenation
Using a Pincer-Supported Iron Compound and Lewis Acid Co-
catalyst. ACS Catal. 2015, 5, 2404−2415. (h) Pena-Lopez, M.;
̃
́
Neumann, H.; Beller, M. Iron(II) Pincer-Catalyzed Synthesis of
Lactones and Lactams through a Versatile Dehydrogenative Domino
I
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Sequence. ChemCatChem 2015, 7, 865−871. (i) Sharninghausen, L.
S.; Mercado, B. Q.; Crabtree, R. H.; Hazari, N. Selective Conversion
of Glycerol to Lactic Acid with Iron Pincer Precatalysts. Chem.
Commun. 2015, 51, 16201−16204. (j) Anke, F.; Han, D.; Klahn, M.;
Spannenberg, A.; Beweries, T. Formation of High-Molecular Weight
Polyaminoborane by Fe Hydride Catalysed Dehydrocoupling of
Methylamine Borane. Dalton Trans 2017, 46, 6843−6847. (k) Lane,
E. M.; Uttley, K. B.; Hazari, N.; Bernskoetter, W. Iron-Catalyzed
Amide Formation from the Dehydrogenative Coupling of Alcohols
and Secondary Amines. Organometallics 2017, 36, 2020−2025.
(12) (a) Chakraborty, S.; Dai, H.; Bhattacharya, P.; Fairweather, N.
T.; Gibson, M. S.; Krause, J. A.; Guan, H. Iron-Based Catalysts for the
Hydrogenation of Esters to Alcohols. J. Am. Chem. Soc. 2014, 136,
7869−7872. (b) Werkmeister, S.; Junge, K.; Wendt, B.; Alberico, E.;
Jiao, H.; Baumann, W.; Junge, H.; Gallou, F.; Beller, M. Hydro-
genation of Esters to Alcohols with a Well-Defined Iron Complex.
Angew. Chem., Int. Ed. 2014, 53, 8722−8726.
(13) Bornschein, C.; Werkmeister, S.; Wendt, B.; Jiao, H.; Alberico,
E.; Baumann, W.; Junge, H.; Junge, K.; Beller, M. Mild and Selective
Hydrogenation of Aromatic and Aliphatic (Di)nitriles with a Well-
Defined Iron Pincer Complex. Nat. Commun. 2014, 5, 4111.
(
l) Bellows, S. M.; Chakraborty, S.; Gary, J. B.; Jones, W. D.;
(14) (a) Schneck, F.; Assmann, M.; Balmer, M.; Harms, K.; Langer,
Cundari, T. R. An Uncanny Dehydrogenation Mechanism: Polar
Bond Control over Stepwise or Concerted Transition States. Inorg.
Chem. 2017, 56, 5519−5524. (m) Nguyen, D. H.; Trivelli, X.; Capet,
R. Selective Hydrogenation of Amides to Amines and Alcohols
Catalyzed by Improved Iron Pincer Complexes. Organometallics 2016,
35, 1931−1943. (b) Rezayee, N. M.; Samblanet, D. C.; Sanford, M. S.
F.; Swesi, Y.; Favre-Reguillon, A.; Vanoye, L.; Dumeignil, F.; Gauvin,
́
Iron-Catalyzed Hydrogenation of Amides to Alcohols and Amines.
ACS Catal. 2016, 6, 6377−6383. (c) Jayarathne, U.; Zhang, Y.;
Hazari, N.; Bernskoetter, W. H. Selective Iron-Catalyzed Deaminative
Hydrogenation of Amides. Organometallics 2017, 36, 409−416.
R. M. Deeper Mechanistic Insight into Ru Pincer-Mediated
Acceptorless Dehydrogenative Coupling of Alcohols: Exchanges,
Intermediates, and Deactivation Species. ACS Catal. 2018, 8, 4719−
4
(
734.
6) Abdur-Rashid, K. C.; Clapham, S. E.; Hadzovic, A.; Harvey, J.
