Mechanistic studies on the alkylation of amines with alcohols catalyzed by a bifunctional iridium complex

Article abstract of DOI:10.1021/acscatal.5b00645

The mechanism of the N-alkylation of amines with alcohols catalyzed by an iridium complex containing an N-heterocyclic carbene (NHC) ligand with a tethered alcohol/alkoxide functionality was investigated by a combination of experimental and computational methods. The catalyst resting state is an iridium-hydride species containing the amine substrate as a ligand, and decoordination of the amine, followed by coordination of the imine intermediate to the iridium center, constitute the rate-determining step (rds) of the catalytic process. The alcohol/alkoxide that is tethered to the NHC participates in every step of the catalytic cycle by accepting or releasing protons and forming hydrogen bonds with the reacting species. Thus, the iridium complex with the alcohol/alkoxide tethered to the N-heterocyclic carbene ligand acts as a bifunctional catalyst.

Full text of DOI:10.1021/acscatal.5b00645

Research Article
pubs.acs.org/acscatalysis
Mechanistic Studies on the Alkylation of Amines with Alcohols
Catalyzed by a Bifunctional Iridium Complex
Agnieszka Bartoszewicz,† Greco Gonzal
,‡,⊥
ez Miera,
†,‡,⊥
Rocıo
Marcos,
†,‡,⊥
Per-Ola Norrby,§,∥
́
́
,†,‡
́
and Belen Martın-Matute*
́
†
‡
Organic Chemistry Department, Arrhenius Laboratory, and Berzelii Center EXSELENT on Porous Materials, Stockholm
University, SE-106 91 Stockholm, Sweden
§
Chemistry and Molecular Biology Department, Gothenburg University, Kemigar
̊
den 4, SE-412 96 Gothenburg, Sweden
∥
Pharmaceutical Development, AstraZeneca, Pepparedsleden 1, SE-431 83 Mo
̈
lndal, Sweden
*
S Supporting Information
ABSTRACT: The mechanism of the N-alkylation of amines with alcohols catalyzed by an iridium complex containing an N-
heterocyclic carbene (NHC) ligand with a tethered alcohol/alkoxide functionality was investigated by a combination of
experimental and computational methods. The catalyst resting state is an iridium−hydride species containing the amine substrate
as a ligand, and decoordination of the amine, followed by coordination of the imine intermediate to the iridium center, constitute
the rate-determining step (rds) of the catalytic process. The alcohol/alkoxide that is tethered to the NHC participates in every
step of the catalytic cycle by accepting or releasing protons and forming hydrogen bonds with the reacting species. Thus, the
iridium complex with the alcohol/alkoxide tethered to the N-heterocyclic carbene ligand acts as a bifunctional catalyst.
KEYWORDS: iridium, amine alkylation, Hammett plots, kinetics, DFT calculations
INTRODUCTION
alkylation of amines with alcohols under base-free and relatively
mild reaction conditions (Scheme 1). This is the first complex
with an aliphatic alcohol functionalized bidentate ligand that
has been successfully used in hydrogen-transfer reactions; the
■
The condensation of alcohols with amines catalyzed by
transition-metal complexes has, in the past decade, been
1
shown to be a very powerful method for synthesizing N-
2
substituted amines. The transformation is highly atom
majority of complexes used in such reactions have relied on
economical and environmentally friendly, since water is
produced as the sole byproduct. Moreover, alcohols, the
substrates for the reaction, are readily available in great
numbers and variety. Successful examples have been reported
13,14
amine/amido
and phenol/carbonyl moieties as the donor
1
5−17
groups.
Importantly, to the best of our knowledge, the
metal−ligand bifunctionality of such systems containing a
3
−12
18
using complexes of Ir, Ru, Pd, and Fe among others.
tertiary alcohol group has not yet been demonstrated.
We have reported the synthesis of an iridium complex (1)
that features an N-heterocyclic carbene (NHC) ligand
functionalized with an additional coordinating group, a tertiary
Received: March 26, 2015
Revised: May 4, 2015
1
2
alcohol. Complex 1 showed remarkably high activity in the
©
XXXX American Chemical Society
3704
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
+
Scheme 1. Alkylation of Amines with Alcohols Catalyzed by
Complex 1
substrate, [Cp*Ir(NH Ph)X] (3), is the key catalytic species
2
7c
(Scheme 2a). The conclusions of these reports differ in a
number of important aspects, including the identity of the rate-
determining step (rds).
In this paper, we present detailed experimental and
theoretical mechanistic studies on the N-alkylation of amines
with alcohols catalyzed by iridium complex 1 (Scheme 1). We
have investigated the existence of a cooperative interaction of
the metal center and the ligand in this catalytic system and have
identified the rds of the reaction. Various experimental
techniques, such as reaction-order studies, kinetic isotope effect
measurements, and Hammett studies, gave information about
the structure of the active catalyst, as well as the relative rates of
the individual mechanistic steps. Theoretical investigations
allowed the elucidation of the fine details of the catalytic cycle.
A number of interesting features of the mechanism that
distinguish it from those studied before were discovered. In
particular, it was found that the resting state of the catalyst
consists of an iridium−hydride species coordinated to the
amine substrate and also that imine coordination is the rds. The
theoretical studies presented here show that the tertiary
alcohol/alkoxide moiety of the ligand plays an important role
in the catalytic cycle and can be considered a novel type of
functional group in metal−ligand bifunctional systems.
The mechanism of the alkylation of amines with alcohols has
been studied both experimentally and computationally for some
catalytic systems, with the following common general
7
c,19,20
steps:
(i) oxidation (dehydrogenation) of the alcohol
21
substrate to form an aldehyde, (ii) condensation of the
aldehyde with the amine substrate to give an imine
intermediate, and (iii) reduction of the imine. The catalyst
oscillates between two forms: dehydrogenated and hydro-
genated. Metal complexes that are used in well-established
hydrogen-transfer processes (e.g., the reduction of ketones
using isopropyl alcohol as the hydrogen donor) have also
22
shown good activity in the alkylation of amines with alcohols.
This is to be expected, since the key mechanistic steps of the
RESULTS AND DISCUSSION
■
2
3
Experimental Investigations. Reaction Order. We began
our investigations by determining the order of reaction for each
of the reactants (benzyl alcohol 5a and aniline 6a), for the
two reactions are conceptually related. The extensive
mechanistic investigations that have been carried out on
transfer hydrogenation reactions give very relevant insights into
2
4
products (7a and H O), and for the catalyst (1) in the reaction
2
the mechanism of the alkylation of amines with alcohols.
20
shown in Scheme 3, using the method of initial rates.
Two reports by Eisenstein and Crabtree and by Madsen
and Fristrup addressed the details of the catalytic cycle in the
reaction catalyzed by iridium complexes. In both studies, the
7
d
Scheme 3. Model Reaction for Experimental Investigation of
Reaction Orders
(
)
mechanism of the reaction catalyzed by [Cp * IrCl ] was
2
2
investigated using a combination of experimental and
theoretical approaches. The key intermediates proposed in
the two papers are different. Eisenstein, Crabtree, and co-
workers pointed out the important role of the carbonate
additive, which can act as a participating ligand, and proposed
20
[
CpIr(CO )] (2) to be the active catalyst (Scheme 2a). On
3
the other hand, Madsen, Fristrup, and co-workers suggested
that a complex formed upon coordination of the amine
Scheme 2. Presumed Active Species Generated in the N-
Alkylation of Amines with Alcohols: (a) Proposals by
Eisenstein and Crabtree and by Fristrup and Madsen; (b)
This Work
Additionally, the influence of the putative imine intermediate
(
benzylidenaniline 8) on the rate was also investigated. All
reactions were carried out in benzene-d6 at 75 °C and
1
monitored by H NMR spectroscopy.
