Simple Ligand Modifications with Pendent OH Groups Dramatically Impact the Activity and Selectivity of Ruthenium Catalysts for Transfer Hydrogenation: The Importance of Alkali Metals

Article abstract of DOI:10.1021/acscatal.6b00229

Remarkable differences in selectivity and activity for ruthenium-catalyzed transfer hydrogenation are described that are imparted by pendent OH groups. Kinetic experiments, as well as the study of control complexes devoid of OH groups, reveal that the pendent OH groups serve to orient the ketone substrate through ion pairing with an alkali metal under basic conditions. The deprotonation of the OH groups was found to modulate the electronics at the metal center, providing a more electron rich ruthenium center. The effects of the ion pairing between alkali metals and the pendent alkoxide groups were highlighted by demonstrating chemoselective transfer hydrogenation of ketones in the presence of olefins. The results illustrate that a simple ligand modification (installation of OH groups) imparts dramatic changes to catalysis. Pendent OH groups turn on catalysis through electronic perturbations at the metal site under basic conditions and can also change the mechanism of catalysis, the latter of which can be used to promote chemoselective reductions.

Full text of DOI:10.1021/acscatal.6b00229

Research Article
pubs.acs.org/acscatalysis
Simple Ligand Modifications with Pendent OH Groups Dramatically
Impact the Activity and Selectivity of Ruthenium Catalysts for
Transfer Hydrogenation: The Importance of Alkali Metals
Cameron M. Moore, Byongjoo Bark, and Nathaniel K. Szymczak*
Department of Chemistry, University of Michigan, 930 North University, Ann Arbor, Michigan 48109, United States
S
* Supporting Information
ABSTRACT: Remarkable differences in selectivity and activity
for ruthenium-catalyzed transfer hydrogenation are described
that are imparted by pendent OH groups. Kinetic experiments,
as well as the study of control complexes devoid of OH groups,
reveal that the pendent OH groups serve to orient the ketone
substrate through ion pairing with an alkali metal under basic
conditions. The deprotonation of the OH groups was found to
modulate the electronics at the metal center, providing a more
electron rich ruthenium center. The effects of the ion pairing
between alkali metals and the pendent alkoxide groups were
highlighted by demonstrating chemoselective transfer hydrogenation of ketones in the presence of olefins. The results illustrate
that a simple ligand modification (installation of OH groups) imparts dramatic changes to catalysis. Pendent OH groups turn on
catalysis through electronic perturbations at the metal site under basic conditions and can also change the mechanism of catalysis,
the latter of which can be used to promote chemoselective reductions.
KEYWORDS: secondary coordination sphere, alkali metals, ion pairing, ruthenium, transfer hydrogenation
that operate through metal/ligand bifunctional pathways have
recently seen a surge in activity.3 However, the successful
incorporation of appended functionality to impart efficient
catalysis has been less straightforward, and multifunctional
complexes that also feature high catalytic activity have been
challenging to predict a priori. In many cases, an active catalyst
undergoes a series of steric and electronic changes imparted by
the reaction media, which in turn afford an active catalyst
structure that may or may not bear a high resemblance to the
starting state.4 Proximity, high directionality, and electronic
effects are all required to work synergistically in order to effect
efficient catalysis.5
Although the benefits of ligand NH2 groups on catalytic
activity were realized in the case of Noyori’s catalyst, the
intimate details of the “NH effect”6 in this complex have not
been well understood until recently. During hydrogenation
catalysis, the deprotonated NH group of the diamine ligand in
Noyori’s catalyst was originally proposed to be directly
responsible for H2 heterolysis: proceeding through a four-
membered transition state in which a hydride (H−) is
transferred to the metal center and H+ is transferred to the
amine ligand (Figure 2A).7
INTRODUCTION
■
Metal−ligand bifunctionality has become a powerful design
concept upon which efficient catalysts used in a wide variety of
chemical transformations have been constructed.1 Key to the
function is the ability of the ligand framework to adopt two (or
more) distinct states which can modulate the electronic
configuration of the metal center, accept/donate protons,
engage in ion pairing, and induce conformational changes to
the primary coordination sphere (Figure 1).1 These concepts
are perhaps best illustrated with Noyori’s hydrogenation
catalysts, which feature ruthenium amine units to facilitate
cooperative H2 activation and transfer.2 Following blueprints
outlined by Noyori and related outer-sphere catalysts, systems
More recently, however, this supposition has been debated in
the literature and contemporary experimental evidence suggests
that a four-membered transition state for H2 heterolysis is
Figure 1. Comparison of metal−ligand bifunctionality with primary
amine (A) and hydroxypyridine ligand motifs (B).
Received: January 23, 2016
© XXXX American Chemical Society
1981
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
providing a precise spatial orientation of the substrate, directed
by ion-pair interactions with alkali metals. Moreover, we
demonstrate that both the presence and location of the pendent
OH groups serve to dictate the activity and selectivity of ketone
reduction and provide experimental evidence for a general
alkoxide effect during catalysis in the presence of alkali metals.
RESULTS AND DISCUSSION
■
Pendent OH Groups Amplify Catalysis. We previously
reported the synthesis and catalytic activity of a ruthenium
complex supported by dhtp and two mutually trans PPh3
ligands (trans-RuCl(dhtp)(PPh3)2PF6, 1).11a This complex
catalyzes transfer hydrogenation of a variety of ketones in the
presence of several functional groups. Moreover, we demon-
strated that the active species during catalysis is homogeneous
and retains both PPh3 ligands.11a In order to interrogate the
precise role of appended OH groups on catalysis, we
investigated the “parent” terpyridine complex (trans-RuCl-
(terpy)(PPh3)2PF6)15 for acetophenone transfer hydrogena-
tion.
Figure 2. Original (A) and revised (B) proposals for H2 cleavage by
Noyori’s catalyst during catalysis.
unlikely to be operative under basic conditions. In fact, later
studies revealed that an alkali-metal “cocatalyst” is required to
achieve high activity for ketone hydrogenation, and the activity
of Noyori’s catalyst is dependent on the identity of alkali metals
added to the reaction mixture.8 Subsequent studies implicated a
mechanism in which the alkali metal is bound to the
deprotonated amine ligand during catalysis, generating an ion
pair (Figure 2B).9 Importantly, recent computational analyses
demonstrated that a four-membered transition state for H2
heterolysis is not the lowest energy pathway under simulated
catalytic conditions.10 Instead, H2 cleavage was proposed to
occur through an ion-pair alkoxide formed after hydride
transfer to the ketone substrate. These studies demonstrate
the complexity associated with predicting an active catalyst
state, even from a deceptively simple ligand arrangement in the
precatalyst.10
Under identical conditions (0.5 mol % of catalyst, 10 mol %
i
of KOtBu, PrOH, 80 °C), the parent terpyridine complex
exhibits diminished catalysis in comparrison to the dhtp
complex 1 (Figure 3A). While complex 1 provided
We, among others, have been working to design metal−
ligand constructs with pendent functionality that can be widely
deployed to facilitate an array of chemical transformations,
including catalysis involving proton transfer.5a,c,11 To clearly
delineate many of the key contributors to a productive
cooperative catalyst, our group has undertaken efforts to
understand how the precise structural, electronic, and
cooperative modes of a metal’s secondary coordination sphere
can be used to regulate reactivity.11o Our approach is to adapt
key structural design criteria from metalloenzymes,12 where the
concept of metal−ligand bifunctionality has long been
established to be vital for function.13 The 2-hydroxypyridine
motif provides an ideal ligand platform for studying metal−
ligand bifunctional reactivity, since the 4-hydroxypyridine
isomer can be used as a control to decouple electronic and
positional effects (Figure 1B). This strategy has previously been
used to understand how pendent OH groups on bipyridine
fragments effect oxidative and reductive catalysis.5a,b,d,11j,k,14 For
catalysis applications, we developed a rigid terpyridine pincer
framework incorporating the 2-hydroxypyridine motif (6,6′-
dihydroxyterpyridine, dhtp),11a which we envisioned could
exploit appended O(H) groups to facilitate cooperative
substrate reduction.
Figure 3. Comparison of dhtp and terpy for transfer hydrogenation
(A), change of ligand classification upon deprotonation of dhtp (B),
and observed changes in carbonyl stretching frequency upon addition
of base to 6 in methanol solvent (C; L is most likely coordinated
CH3OH).
approximately 70% conversion of acetophenone to 1-phenyl-
ethanol (corresponding to 141 turnovers) in 1 h, the parent
terpyridine complex only provided ∼10% conversion (21
turnovers) over the same time period (Figure 3A). This
amplification of catalysis highlights the regulatory role of
pendent OH groups, and we hypothesized that the enhanced
catalysis may be due to electronic differences between the two
ligands under the reaction conditions. In the starting complex,
In this article, we build on our initial report and present
transfer hydrogenation catalysis that is substantially amplified
using well-positioned pendent OH groups on a terpyridine
scaffold. The OH groups serve to turn on catalytic activity and
chemoselective reduction reactions by modifying the electronic
environment of the metal center upon deprotonation and
1982
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
dhtp presents a neutral, L3-type ligand; however, upon
deprotonation and tautomerization, the dhtp ligand presents
the metal center with a dianionic LX2-type donor (Figure 3B).
To evaluate this electronic effect, we examined the IR spectral
changes for a variant of complex 1, RuCl(CO)dhtp(PPh3)PF6
(complex 6, vide infra), in which a PPh3 ligand is substituted for
CO. The νCO band of complex 6 was found to be sensitive to
the presence of added base. For instance, the IR spectrum of 6
reveals a carbonyl stretching frequency of 2046 cm−1. Upon
treatment with 4 equiv of a strong, neutral base (Verkade’s
b a s e ; 2 , 8 , 9 - t r i i s o p r o p y l - 2 , 5 , 8 , 9 - t e t r a a z a - 1 -
phosphabicyclo[3.3.3]undecane) in methanol solution, a shift
of the carbonyl stretching frequency to 1969 cm−1 was
observed (Figure 3C). The bathochromic shift is consistent
with increased electron density at the metal center and greater
donation into the CO π* orbitals upon ligand deprotonation.
The bathochromic shift (Δν = −77 cm−1) is greater in
magnitude than that observed for a copper(I)−carbonyl
complex featuring a single 2-hydroxypyridine group (Δν =
−26 cm−1)16 but similar to that for a deprotonation of two
alcohol ligands on iron (Δν = −69 cm−1)17 and is therefore
consistent with two deprotonation events for the dhtp ligand.
PPh3 Ligand Substitution Modulates Catalyst Stability
and Activity. During the course of our subsequent studies on
ketone transfer hydrogenation catalyzed by 1, it became
apparent that the efficiency of 1 was impeded by catalyst
deactivation. This was evident by the formation of a deeply
colored precipitate emerging as the reaction neared completion.
Fortuitously, we were able to isolate this material and identify
the structure of the deactivation product through solution
(NMR) and solid-state (X-ray) techniques (Figure 4).
since this deactivation product is highly stable and inert to
ligand substitution.18
We hypothesized that catalyst activity could be optimized by
modifying the auxiliary ligands on ruthenium and therefore
prepared a series of complexes with varied donor sets. Two
classes of complexes were synthesized (Scheme 1): one class
Scheme 1. Synthesis of Ruthenium-dhtp Adducts
featuring identical auxiliary ligands (3 and 4) and the other with
dissimilar auxiliary ligands (5 and 6). Complexes 3 and 4 were
prepared by heating benzene solutions of 1 with the
appropriate PR3 ligand in a sealed vessel. Complexes 5 and 6
were prepared by allowing complex 2 (prepared from the direct
reaction of dhtp with RuCl2(PPh3)3 in a nonpolar solvent at
high temperature) to react with TlPF6 and an excess amount of
the desired ligand in dichloromethane solution. Both routes
provided the target complexes in good yield and purity,
allowing for their characterization by NMR (1H and 31P) and
IR spectroscopy.
Additionally, crystals of complexes 4−6 suitable for X-ray
diffraction were obtained and their solid-state structures are
presented in Figure 5, along with the structure of complex 2.19
With a series of complexes featuring varying auxiliary ligands in
hand, we examined the performance of the ruthenium
complexes as catalysts for the transfer hydrogenation of
acetophenone. The conditions previously employed for
complex 1 were used, and catalysis was allowed to proceed
for 2 h. The amount of product formed at 1 and 2 h was
quantified using 1H NMR spectroscopy, and the turnover
numbers (TON) at each time point are presented in Table 1.
The data in Table 1 show that complexes 1, 2, and 5 display the
highest activity for acetophenone transfer hydrogenation under
basic conditions; the activities for complexes 4 and 6 were
marginally lower. Interestingly, complex 3 is almost completely
inactive under these conditions for transfer hydrogenation.
Similar to catalysis promoted by 1, where precipitation of 1-d
emerges in the later stages of the reaction, catalytic experiments
with 2 and 5 became purple and cloudy, consistent with
formation of dimer 1-d. In contrast to the instability noted
above, experiments performed with complexes 4 and 6
maintained homogeneity during catalysis and remained as
bright orange-red solutions throughout the duration of the
Figure 4. Synthesis and solid-state structure of 1-d (30% probability
ellipsoids; PPh3 C atoms, H atoms not involved in hydrogen bonding,
and solvent omitted for clarity).
The deactivation product (1-d) is a ruthenium aquo-bridged
dimer in which the Ru−(OH2)−Ru unit is stabilized by
intramolecular hydrogen-bonding interactions with the depro-
tonated dhtp ligand scaffold. Complex 1-d could alternatively
be prepared by allowing cis-RuCl2(dhtp)PPh3 (2) to react with
NaOH (Figure 4). Although the solvent used during catalysis
(iPrOH) was distilled from CaH2 and dried over activated
molecular sieves prior to use, any adventitious water in solution
presents a significant hurdle for achieving efficient catalysis,
1983
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
Figure 5. Solid-state structures of 2 (A), 3 (B), 5 (C), and 6 (D). (30% probability ellipsoids; PR3 C atoms, H atoms not involved in hydrogen
bonding, solvent, and counteranions omitted for clarity).
catalytically active species which retains the auxiliary ligands
coordinated to ruthenium.
Table 1. Acetophenone Transfer Hydrogenation Catalyzed
by Complexes 1−6
a
Outer-Sphere Hydride Transfer Is the Rate-Limiting
Step during Catalysis. On the basis of prior precedent, we
envisioned two common, limiting mechanistic scenarios for
ketone transfer hydrogenation catalyzed by complex 6: (1) an
inner-sphere hydride insertion pathway to coordinated
substrate and (2) an outer-sphere delivery of hydride without
substrate coordination.21 To distinguish between these two
distinct mechanistic possibilities for catalysis, we performed
TON
catalyst
1 h
2 h
1
2
3
4
5
6
141
108
9
159
124
16
1
kinetic analyses by monitoring reaction progress via H NMR
spectroscopy; either in situ or by collecting aliquots from
reaction solutions at regular time intervals (see the Supporting
Information for more details). Product (either acetone or 1-
phenylethanol) concentrations were determined by integration
using trimethyl(phenyl)silane as an internal standard, and all
experiments were performed in triplicate to establish error.
A linear free energy analysis for transfer hydrogenation was
conducted to first substantiate a rate-determining hydride
transfer step and also to investigate the electronic character of
the transition state during catalysis. A Hammett plot (Figure 6)
was constructed by measuring the initial rates of reduction of
para-substituted acetophenone substrates, which yielded a ρ
value of +1.38(17). The dependence of the rate of ketone
reduction on the σ value implicates a rate-determining hydride
transfer to the ketone substrate. Moreover, the positive ρ value
indicates buildup of negative charge on the ketone substrate
91
106
138
107
120
62
a
Conditions: Ru:KOtBu:ketone = 1:20:200, [iPrOH]0 = 12.8 M,
[acetophenone]0 = 0.1 M, 80 °C.
experiment. Taken together, the data suggest that ruthenium-
dhtp adducts featuring π-acidic auxiliary ligand sets display
higher stability during catalysis. On the other hand, σ-donating
ligand sets offer a catalytic system which has a higher activity for
acetophenone reduction and suggests that an electron-rich
ruthenium hydride species is required to facilitate insertion into
the ketone substrate.
Although the σ-donating frameworks provide faster initial
catalysis, they are plagued by the formation of 1-d, due to the
lability of the supporting ligands on ruthenium under basic
conditions. This is likely a consequence of weak metal−ligand
bonds between the electron-rich metal center (formed upon
deprotonation of the dhtp ligand) and the σ-donating ligands.20
One of the hallmarks of a desirable catalyst is stability, and
therefore we narrowed our attention to complexes featuring π-
acidic supporting ligands, which displayed high stability during
catalysis. From the two complexes which fit this criterion (4
and 6), we chose to pursue mechanistic experiments with
complex 6, since the CO ligand provides a valuable signature
for IR spectroscopy which could be used to gain insight into
changes of electron density at the metal center (vide supra).
To further distinguish the reactivity profiles of complexes 1
and 6, we examined the reduction of benzophenone catalyzed
by these two complexes. We previously reported that complex
1 is unable to reduce this substrate (no conversion was
observed after 12 h using our standard conditions), and we
hypothesized this was due to steric constraints between the
PPh3 ligands and the incoming sterically imposing substrate.11a
Complex 6, on the other hand, is able to reduce benzophenone
to diphenylmethanol, providing 67% NMR yield after 16 h
under identical conditions. These findings are consistent with a
Figure 6. Hammett plot for transfer hydrogenation of substituted
acetophenones catalyzed by 6. Conditions: Ru:KOtBu:ketone =
1:20:200, [Ru] = 0.5 mM, [iPrOH]0 = 12.4 M, [ketone]0 = 0.1 M,
[acetone]0 = 0.5 M, 80 °C.
1984
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
during the rate-determining step. The magnitude of ρ is
consistent with hydride delivery to the ketone substrate in the
transition state. This value is lower than that for the reduction
of substituted acetophenones by NaBH4 at 30 °C (ρ = +2.02,
+3.06),22 reflecting a more anionic nature of the transition state
catalytic rate is achieved.4b,24 To determine whether base
activation via deprotonation of the pendent OH groups
precedes catalysis, we investigated the base dependence of
acetophenone reduction using complex 6. As can be seen in
Figure 7B, saturation behavior is observed and the maximum
rate of acetophenone reduction is achieved at a ruthenium:base
ratio of approximately 1:10.25 Notably, no catalysis was
observed in the absence of base. The saturation behavior
−
for hydride transfer in the case of BH4 reduction.
To further evaluate a rate-limiting hydride transfer step, we
performed kinetic analyses during catalysis. For a rate-
determining hydride insertion pathway into coordinated
substrate, a zero-order dependence of substrate concentration
on the catalytic rate would be expected, since this step would
involve a unimolecular insertion of hydride into substrate.
Alternatively, for an outer-sphere hydride transfer, a first-order
dependence on substrate would be predicted, since the transfer
would involve a bimolecular reaction between the hydride
source (ruthenium complex) and substrate. As such, we
examined the effect of varying substrate concentration on the
initial rate of acetophenone reduction to experimentally
distinguish between these two limiting mechanistic regimes.
As shown in Figure 7A, a first-order dependence on
t
could alternatively be explained by inhibition from BuOH
(generated from KOtBu); however, added tBuOH has no effect
on the reaction profile (Figure S4 in the Supporting
Information). The base dependence is therefore most
consistent with catalyst activation via deprotonation of the
two OH groups on the dhtp scaffold, which turns on catalysis.
In order to analyze the activation parameters for catalysis and
clarify the nature of the rate-determining transition state, an
Eyring analysis was performed. For this study, 4-
(trifluoromethyl)acetophenone was used as the ketone
substrate, since appreciable rates for acetophenone reduction
could not be observed at temperatures below 80 °C using our
standard conditions.26 The reaction rates were measured over a
40 °C window and fit to the Eyring equation (Figure 8) using
Figure 8. Eyring plot for the transfer hydrogenation of 4-
(trifluoromethyl)acetophenone catalyzed by 6. Conditions: Ru:-
KOtBu:4-trifluoromethylacetophenone = 1:20:200, [Ru] = 0.5 mM,
[iPrOH] = 12.8 M, [4-trifluoromethylacetophenone]0 = 0.1 M,
temperatures 50, 60, 70, and 80 °C.
the maximum linear, initial rate. This treatment provided ΔH⧧
= 15(1) kcal/mol and ΔS⧧ = −30(3) eu.27 At 80 °C, these
activation parameters correspond to a free energy of activation
(ΔG⧧) of 25.3(1) kcal/mol. The negative value of the
activation entropy implies that an associative, higher-order
process is operative during the rate-determining step.20 In
addition, the magnitude of the activation entropy is suggestive
of a highly organized transition state involving association of
substrate with the catalytically active species. In combination
with the first-order substrate dependence and the Hammett
studies, the kinetic data are consistent with rate-determining
hydride transfer to the ketone substrate through a transition
state which has a highly organized structure.
Figure 7. Dependence of acetophenone (A) and KOtBu (B)
concentrations on the initial rate of acetophenone transfer hydro-
genation catalyzed by 6. Conditions: [Ru] = 0.5 mM, [acetone]0 = 0.5
M, 80 °C.
acetophenone concentration is observed, which is consistent
with a hydride transfer step from a ruthenium hydride species
to a ketone substrate not located in the primary coordination
sphere of the metal complex.
Base activation is a common step to initiate catalysis for
transfer hydrogenation in order to generate catalytically active
hydride species.23 This commonly involves deprotonation of
acidic groups on the ligand, such as the NH2 groups for
Noyori’s catalyst. Catalytic cycles that are dependent on base
activation typically exhibit saturation behavior; whereupon
reaching a threshold of base concentration, the maximum
Alkali-Metal Cations Dictate the Rate of Acetophe-
none Reduction. The base employed in our studies (KOtBu)
contains an alkali-metal cation, which may or may not
contribute to catalysis. Based on precedent from Noyori’s
system, which indicated a crucial role of alkali metals to achieve
1985
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
efficient catalysis,8 we evaluated the potential role of the alkali-
metal cation with complex 6. In order to decouple the role of
the base from that of the alkali-metal cation, we required a
i
strong, neutral base which would generate PrO− anion in
solution without any added metal cations. These conditions
would allow us to directly probe whether or not alkali metals
influence catalysis. To meet these criteria, we employed
Verkade’s base (2,8,9-triisopropyl-2,5,8,9-tetraaza-1-
i
phosphabicyclo[3.3.3]undecane) to generate PrO− anion in
solution and salts of the noncoordinating anions B(C6F5)4− and
−
B(C8H3F6)4 to introduce alkali-metal cations. Verkade’s base
i
is ideal for these purposes, since dissolution in PrOH
quantitatively generates the corresponding phosphonium salt
(R3PH+, as determined by 31P NMR; Figure S5 in the
i
Supporting Information), and presumably the PrO− counter-
anion. Moreover, the noncoordinating tetraarylborate anions
should also be stable under the reaction conditions during
catalysis.
Prior to performing catalysis, we investigated the reactivity of
complex 6 with Verkade’s base in iPrOH. When a sample of is 6
i
suspended in PrOH (nondeuterated) and treated with excess
(20 equiv) Verkade’s base, the reaction mixture rapidly
1
becomes homogeneous and orange. Analysis by H and 31P
NMR spectroscopy revealed that a new ruthenium hydride
1
complex (6-h) formed (eq 1). The H NMR spectrum of the
Figure 9. Binding of cations to 6-h: hydride region of the 700 MHz 1H
NMR spectra of 6-h in the presence of added cations (A) and initial
reaction profiles for acetophenone transfer hydrogen hydrogenation
catalyzed by 6 in the presence of Verkade’s base and alkali metals (B;
conditions Ru:Verkade’s base:cation:acetophenone = 1:20:20:200,
[Ru] = 0.5 mM, [iPrOH] = 12.8 M, [acetophenone]0 = 0.1 M, 80
°C). Note that we do not imply a specific number of bound cations
and the representation in this figure serves to illustrate how a single
cation may associate with 6-h.
resulting solution shows that the resonances for the OH groups
of the dhtp ligand have disappeared (which should be otherwise
visible in nondeuterated, alcohol solvent in the absence of base)
and a new doublet (JHP = 106 Hz) hydride resonance at −8.91
ppm is present (Figure 9A, bottom spectrum). In addition, the
31P NMR spectrum of this reaction mixture exhibits two
resonances (in addition to the PF6− resonance): a doublet (JPH
= 496 Hz) at −11.73 ppm, corresponding to protonated
Verkade’s base, and a doublet (JPH = 105 Hz) at 33.52 ppm,
corresponding to the new ruthenium hydride species (Figure
S7 in the Supporting Information). Note that free PPh3 was not
observed during the reaction of 6 with Verkade’s base,
consistent with retention of the PPh3 ligand on ruthenium.
On the basis of the magnitude of the coupling constant
between the hydride and PPh3 ligands, and the absence of dhtp
OH resonances, the product is formulated as an anionic
ruthenium hydride complex, in which the hydride ligand is
trans-disposed to the PPh3 ligand.28 The formation of
ruthenium hydride 6-h likely proceeds through a transient
five-coordinate intermediate (generated upon dehydrohaloge-
suggests that the electronic character of the hydride is
perturbed by the presence of alkali-metal cations. This finding
is noteworthy, given the fact that these experiments are
conducted in alcohol media, where alkali-metal cations are
expected to be highly solvated.29
In addition to altering the room-temperature NMR signature
of the ruthenium hydride complex, alkali-metal cations
dramatically impact the rate of acetophenone transfer hydro-
i
genation. When acetophenone is heated to 80 °C in an PrOH
solution containing 0.5 mol % of 6, 10 mol % of Verkade’s base,
and 10 mol % of tetraarylborate salt, the rate of acetophenone
reduction is dependent on the identity of the alkali-metal cation
of the added salt. Salts of Li+, Na+, K+, and Cs+ were analyzed,
and the initial reaction profiles are presented in Figure 9B.
Comparison of the maximum initial rates for acetophenone
reduction using Verkade’s base/K+ cations (ki = [6.3(6)] ×
10−3 M/min) and KOtBu (ki = [7.0(3)] × 10−4 M/min) under
identical conditions reveals that the combination of Verkade’s
base and K+ cations results in a significantly higher maximum
initial rate, which may be due to differences in solution
aggregation of KOtBu or an increase in ionic strength.30 The
results presented in Figure 9B clearly demonstrate that the
i
nation of 6) that readily abstracts hydride from PrO− anion in
solution.
The position of the hydride resonance of 6-h in the 1H NMR
spectrum is significantly affected by added alkali metals. For
instance, addition of 20 equiv of Cs+ to a solution of 6-h causes
a downfield shift of ca. 0.3 ppm, whereas the addition of 20
equiv of Li+ to 6-h induces an upfield shift of ca. 0.2 ppm
(Figure 9A). The distinct shifting of the hydride resonance
observed for each alkali metal provides support for alkali-metal
association with the ruthenium complex in solution and
1986
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
identity of the metal cation dictates the rate of reduction in the
order Li < Na < K < Cs. This result has an important parallel to
Noyori’s catalyst, which also showed attenuated catalysis with
increased cation Lewis acidity. In contrast, while the rate of
catalysis of the Noyori system was greatest with K+
(rationalized by substrate−cation−π interactions),8 this was
not the case for 6, whose rate further increased from K+ to Cs+.
Of additional importance, the maximum rate enhancement for
complex 6 using Cs+ (ca. 15-fold) exceeds that of Noyori’s
catalyst (ca. 4-fold),8 which strongly indicates that alkali-metal
coordination plays a vital role in the present system.31 The rate
enhancement is consistent with a change in electron density at
the metal center upon binding the cation via an inductive effect,
where the least electropositive cation provides the most
electron rich metal center along the series. This notion is
supported by the changes in chemical shift of the hydride
resonance for 6-h as a function of added alkali metal noted
above.
Scheme 2. Synthesis of 4-dhtp and Ruthenium Complexes 8
and 9
Selectivity and Activity of the Catalyst Are Deter-
mined by the Location of the OH Groups. The dramatic
cation dependence on catalytic activity implicates a transition
state for hydride transfer that involves association of catalyst,
cation, and substrate. The cation can bind to either the
substrate or the catalyst, and the enhanced catalysis might be
rationalized by distinct mechanistic scenarios to explain the
experimental results: (1) the deprotonated alkoxide groups
bind cations and organize substrate for hydride delivery within
the secondary coordination sphere of the metal complex or (2)
the excess cations in solution activate the ketone substrate for
hydride attack in solution removed from the metal center. The
former of these two limiting scenarios implies that the
orientation of the OH groups on the ligand scaffold should
affect the efficiency of catalysis, while if the latter were
operative, the substitution pattern of the OH groups on the
ligand scaffold should not significantly affect catalysis. In order
to provide experimental evidence for either of these two
scenarios, and clarify the proximity requirement of the alkoxide
groups for catalysis, complexes analogous to 6 which (a) do not
contain OH groups on the terpyridine ligand or (b) contain
OH groups in a different position on the ligand were prepared.
We examined two complexes identical with 6 which varied only
with respect to the dhtp scaffold: the terpyridine complex
(RuCl(CO)terpy(PPh3)PF6 (7) and a new complex (9)
featuring the 4,4′-dihydroxyterpyridine (4-dhtp) ligand
(Scheme 2).
We employed IR spectroscopy to understand how the
position of the OH groups on the terpyrdine ligand would
affect the electronics at the ruthenium center. Examination of
the carbonyl stretching frequencies in the IR spectra of
complexes 6 (2046 cm−1) and 9 (2020 cm−1) revealed that the
protonated complexes present slightly different ligand fields to
the ruthenium center, with complex 9 exhibiting a somewhat
stronger ligand field. In contrast, the IR spectra in solution
(CH3OH) with excess base (Verkade’s base, four equivalents)
reveal that complexes 6 (νCO 1969 cm−1) and 9 (νCO 1970
cm−1) have nearly identical electronic environments at the
ruthenium center (Figure S8 in the Supporting Information).
These findings suggest that in basic, alcohol media, the ligand
fields presented to the ruthenium center by dhtp and 4-dhtp are
identical and that differences in catalytic activity under similar
conditions are likely not attributable to electronic effects.
Although complexes 6 and 9 display comparable electronic
environments in basic alcohol media, their activities for
acetophenone transfer hydrogenation are dissimilar. When
acetophenone is heated to 80 °C in PrOH in the presence of
0.5 mol % of 6, 10 mol % of Verkade’s base and 10 mol % of
KB(C6F5)4, nearly complete conversion (191 turnovers) to 1-
phenylethanol is achieved in 2 h. Under identical conditions,
complex 9 provides 104 turnovers and complex 7, featuring the
parent terpyridine ligand, yields 34 turnovers (Table 2). These
i
Table 2. Acetophenone Transfer Hydrogenation Catalyzed
a
by Complexes 6, 7, and 9
TON
catalyst
1 h
2 h
6
7
9
163
17
191
34
83
104
a
Conditions: Ru:Verkade’s base:KB(C6F5)4:acetophenone =
1:20:20:200, [iPrOH]0 = 12.8 M, [acetophenone]0 = 0.1 M, 80 °C.
results unambiguously establish that the presence of the OH
groups on the terpyridine ligand increases catalytic activity and,
moreover, that the position of the OH groups on the ligand
dictates the overall efficiency.
Chemoselective reduction of ketone substrates which contain
other sites of unsaturation is a hallmark of outer-sphere
catalysis.21,32 To provide further support for a reduction
mechanism in which alkali-metal binding activates and directs
hydride insertion into the ketone, we examined the chemo-
selectivity of catalytic reductions. Previously, we reported that
complex 1 was an inefficient catalyst for selectively reducing 5-
hexen-2-one to 5-hexen-2-ol, providing only 44% selectivity33
for the desired product. We hypothesized that this substrate
would be ideal for benchmarking complexes 6, 7, and 9, since
product distributions can be used to assess general inner- or
outer-sphere pathways for substrate reduction, where alkene
reduction products likely arise from competitive inner-sphere
pathways (note that we define an outer-sphere pathway as one
that does not require substrate coordination directly to the
1987
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
Figure 10. Chemoselectivity of transfer hydrogenation for complexes 6, 7, and 9 (conditions: Ru:Verkade’s base:KB(C6F5)4:ketone = 1:20:20:200,
[iPrOH]0 = 12.8 M, [ketone]0 = 0.1 M, 80 °C).
Figure 11. Proposed catalytic cycle for the transfer hydrogenation of ketones catalyzed by 6 under basic conditions in the presence of K+ ions.
metal center). When 5-hexen-2-one is subjected to the above
transfer hydrogenation conditions using Verkade’s base and
KB(C6F5)4 over 3 h, complexes 6, 7, and 9 display distinctive
reactivity profiles. In the case of the parent terpyridine complex
7, 13% conversion of the starting ketone was noted and a
mixture of carbonyl and olefin reduction products (A and B,
respectively; Figure 10) was obtained in a 8:5 ratio,
respectively.34 However, in the case of the 4-dhtp complex 9,
only the olefin reduction product and over-reduction product
(C) were obtained in a 3:1 ratio with a total conversion from
starting material of 12%. These findings are in stark contrast to
reduction of this substrate by complex 6, which proceeded to
96% conversion, yielding exclusively the carbonyl reduction
product (A).
transfer to outer-sphere substrate) are operative (Figure 10
inset). In combination with the alkali-metal dependence, the
selectivity of complex 6, on the other hand, suggests that, when
deprotonated, the pendent alkoxide groups are responsible for
orienting substrate for hydride transfer in an outer-sphere
pathway.
Ketone Reduction Involves Preorganization of Sub-
strate through Alkali-Metal Coordination. A unified
mechanism for ketone transfer hydrogenation catalyzed by
complex 6 is proposed in Figure 11. Entrance into the catalytic
cycle begins with the formation of the hydride complex (A), via
a dehydrohalogenation reaction, upon introduction of base and
i
cation (in this case K+) in PrOH solution. This step is
supported by NMR characterization of complex 6-h, as well as
While the position of the OH groups on the ligand scaffold
dictated the efficiency of catalysis in the case of acetophenone,
the OH groups similarly dictate the selectivity of catalysis when a
substrate such as 5-hexen-2-one is used. This result further
exemplifies the utility of the 2-hydroxypyridine motif for
catalysis. The chemoselectivity data presented here are
consistent with two distinct mechanistic regimes operative for
complex 6 and complexes 7 and 9. The ability of complexes 7
and 9 to reduce olefin substrates suggests that hydride insertion
into substrate is nonselective and that two competing
mechanisms (insertion into coordinated substrate and hydride
through the demonstration of alkali-metal association to this
species in PrOH solution. Following hydride formation, attack
i
on the substrate proceeds through an intermediate (B) in which
the carbonyl group is polarized through interaction with K+.
Key data to support this supposition are the observed cation
effects on reduction rate and large, negative entropy of
activation (−30 eu), supporting a highly organized, associative
transition state during hydride delivery.35 Upon hydride attack
on the carbonyl, we postulate that a K+-bound alkoxide (C) is
formed, in analogy to Noyori’s catalyst as described by several
research groups.9,10 Protonolysis of the alkoxide C with PrOH
i
1988
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
(2) (a) Ohkuma, T.; Ooka, H.; Hashiguchi, S.; Ikariya, T.; Noyori, R.
J. Am. Chem. Soc. 1995, 117, 2675−2676. (b) Ohkuma, T.; Ooka, H.;
Ikariya, T.; Noyori, R. J. Am. Chem. Soc. 1995, 117, 10417−10418.
(c) Noyori, R.; Sandoval, C. A.; Muniz, K.; Ohkuma, T. Philos. Trans.
R. Soc., A 2005, 363, 901−912. (d) Noyori, R.; Yamakawa, M.;
Hashiguchi, S. J. Org. Chem. 2001, 66, 7931−7944. (e) Noyori, R.
Angew. Chem., Int. Ed. 2002, 41, 2008−2022.
to release substrate provides D and the catalytic cycle is closed
following hydride abstraction from isopropyl alcohol in
complex D, liberating acetone. In this catalytic cycle, the cation
plays a key role in increasing activity in two respects: (1)
polarizing the carbonyl group of the ketone through
coordination and (2) orienting the ketone substrate for hydride
transfer.
(3) For selected recent examples, see: (a) Schmeier, T. J.;
Dobereiner, G. E.; Crabtree, R. H.; Hazari, N. J. Am. Chem. Soc.
2011, 133, 9274−9277. (b) Bartoszewicz, A.; Marcos, R.; Sahoo, S.;
Inge, A. K.; Zou, X.; Martin-Matute, B. Chem. - Eur. J. 2012, 18,
14510−14519. (c) Kashiwame, Y.; Kuwata, S.; Ikariya, T. Organo-
metallics 2012, 31, 8444−8455. (d) Yoshinari, A.; Tazawa, A.; Kuwata,
S.; Ikariya, T. Chem. - Asian J. 2012, 7, 1417−1425. (e) Bichler, B.;
Holzhacker, C.; Stoeger, B.; Puchberger, M.; Veiros, L. F.; Kirchner, K.
Organometallics 2013, 32, 4114−4121. (f) Hulley, E. B.; Helm, M. L.;
Bullock, R. M. Chem. Sci. 2014, 5, 4729−4741. (g) Kuwata, S.; Ikariya,
T. Chem. Commun. 2014, 50, 14290−14300. (h) Toda, T.; Kuwata, S.;
Ikariya, T. Chem. - Eur. J. 2014, 20, 9539−9542. (i) Marelius, D. C.;
Darrow, E. H.; Moore, C. E.; Golen, J. A.; Rheingold, A. L.; Grotjahn,
D. B. Chem. - Eur. J. 2015, 21, 10988−10992. (j) Grotjahn, D. B.;
Kragulj, E. J.; Zeinalipour-Yazdi, C. D.; Miranda-Soto, V.; Lev, D. A.;
Cooksy, A. L. J. Am. Chem. Soc. 2008, 130, 10860−10861.
(k) Kawahara, R.; Fujita, K.-i.; Yamaguchi, R. Angew. Chem., Int. Ed.
2012, 51, 12790−12794.
(4) (a) Hayes, J. M.; Deydier, E.; Ujaque, G.; Lledos, A.; Malacea-
Kabbara, R.; Manoury, E.; Vincendeau, S.; Poli, R. ACS Catal. 2015, 5,
4368−4376. (b) Mikhailine, A. A.; Maishan, M. I.; Lough, A. J.;
Morris, R. H. J. Am. Chem. Soc. 2012, 134, 12266−12280. (c) Zhang,
G.; Vasudevan, K. V.; Scott, B. L.; Hanson, S. K. J. Am. Chem. Soc.
2013, 135, 8668−8681. (d) Lagaditis, P. O.; Sues, P. E.; Lough, A. J.;
Morris, R. H. Dalton Trans. 2015, 44, 12119−12127. (e) Sonnenberg,
J. F.; Coombs, N.; Dube, P. A.; Morris, R. H. J. Am. Chem. Soc. 2012,
134, 5893−5899.
(5) (a) Suna, Y.; Ertem, M. Z.; Wang, W.-H.; Kambayashi, H.;
Manaka, Y.; Muckerman, J. T.; Fujita, E.; Himeda, Y. Organometallics
2014, 33, 6519−6530. (b) Wang, W.-H.; Hull, J. F.; Muckerman, J. T.;
Fujita, E.; Himeda, Y. Energy Environ. Sci. 2012, 5, 7923−7926.
(c) Zeng, G.; Sakaki, S.; Fujita, K.-i.; Sano, H.; Yamaguchi, R. ACS
Catal. 2014, 4, 1010−1020. (d) Wang, W.-H.; Muckerman, J. T.;
Fujita, E.; Himeda, Y. ACS Catal. 2013, 3, 856−860.
(6) (a) Noyori, R.; Kitamura, M.; Ohkuma, T. Proc. Natl. Acad. Sci. U.
S. A. 2004, 101, 5356−5362. (b) Phillips, S. D.; Fuentes, J. A.; Clarke,
M. L. Chem. - Eur. J. 2010, 16, 8002−8005. (c) Zhao, B.; Han, Z.;
Ding, K. Angew. Chem., Int. Ed. 2013, 52, 4744−4788.
CONCLUSIONS
■
In conclusion, we have demonstrated remarkable differences in
selectivity and activity for transfer hydrogenation that are
imparted by pendent OH groups. Deprotonation of the OH
groups modulates the electronics at the metal center, providing
a more electron rich ruthenium center under basic conditions.
The location of the resulting alkoxide groups serves to orient
the ketone substrate through ion pairing with alkali metals. This
ion pairing facilitates chemoselective transfer hydrogenation of
ketones in the presence of olefins. We note that the presently
observed role of cations may well be operative in other 2-
hydroxypyridine catalysts that function under highly basic
conditions in the presence of high concentrations of alkali
metals.11j,m,36 The results here illustrate that a simple ligand
modification (installation of OH groups in the 2-position of a
pyridine ring) imparts dramatic changes to catalysis. They turn
on catalysis through electronic perturbations at the metal site
and, importantly, can change the mechanism of catalysis
(switching from inner-sphere to outer-sphere pathways). Thus,
we term the modification as a general alkoxide effect and note
that the design principles described herein should be valuable
tools when constructing future ligand ensembles for catalysis.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.6b00229.
■
S
Synthetic and experimental procedures, NMR spectra,
and representative kinetic data (PDF)
Crystallographic data (CIF)
Crystallographic data (CIF)
Crystallographic data (CIF)
Crystallographic data (CIF)
Crystallographic data (CIF)
(7) Sandoval, C. A.; Ohkuma, T.; Muniz, K.; Noyori, R. J. Am. Chem.
Soc. 2003, 125, 13490−13503.
(8) Hartmann, R.; Chen, P. Angew. Chem., Int. Ed. 2001, 40, 3581−
3585.
AUTHOR INFORMATION
Corresponding Author
*E-mail for N.K.S.: nszym@umich.edu.
Notes
■
(9) (a) Hamilton, R. J.; Leong, C. G.; Bigam, G.; Miskolzie, M.;
Bergens, S. H. J. Am. Chem. Soc. 2005, 127, 4152−4153. (b) Hamilton,
R. J.; Bergens, S. H. J. Am. Chem. Soc. 2006, 128, 13700−13701.
(c) Hamilton, R. J.; Bergens, S. H. J. Am. Chem. Soc. 2008, 130,
11979−11987. (d) Takebayashi, S.; Dabral, N.; Miskolzie, M.;
Bergens, S. H. J. Am. Chem. Soc. 2011, 133, 9666−9669. (e) John, J.
M.; Takebayashi, S.; Dabral, N.; Miskolzie, M.; Bergens, S. H. J. Am.
Chem. Soc. 2013, 135, 8578−8584.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the University of Michigan
Department of Chemistry, an NSF-CAREER grant (CHE-
1350877), an NSF-GRFP grant (C.M.M.), the UM Rackham
Graduate School (C.M.M.), and the NSF (CHE-0840456) for
X-ray instrumentation. N.K.S. is a Dow Corning Assistant
Professor and Alfred P. Sloan Research Fellow. We thank Dr
Jeff W. Kampf for X-ray assistance and Boulder Scientific Co.
for a generous donation of LiB(C6F5)4.
(10) Dub, P. A.; Henson, N. J.; Martin, R. L.; Gordon, J. C. J. Am.
Chem. Soc. 2014, 136, 3505−3521.
(11) (a) Moore, C. M.; Szymczak, N. K. Chem. Commun. 2013, 49,
400−402. (b) Tutusaus, O.; Ni, C.; Szymczak, N. K. J. Am. Chem. Soc.
2013, 135, 3403−3406. (c) Moore, C. M.; Szymczak, N. K. Chem.
Commun. 2015, 51, 5490−5492. (d) Wang, W.-H.; Muckerman, J. T.;
Fujita, E.; Himeda, Y. New J. Chem. 2013, 37, 1860−1866. (e) Fujita,
K.-i.; Ito, W.; Yamaguchi, R. ChemCatChem 2014, 6, 109−112.
(f) Fujita, K.-i.; Tanaka, Y.; Kobayashi, M.; Yamaguchi, R. J. Am. Chem.
Soc. 2014, 136, 4829−4832. (g) Wang, W.-H.; Himeda, Y.;
Muckerman, J. T.; Fujita, E. Adv. Inorg. Chem. 2014, 66, 189−222.
(h) Nieto, I.; Livings, M. S.; Sacci, J. B.; Reuther, L. E.; Zeller, M.;
REFERENCES
■
(1) (a) Crabtree, R. H. New J. Chem. 2011, 35, 18−23.
(b) Gunanathan, C.; Milstein, D. Acc. Chem. Res. 2011, 44, 588−602.
1989
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990

ACS Catalysis
Research Article
Papish, E. T. Organometallics 2011, 30, 6339−6342. (i) DePasquale, J.;
Nieto, I.; Reuther, L. E.; Herbst-Gervasoni, C. J.; Paul, J. J.; Mochalin,
V.; Zeller, M.; Thomas, C. M.; Addison, A. W.; Papish, E. T. Inorg.
Chem. 2013, 52, 9175−9183. (j) Gerlach, D. L.; Bhagan, S.; Cruce, A.
A.; Burks, D. B.; Nieto, I.; Truong, H. T.; Kelley, S. P.; Herbst-
Gervasoni, C. J.; Jernigan, K. L.; Bowman, M. K.; Pan, S.; Zeller, M.;
Papish, E. T. Inorg. Chem. 2014, 53, 12689−12698. (k) Marelius, D.
C.; Bhagan, S.; Charboneau, D. J.; Schroeder, K. M.; Kamdar, J. M.;
McGettigan, A. R.; Freeman, B. J.; Moore, C. E.; Rheingold, A. L.;
Cooksy, A. L.; Smith, D. K.; Paul, J. J.; Papish, E. T.; Grotjahn, D. B.
Eur. J. Inorg. Chem. 2014, 2014, 676−689. (l) Fujita, K.-i.; Kawahara,
R.; Aikawa, T.; Yamaguchi, R. Angew. Chem., Int. Ed. 2015, 54, 9057−
9060. (m) Zhang, T.; Wang, C.; Liu, S.; Wang, J.-L.; Lin, W. J. Am.
Chem. Soc. 2014, 136, 273−281. (n) Dimondo, D.; Thibault, M. E.;
Britten, J.; Schlaf, M. Organometallics 2013, 32, 6541. (o) Geri, J. B.;
Szymczak, N. K. J. Am. Chem. Soc. 2015, 137, 12808−12814.
(p) Chakraborty, S.; Piszel, P. E.; Brennessel, W. W.; Jones, W. D.
Organometallics 2015, 34, 5203−5206.
(32) (a) Noyori, R.; Ohkuma, T. Angew. Chem., Int. Ed. 2001, 40,
40−73. (b) Casey, C. P.; Guan, H. J. Am. Chem. Soc. 2007, 129, 5816−
5817. (c) Mikhailine, A.; Lough, A. J.; Morris, R. H. J. Am. Chem. Soc.
2009, 131, 1394−1395.
(33) The selectivity could be increased by adding PPh3 to the
reaction mixture, which likely suppresses a competitive inner-sphere
reduction pathway.11a
(34) Note that CC bond migration from product A may provide
product B. At this time we cannot unequivocally rule out this pathway.
However, such a pathway implies an inner-sphere pathway of hydride
delivery to substrate for complexes 7 and 9.
(35) For a related proposed transition state involving an alkali-metal
cation during transfer hydrogenation, see: Wettergren, J.; Buitrago, E.;
Ryberg, P.; Adolfsson, H. Chem. - Eur. J. 2009, 15, 5709−5718.
(36) Hull, J. F.; Himeda, Y.; Wang, W.-H.; Hashiguchi, B.; Periana,
R.; Szalda, D. J.; Muckerman, J. T.; Fujita, E. Nat. Chem. 2012, 4, 383−
388.
(12) Moore, C. M.; Dahl, E. W.; Szymczak, N. K. Curr. Opin. Chem.
Biol. 2015, 25, 9−17.
(13) Lancaster, K. M. Struct. Bonding (Berlin, Ger.) 2011, 142, 119−
153.
(14) Manbeck, G. F.; Muckerman, J. T.; Szalda, D. J.; Himeda, Y.;
Fujita, E. J. Phys. Chem. B 2015, 119, 7457−7466.
(15) Sullivan, B. P.; Calvert, J. M.; Meyer, T. J. Inorg. Chem. 1980, 19,
1404−1407.
(16) Letko, C. S.; Rauchfuss, T. B.; Zhou, X.; Gray, D. L. Inorg. Chem.
2012, 51, 4511−4520.
(17) Chu, W.-Y.; Zhou, X.; Rauchfuss, T. B. Organometallics 2015,
34, 1619−1626.
(18) Note that heating 1-d with excess PPh3 at 80 °C did not result
in any reaction, highlighting the stability of this complex.
(19) Please see the Supporting Information for more details.
(20) Hartwig, J. F. Organotransition Metal Chemistry; University
Science Books: Mill Valley, CA, 2010.
(21) Clapham, S. E.; Hadzovic, A.; Morris, R. H. Coord. Chem. Rev.
2004, 248, 2201−2237.
(22) (a) Bowden, K.; Hardy, M. Tetrahedron 1966, 22, 1169−1174.
(b) Mullins, R. J.; Vedernikov, A.; Viswanathan, R. J. Chem. Educ.
2004, 81, 1357.
(23) Abdur-Rashid, K.; Clapham, S. E.; Hadzovic, A.; Harvey, J. N.;
Lough, A. J.; Morris, R. H. J. Am. Chem. Soc. 2002, 124, 15104−15118.
(24) Baratta, W.; Siega, K.; Rigo, P. Chem. - Eur. J. 2007, 13, 7479−
7486.
(25) Note that, at high base concentrations, the background KOtBu-
catalyzed Meerwein−Ponndorf−Verley reduction of acetophenone
was subtracted.
(26) Note that a linear Hammett plot indicates that the rate-limiting
step during catalysis is unchanged upon modifying the substrate.
(27) Morse, P. M.; Spencer, M. D.; Wilson, S. R.; Girolami, G. S.
Organometallics 1994, 13, 1646−1655.
(28) Reade, S. P.; Nama, D.; Mahon, M. F.; Pregosin, P. S.;
Whittlesey, M. K. Organometallics 2007, 26, 3484−3491.
(29) Note that differences in coordination environment across the
alkali metals are likely, and the precise aggregation/solvation state may
be complex.
(30) Protonated Verkade’s base (R3PH+) is unlikely to perturb the
reaction profile, given the stability of the P−H bond and the steric
i
constraints imposed by the Pr groups. A search of the Cambridge
Structural Database for similar R3PH+ cations revealed that the P−H
units in these structures do not engage in intramolecular interactions
(i.e. hydrogen bonds) with hydrogen bond acceptors such as Cl− and
phosphonates. For an example, please see: Chintareddy, V. R.; Ellern,
A.; Verkade, J. G. J. Org. Chem. 2010, 75, 7166−7174.
(31) The catalytic profile for Li+ is distinct among the other alkali
metals, similar to the 1H NMR signatures of the hydride noted above,
and this may be a further consequence of differences in solution
aggregation/solvation.
1990
DOI: 10.1021/acscatal.6b00229
ACS Catal. 2016, 6, 1981−1990