N.; Lough, A. J.; Morris, R. H. J. Am. Chem. Soc. 2002, 124, 15104−
5118. Here, the kinetics of a catalytic hydrogenation reaction, in
(15) Xu, R.; Chakraborty, S.; Bellows, S. M.; Yuan, H.; Cundari, T.
R.; Jones, W. D. Iron-Catalyzed Homogeneous Hydrogenation of
Alkenes under Mild Conditions by a Stepwise, Bifunctional
Mechanism. ACS Catal. 2016, 6, 2127−2135.
1
which the 1,2-addition of H is proposed to be the turnover-limiting
2
(16) Zhang, Y.; MacIntosh, A. D.; Wong, J. L.; Bielinski, E. A.;
step, are determined.
Williard, P. G.; Mercado, B. Q.; Hazari, N.; Bernskoetter, W. H. Iron
(
7) (a) Iron, M. A.; Ben-Ari, E.; Cohen, R.; Milstein, D. Metal−
Catalyzed CO Hydrogenation to Formate Enhanced by Lewis Acid
2
Ligand Cooperation in the Trans Addition of Dihydrogen to a Pincer
Co-Catalysts. Chem. Sci. 2015, 6, 4291−4299.
Ir(I) Complex: A DFT Study. Dalton Trans 2009, 9433−9439.
(17) Bielinski, E. A.; Lagaditis, P. O.; Zhang, Y.; Mercado, B. Q.;
(
b) Chakraborty, S.; Lagaditis, P. O.; Forster, M.; Bielinski, E. A.;
Wurtele, C.; Bernskoetter, W. H.; Hazari, N.; Schneider, S. Lewis
Acid-Assisted Formic Acid Dehydrogenation Using a Pincer-
Supported Iron Catalyst. J. Am. Chem. Soc. 2014, 136, 10234−10237.
Hazari, N.; Holthausen, M. C.; Jones, W. D.; Schneider, S. Well-
Defined Iron Catalysts for the Acceptorless Reversible Dehydrogen-
ation-Hydrogenation of Alcohols and Ketones. ACS Catal. 2014, 4,
(18) Lee, C.; Xu, L.; Singhal, M.; Mendes, P.; Hoops, S.; Pahle, J.;
3
994−4003. (c) Yang, X. Mechanistic Insights into Ruthenium-
Simus, N.; Gauges, R.; Sahle, S.; Kummer, U. COPASIa COmplex
Catalyzed Production of H and CO from Methanol and Water: A
2
2
PAthway SImulator. Bioinformatics 2006, 22, 3067−3074.
DFT Study. ACS Catal. 2014, 4, 1129−1133. (d) Dub, P. A.; Henson,
N. J.; Martin, R. L.; Gordon, J. C. Unravelling the Mechanism of the
A s y m m e t r i c H y d r o g e n a t i o n o f A c e t o p h e n o n e b y
(19) In all likelihood the reaction is a multistep process, given the
low statistical probability of a truly trimolecular process. Nevertheless,
our kinetic model does not enable us to differentiate between these
two pathways.
[
RuX (diphosphine)(1,2-diamine)] Catalysts. J. Am. Chem. Soc.
2
2014, 136, 3505−3521. (e) Dub, P. A.; Scott, B. L.; Gordon, J. C.
(
20) Waterman, R. σ-Bond Metathesis: A 30-Year Retrospective.
Organometallics 2013, 32, 7249−7263.
21) (a) Lin, Z.; Marks, T. J. Metal, Bond Energy, and Ancillary
Why Does Alkylation of the N−H Functionality within M/NH
Bifunctional Noyori-Type Catalysts Lead to Turnover? J. Am. Chem.
Soc. 2017, 139, 1245−1260. (f) Dub, P. A.; Gordon, J. C. The Role of
the Metal-Bound N−H Functionality in Noyori-Type Molecular
Catalysts. Nature Rev. Chem. 2018, 2, 396−408.
(
Ligand Effects on Actinide-Carbon 2-Bond Hydrogenolysis. A Kinetic
and Mechanistic Study. J. Am. Chem. Soc. 1987, 109, 7979−7985.