We found that the initial rate is not influenced by the
concentration of benzyl alcohol (5a; Figure 1), but it is
inversely proportional to the concentration of aniline (6a;
Figure 2). The fact that the rate is independent of the
concentration of 5a suggests that the dehydrogenation step
cannot be the rds of the catalytic cycle. On the other hand, the
inhibition of the reaction by 6a indicates the existence of a
resting state where 6a acts as a ligand. Regarding the products
of the reaction, it was established that the concentration of 7a
does not affect the initial rate (see Figure S4 in the Supporting
Information). As shown in Figure 3, the initial rate is inversely
proportional to the concentration of water. Since it is unlikely
that water is a good ligand for iridium (or is at least a better
ligand than the amines present in the reaction mixture), its
inhibitory effect may be ascribed to a shifting backward of the
3
705
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
Figure 1. Dependence of the initial rate on benzyl alcohol (5a)
concentration.
Figure 4. Dependence of the initial rate on imine (8) concentration.
Figure 2. Dependence of the reciprocal initial rate on aniline (6a)
concentration.
Figure 5. (a) Dependence of the initial rate on the square of catalyst
concentration. (b) Dependence of the initial rate on the square of
catalyst concentration in experiments with 15 mol % of imine 8.
Figure 3. Dependence of the reciprocal initial rate on water
concentration.
which in turn could affect the reaction rate. Alternatively, if we
assume that the alcohol dehydrogenation and the imine
formation are both rapid processes, the reaction rate will be
controlled by the bimolecular reaction between 8 and iridium
hydride species. The concentrations of both of these are
proportional to the initial concentration of the catalyst, which
could explain the observed high reaction order in catalyst. We
therefore investigated the catalyst reaction order in the
presence of 15 mol % of imine 8, which indeed reduced the
apparent reaction order in catalyst significantly (Figure 5b and
Figures S7b−d in the Supporting Information). We cannot rule
out any of these or other similar hypotheses, but we can state
that addition of imine lowers the apparent reaction rate in
catalyst.
equilibrium of imine 8 formation. This in turn implies that the
imine intermediate 8 may be involved in the rds. To verify this,
we evaluated whether the addition of 8 to the reaction mixture
would have an impact on the initial reaction rate. Indeed, it was
found that the reaction is first order in imine 8 (Figure 4).
Finally, the reaction was found to have a higher order in
catalyst than expected: at least 2 (Figure 5a and Figure S7b−d
in the Supporting Information). We cannot fully explain this
but can advance a number of possible explanations for the
discrepancy. Our plots are based on total amount of added
catalyst; the actual concentration of active catalyst could be
lower: for example, due to the presence of some unknown
inhibitor. Alternatively, the catalyst could have secondary
effects, such as changing the pH of the reaction medium,
In summary, the reaction-order studies indicate that the rds
of the overall transformation lies in the last part of the
3
706
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
mechanism (i.e., imine reduction, step iii; vide supra).
Furthermore, a complex formed by coordination of aniline to
an iridium−hydride species seems to be a likely resting state of
Scheme 4. Crossover Experiment
43
the catalyst on the basis of the observed inhibition by aniline.
In Situ Characterization of the Catalyst Resting State. In
an attempt to characterize the postulated resting state, the
reaction between benzyl alcohol (5a) and aniline (6a) catalyzed
1
by iridium (1/AgBF , 2 mol %) was monitored by H NMR
4
spectroscopy (Figure 6 and Figure S8 in the Supporting
reaction was monitored by H and quantitative 13C NMR
spectroscopy (Scheme S1 and Figure S10 in the Supporting
Information) until approximately 80% conversion was reached
1
(
at higher conversions, increased loss of deuterium was
25
observed ). The formation of 7a-d (for example, 13% after
1
1
Figure 6. Selected H NMR chemical shifts (in ppm) of Ir−hydride
resting state A1 and imine 8 observed in the catalytic reactions.
2
h) implies that at least some portion of the intermediate
aldehyde (9a, 9a-d ) produced in the dehydrogenation step
1
leaves the coordination sphere of the iridium hydride/deuteride
intermediate. This is because, after condensation of the
aldehyde with aniline in solution, the resulting imine may be
hydrogenated by an iridium−hydride species containing a
different isotope of hydrogen, thus leading to the mono-
Information). In addition to the signals originating from the
substrates and products, we could observe the imine
intermediate 8. Furthermore, an additional clear signal at high
field (−13.27 ppm) indicated the presence of an iridium−
hydride species. By integration of the resonances, it was
established that the concentration of 8 was comparable to that
of the hydride species and that both corresponded well with the
initial catalyst concentration (i.e., 2 mol %). The results of the
kinetic studies, along with the observation of the Ir−hydride
complex, suggest that the resting state of the catalyst is an
iridium−hydride complex with aniline coordinated (A1) (see
Figure 6 for characteristic shifts).
deuterated product 7-d . The formation of a small amount
1
(
3%) of monodeuterated benzyl alcohol 5a-d1 was also
observed. This shows that the alcohol oxidation is effectively
almost irreversible.
We also observed that 5a was consumed more quickly than
5
a-d under competition conditions. The KIE obtained for the
2
26
alcohol dehydrogenation step was 2.42 ± 0.07 (Figure 7).
Kinetic Isotope Effect (KIE) Investigations. Kinetic experi-
ments with deuterium-labeled benzylic alcohols were carried
out to investigate the relative rates of the alcohol oxidation and
imine reduction steps. The progress of the reactions between
benzyl alcohol (5a) or α-d -benzyl alcohol (5a-d ) and aniline
2
2
(
6a) was followed by NMR spectroscopy. The decay of the
concentration with time was similar for the two substrates (5a
and 5a-d ) (Figure S9 in the Supporting Information), and the
2
25
calculated KIE was 1.24 ± 0.35. This low KIE value suggests
that the alcohol oxidation step is not the rds of the reaction.
26,27
During the reduction of 8 by the iridium−hydride, an Ir−H
bond is broken. Although the KIE resulting from this step
should be of a smaller magnitude than that of a KIE due to the
cleavage of a C−H bond (the Ir−H bond has a lower force
constant), the low KIE determined for the overall reaction is
somewhat surprising. It indicates that the breaking of neither
the C−H bond nor the Ir−H bond is involved in the rds.
From the kinetic studies on the reaction orders (Figures
Figure 7. ln([C5a-d2(0)]/[C5a-d2(t)]) vs ln([C5a(0)]/[C5a(t)]) dependence
0−3 h) in a crossover experiment; [C5a-d2(0)] is the initial
concentration, and [C5a-d2(t)] is the concentration of α-d -benzyl
(
2
alcohol at a given time; [C5a(0)] is the initial concentration, and [C5a(t)
]
is the concentration of benzyl alcohol at a given time.
1−5), we had concluded that imine 8 and iridium hydride A
react in the rds. Together with the small KIE measured, it
seems reasonable to propose that coordination of the imine to
the hydride complex, rather than the subsequent imine
reduction step, constitutes the rds. Hammett investigations
and theoretical studies (vide infra) will reinforce this
conclusion.
To test whether formation of imine 8 takes place within the
coordination sphere of iridium, we carried out a crossover
experiment (Scheme 4). Thus, a 1:1 mixture of 5a and 5a-d2
was subjected to the reaction with 6a. The progress of the
This competition KIE together with the low KIE (1.24 ± 0.35)
measured from initial rates of two parallel reactions is a clear
indication that the alcohol oxidation is not rate limiting and
26
occurs before formation of the resting state.
Reaction of Enantiopure Phenylethanol ((R)-10) with p-
Chloroaniline (6b). When (R)-1-phenylethanol ((R)-10, >99%
ee) was used (Scheme 5 and Figure 8), it underwent only slow
racemization under the reaction conditions (77% ee at 85%
conversion). This shows that ketones as well as aldehydes
3
707
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
Scheme 5. Reaction of an Enantiopure Alcohol
Scheme 6. Reaction Conditions Used in the One-Pot
Hammett Competition Study of Para-Substituted Anilines
Figure 8. Slow racemization of (R)-10 (black ◆) and lack of
stereospecificity in the formation of 11 (red ●).
prefer the forward reaction to imine and further reduction over
back-reaction between the carbonyl and the iridium−hydride.
Interestingly, the product of this reaction (11) was formed as a
racemate, indicating that the iridium hydride intermediate,
which is formally chiral at iridium, racemizes efficiently. An
alternative rationalization would be that the chirality transfer
from alcohol to iridium and further on to imine is so inefficient
that no enantiomeric excess can be detected in the final
product.
Hammett Studies. Hammett studies were carried out to
28
obtain additional insights into the mechanism of the reaction.
The electronic influence of the para substituents on the alcohol
and aniline substrates was investigated using “one-pot”
Figure 9. (a) Hammett plot constructed from the relative rates in
competition experiments with para-substituted anilines 6a−e and
29,30
competition experiments
and noncompetition experiments.
2
benzyl alcohol 5a (y = −1.69x (±0.13) − 0.01 (±0.03); R = 0.98).
The ρ values obtained in the Hammett competition
experiments correspond to the magnitude of charge (positive
or negative) that is built up in the transition state in the first
irreversible step of the mechanism that involves the substrate
under investigation. Thus, these Hammett studies conducted
using competition experiments give information about the step
in which selectivity between the competing substrates is
decided and not about the overall rds of the mechanism. On
the other hand, in the separate experiments ρ is related to the
change of charge on the transition states involving the studied
substrates during the rds.
(b) Hammett plot constructed from the relative rates obtained from
noncompetition experiments with para-substituted anilines 6a−e and
2
benzyl alcohol 5a (y = 2.29x (±0.13) − 0.03 (±0.03), R = 0.99).
consistent with the result of KIE studies described in the
previous section.
Interestingly, the relative reactivity of the anilines in separate
experiments were opposite (Figure 9b and Figure S20 in the
Supporting Information) to those observed in the competition
experiment. In the former, electron-poor anilines reacted more
quickly than electron-rich anilines. These experiments both
measure the same step of the mechanism, i.e. coordination of
imine 8 to the iridium−hydride intermediate (Figure 6), which
consists of dissociation of aniline from A1 (L = aniline) and
subsequent reaction of the resulting complex with 8. This step
is the rds of the overall reaction, and as such, it determines the
rates of the separate experiments with different anilines (Figure
9b). On the other hand, it is also the selectivity-determining
Hammett Studies with Anilines. First, we performed
Hammett studies under competition conditions with a series
of para-substituted anilines (Scheme 6). Assuming that
equilibration of the different imines is much faster than the
3
1
subsequent hydrogenation, i.e., Curtin−Hammett conditions,
the experiment should reveal the selectivity-determining step
during the imine hydrogenation. A linear plot of [log(k /k )]
X
H
vs neutral ρ value was obtained with a good correlation (Figure
a and Figures S11 and S13 in the Supporting Information).
26
9
step for the Hammett competition experiment with anilines.
The negative reaction constant (ρ) indicates that positive
charge is built up in the imine moiety in the selectivity-
determining transition state. This agrees with imine coordina-
tion to the iridium−hydride being an irreversible step and is
This discrepancy in the relative reactivities of different anilines
under competition and noncompetition conditions can be
justified. In the noncompetition experiments, both the ground
states (Ir−H complexes with the corresponding aniline
3
708
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
substrate coordinated) and the transition states (imine
coordination) vary and, apparently, the different energies in
the ground states have a larger influence on the reaction rate
negative ρ values (Figure 10 and Figures S32 and S34 in the
Supporting Information) were obtained, indicating that a
positive charge is built up at the benzylic carbon in the
selectivity-determining transition state. This is consistent with
the alcohol oxidation being the first effectively irreversible step,
which is in agreement with the KIE investigations (vide supra).
No significant differences among the slopes of the three
independent Hammett plots were observed, indicating that the
alcohol oxidation is likely to be the dominant selectivity-
determining step. The ρ value obtained in the experiment with
(
in other words, the inhibitory power of electron rich anilines
outcompetes their high inherent reactivity; Scheme S4 in the
Supporting Information). Thus, the reactions involving better-
coordinating, electron-rich anilines are slower under non-
competition conditions. However, in the competition experi-
ment, there is a common ground state for all of the reactions,
the iridium−hydride complex with the most electron rich
aniline coordinated, and thus only the transition-state energies
for the coordination of the various imines differ. Hence, in this
case, the imines derived from the electron-rich anilines react
more quickly.
Hammett Studies with Benzyl Alcohols. In the second
Hammett competition study, the relative reactivities of several
para-substituted benzyl alcohols 5a−d,f were compared. Since
in this experiment the para-substituted aryl moieties originally
in the alcohol substrates are present throughout the
mechanistic cycle, this study will give information about both
alcohol oxidation and imine coordination/reduction. Both of
these steps could contribute to the selectivity of the reaction.
To determine which of the steps has a dominant influence on
the selectivity, three competition experiments were carried out.
In each experiment, the same set of alcohols 5a−d,f was reacted
with a different aniline (6a,c,f; Scheme 7). Hammett plots with
p-CF -aniline (6f) is, however, slightly more negative than
3
those obtained for p-OMe-aniline (6c) and for aniline (6a).
This suggests that in those cases in which electron-poor
imines are formed as intermediates, and thus the imine
reduction is slower, the alcohol oxidation becomes slightly
more reversible and the imine reduction step also contributes
to the selectivity of the reaction. Thus, the product-determining
step will be highly dependent on the electronic properties of
the alcohol and amine substrates and may differ for different
reactants.
The trend in the relative reactivity of electronically different
benzyl alcohols determined in noncompetition experiments
(
Figure S21 in the Supporting Information) was found to be
the same as that observed in the competition experiments
Figure 10): the reactions proceeded more quickly for electron-
(
rich alcohols than for electron-poor ones. This result is
consistent with the imine coordination step being the slowest
step in the reaction.
Scheme 7. Reaction Conditions Used in Three One-Pot
Hammett Competition Studies of Para-Substituted Benzyl
Alcohols
Summary of the Experimental Mechanistic Studies. On
the basis of the results of the experiments described above, we
propose a plausible reaction mechanism (Scheme 8). In the first
step (step i) of the catalytic cycle, benzyl alcohol (5a) is
oxidized to give benzaldehyde (9a) and an iridium−hydride
complex (A). The alcohol oxidation was found to be practically
irreversible. After dissociation of aldehyde 9a, aniline (6a)
coordinates to the iridium−hydride complex A to give hydridic
complex A1 (step iii), the resting state of the catalyst. In an
Scheme 8. Summary of Experimental Mechanistic
Investigations
Figure 10. Plots of the Hammett competition studies of the reactions
between a set of para-substituted benzyl alcohols (5a−5d,f) with three
different anilines: PhNH (6a, Hammett plot a; y = −1.50x (±0.06) −
2
2
0
−
.03 (±0.02), R = 0.99), p-MeOC H NH (6c, Hammett plot b; y =
6
4
2
2
1.38x (±0.11) − 0.08 (±0.03); R = 0.98), and p-CF C H NH (6f,
3
6
4
2
2
Hammett plot c; y = −1.81x (±0.08) − 0.04 (±0.08), R = 0.99).
3
709
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
independent reaction (step ii), 9a reacts with aniline (6a) to
give imine 8 and water. The inhibiting effect of water on the
reaction rate suggests that the imine formation is a reversible
process under the reaction conditions. The next step consists of
two substeps: namely dissociation of aniline 6a from the
iridium−hydride complex A1 and subsequent coordination of
imine 8 (step iv) to give imine−iridium complex B. KIE studies,
as well as Hammett investigations, provide evidence that
coordination of imine is the rds. After effectively irreversible
formation of complex B, imine reduction takes place. The sec-
amine 7 produced as the product of this process does not
inhibit the reaction: i.e., it does not form strong complexes with
iridium at any stage of the catalytic cycle. The overall proposed
catalytic cycle yields a rate law that is consistent with the results
of all the experimental mechanistic studies described in the
previous sections (see the Supporting Information for details).
Theoretical Investigations. Density functional theory
forming INT-2. At this point, the outer- and inner-sphere
pathways diverge but merge again later at the stage of the
hydrogenated catalyst (INT-9) (Figure 12). From INT-2, the
outer-sphere route involves a single step (TS ), in which the
oxidation of 5a occurs by concerted synchronous transfer of the
α-hydrogen and the proton from the alcohol to the metal
center and the tethered alkoxide, respectively, in a cyclic six-
membered transition state. The overall barrier for the outer-
sphere dehydrogenation is 69 kJ/mol relative to INT-1A.
The inner-sphere mechanism consists of several steps. The
first step is the coordination of benzyl alcohol 5a to iridium
2
‑9
(
INT-3), followed by a proton transfer from 5a to the tethered
alkoxide (TS ). Upon protonation, the tethered alcohol
3
‑4
dissociates from iridium (TS ) to give INT-5. This complex,
4
‑5
like INT-1, contains only three ligands, and the alkoxide shows
strong π donation to iridium (Ir−O distance 1.97 Å). The
calculations revealed that, for the subsequent β-hydride
elimination to occur, INT-5 must first undergo a rearrangement
into intermediate INT-6, in which there is an agostic
(
DFT) calculations were used to study the fine details of the
catalytic cycle and to investigate whether the alkoxide/alcohol
functional group in the carbene ligand plays an active role in
the mechanism: i.e., whether complex 1 works as a metal−
ligand bifunctional catalyst. Additionally, we aimed to
investigate whether the mechanism follows an inner- or an
outer-sphere pathway.
interaction between the metal center and one of the benzylic
20,33
C−H bonds.
Formation of the agostic intermediate is
facilitated by a hydrogen bond between the substrate alkoxide
ligand and the hydroxyl group on the carbene ligand. This
diminishes the π donation from the alkoxide to the iridium
center, which results in a longer Ir−O distance (2.09 Å) for this
intermediate. From INT-6, the inner-sphere β-hydride
elimination (TS ) takes place with a negligible energy barrier
Alcohol Oxidation. On the basis of our earlier experimental
1
2
results, the cationic iridium complex INT-1 (Figure 11),
6
‑7
of ∼1 kJ/mol. Hence, formation of INT-6 represents the major
contribution to the overall energy barrier in the alcohol
oxidation via the inner-sphere mechanism (69 kJ/mol from
INT-1A to TS ). INT-7, the product of the β-hydride
6
‑7
elimination, contains a molecule of benzaldehyde π bonded to
iridium. After a change in the coordination mode from π to σ
bonding (TS ), the benzaldehyde is replaced by the tethered
7
‑8
alcohol (TS ).
8
‑9
Overall, since the energy barriers for the inner- and outer-
sphere mechanisms of the alcohol oxidation are calculated to be
practically the same (69 kJ/mol relative to INT-1A), it can be
concluded that both pathways might operate concurrently or
that there is a small difference that cannot be ascertained by the
Figure 11. Optimized structures of an iridium−alkoxide active species
(
INT-1) and of a related complex formed upon coordination of aniline
INT-1A). Distances are given in Å.
(
27,34
current level of modeling.
For the inner-sphere mechanism,
the highest-energy TS is seen for the formation of the agostic
containing a Cp* ligand and a bidentate carbene ligand
functionalized with an alkoxide moiety, was chosen to be the
starting point of the catalytic cycle. For the computational
studies, the structure of the carbene ligand was simplified by
replacing the butyl group with an ethyl substituent. No further
simplifications of the catalyst structure were introduced. The
free energy profiles were calculated for benzyl alcohol and
aniline as model substrates.
bond (TS ), with only a marginal elongation of the C−H
5
‑6
bond whereas, in the outer-sphere mechanism, both the C−H
and O−H bonds are elongated in the TS (TS , Figure 13).
2
‑9
Therefore, we would expect a strong KIE from the outer-sphere
mechanism but not from the inner-sphere mechanism. The
calculated KIEs (at 80 °C) for those steps were 3.3 and 1.7 for
the outer- and inner-sphere mechanisms, respectively (see the
Supporting Information for the details). The competition
experiment (vide supra) gave a KIE of 2.42 ± 0.07, which, in
combination with the computational results, suggests that both
calculated mechanisms can contribute in the alcohol dehydro-
genation step.
The Ir−O distance in INT-1 is calculated to be only 1.95 Å,
which suggests that there is a significant π donation from the
32
alkoxide group to the electron-poor metal center (see Figures
1, 13, 14, and 16 for optimized structures of this and other key
1
intermediates and transition states; the calculated structures of
the other stationary points can be found in the Supporting
Information). Coordination of aniline (6a) to INT-1 results in
the formation of INT-1A, which is more stable than INT-1 by
Formation of the Imine and the Iridium−Hydride−Aniline
Complex. Aldehyde 9 dissociates from INT-9, giving iridium
hydride INT-10. This is an endergonic process by 8 kJ/mol
(Figure 12). The results of the experimental studies indicated
that the aldehyde does dissociate. INT-10 may coordinate a
molecule of aniline (6a) to give INT-10A (Figure 14), whose
calculated energy is lower than that of INT-9. Formation of an
iridium−hydride complex with aniline coordinated was
1
9 kJ/mol (Figure 11). INT-1A is the lowest energy non-
hydridic complex, but it is not the resting state of the catalyst
(
vide infra), in agreement with the experimental results.
The catalytic cycle begins with formation of a hydrogen bond
between benzyl alcohol 5a and the alkoxide ligand of INT-1,
3
710
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
Figure 12. Calculated free-energy profile for the dehydrogenation of benzyl alcohol with the bifunctional iridium catalyst. All energies are given
relative to INT-1 + 5a + 6a in kJ/mol.
Figure 13. Selected transition states for the alcohol oxidation via (a) outer-sphere and (b, c) inner-sphere mechanisms. Distances are given in Å.
imine formation, which can be mediated by either an iridium
7c
species or an anilinium salt (formed during the catalyst
activation) acting as a Lewis or a Brønsted acid catalyst. These
processes are expected to be rapid and reversible, and the
mechanism, which includes acid−base equilibria and potentially
zwitterionic intermediates, is not expected to be well modeled
34
by standard DFT methods as employed here and therefore
were not studied explicitly beyond calculation of the
thermodynamics of the overall imine formation. Assuming a
rapid endergonic equilibrium (Curtin−Hammett conditions),
the imine concentration (and therefore also the observed rate)
will depend on the concentrations of all species in the
equilibrium, in nice agreement with the observed inhibitory
effect of water.
Imine Hydrogenation. For the hydrogenation of imine 8 to
take place, INT-10A must first release the aniline ligand and
subsequently bind imine 8 by hydrogen bonding to the
tethered alcohol (Figure 15). From the resulting intermediate
Figure 14. Optimized structures of iridium hydride INT-10 and
resting state INT-10A. Distances are given in Å.
1
observed by H NMR spectroscopy (vide supra), and it was
44
suggested that this species is the resting state of the catalyst.
The condensation between benzaldehyde (9a) and aniline
6a) was calculated to be endergonic by +9 kJ/mol (Figure
(
4
5
15). There are a number of mechanistic possibilities for the
3
711
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
Figure 15. Calculated free-energy profile for hydrogenation of the imine with the bifunctional iridium catalyst. All energies are given relative to INT-
+ 5a + 6a. Observe that the energy of INT-10 (−11 kJ/mol) here includes the imine and therefore is 9 kJ/mol higher than that given in Figure 12.
1
Figure 16. Transition states for imine hydrogenation via the (a) outer-sphere and (b, c) inner-sphere mechanisms. Distances are given in Å.
INT-11, the outer- and inner-sphere mechanisms diverge.
Similarly to the alcohol oxidation, the former alternative
involves only a single step leading directly to INT-17, which
contains the final product (i.e., sec-amine 7) hydrogen-bonded
to the tethered alkoxide. The corresponding concerted
insertion) occurs via TS12−13 (Figure 16). The addition is
facilitated by an increased electrophilicity of the imine as a
result of its coordination to iridium, as well as hydrogen
bonding to the alcohol ligand.
TS12‑13 leads to transient intermediate INT-13, which has an
agostic interaction between the Ir center and an α-C−H bond
of the amine. The transformation of INT-13 into INT-14, in
which the final reaction product is coordinated to iridium, is
strongly exergonic. Similarly to INT-5, the electron-deficient
metal center in INT-14 is stabilized by π donation from the N
transition state (TS11‑17, Figure 16) is analogous to TS , but
2‑9
in this case, proton and hydride are transferred asynchro-
3
5
nously. The overall barrier for this step is 91 kJ/mol (from
INT-10A to TS11‑17).
The inner-sphere route for the imine hydrogenation consists
of steps similar to the reverse of the alcohol oxidation.
Specifically, first the tethered alcohol decoordinates from
iridium to allow complexation of the imine to the metal center
36
of the amido ligand. INT-14 undergoes sequential coordina-
tion of the tethered alcohol to iridium (TS14‑15), followed by
proton transfer (TS15‑16). Dissociation of the amine product
from the iridium center forms INT-17, in which the amine
product remains hydrogen bonded.
46
through a π-bond (TS11−12, Figure 16). From the resulting
intermediate INT-12, the hydride addition (migratory
3
712
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
Scheme 9. Mechanism for the Alkylation of Aniline with Benzyl Alcohol Catalyzed by Bifunctional Complex 1
In contrast to the alcohol oxidation, in the last part of the
mechanism (i.e., imine hydrogenation), the inner-sphere
mechanism is clearly favored over the outer-sphere mechanism
by 10 kJ/mol (TS11‑12 or TS12‑13 vs TS11‑17) for the catalyst
studied. This agrees well with our previous investigations
showing that a related catalyst bearing an ether substituent on
the carbene ligand was also able to afford good conversion;
however the latter occurred under harsher reaction conditions
Summary of the Theoretical Mechanistic Studies. It
can be concluded that the alcohol oxidation step can occur by
either an inner-sphere or an outer-sphere mechanism, which
suggests that the electronic and steric properties of the
particular substrate will determine which mechanism operates
in each case. For the imine hydrogenation step the inner-sphere
mechanism is favored, with a 10 kJ/mol difference in energy
between the highest transition states for the outer-sphere
(TS11‑17) and inner-sphere (TS11‑12 or TS12‑13) pathways.
The highest energy span for the imine hydrogenation part of
the mechanism (via the inner-sphere pathway) is located
between INT-10A and TS11‑12 (81 kJ/mol). It is higher than
the overall barrier for the alcohol oxidation (69 kJ/mol), which
is consistent with the experimental results, and indicates that
the rds is located in the second part of the mechanism (imine
reduction) and not in the alcohol oxidation part. INT-10A is
thus identified as the resting state of the catalyst, and its
structure correlates with that of aniline coordinated hydride A1
12
with a higher catalyst loading.
The calculated KIEs for the imine coordination step
(TS11‑12) and inner sphere hydride transfer from iridium to
imine (TS12‑13) are 1.0 and 1.6, respectively. The first value is
closer to the experimental KIE of 1.24 ± 0.35, which further
supports that imine coordination is the rds (see the Supporting
Information for more details). The catalytic cycle is closed by
the release of amine 7a and the regeneration of INT-1. The
overall reaction is exergonic by 21 kJ/mol.
3
713
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
proposed in the experimental investigations. The rds involves
studies. These conclusions were well supported computation-
ally. In addition, the theoretical investigations also showed that
the alcohol oxidation step follows either an outer- or an inner-
sphere pathway, while the inner-sphere mechanism seems
dominant for the reduction of the imine intermediate by the
iridium hydride. We have also investigated the role of the
alcohol-functionalized carbene ligand in complex 1 and showed
that it participates in almost every step of the catalytic cycle by
donating or accepting protons to or from the substrates, as well
as by hydrogen bonding to substrates and intermediates,
facilitating the reaction. Thus, the efficient catalytic behavior of
complex 1 (i.e., additional base is not needed, an excess of any
of the substrates is not required, and the reaction can proceed
the dissociation of aniline from iridium−hydride complex INT-
10A followed by coordination of imine 8. This agrees well with
the experimentally observed inhibiting effects of aniline (6a)
and water and with the accelerating effect of added imine 8.
Transition state TS11‑12 does not involve a hydrogen transfer,
which is consistent with the experimentally observed low KIE
in the reaction. Hydride transfer transition state TS12‑13 is
calculated to be isoergic with TS11‑12, but in this case, the
results of the Hammett studies show that the rate-limiting step
is indeed coordination, TS11‑12. The real energy difference
between these two steps is simply too small to be captured
34
accurately by the DFT methods used here. The highest
calculated energy for the alcohol dehydrogenation (TS2‑9 and
TS5‑6 at 50 kJ/mol) is slightly lower than the highest calculated
energy in the inner-sphere imine hydrogenation sequence
12
at mild temperatures, 50 °C) can be ascribed to the presence
of a bidentate and bifunctional nature of this unique iridium
carbene complex 1.
(
TS11‑12 and TS12‑13 at 52 kJ/mol. This should indicate that the
former step is reversible. This contrasts with the experimental
results obtained in the competition crossover experiment
EXPERIMENTAL SECTION
■
General Considerations. All reactions were carried out
under an argon atmosphere in oven-dried glassware. The
sealable tubes used were Biotage microwave vials. Reagents
were of analytical grade and were obtained from commercial
suppliers and used as purchased. CH Cl was dried over CaH
(
5
Scheme 5), in which a very small amount of monodeuterated
-d was produced, and also with the observed stereochemical
1
integrity of chiral alcohol (R)-10 (Scheme 6). However, it must
be noted that all calculations deliver results at the standard
state, 1 M solution of all participating species and that the
reaction is run at infinite dilution in the solvent of choice. Imine
formation, disfavored under standard conditions according to
the calculated energies, is actually favored under real
experimental reaction conditions, as was observed during the
kinetic experiments. The concentration of aniline is higher than
that of water, especially in the initial phase of reaction, shifting
the equilibrium toward imine formation. Hence, on the basis of
experimental studies, we can presume that the imine
coordination step (TS11‑12) is slightly lower in energy than
2
2
2
and distilled under nitrogen. Anhydrous toluene was obtained
1
13
using a VAC solvent purifier system. H and C NMR spectra
were recorded at 500 and 125 MHz, respectively, using a
Bruker Avance spectrometer. Chemical shifts (δ) are reported
in ppm, using the residual solvent peak in CDCl (δH 7.26
3
ppm) as a reference. Enantiomeric excess was determined by
HPLC analysis (Waters 2695).
Computational Details. Calculations were carried out
37
using the Jaguar package. Geometries were optimized using
38
39
the B3LYP functional with the LACVP* basis set. These
were then characterized with frequency calculations to confirm
their character as minima (no imaginary frequencies) or
transition states (a single imaginary frequency). The con-
nections between transition states and their corresponding
reactants and products were verified by fully optimizing
structures derived from the transition states with a small
the alcohol oxidation step (TS5‑6 or TS ) in the initial phase of
2‑9
the reaction but becomes competitive eventually, when
concentrations have shifted sufficiently to allow back-reaction
to alcohol.
In addition, the theoretical investigations support the
importance of the bifunctional character of the hydroxyl-
functionalized carbene ligand and its participation throughout
the catalytic cycle. This is in agreement with the greatly reduced
activity observed when a related complex, in which the hydroxyl
functionality had been replaced by a methyl ether group, was
displacement following the transition vector in both directions
40
(
QRC). The final Gibbs free energies were obtained from
41
single-point calculations using the M06 functional with the
larger basis set LACV3P** , corrected for zero-point and
thermal effects at 353.0 K and 325.0 atm from the frequency
+
1
2
used.
By combining the results from the experimental inves-
tigations with those obtained computationally, we propose the
catalytic cycle depicted in Scheme 9.
33b
calculations, and solvation free energies. The effect of the
solvent was calculated using a self-consistent reaction field
(
SCRF) polarized continuum model (PBF) with parameters for
42
benzene, implemented in Jaguar.
CONCLUSIONS
The rate of the alkylation of aniline (6a) with benzyl alcohol
5a) catalyzed by iridium(III) complex 1 was established
■
(
ASSOCIATED CONTENT
■
*
S
Supporting Information
experimentally to be zero order with respect to benzyl alcohol
5a), second order in catalyst 1, and negative first order in
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.5b00645.
(
water and aniline (6a). The rate is also zero order with respect
to the product of the reaction (benzylaniline 7a). Furthermore,
as supported by Hammett studies and kinetic isotope
investigations, the coordination of the imine intermediate (8)
to an iridium hydride species constitutes the rate-determining
step of the reaction. The resting state is in the form of an
iridium hydride with the aniline substrate coordinated, as
observed by NMR spectroscopy. Further evidence suggesting a
resting state formed upon coordination of the aniline substrate
to iridium was obtained by one-pot Hammett competition
Procedures, full kinetic analysis, and further experimental
details (PDF)
Cartesian coordinates and energies of the calculated
geometries (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*belen.martin.matute@su.se.
3
714
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
Author Contributions
A.B., G.G.M., and R.M. contributed equally.
Tetrahedron 2011, 67, 3140−3149. (d) Martın
Ramon, D. J. Synthesis 2011, 22, 3730−3740.
9) (a) Prades, A.; Corberan, R.; Poyatos, M.; Peris, E. Chem. - Eur. J.
008, 14, 11474−11479. (b) Segarra, C.; Mas-Marza, E.; Mata, J. A.;
Peris, E. Adv. Synth. Catal. 2011, 353, 2078−2084.
10) For recent examples, see: (a) Li, J.-Q.; Andersson, P. G. Chem.
́
ez-Asencio, A.; Yus, M.;
⊥
́
(
2
́
Notes
́
The authors declare no competing financial interest.
(
ACKNOWLEDGMENTS
Commun. 2013, 49, 6131−6133. (b) Wang, D.; Zhao, K.; Xu, C.;
Miao, H.; Ding, Y. ACS Catal. 2014, 4, 3910−3918. (c) Enyong, A. B.;
Moasser, B. J. Org. Chem. 2014, 79, 7553−7563. (d) Zhang, Y.; Lim,
C.-S.; Sim, D. S. B.; Pan, H.-J.; Zhao, Y. Angew. Chem. 2014, 126,
■
Financial support was obtained from the Swedish Research
Council (VR), the Swedish Governmental Agency for
Innovation Systems (VINNOVA) through the Berzelii Center
EXSELENT, the Knut and Alice Wallenberg Foundation, and
the Wenner-Gren Foundation. B.M.-M. also thanks VINNOVA
for a VINNMER grant.
1
423−1427; Angew. Chem., Int. Ed. 2014, 53, 1399−1403. (e) Old-
enhuis, N. J.; Dong, V. M.; Guan, Z. J. Am. Chem. Soc. 2014, 136,
2548−12551. (f) Yan, T.; Feringa, B. L.; Barta, K. Nat. Commun.
2014, 5, 5602. (g) Rawlings, A. J.; Diorazio, L. J.; Wills, M. Org. Lett.
015, 17, 1086−1089.
(11) (a) Cumpstey, I.; Agrawal, S.; Borbas, E. K.; Martın
Chem. Commun. 2011, 47, 7827−7829. (b) Agrawal, S.; Lenormand,
M.; Martın-Matute, B. Org. Lett. 2012, 14, 1456−1459.
12) Bartoszewicz, A.; Marcos, R.; Sahoo, S.; Inge, A. K.; Zou, X.;
Martın-Matute, B. Chem. - Eur. J. 2012, 18, 14510−14519.
13) Selected reviews: (a) Muniz, K. Angew. Chem., Int. Ed. 2005, 44,
1
2
REFERENCES
́
-Matute, B.
■
(
1) (a) Hamid, M. H. S. A.; Slatford, P. A.; Williams, J. M. J. Adv.
Synth. Catal. 2007, 349, 1555−1575. (b) Nixon, T. D.; Whittlesey, M.
K.; Williams, J. M. J. Dalton Trans. 2009, 38, 753−762. (c) Guillena,
́
(
́
G.; Ramon
́
, D. J.; Yus, M. Chem. Rev. 2010, 110, 1611−1641.
(
6
2
(
6
(
̃
(
7
6
d) Dobereiner, G. E.; Crabtree, R. H. Chem. Rev. 2010, 110, 681−
622−6627. (b) Ikariya, T.; Kuwata, S.; Kayaki, Y. Pure Appl. Chem.
03. (e) Watson, A. J. A.; Williams, J. M. J. Science 2010, 329, 635−
36. (f) Bahn, S.; Imm, S.; Neubert, L.; Zhang, M.; Neumann, H.;
010, 82, 1471−1483.
̈
14) Gunanathan, C.; Milstein, D. Acc. Chem. Res. 2011, 44, 588−
Beller, M. ChemCatChem 2011, 3, 1853−1864. (g) Pan, S.; Shibata, T.
02.
15) Selected examples: (a) Shvo, Y.; Czarkie, D.; Rahamim, Y.;
ACS Catal. 2013, 3, 704−712.
(
2) (a) Mu
b) Mu
Chem. Rev. 2008, 108, 3795−3892. (c) Nugent, T. C.; El-Shazly, M.
ger, K.; Tillack, A.;
̈
ller, T. E.; Beller, M. Chem. Rev. 1998, 98, 675−703.
Chodosh, D. F. J. J. Am. Chem. Soc. 1986, 108, 7400−7402. (b) M. J.
Mays, M. J.; Morris, M. J.; Raithby, P. R.; Shvo, Y.; Czarkie, D.
Organometallics 1989, 8, 1162−1167. (c) Casey, C. P.; Singer, S. W.;
Powell, D. R.; Hayashi, R. K.; Kavana, M. J. Am. Chem. Soc. 2001, 123,
(
̈
ller, T. E.; Hultzsch, K. C.; Yus, M.; Foubelo, F.; Tada, M.
Adv. Synth. Catal. 2010, 352, 753−819. (d) Kru
̈
Beller, M. ChemSusChem 2009, 2, 715−717. (e) Ward, J.;
Wohlgemuth, R. Curr. Org. Chem. 2010, 14, 1914−1927. (f) Crozet,
D.; Urrutigoïty, M.; Kalck, P. ChemCatChem 2011, 3, 1102−1118.
1
090−1100. (d) Casey, C. P.; Johnson, J. B.; Singer, S. W.; Cui, Q. J.
Am. Chem. Soc. 2005, 127, 3100−3109. (e) Yamaguchi, R.; Ikeda, C.;
Takahashi, Y.; Fujita, K-i. J. Am. Chem. Soc. 2009, 131, 8410−8412.
(
(
1
g) Gunanathan, C.; Milstein, D. Science 2013, 341, 249−261.
h) Banerjee, D.; Junge, K.; Beller, M. Angew. Chem. 2014, 126,
(f) Conley, B. L.; Pennington-Boggio, M. K.; Boz, E.; Williams, T. J.
Chem. Rev. 2010, 110, 2294−2312. (g) Bauer, G.; Kirchner, A. K.
3265−13269; Angew. Chem., Int. Ed. 2014, 53, 13049−13053.
3) Early reports: (a) Watanabe, Y.; Tsuji, Y.; Ohsugi, Y. Tetrahedron
Lett. 1981, 22, 2667−2670. (b) Grigg, R.; Mitchell, T. R. B.;
Sutthivaiyakit, S.; Tongpenyai, N. J. Chem. Soc., Chem. Commun. 1981,
Angew. Chem. 2011, 123, 5918−5920; Angew. Chem., Int. Ed. 2011, 50,
(
5
(
798−5800.
16) Noyori, R.; Yamakawa, M.; Hashiguchi, S. J. Org. Chem. 2001,
6, 7931−7944.
17) Blum, Y.; Czarkie, D.; Rahamim, Y.; Shvo, Y. Organometallics
985, 4, 1459−1461.
18) For a pendant alkoxide/alkoxide group in the Pt-catalyzed
6
(
1
(
6
(
11−612.
4) (a) Yamaguchi, R.; Mingwen, Z.; Kawagoe, S.; Asai, C.; Fujita, K.-
i. Synthesis 2009, 7, 1220−1223. (b) Kawahara, R.; Fujita, K.-i.;
Yamaguchi, R. J. Am. Chem. Soc. 2010, 132, 15108−15111.
hydration of nitriles, see; (a) Ghaffar, T.; Parkins, A. W. Tetrahedron
Lett. 1995, 36, 8657−8660. (b) Ghaffar, T.; Parkins, A. W. J. Mol.
Catal. A: Chem. 2000, 160, 249−261. (c) Jiang, X.-B.; Minnaard, A. J.;
Feringa, B. L.; De Vries, J. G. J. Org. Chem. 2004, 69, 2327−2331. A
catalyst that could be considered to be a metal−ligand bifunctional
(
2
(
c) Kawahara, R.; Fujita, K.-i.; Yamaguchi, R. Adv. Synth. Catal.
011, 353, 1161−1168.
5) (a) Bahn, S.; Tillack, A.; Imm, S.; Mevius, K.; Michalik, D.;
Hollmann, D.; Neubert, L.; Beller, M. ChemSusChem 2009, 2, 551−
57. (b) Bahn, S.; Imm, S.; Mevius, K.; Neubert, L.; Tillack, A.;
Williams, J. M. J.; Beller, M. Chem. - Eur. J. 2010, 16, 3590−3593.
c) Imm, S.; Bahn, S.; Zhang, M.; Neubert, L.; Neumann, H.;
Klasovsky, F.; Pfeffer, J.; Haas, T.; Beller, M. Angew. Chem. 2011, 123,
741−7745; Angew. Chem., Int. Ed. 2011, 50, 7599−7603. (d) Zhang,
M.; Imm, S.; Bahn, S.; Neumann, H.; Beller, M. Angew. Chem. 2011,
23, 11393−11397; Angew. Chem., Int. Ed. 2011, 50, 11197−11201.
6) (a) Blank, B.; Michlik, S.; Kempe, R. Chem. - Eur. J. 2009, 15,
790−3799. (b) Blank, B.; Michlik, S.; Kempe, R. Adv. Synth. Catal.
̈
5
̈
catalyst: (d) Pen
Olivan, M. Organometallics 2012, 31, 6154−6161. Work suggesting
potential use of tertiary alkoxides ligands in bifunctional catalysts:
e) Schley, N. D.; Halbert, S.; Raynaud, C.; Eisenstein, O.; Crabtree, R.
̃
afiel, I.; Pastor, I. M.; Yus, M.; Esteruelas, M. A.;
́
(
̈
(
7
H. Inorg. Chem. 2012, 51, 12313−12323.
̈
(19) (a) Liu, C.; Liao, S.; Li, Q.; Feng, S.; Sun, Q.; Yu, X.; Xu, Q. J.
1
(
3
2
2
1
́
Org. Chem. 2011, 76, 5759−5773. (b) Martınez-Asencio, A.; Ramon,
́
D. J.; Yus, M. Tetrahedron 2011, 67, 3140−3149. (c) Li, Q.; Fan, S.;
Sun, Q.; Tian, H.; Yu, X.; Xu, Q. Org. Biomol. Chem. 2012, 10, 2966−
2972. (d) Zhao, Y.; Foo, S. W.; Saito, S. Angew. Chem. 2011, 123,
3062−3065; Angew. Chem., Int. Ed. 2011, 50, 3006−3009. (e) Zhang,
Y.; Qi, X.; Cui, X.; Shi, F.; Deng, Y. Tetrahedron Lett. 2011, 52, 1334−
1338. (f) Weickmann, D.; Frey, W.; Plietker, B. Chem. - Eur. J. 2013,
19, 2741−2748.
009, 351, 2903−2911. (c) Blank, B.; Kempe, R. J. Am. Chem. Soc.
010, 132, 924−925. (d) Michlik, S.; Kempe, R. Chem. - Eur. J. 2010,
6, 13193−13198. (e) Michlik, S.; Hille, T.; Kempe, R. Adv. Synth.
Catal. 2012, 354, 847−862. (f) Ruch, S.; Irrgang, T.; Kempe, R. Chem.
Eur. J. 2014, 20, 13279−13285.
7) (a) Tursky, M.; Lorentz-Petersen, L. L. R.; Olsen, L. B.; Madsen,
-
(
R. Org. Biomol. Chem. 2010, 8, 5576−5582. (b) Monrad, R. N.;
Madsen, R. Org. Biomol. Chem. 2011, 9, 610−615. (c) Fristrup, P.;
Tursky, M.; Madsen, R. Org. Biomol. Chem. 2012, 10, 2569−2577.
(20) Bacells, D.; Nova, A.; Clot, E.; Gnanamgari, D.; Crabtree, R. H.;
Eisenstein, O. Organometallics 2008, 27, 2529−2535.
(21) For experimental and theoretical studies on transfer hydro-
genation with other Ir systems: (a) Campos, J.; Hintermair, U.;
Brewster, T. P.; Takase, M. K.; Crabtree, R. H. ACS Catal. 2014, 4,
973−985. (b) Fujita, K.-i.; Yoshida, T.; Imori, Y.; Yamaguchi, R. Org.
Lett. 2011, 13, 2278−2281. (c) Li, H.; Lu, G.; Jiang, J.; Huang, F.;
Wang, Z.-X. Organometallics 2011, 30, 2349−2363. (d) Albrecht, M.;
(
d) Lorentz-Petersen, L. L. R.; Nordstrøm, L. U.; Madsen, R. Eur. J.
Org. Chem. 2012, 34, 6752−6759.
8) (a) Martınez, R.; Ramon, D. J.; Yus, M. Org. Biomol. Chem. 2009,
, 2176−2181. (b) Cano, R.; Ramon, D. J.; Yus, M. J. Org. Chem. 2011,
6, 5547−5557. (c) Martınez-Asencio, A.; Ramon, D. J.; Yus, M.
(
7
7
́
́
́
́
́
3
715
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716

ACS Catalysis
Research Article
Miecznikowski, J. R.; Samuel, A.; Faller, J. W.; Crabtree, R. H.
(38) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (b) Becke,
A. D. J. Chem. Phys. 1993, 98, 1372−1377. (c) Lee, C.; Yang, W.; Parr,
R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789.
(39) LACVP* uses the 6-31G* basis set for all light elements and the
Hay−Wadt ECP and basis set for Ir, see: Hay, P. J.; Wadt, W. R. J.
Chem. Phys. 1985, 82, 299−310.
Organometallics 2002, 21, 3596−3604. (e) Hanasaka, F.; Fujita, K.-i.;
̀
Yamaguchi, R. Organometallics 2005, 24, 3422−3433. (f) Pamies, O.;
Bac
22) (a) Royo, B.; Peris, E. Eur. J. Inorg. Chem. 2012, 9, 1309−1318.
b) da Costa, A. P.; Viciano, M.; Sanau, M.; Merino, S.; Tejeda, J.;
̈
kvall, J.-E. Chem. - Eur. J. 2001, 7, 5052−5058.
(
(
́
(
40) Goodman, J. M.; Silva, M. A. Tetrahedron Lett. 2003, 44, 8233−
236.
41) (a) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−
41. (b) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167.
42) (a) Tannor, D. J.; Marten, B.; Murphy, R.; Friesner, R. A.;
Peris, E.; Royo, B. Organometallics 2008, 27, 1305−1309. (c) Liu, C.;
Liao, S.; Li, Q.; Feng, S.; Sun, Q.; Yu, X.; Xu, Q. J. Org. Chem. 2011,
8
(
2
(
7
6, 5759−5773. (d) Mehta, A.; Thaker, A.; Londhe, V.; Nandan, S. R.
Appl. Catal., A 2014, 478, 241−251. (e) Satyanarayana, P.; Reddy, G.
M.; Maheswaran, H.; Kantam, M. L. Adv. Synth. Catal. 2013, 355,
Sitkoff, D.; Nicholls, A.; Ringnalda, M.; Goddard, W. A., III; Honig, B.
J. Am. Chem. Soc. 1994, 116, 11875−11882. (b) Marten, B.; Kim, K.;
Cortis, C.; Friesner, R. A.; Murphy, R. B.; Ringnalda, M. N.; Sitkoff,
D.; Honig, B. J. Phys. Chem. 1996, 100, 11775−11788.
1
(
(
859−1867.
23) (a) Bullock, R. M. Chem. - Eur. J. 2004, 10, 2366−2374.
b) Samec, J. S. M.; Backvall, J.-E.; Andersson, P. G.; Brandt, P. Chem.
Soc. Rev. 2006, 35, 237−248.
(
43) With [Cp*IrCl ] as the catalyst, inhibition by the aniline
́
2 2
(
24) (a) Samec, J. S. M.; Ell, A. H.; Åberg, J. B.; Privalov, T.;
Eriksson, L.; Backvall, J.-E. J. Am. Chem. Soc. 2006, 128, 14293−14305.
b) Clapham, S. E.; Hadzovic, A.; Morris, R. H. Coord. Chem. Rev.
substrate was attributed to disfavored coordination of the alcohol
substrate, rather than disfavored coordination of the imine
intermediate. It was also shown that this inhibition is dependent on
̈
(
2
004, 248, 2201−2237. (c) Casey, C. P.; Johnson, J. B. J. Am. Chem.
20
the structure of the amine substrate and/or product.
Soc. 2005, 127, 1883−1894. (d) Casey, C. P.; Bikzhanova, G. A.; Cui,
Q.; Guzei, I. A. J. Am. Chem. Soc. 2005, 127, 14062−14071. (e) Casey,
C. P.; Johnson, J. B. Can. J. Chem. 2005, 83, 1339−1346. (f) Noyori,
R.; Hashiguchi, S. Acc. Chem. Res. 1997, 30, 97−102. (g) Dorta, R.;
Broggini, D.; Kissner, R.; Togni, A. Chem. - Eur. J. 2004, 10, 4546−
(44) The coordination of 7a to INT-10 was also studied. The
resulting complex, INT-10B, was found to have an energy similar to
that of INT-10 (−21 kJ/mol relative to INT-1). Hence, the binding of
7
a to the iridium hydride is much weaker than that of 6a, which agrees
well with the experimental findings.
45) 9 kJ/mol refers to the formation of an (E)-imine. The formation
4
(
2
555.
25) Feng, Y.; Jiang, B.; Boyle, P. A.; Ison, E. A. Organometallics 2010,
̈
9, 2857−2867. (b) Neubert, L.; Michalik, D.; Bahn, S.; Imm, S.;
(
of the corresponding (Z)-imine was also evaluated, but it was found to
be higher in energy. Importantly, both isomers of the imine were
tested in the subsequent hydrogenation step (via both the inner- and
outer-sphere mechanisms), and the calculated energies for transition
states involving the (E)-imine were uniformly lower. Therefore, the
involvement of the (Z)-imine was ruled out.
Neumann, H.; Atzrodt, J.; Derdau, V.; Holla, W.; Beller, M. J. Am.
Chem. Soc. 2012, 134, 12239−12244.
(
26) Simmons, E. M.; Hartwig, J. F. Angew. Chem., Int. Ed. 2012, 51,
3
(
066−3072.
27) General perspective: Synergy between experiment and theory:
Lupp, D.; Christensen, N. J.; Fristrup, P. Dalton Transactions 2014, 43,
1093−11105.
28) (a) Hammett, L. P. J. Am. Chem. Soc. 1937, 59, 96−103.
b) Ingold, C. K.; Shaw, F. R. J. Chem. Soc. 1927, 2918−2926.
29) For ‘one pot’ Hammett investigations of stoichiometric
reactions, see: Yau, H. M.; Croft, A. K.; Harper, J. B. Chem. Commun.
012, 48, 8937−8939.
30) For examples of other competition Hammett investigations of
(46) An alternative pathway from INT-11 to INT-12 via INT-11A,
which contains a σ-bonded imine, was calculated. Relative Gibbs free
energies of 42, 4, and 67 kJ/mol were found for TS11‑11A, INT-11A,
and TS11A‑12, respectively. The TS energy was 42 kJ/mol.
1
(
(
(
2
(
catalytic reactions, see: (a) Desai, L. V.; Strowers, K. J.; Sanford, M. S.
J. Am. Chem. Soc. 2008, 130, 13285−13293. (b) Fristrup, P.; Kreis, M.;
Palmelund, A.; Norrby, P.-O.; Madsen, R. J. Am. Chem. Soc. 2008, 130,
5
206−5215. (c) Keinicke, L.; Fristrup, P.; Norrby, P.-O.; Madsen, R. J.
Am. Chem. Soc. 2005, 127, 15756−15761. (d) Hedstrom, A.;
Bollmann, U.; Bravidor, J.; Norrby, P.-O. Chem. - Eur. J. 2011, 17,
̈
1
(
(
1991−11993.
31) Curtin, D. Y. Rec. Chem. Prog. 1954, 15, 111−128.
32) For Ir−alkoxide complexes, see ref 18e. Ru−alkoxide
complexes: (a) Johnson, T. J.; Huffman, J. C.; Caulton, K. G. J. Am.
Chem. Soc. 1992, 114, 2725−2727. (b) Caulton, K. G. New J. Chem.
1
994, 18, 25−41. (c) Johnson, T. J.; Folting, K.; Streib, W. E.; Martin,
J. D.; Huffman, J. C.; Jackson, S. A.; Eisenstein, O.; Caulton, K. G.
Inorg. Chem. 1995, 34, 488−499. (d) Stewart, B.; Nyhlen, J.; Martın
Matute, B.; Backvall, J.-E.; Privalov, T. Dalton Trans. 2013, 42, 927−
34.
33) (a) Makarov, I. S.; Fristrup, P.; Madsen, R. Chem. - Eur. J. 2012,
8, 15683−15692. (b) Sieffert, N.; Buhl, M. J. Am. Chem. Soc. 2010,
32, 8056−8070.
34) Plata, R. E.; Singleton, D. A. J. Am. Chem. Soc. 2015, 137, 3811−
826.
35) (a) Chen, Y.; Liu, S.; Lei, M. J. Phys. Chem. C 2008, 112,
3524−13527. (b) Zhang, X.; Guo, X.; Chen, Y.; Tang, Y.; Lei, M.;
Fang, W. Phys. Chem. Chem. Phys. 2012, 14, 6003−6012.
36) (a) Glueck, D. S.; Wu, J. X.; Hollander, F. J.; Bergman, R. G. J.
́
́
-
̈
9
(
1
1
̈
(
3
(
1
(
Am. Chem. Soc. 1991, 113, 2041−2054. (b) Macgregor, S. A.; Vadivelu,
P. Organometallics 2007, 26, 3651−3659.
(
̈
37) Jaguar, version 7.8; Schrodinger, LLC, New York, 2011.
3
716
DOI: 10.1021/acscatal.5b00645
ACS Catal. 2015, 5, 3704−3716