(b) Hill, C. L. Activation and Functionalization of Alkanes; Wiley: New
(
8) Friedrich, A.; Drees, M.; Schmedt auf der Gunne, J.; Schneider,
̈
Jersey, USA, 1989. (c) Wolczanski, P. T. Activation of Carbon−
Hydrogen Bonds via 1,2-RH-Addition/-Elimination to Early Tran-
sition Metal Imides. Organometallics 2018, 37, 505−516.
S. J. Am. Chem. Soc. 2009, 131, 17552. Here, the impact of water on
chemical exchange between the hydrido ligands and the proton
iPr
H
associated with the N−H group in [( PN P)Ru(H) (PMe) ] using
2
3
(22) (a) Parkin, G. Temperature-Dependent Transitions Between
1
H NMR spectroscopy is investigated. They conclude that the rate of
Normal and Inverse Isotope Effects Pertaining to the Interaction of
H-H and C-H Bonds with Transition Metal Centers. Acc. Chem. Res.
exchange increases in the presence of water and suggest that proton
shuttling is involved in the exchange.
2
009, 42, 315−325. (b) Gomez-Gallego, M.; Sierra, M. A. Kinetic
́
(
9) In ref 5a, Casey and co-workers measured the kinetics of H2
Isotope Effects in the Study of Organometallic Reaction Mechanisms.
elimination in the presence of H O and EtOH from a complex related
2
Chem. Rev. 2011, 111, 4857−4963.
to Shvo’s catalyst. However, because another molecule of the complex
(23) Essentially, the large KIE occurs because there is a much greater
can assist in H elimination by acting as a proton shuttle, they were
2
difference between the relative strength of O−D and O−H bonds
unable to measure the rate of the reaction in the absence of a shuttle.
compared to the difference between N−D and N−H bonds.
(
10) (a) Shvo, Y.; Czarkie, D.; Rahamim, Y.; Chodosh, D. F. A New
(24) Jeffrey, G. A. An Introduction to Hydrogen Bonding; Oxford
Group of Ruthenium Complexes: Structure and Catalysis. J. Am.
Chem. Soc. 1986, 108, 7400−7402. (b) Gao, J.-X.; Ikariya, T.; Noyori,
R. A Ruthenium(II) Complex with a C2-Symmetric Diphosphine/
Diamine Tetradentate Ligand for Asymmetric Transfer Hydro-
genation of Aromatic Ketones. Organometallics 1996, 15, 1087−
University Press: New York, 1997.
(25) This assumes that the interaction between the N−H of the
ancillary ligand and the O of the proton shuttle is the dominant
interaction in the hydrogen bonding between 2 and the shuttle. There
is additionally likely an interaction between the O−H of the shuttle
and the Fe−H; in fact this is required for the shuttle to operate, but
energetically it is likely less significant.
1
089. (c) Ikariya, T. In Bifunctional Molecular Catalysis; Ikariya, T.,
Shibasaki, M., Eds.; Springer Berlin Heidelberg: Berlin, Heidelberg,
011.
11) Zell, T.; Milstein, D. Hydrogenation and Dehydrogenation Iron
2
(
(26) Artus Suarez, L.; Culakova, Z.; Balcells, D.; Bernskoetter, W.
́
̀
H.; Eisenstein, O.; Goldberg, K. I.; Hazari, N.; Tilset, M.; Nova, A.
The Key Role of the Hemiaminal Intermediate in the Iron-Catalyzed
Deaminative Hydrogenation of Amides. ACS Catal. 2018, 8, 8751−
8762.
Pincer Catalysts Capable of Metal−Ligand Cooperation by
Aromatization/Dearomatization. Acc. Chem. Res. 2015, 48, 1979−
1
994.
J
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(
27) (a) Nielsen, M.; Junge, H.; Kammer, A.; Beller, M. Towards a
Green Process for Bulk-Scale Synthesis of Ethyl Acetate: Efficient
Acceptorless Dehydrogenation of Ethanol. Angew. Chem., Int. Ed.
2012, 51, 5711−5713. (b) Ziebart, C.; Jackstell, R.; Beller, M.
Selective Catalytic Hydrogenation of Diethyl Oxalate and Related
Esters. ChemCatChem 2013, 5, 3228−3231. (c) Nielsen, M.; Alberico,
E.; Baumann, W.; Drexler, H.-J.; Junge, H.; Gladiali, S.; Beller, M.
Low-Temperature Aqueous-Phase Methanol Dehydrogenation to
Hydrogen and Carbon Dioxide. Nature 2013, 495, 85. (d) Sponholz,
P.; Mellmann, D.; Cordes, C.; Alsabeh, P. G.; Li, B.; Li, Y.; Nielsen,
M.; Junge, H.; Dixneuf, P.; Beller, M. Efficient and Selective
Hydrogen Generation from Bioethanol using Ruthenium Pincer-
type Complexes. ChemSusChem 2014, 7, 2419−2422. (e) Liu, Q.;
Wu, L.; Gulak, S.; Rockstroh, N.; Jackstell, R.; Beller, M. Towards a
̈
Sustainable Synthesis of Formate Salts: Combined Catalytic Methanol
Dehydrogenation and Bicarbonate Hydrogenation. Angew. Chem., Int.
Ed. 2014, 53, 7085−7088. (f) Zhang, L.; Raffa, G.; Nguyen, D. H.;
Swesi, Y.; Corbel-Demailly, L.; Capet, F.; Trivelli, X.; Desset, S.; Paul,
S.; Paul, J.-F.; Fongarland, P.; Dumeignil, F.; Gauvin, R. M.
Acceptorless Dehydrogenative Coupling of Alcohols Catalysed by
Ruthenium PNP Complexes: Influence of Catalyst Structure and of
Hydrogen Mass Transfer. J. Catal. 2016, 340, 331−343. (g) Zhang,
L.; Nguyen, D. H.; Raffa, G.; Trivelli, X.; Capet, F.; Desset, S.; Paul,
S.; Dumeignil, F.; Gauvin, R. M. Catalytic Conversion of Alcohols
into Carboxylic Acid Salts in Water: Scope, Recycling, and
Mechanistic Insights. ChemSusChem 2016, 9, 1413−1423. (h) Jiao,
H.; Junge, K.; Alberico, E.; Beller, M. A Comparative Computational
Study About the Defined M(II) Pincer Hydrogenation Catalysts (M
=
Fe, Ru, Os). J. Comput. Chem. 2016, 37, 168−176. (i) Glu
̈
er, A.;
Schweizer, J. I.; Karaca, U. S.; Wurtele, C.; Diefenbach, M.;
̈
Holthausen, M. C.; Schneider, S. Hydrosilane Synthesis by Catalytic
Hydrogenolysis of Chlorosilanes and Silyl Triflates. Inorg. Chem.
2
(
018, 57, 13822−13828.
28) Webb, J. R.; Burgess, S. A.; Cundari, T. R.; Gunnoe, T. B.
Activation of Carbon−Hydrogen Bonds and Dihydrogen by 1,2-CH-
Addition Across Metal−Heteroatom Bonds. Dalton Trans 2013, 42,
1
(
6646−16665.
29) (a) Khaskin, E.; Iron, M. A.; Shimon, L. J. W.; Zhang, J.;
Milstein, D. N-H Activation of Amines and Ammonia by Ru via
Metal-Ligand Cooperation. J. Am. Chem. Soc. 2010, 132, 8542−8543.
(
b) Gunanathan, C.; Milstein, D. Metal−Ligand Cooperation by
Aromatization−Dearomatization: A New Paradigm in Bond Activa-
tion and “Green” Catalysis. Acc. Chem. Res. 2011, 44, 588−602.
(
c) Montag, M.; Zhang, J.; Milstein, D. Aldehyde Binding through
Reversible C−C Coupling with the Pincer Ligand upon Alcohol
Dehydrogenation by a PNP−Ruthenium Catalyst. J. Am. Chem. Soc.
2012, 134, 10325−10328.
K
DOI: 10.1021/jacs.9b09062
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX