Unravelling the Mechanism of Basic Aqueous Methanol Dehydrogenation Catalyzed by Ru-PNP Pincer Complexes

Article abstract of DOI:10.1021/jacs.6b05692

Ruthenium PNP complex 1a (RuH(CO)Cl(HN(C₂H₄Pi-Pr₂)₂)) represents a state-of-the-art catalyst for low-temperature (<100 °C) aqueous methanol dehydrogenation to H₂ and CO₂. Herein, we describ

Full text of DOI:10.1021/jacs.6b05692

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Unravelling The Mechanism of Basic Aqueous Methanol
Dehydrogenation Catalyzed By Ru-PNP Pincer Complexes
Elisabetta Alberico, Alastair J. J. Lennox, Lydia K. Vogt, Haijun Jiao, Wolfgang Baumann, Hans-Joachim
Drexler, Martin Nielsen, Anke Spannenberg, Marek P. Checinski, Henrik Junge, and Matthias Beller
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b05692 • Publication Date (Web): 19 Oct 2016
Downloaded from http://pubs.acs.org on October 19, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 18
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TOC
85x42mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Unravelling The Mechanism of Basic Aqueous Methanol Dehydro-
genation Catalyzed By Ru-PNP Pincer Complexes
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
‡
a,b
‡a
a
a
a
Elisabetta Alberico, Alastair J. J. Lennox, Lydia K. Vogt, Haijun Jiao, Wolfgang Baumann,
a
c
a
d
a
HansꢀJoachim Drexler, Martin Nielsen, Anke Spannenberg, Marek P. Checinski, Henrik Junge,
a
Matthias Beller *
a
Leibniz Institute for Catalysis at the University of Rostock, Albert EinsteinꢀStraße 29a, 18059 Rostock, Germany
b
Istituto di Chimica Biomolecolare, Consiglio Nazionale delle Ricerche, tr. La Crucca 3, 07100 Sassari, Italy
c
Centre for Catalysis and Sustainable Chemistry, Department of Chemistry, Technical University of Denmark, Kemitorvet
07, 2800 Kgs. Lyngby, Denmark
2
d
CreativeQuantum GmbH, Wegedornstr. 32, 12524 Berlin, Germany
ABSTRACT: Ruthenium PNP complex 1a (RuH(CO)Cl(HN(C H PiꢀPr ) )) represents a stateꢀofꢀtheꢀart catalyst for lowꢀ
2
4
2 2
temperature (<100 °C) aqueous methanol dehydrogenation to H and CO . Herein, we describe an investigation that combines exꢀ
2
2
periment, spectroscopy and theory to provide a mechanistic rationale for this process. During catalysis, the presence of two anionic
-
-
resting states was revealed, Ruꢀdihydride (3 ) and Ruꢀmonohydride (4 ) that are deprotonated at nitrogen in the pincer ligand backꢀ
-
bone. DFT calculations showed that Oꢀ and CHꢀ coordination modes of methoxide to ruthenium compete, and form complexes 4
-
-
-
and 3 , respectively. Not only does the reaction rate increase with increasing KOH, but the ratio of 3 /4 increases, demonstrating
-
that the “innerꢀsphere” CꢀH cleavage, via CꢀH coordination of methoxide to Ru, is promoted by base. Protonation of 3 liberates H
2
gas and formaldehyde, the latter of which is rapidly consumed by KOH to give the corresponding gemꢀdiolate and provides the
overall driving force for the reaction. Full MeOH reforming is achieved through the corresponding steps that start from the gemꢀ
diolate and formate. Theoretical studies into the mechanism of the catalyst Me-1a (Nꢀmethylated 1a) revealed that CꢀH coordinaꢀ
tion to Ru setsꢀup CꢀH cleavage and hydride delivery; a process that is also promoted by base, as observed experimentally. Howevꢀ
er, in this case, Ruꢀdihydride Me-3 is much more stable to protonation and can even be observed under neutral conditions. The
greater stability of Me-3 rationalizes the lower rates of Me-1a compared to 1a, and also explains why the reaction rate then drops
with increasing KOH concentration.
reverse waterꢀgasꢀshift reaction, the high temperatures favor
CO formation, which is incompatible with current fuel cell
technologies.
INTRODUCTION
Concerns over depleting fossil fuels and the negative efꢀ
fects of increasing CO emissions have stimulated the search
2
1
Scheme 1. Methanol steam reforming (A) and liquidꢀphase dehyꢀ
drogenation (B)
for more sustainable energy sources. Hydrogen has been idenꢀ
2
tified as a possible alternative source. Its highꢀenergy comꢀ
3
bustion or use in fuel cells generates water as the sole byꢀ
product. However, the physical and chemical properties of H2
gas do not render it an ideal energy vector. With a limited volꢀ
umetric energy density, it must be either compressed at very
high pressure (350ꢀ700 bars) or liquefied at very low temperaꢀ
ture (ꢀ253 °C). In addition, H is highly flammable and can
2
2
diffuse through several metals and materials. Thus, the chemꢀ
In efforts towards tackling these problems, we recently
demonstrated the first low temperature dehydrogenation of
aqueous methanol to H and CO with almost no trace of CO
ical storage of H in solid or liquid compounds is currently inꢀ
2
4
5
tensively investigated. In particular, alcohols constitute suitꢀ
2
2
able H carriers. Among these, methanol is considered to be
2
9
6
contamination, Scheme 1B. First reported for acceptorless deꢀ
the most viable option, as it is a liquid at room temperature
10a,b
hydrogenation of secondary alcohols,
we identified the
and has a comparatively high H content (12.6 wt%), which
2
7
homogeneous rutheniumꢀbased PNPꢀpincer complex 1a, Figꢀ
can be released through steam reforming, Scheme 1A.
ure 1, suitable for the aqueous reforming of methanol under
In general, this reaction is performed using either copperꢀ
9
basic conditions. Originally developed for ester hydrogenaꢀ
based (CuO/ZnO/Al O ) or group 8–10 metalꢀbased (PdꢀZn
2
3
10c,d
tion,
catalyst 1b was also found to facilitate this transforꢀ
alloys) heterogeneous catalysts that operate at high temperaꢀ
8
mation. Notably, employing <1 ppm of 1a below 100 °C TOFs
ture (200ꢀ300 °C). Although highly active and selective, the
ꢀ1
up to = 4720 hr were observed (40 mL MeOH, 8 M KOH,
copperꢀbased catalysts are pyrophoric and deactivate due to
metal particle sintering above 300 °C. Moreover, through a
1
.58 μmol 1a). This complex proved to be highly stable (23
ACS Paragon Plus Environment

Page 3 of 18
Journal of the American Chemical Society
1
3
days) affording TONs > 350,000 (40 mL MeOH:H O (9:1), 8
M KOH, 0.88 μmol 1a), which represents the most active and
innocent”1 ligands participate in metalꢀligand bifunctional
2
4
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
catalysis. Other catalysts that do not bear pincer ligands have
productive low temperature methanol reforming system develꢀ
oped to date. Since then, other catalysts based on Ru and Fe
appeared in the literature, Figure 1, clearly illustrating the poꢀ
tential for this strategy. Multidentate pincer ligands are comꢀ
mon to these complexes, as they secure high thermal stability
and are uniquely involved in the catalytic cycle. These “nonꢀ
also been reported, including Grützmacher’s 15and coꢀ
1
1
12
11a
workers', which engages in ligand cooperativity, and two
16,17
others based on iridium,
which, like the catalysts used by
us, are active without additional solvent.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. A comparison of a selection of catalysts able to engender full methanol dehydrogenation. *Syringeꢀpump addition of
ꢀ
1
MeOH/H O/NaOH (0.6/0.6/0.001 mmol hr ).
2
20
Complete methanol reforming requires a catalyst that can
workers. Despite providing considerably different absolute
energetics for each step, both theoretical studies agree on the
relative freeꢀenergy barriers of the three consecutive steps;
with gemꢀdiol and formic acid dehydrogenation being the
facilitate three consecutive steps, Scheme 2. The first H reꢀ
2
lease originates from the dehydrogenation of methanol to forꢀ
maldehyde. Subsequent reaction with water gives a gemꢀdiol,
which liberates the second equivalent of H upon its dehydroꢀ
most and least facile step, respectively. This prediction is in
2
9
genation to formic acid. Final dehydrogenation to CO releases
line with experimental observations,
as formaldeꢀ
2
the third H molecule.
hyde/gemdiol escape detection, whilst formate steadily builds
2
1
up in solution ( H NMR).
Scheme 2. Three steps for methanol dehydrogenation.
Most catalysts developed so far for methanol dehydroꢀ
genation require either a base or acid additive to secure signifꢀ
icant activity. A Lewis acid promotes precatalyst decarboxylaꢀ
12b
tion in 1e, however, base is required for those described by
9,12a
11c
our group (1a-d),
as well as Milstein and coꢀworkers and
17
Fujita/Yamaguchi and coꢀworkers. Notably, present mechaꢀ
nistic and theoretical investigations of the stateꢀofꢀtheꢀart cataꢀ
lyst 1a do not provide a reasonable rationale for the necessity
of the very high base concentrations employed. Herein, we
report for the first time on a detailed mechanistic investigation
of this process employing a mixture of experimental, spectroꢀ
scopic and theoretical tools.
Through a qualitative analysis of the data, a common
mechanism for all three steps using 1a was proposed, involvꢀ
ing an “outerꢀsphere” concerted association of methanol to the
coordinatively unsaturated amido complex 2, Scheme 3.
Transfer of a proton to nitrogen and a hydride to Ru generates
18
dihydride complex 3, from which a solvent assisted liberaꢀ
tion of H gas ensues. Support for this hypothesis came from
2
1
9a
DFT calculations by Yang, according to whom the key CꢀH
cleavage step, Scheme 3, is a stepꢀwise process, wherein hyꢀ
dride is transferred from an uncoordinated methoxide to the
cationic Ru centre. This proposal was supported by Lei and
coꢀworkers, who suggested that the dehydrogenation of forꢀ
mate could occur either via an outer or inner sphere mechaꢀ
Scheme 3. Previously proposed direct outerꢀsphere addition to
form intermediate 3. P = PiꢀPr2.
19b
nism. The outerꢀsphere mechanism was proposed to proceed
via the same transition state to our prior suggestions, Scheme
3
, whereas the inner sphere followed a nonꢀclassical hydride
elimination, much like that proposed by Milstein and coꢀ
2
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 18
O
R
H
methyl: syn or anti to the hydride on Ru, and have very similar
chemical shifts. Xꢀray analysis from a single crystal was conꢀ
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
H
N
OH
R
+
P
sistent with NMR and confirmed the CO to be trans to the niꢀ
trogen of the meridional coordinated PNP ligand, Figure 2.
Unfortunately, it was not possible to correlate the solid
H
RuII
Cl
H
N
H
H
N
P
CO
P
N
P
P
H
P
RuII
H
RuII
RuII
structure to one of the two isomers in solution and there was
P
CO
CO Step I R = CH2
Step II R = CHOH
Step III R = CO
P
CO
1
H
2
H
3
no evidence of a spatially relevant crossꢀpeak in the
H
1a
NOESY spectrum. However, both isomers were computationꢀ
ally located: syn Me-1a is more stable than anti Me-1a by
ꢀ
1
0
.50 kcal mol , corresponding to an isomeric ratio of 70 to 30,
which is in good agreement with that observed experimentally
80:20). When the catalytic activity of Me-1a was tested under
RESULTS AND DISCUSSION
Inner-sphere vs. outer-sphere
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
the optimized dehydrogenation conditions, Figure 3 (See SI
2.2), we were surprised to observe that the rate was only 2.4
times lower than using 1a. Considering a difference of two
orders of magnitude was observed for catalysts undergoing the
same mechanism in the case of ammoniaꢀborane dehydrocouꢀ
We began by investigating the notion of an “outerꢀsphere”
mechanism in order to establish whether the ligand is truly
nonꢀinnocent. Replacing HꢀN with MeꢀN on the backbone is a
common strategy to probe ligand cooperativity. Metal comꢀ
2
1
22
23
24
21e
plexes, e.g., Ru, Fe, Co, Ni with the aliphatic pincer ligꢀ
ands (HN(CH CH PR ) , R = alkyl or Ph) have been applied
pling, it was conceivable that 1a and Me-1a followed the
same mechanism.
2
2
2 2
2
1a,22b
21b
to the hydrogenation of CO ,
bicarbonate, cyclic carꢀ
2
2
1f
21c,22c
22d
23b 22a,23b
bonates, nitriles,
esters, ketones , alkenes.
Nꢀheterocycles. Use of the corresponding Nꢀmethylated catꢀ
alysts results in no activity in all cases of hydrogenaꢀ
except that of bicarbonate , CO promoted
by FeꢀMePNP complexes in the presence of Lewis acid coꢀ
catalysts, and olefins catalysed by Co complexes.
ever, Nꢀmethylation of the ligand in Co and Ru complexes
and
23d
1
a
600
Me-1a
2
1a,c,f,22aꢀc,23b,d
21b
tion,
2
5
4
3
2
00
00
00
00
2
2b
23b,6
Howꢀ
21d,22e,23b
furnish good yields in the dehydrogenation of alcohols
23d
and Nꢀheterocycles as well as in the transfer hydrogenation
23c
23c
23e
of ketones , imines and olefins. However, rates of these
reactions were not measured and cannot be compared to those
using the nonꢀmethylated catalysts. In ammoniaꢀborane dehyꢀ
drocoupling, the Nꢀmethylated Ru complex results in a rate
that is two orders of magnitude lower than the corresponding
100
0
0
.0
0.5
1.0
1.5
2.0
2.5
3.0
2
1e
t [hr]
NꢀH complex. Gas phase calculations of this system predictꢀ
ed the same general mechanism, but with higher energy barriꢀ
ers.
1a or Me-1a (0.42 mM)
MeOH + H O
3H2 + CO2
2
ο
KOH (8 M), 91.0 C (set)
P
Cl
RN Ru CO 1a
R = H
H
Me-1a R = Me
P
Figure 3. A comparison of rates in the aqueous methanol reformꢀ
ing promoted by parent catalyst 1a and its Nꢀmethylated derivaꢀ
tive Me-1a. P = PiꢀPr2.
In order to further understand this unexpected reactivity,
Figure 2. ORTEP view of anti Me-1a with thermal ellipsoids
drawn at the 30% probability level. H atoms (except H1R) are
omitted for clarity.
25
we measured the kinetic isotope effects (KIEs) with each catꢀ
alyst, 1a and Me-1a, using both fully deuterated and undeuterꢀ
ated solvents and base (See SI 2.3). A striking difference in
the KIEs between the two catalysts was observed. 1a provided
a substantial isotope rate ratio of 7.07 compared to only 1.76
recorded with Me-1a, Scheme 4. It is nonꢀtrivial to directly
assess the implications of these values, as the rate measureꢀ
In order to probe bifunctional reactivity of 1a, we prepared
the corresponding Nꢀmethylated complex (Meꢀ1a)(See SI 2.1).
The complex was obtained as a mixture of two isomers Me-1a
80%) and Me-1a' (20%), both containing equivalent phosꢀ
phorus donors ( P NMR (C D ): Me-1a δ = 71.05 and Me-1a'
δ = 73.63) (See SI 2.1). The relative J coupling constants of
the corresponding triplet in the H NMR spectrum ( H NMR
C D ): Me-1a δ = ꢀ 15.33 ( J = 18.2 Hz), Me-1a' δ = ꢀ
5.25 ( J = 18.6 Hz)) indicate the hydride is cis to both Pꢀ
(
31
6
6
ments are from a product (H ) that evolves over three separate
2
2
HP
reactions, Scheme 2. Absolute rates before the steady state
stage are interconnected, because the product of each step is
the starting material of the following one. Once at steadyꢀstate
and “real” reforming proceeds, rates for each step do not
1
1
2
(
1
6
6
HP
2
HP
atoms. The two isomers are due to the relative orientation of
3
ACS Paragon Plus Environment

Page 5 of 18
Journal of the American Chemical Society
To further understand the influence of base, we measured
change, however, the KIE is still a composite from all steps
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and is therefore difficult to disentangle.
rates at different KOH concentrations at constant temperature.
It was necessary to modify the setꢀup, because at low concenꢀ
trations of base it is not possible to raise the temperature above
the boiling point of the solvent. Hence, we used an autoclave
to which an overꢀpressure was applied in order to increase the
boiling point and to ensure the solvent remained in the liquid
phase (See SI 1.2). However, hydrogen evolution was considꢀ
erably lower than that observed using our original burette setꢀ
up in an open system (See SI 1.1). Several control reactions
were undertaken that confirmed the attenuation to arise from
the reverse reaction: specifically, H and CO inserting into
The evolution of gas is observed to proceed through pseuꢀ
do zero order kinetics, Figure 3. At the levels of conversion
reached over the time periods studied, the concentrations of
MeOH, water and catalyst remain effectively constant. Thus,
additional information from the decay curves of limiting reaꢀ
gents or intermediates cannot be gathered under normal cataꢀ
26
lytic conditions. The TOF decreased with increasing catalyst
loading (See SI 2.4), an interesting effect similarly observed
2
1g
by Gusev and coꢀworkers
and more recently by
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21h
Gauvin/Dumeignil and coꢀworkers. Indeed, the reaction orꢀ
der with respect to catalyst was measured to be below 1. These
data may be attributed to an offꢀcycle Ruꢀdimerization at
higher loadings, although no such intermediates could be deꢀ
2
2
catalytic intermediates (vide infra). This complication was
controlled by leaking the evolved gas through a valve and reꢀ
taining an over pressure of 0.6 bar (See SI 1.3). Despite
31
ꢀ1
tected by P NMR. Alternatively, the massꢀtransfer of H out
providing lower rates than in an open vessel (TOF = 550 hr
2
ꢀ
1
of the system might limit the rate at higher Ruꢀloadings, as the
reaction is reversible and hydrogenation processes can occur at
higher H2 concentrations (pressures), vide infra. Providing
support for this hypothesis, the rate of reaction was found to
be dependent on the stirring rate (See SI 2.5), as also observed
vs 1770 hr , 8 M KOH), this adjustment was sufficient to
reach 90 °C without suffering from the strong attenuation obꢀ
served with full pressurization. Using this modified setꢀup, it
was possible to measure the rate using 4ꢀ8 M KOH with both
1a and Me-1a catalysts, albeit with an increased loading of
Me-1a. To record the rates below 4 M KOH, it was necessary
to decrease the temperature to 60 °C. At this temperature, the
regular burette setꢀup was employed (See SI 1.1). For catalyst
1a, a first order dependence was observed at 60 °C, Figure 5.
At 90 °C, saturation kinetics appear at KOH concentrations
above 7 M, which may be due to the massꢀtransport limiting
2
1h
by Gauvin/Dumeignil and coꢀworkers.
Deuteration decreased the reaction rate seven fold in the
case of 1a, but not even two fold in the case of Me-1a. Thus,
despite Me-1a being active, the magnitude of the KIE differꢀ
ence implies different operating mechanisms for the two cataꢀ
lysts. Apparently for Me-1a, in contrast to 1a, the turnover
limiting step(s) does not significantly involve the cleavage of
an XꢀH (X = C, O, Ru) bond.
loss of H from the system at high reaction rates, vide supra.
2
For catalyst Me-1a, a peak in rate was observed at 4 M KOH,
after which the rate dropped.
Scheme 4. KIE measurements for catalyst 1a and Me-1a.
-10
-12
-14
-16
-18
-20
-
1
E = 82.4 kJ mol
a
6
-1
A = 1.2 10 mol s
×
Temperature and base dependency measurements
During reaction optimization, we observed that high base
concentrations increased the catalytic activity significantly. An
optimal operational pH has also been reported for bifunctional
2.75 2.80 2.85 2.90 2.95 3.00 3.05 3.10
/T [1000/K]
2
7a,b,c
Ru catalyzed hydrogenation
tion
and transfer hydrogenaꢀ
of ketones. Initially, we speculated that the increased
1
27b,d
Figure 4. Arrhenius plot to show the dependency of activity
on temperature for aqueous methanol reforming promoted by
activity was due to the inflated temperature over the boiling
point of the solvent, which could be reached due to the saltꢀ
effect from the high KOH concentration. To test this, we atꢀ
tempted to replace KOH with innocent salts (e.g., KNO3,
catalyst 1a. Conditions: 8 M KOH, MeOH:H O (9:1, 10 mL),
2
4
.2 ꢀmol 1a, 3 hr.
KPF ) to effect the same temperature increase. However, all
6
the salts tested did not provide a homogeneous solution and so
it could not be fairly examined this way. In addition, mixing
methanol with higher boiling point coꢀsolvents, e.g. Nꢀ
methylpyrrolidine or tꢀBuOH, shut down the reactivity. Thus,
we examined the rate of methanol dehydrogenation systematiꢀ
cally at a range of lower temperatures (50ꢀ90 °C). The average
rate measured over a threeꢀhour period increased exponentialꢀ
ly with temperature, and produced a linear Arrhenius plot,
Figure 4.
4
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
50
00
1a, 60 °C
Me-1a, 60 °C
Me-1a, 90 °C
700
600
500
400
300
200
3
1a, 90 °C
250
200
150
100
Figure 6. ORTEP view of syn 1a with thermal ellipsoids drawn at
the 30% probability level. H atoms, other than H1 and H1R, are
omitted for clarity.
5
0
100
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0
0
2
4
6
8
Scheme 5. Generation of 2 by dehydrochlorination of 1a with
base. P = PiꢀPr2.
KOH [M]
Figure 5. The influence of KOH concentration on the activity of
catalysts 1a and Me-1a in methanol reforming at 60 °C and 90
°C. Conditions at 60 °C: 10 mL MeOH:H O (9:1) and 1a
(8.41 ꢀmol, blue dots) or Me-1a (8.41 ꢀmol, blue triangles) using
the “regular” burette setꢀup. Conditions at 90 °C: 20 mL
MeOH:H O (9:1) and 1a (8.41 ꢀmol, red dots) or Me-1a
16.82 ꢀmol, red triangles) in a leaking autoclave set to an over
pressure of 0.6 bar. Lines are solely a guide for the eye.
HBase+
P
P
P
2
- Cl-
H
H
H
Base
-
HN Ru CO
Cl
N Ru CO
Cl
N Ru CO
1
a
1a-
P
2
P
P
2
(
Dehydrochlorination of 1a to Ruꢀamido 2 must occur beꢀ
fore catalysis proceeds (Scheme 5). Initiation of this activation
process by base was tested by treatment with one equivalent of
t-BuOK in diethyl ether. A yellow, air sensitive, complex was
formed, whose spectral properties are consistent with that of
Stoichiometric studies
To further deconvolute the mechanism, we prepared and charꢀ
acterized the reactivity of a number of Ru complexes that are
possible intermediates in the catalytic cycle.
9
amido complex 2, Scheme 5. The two phosphorus donors of 2
31
are again equivalent, ( P NMR (THFꢀd ): δ = 93.8, one sigꢀ
8
nal)), although the peak is shifted to a lower field than 1a,
consistent with a more deshielded Pꢀnucleus. The hydride ligꢀ
and is cis to both phosphorus atoms, as judged by the relative
Activation of 1a
10a
1
a was prepared following a published procedure, and,
1
8
like Me-1a, was obtained as a mixture of two stereoisomers,
syn 1a and anti 1a, which differ only in the relative orientation
of the chloride ligand and the hydrogen on nitrogen. Both
isomers are spectroscopically ( H, P NMR) similar, and have
equivalent phosphorus donors indicated by singlet peaks
JHP coupling constants in the triplet peak ( H NMR (THFꢀd ):
29
δ = ꢀ19.0 (JHP = 17.1 Hz)). Crystals of 2 suitable for Xꢀray
analysis were grown from nꢀheptane at ꢀ78 °C, Figure 7. The
28
1
31
Ru(II) complex 2 displays a Yꢀshaped distorted trigonal biꢀ
pyramidal coordination geometry, where N1, H1R and C17
(from the CO ligand) define the equatorial plane.
31
1
( P{ H} NMR (THFꢀd ) syn 1a δ = 75.8 and anti 1a δ = 76.3)
8
(See SI 3.1). The coupling constant, J , of the hydride triplet
HCl elimination from 1a proceeds according to a dissociaꢀ
HP
1
31
in the H NMR spectrum indicates a cis relationship to both
tive conjugate base mechanism (Dcb mechanism). Base abꢀ
1
phosphorus atoms ( H NMR: syn 1a δ = ꢀ15.7 (JHP = 19.2 Hz,
stracts the acidic proton on nitrogen to generate a stabilized
anion, prompting chloride to leave and the formal formation of
an NꢀRu double bond. The lifeꢀtime of the anionic intermediꢀ
RuH) and anti 1a δ = ꢀ16.0 (JHP = 17.9 Hz, RuH)). Crystals
suitable for Xꢀray analysis were obtained from the slow diffuꢀ
sion between a dilute diglyme/diethylether solution, Figure 6.
The diffraction data allowed for independent location and reꢀ
finement of both the hydride ligand and the hydrogen on niꢀ
trogen. The solidꢀstate structure confirmed that the ligand is
coordinated to Ru in a meridional fashion with CO ligated
trans to nitrogen. The hydride is disposed cis to both phosphoꢀ
rus donors and also syn to the hydrogen on nitrogen, thereby
confirming it to be syn 1a, which was shown to be the minor
-
ate (1a ) is too short for detection, indicating rapid chloride
elimination. Coordination to Ru and the presence of the trans,
strongly πꢀaccepting, CO ligand, increase the acidity of the
amino group. The reversibility of this activation mechanism
was tested by employing a weaker base. Thus, anti 1a was
treated with 1 ꢀ 3 equivalents of triethylamine at room temperꢀ
1
ature in THFꢀd
and monitored by H NMR (See SI 3.2). Inꢀ
8
deed, isomerization between anti and syn was observed,
demonstrating rapid reversibility in the process, Scheme 5.
However, under catalytic conditions, the rate of methanol deꢀ
hydrogenation was unaffected by a 10ꢀfold excess of additionꢀ
al KCl (See SI 3.3), thus indicating the reverse process is kiꢀ
netically irrelevant. 2 was also never observed in solution unꢀ
der catalyticꢀlike conditions. In order to understand the reacꢀ
tivity of this highly sensitive complex, its interactions with
methanol, water and formic acid were examined under inert
conditions.
1
28
isomer that appears in solution ( H NMR). We also located
both isomers computationally; in line with our NMR observaꢀ
tions, anti 1a is more stable than syn 1a by 3.26 kcal mol .
ꢀ
1
5
ACS Paragon Plus Environment

Page 7 of 18
Journal of the American Chemical Society
H
2
1
2
3
4
5
6
7
8
9
MeOH
P
P
P
CH
THF-d
3
OH (2 equiv.)
, RT
H
H
H
8
N Ru CO
H
N
Ru CO
+
H
N
Ru CO
H
H CO
3
P
P
P
2
3
base
4
2
CH O
An apparent equilibrium exists between the corresponding
mono and dihydride Ru complexes, Scheme 6, which is afꢀ
fected by the addition of base or protic species. Hence, when
NaOMe was added to 2, only 3 was detected. In addition,
when 2 equiv. t-BuOK were added to a mixture of 4 and 3, all
material was converted to 3, consistent with a single turnꢀover
in the catalytic cycle, vide infra. Conversely, MeOH addition
induced the conversion of 3 to 4; after 50 equivalents were
added, 78% of 4 had formed from protonation of 3 and release
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. ORTEP view of 2 with thermal ellipsoids drawn at the
0% probability level. Hydrogen atoms other than H1R are omitꢀ
3
36
ted for clarity. The distortion from an ideal trigonal bipyramid
arises from the small H1RꢀRuꢀC17 angle (82.7(7)°) and the large
H1RꢀRuꢀN1 (120.4(7)°) and N1ꢀRuꢀC17 (156.89(6)°) angles. The
N1ꢀRu distance (1.9985(12) Å) in 2 is considerably shorter than in
30
syn 1a (2.1949(18) Å).
Reactions of complex 2
of H gas. We calculated the free energy change of the studied
reactions, Scheme 6. The reaction of 2 with methanol to 3 and
formaldehyde [2 + CH OH = 3 + CH O] is endergonic by 11.4
2
When two equivalents of MeOH in THFꢀd were added to
8
amido complex 2 at room temperature, the characteristic
3
2
bright yellow color instantly disappeared and two major comꢀ
ꢀ1
kcal mol , which is close to that calculated by Yang (9.2 kcal
1
31
plexes were observed ( H and P NMR), Scheme 6 (See SI
.1). Under these aprotic conditions, dihydride complex 3 was
ꢀ1 19a
ꢀ1 19b
mol ) and Lei and coꢀworkers (11.8 kcal mol ). The drivꢀ
ing force for this reaction under these conditions (THF, room
temperature) should be the formation of gaseous formaldehyde
(bp = ꢀ21°C). Formaldehyde dissociation (ꢁG° = ꢀ2.45 kcal
mol ) and trimerization (ꢁG° = ꢀ2.96 kcal mol ) pathways
were also computed but could not account for the energy
4
1
31
detected ( H and P NMR, vide infra) and accounted for 54%
of the total Ru content. This observation is consistent with that
1
0a
made by Gusev and coꢀworkers, where a Ruꢀalkoxide is in
equilibrium with a Ruꢀdihydride complex. The second major
ꢀ1
ꢀ1
species (36%) identified was a monohydride species. The peak
shortfall. The exchange reaction between 3 and 4 [3 + CH OH
1
3
for this monohydride complex ( H NMR (THFꢀd ): δ = ꢀ17.3
ꢀ1
8
=
4 + H ], is endergonic by 3.19 kcal mol . The driving force
2
32
2
(
JHP = 18.9 Hz)) suggests the hydride ligand is trans to a
in this case should be the formation of gaseous H2.
33
weak donor and cis to the two equivalent phosphorus donors.
This is consistent with formation of the Ruꢀmethoxide comꢀ
plex (4); similar complexes have been detected in the hydroꢀ
The methyl peaks of Ruꢀcoordinated
Cꢀenriched methoxide (RuꢀO CH ) and methanol
The addition of gemꢀdiol to 2 was calculated to be exerꢀ
gonic by 1.50 kcal mol or endergonic by 2.61 kcal mol with
or without hydrogen bonding, respectively, and with a low
ꢀ1
ꢀ1
2
7cꢀe,34
genation of ketones.
ꢀ1
barrier of 3.60 kcal mol . Thus, it is to be expected that this
1
3
1
13
3
reaction is also readily reversible.
3
( CH OH) appear as a single doublet in the room temperature
H NMR (See SI 4.1). When the solution was cooled, these
3
1
peaks separated out into two well resolved doublets. Ruꢀ
alkoxide species can be stabilized by intramolecular hydrogen
2
7e
ꢀ1
bonding with the backbone NꢀH, as well as intermolecular
Scheme 7. B3PW91ꢀgenerated Gibbs freeꢀenergies (kcal mol ) of
protic species addition to Ruꢀamido complex 2. P = PiꢀPr2.
35
bonding with excess methanol. To confirm the existence of a
RuꢀO bond, we treated the sample with CO and rapidly deꢀ
2
HR
H
N
P
R
R
tected a Ruꢀmethylcarbonate complex, resulting from the forꢀ
H
N
P
P
RH
N
P
P
P
N
P
mal insertion of CO in the RuꢀO bond (See SI 4.2). Only one
2
Ru
H
Ru
H
Ru
H
Ru
H
CO
CO
CO
P
CO
isomer was formed, in which the backꢀbone NꢀH was strongly
2
shifted to lower fields due to hydrogen bonding to the carꢀ
1
R: OCH3
0.00
0.00
0.00
4.19
2.74
9.37
7.20
4.41
2.65
18.76
3.60
0.89
0.25
-2.31 3
4
5
bonate ( H NMR, THFꢀd , δ = 7.6 (b)). 4 was too unstable to
8
R: OH
R: H
R: OCH
R: OOCH 0.00
be isolated, as it reverted back to Ruꢀamido 2 upon solvent
removal and reduced pressure. This seemingly very facile adꢀ
dition/elimination process is supported by DFT. The addition
OH 0.00
2.61 (-1.50) 7
-19.71 6
2
not located
not located
of methanol across the NꢀRu bond [2 + CH OH = 4] is only
3
ꢀ
1
In addition to methanol, water is a constituent of the mixꢀ
ture under catalytic conditions, and thus its addition to 2 in
THF was investigated. Interestingly, the intense yellow color
of 2 still remained after the addition of one equivalent of waꢀ
slightly endergonic (0.89 kcal mol ) and the Gibbs free energy
ꢀ
1
barrier is 4.41 kcal mol , Scheme 7. This is in line with the
experimental observations and confirms the reaction reversiꢀ
bility.
ter, but completely faded after two equivalents. A monohyꢀ
1
Scheme 6. Reaction of amido complex 2 with methanol and the
equilibrium between the products. P = PiꢀPr2.
dride species was detected ( H NMR (THFꢀd
8
): δ = – 16.5 (t,
): δ = 76.9 (s)), which
is consistent with the formation of Ruꢀhydroxide 5, Scheme 7
2
31
1
J
HP = 18.4 Hz); P{ H} NMR (THFꢀd
8
28
(
See SI 5.1). The coupling constants indicate cis orientation
6
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 18
of the hydride to both P donors and a downꢀfield broad peak solution, showed the NH and the formate group to be properly
1
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
oriented for such an interaction, Figure 8. The H1ꢀO3 distance
of 1.96(6) Å, the N1ꢀO3 distance of 2.793(4) and the N1ꢀH1ꢀ
O3 angle of 163(4)° are all characteristic of Hꢀbonding in the
(
H NMR: δ = 2.9) composed of the resonances of NH and
RuꢀOH protons. When D O was employed under otherwise
2
identical conditions, the intensity of this broad peak dropped
significantly, while the hydride signal (and the signals asꢀ
signed to the ligand backbone protons) remained unchanged.
The chemical shift for RuꢀH in 5 gradually increased as more
water was added, up to a maximum of 4 equivalents. This
spectroscopic evidence suggests a rapid exchange between 5
and Ruꢀamido 2 that is faster than the NMR timeꢀscale. This is
also consistent with the persistence of the characteristic yellow
color of 2 even after the addition of an equivalent of water.
Like RuꢀOMe 4 (vide supra), Ruꢀhydroxide 5 reacted with
41
complex and share similarities to the related ꢀPPh complex
2
(H1ꢀO3 = 2.04(3) Å, N1ꢀO3 = 2.831(3) Å and N1ꢀH1ꢀO3 =
21b
157(3)°), where such an interaction was also suggested. This
intramolecular stabilization will also account for the fact that
only one NH isomer is observed spectroscopically in solution.
6 was found to be thermally stable, at least up to 90 °C for 2
hrs in dioxane, but when treated with one equivalent of tꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
BuOK in THFꢀd , it reverted back to 2 along with small quanꢀ
8
tities of 3 (5%) (See SI 6.2). Interestingly, formation of 6 was
CO via facile insertion into the RuꢀO bond to give a Ruꢀ
also possible by reacting complex 3 with CO , which we calꢀ
culated to be exergonic by 6.72 kcal mol .
2
2
ꢀ1
bicarbonate complex with a strongly downꢀfield shifted NꢀH
1
due to hydrogen bonding with the carbonate ( H NMR, THFꢀ
37
d8, δ = 8.25 (bt, J = 10.7 Hz)) (See SI 5.2). 5 also readily reꢀ
Scheme 8. Synthesis of 6 by reaction of either 2 or 3 with formic
verted back to 2 upon solvent removal. This reversibility, of
acid. P = PiꢀPr .
2
38
which similar examples have been reported elsewhere, was
ꢀ
1
x
supported by theory, Scheme 7. The barrier (2.65 kcal mol )
was found to be lower than with methanol and less endergonic
P
P
P
H
H
H
HCOOH
THF
HCOOH
THF
ꢀ
1
N Ru CO
HN Ru CO
HN Ru CO
H
(0.25 kcal mol ), which is broadly in agreement with the findꢀ
1
9b
ings of Lei and coꢀworkers. However, our results differ
OHCO
- H
2
P
P
P
1
9a
from those of Yang who reported an exergonic reaction (ꢀ
2
6
3
ꢀ
1
ꢀ1
1
1
4
.9 kcal mol ) with a higher barrier (4.8 kcal mol ), which,
96% (by H NMR)
58% (by H NMR)
ꢀ
1
with a backꢀreaction barrier of 9.7 kcal mol , does not support
the experimental observations.
Under the catalytic conditions, formation of complex 5 is
assumed to be an unproductive, offꢀcycle, intermediate, as
there is no reasonable mechanism from which a Ruꢀdihydride
species can be generated. Thus, despite water being necessary
to establish full dehydrogenation, it is unsurprising that lower
reaction rates are observed when higher proportions of water
9
are present. In addition to solubility issues, this rationale proꢀ
vides a reasonable justification for the use of only low
amounts of water in the reaction mixture (9:1 MeOH:water)
under optimal conditions.
Figure 8. ORTEP view of anti 6 with thermal ellipsoids drawn at
the 30% probability level. H atoms other than H1 and H1R have
been omitted for clarity.
During catalysis potassium formate is detected in soluꢀ
tion, and thus we considered the formation of Ruꢀformate 6
9
Reactions of complex 3
(
See SI 6.1). Metalꢀformate complexes have been shown to be
We anticipated that Ruꢀdihydride 3 should be the precursor
39a
key intermediates, both in formic acid dehydrogenation and
in CO hydrogenation to formic acid.
lysts bearing nonꢀinnocent pincer ligands have afforded highly
active systems for these two transformations,
crystal structures of the involved metalꢀformate complexes
have been reported.
for a Ruꢀdihydrogen complex, from which H
rapid. Crystals of 3 were obtained from a reaction solution
(KOH 8 M; MeOH:H O (9:1)). In this complex the PNP pinꢀ
2
release is very
3
9b,c
A number of cataꢀ
2
2
2
1b,22b,40
and the
cer ligand is coordinated in a meridional fashion, with two hyꢀ
drides trans to each other and cis to the two P donors, and CO
sitting trans to nitrogen, Figure 9.
21b,40b,c,e,f
The reaction of 2 with formic acid
[
2 + HCOOH = 6], Scheme 7, was calculated to be more facile
than with water or methanol. The intermediate and transition
state could not be located as the reaction is barrierꢀless and
ꢀ
1
exergonic by 19.71 kcal mol . Indeed, reaction of formic acid
with either 2 or 3 yielded a white powder that corresponded to
6, Scheme 8 (See SI 6.1). Unlike with the addition of MeOH
to 2, which formed a mixture of 4 and 3, formic acid addition
to 2 exclusively gave Ruꢀmonohydride 6, without signs of Ruꢀ
dihydride 3. NMR data for 6 showed characteristic peaks for
1
formate and hydride ligands ( H NMR (THFꢀd ): δ = 8.35 and
8
Figure 9. ORTEP view of 3 with thermal ellipsoids drawn at the
ꢀ
18.14) and confirmed their relative geometry about Ru. The
3
0% probability level. Hydrogen atoms other than H1, H1R and
NH peak of the backbone resonates at a lower field than in 3
H2R are omitted for clarity.
1
(
H NMR (THFꢀd ): δ = 8.61 vs 3.7), indicative of its inꢀ
8
volvement in intramolecular hydrogen bonding. Xꢀray analysis
of crystals obtained by slow evaporation from a diethyl ether
H may be released from 3, (via Ruꢀdihydrogen complex),
2
through a fourꢀmembered transitionꢀstate that yields Ruꢀamido
7
ACS Paragon Plus Environment

Page 9 of 18
Journal of the American Chemical Society
change in solvent environment. However, the multiplicity of
2
. A concerted H addition onto 2 was calculated to be exerꢀ
2
ꢀ
1
ꢀ1
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
gonic by 2.31 kcal mol with a barrier of 18.76 kcal mol ,
Scheme 7. This is consistent with experimental observations
under aprotic conditions, as 3 was found to be relatively stable
and dehydrogenation could only be partially promoted by
the hydride peak is very different and the triplet observed in
basic MeOH:H O (9:1) (See SI 8.2) is inconsistent with the
2
44
multiplet observed in THFꢀd8. To investigate this, methanol
was sequentially added to 3 in THFꢀd , which instigated peak
8
thermal treatment: heating 3 (62 mM) in dioxaneꢀd to 100 °C
for 50 min in a sealed NMR tube furnished only 20% of the
broadening of one of the super imposed Ruꢀhydride signals in
the multiplet, leading to the triplet observed under catalytic
conditions (See SI 7.2). We interpret this as additional eviꢀ
dence for the interaction of methanol and 3 through hydrogenꢀ
bonding, Scheme 9. The NMR signals were also progressively
shifted to higher fields, which is further indication of dihydroꢀ
8
dehydrogenated Ruꢀamido complex 2, in parallel with H evoꢀ
lution. H was then shown to add back onto 2 to form 3 within
2
2
a few hours (See SI 7.1). The observed low conversion, as
well as the slow back reaction are consistent with theory,
which reveals that the reaction is slightly endergonic and can
form an equilibrium under an H atmosphere in favor of 3.
Yang and Lei and coꢀworkers calculated highly exergonic
reactions (11.9 and 7.6 kcal/mol, respectively), which do not
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
45
gen bonding. A comparable effect was reported by Schneider
46
and coꢀworkers.
2
19a
19b
Studies under catalytic conditions
Scheme 10. Possible metalꢀligand bifunctional catalysis through
ligand backbone deprotonation. P = PiꢀPr2.
support reversibility between 2 and 3 under a H atmosphere.
2
Scheme 9. Hydrogen generation via protonation of 3. P = PiꢀPr2.
As many high performing catalysts contain PNP ligands
5
with aliphatic backbones, we investigated the possible inꢀ
volvement of Ruꢀenamido complex 2-E in a mechanism proꢀ
30,47
posed using a related complex,
Scheme 10 (See SI 8.1). No
incorporation of deuterium into the ethylene bridges of the
1
backbone was detected ( H NMR) under the reaction condiꢀ
tions, as would be expected from such a pathway in deuterated
A solvent mediated process may facilitate H release from
2
1
9a
complex 3, where MeOH (or H O) shuttles a proton from
nitrogen to the hydride, presumably proceeding through interꢀ
mediates I and II, Scheme 9. The barrier for the solvent asꢀ
sisted H loss from 3 was then considered: the direct addition
of H to 2 reduced to 13.71 kcal mol with MeOH assistance
compared to 18.76 without. After the addition of one equivaꢀ
lent of MeOH to a solution of 3 in THFꢀd 4 appeared and the
release of H gas was observed (See SI 7.2). However, despite
the relative amount of 4 increasing with further equivalents of
added methanol, incomplete conversion (78%) was observed,
even with 50 equivalents. As indicated above, this reaction [3
+ CH OH = 4 + H ] is endergonic by 3.19 kcal mol , thermoꢀ
dynamically unfavorable, and thus a large excess of methanol
2
solvents. In addition, the isomerization between 2 and 3-I was
ꢀ1
calculated to be endergonic by 8.83 kcal mol . Thus, we conꢀ
4
2
clude this pathway to be unfeasible.
2
Stoichiometric reactions of the Ruꢀamido complex 2
demonstrated that methanol, water and formic acid can readily
add across the RuꢀN bond, affording the corresponding 18ꢀ
electron monohydride species 4, 5 and 6, respectively. The
more acidic the protic compound, the more stable the monoꢀ
hydride adduct was formed. Addition of base to 6 under aproꢀ
tic conditions reversed this reaction back to 2. However, under
ꢀ
1
2
8
2
the basic catalytic dehydrogenation conditions (MeOH:H O
ꢀ
1
2
3
2
(9:1), KOH 8 M) it was never possible to either observe the
characteristic yellow color of 2 or spectroscopically detect this
complex. Two other species were detected under these condiꢀ
tions, albeit using an increased concentration of catalyst 1a (20
mM) and at room temperature (See SI 8.2). The first was
is necessary to fully convert 3 to 4.
4
3
Both the hydricity of the RuꢀH and the acidity of the proꢀ
ton source are important when the rate of H release from 3 is
considered. Consistently, by employing the more acidic formic
acid, quantitative formation of Ruꢀformate 6 was observed
with concomitant evolution of H gas. In agreement, this therꢀ
modynamically favorable reaction [3 + HCOOH ꢀ= 6 + H ]
was calculated to be exergonic by 14.70 kcal mol . In addiꢀ
tion, the cationic monohydride PNP Ru BF4 complex was
rapidly prepared by reaction of 3 with one equivalent of HBF4
2
1
thought to be the Ruꢀmonohydride methoxy species 4 ( H
31
NMR δ = – 18.17; P NMR: δ = 74.05) and the second possiꢀ
2
1
31
bly the Ruꢀdihydride 3 ( H NMR: δ = – 7.20; P NMR: δ =
88.64). Due to the nonꢀtrivial task of characterizing rapidly
equilibrating species under the reaction conditions, where hyꢀ
drogen bonding adds further complication, it was not possible
to unambiguously identify the monohydride species by specꢀ
troscopic means.
2
ꢀ
1
+
ꢀ
and release of H gas (See SI 7.3).
2
1
31
The difference between the H and P NMR chemical
NMR spectra showed that the ratio of 3 and 4 was unꢀ
changed from room temperature to 90 °C. With continued
heating, free formate was detected, alongside monohydride
shifts of complex 3 (in THFꢀd ) and those observed under the
reaction conditions (MeOH:H O (9:1), 8 M KOH, 91 °C) is
within the reasonable range expected from such a dramatic
8
2
13
Ruꢀformate 6 ( P NMR: δ = 74.64) (See SI 8.2). A precipitate
8
ACS Paragon Plus Environment

Journal of the American Chemical Society
was also generated after prolonged heating (60 min), which
Page 10 of 18
1
3
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
was isolated, analyzed ( C NMR) and found to contain a mixꢀ
48
ture of HCOOK and K CO /KHCO , i.e. products of partial
2
3
3
and full methanol dehydrogenation, respectively.
Despite efficient hydrogen release from preꢀformed 3 unꢀ
der acidic conditions, vide supra, catalysis is most efficient
under highly basic conditions, implying a high pH is necessary
to turnꢀover the catalytic cycle. To test for the influence of
KOH on the ratio of catalyst resting states, its concentration
was incrementally increased under catalytic conditions and the
1
proportion of dihydride to monohydride was recorded by H
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NMR (See SI 8.3). 3 only appeared after 3.4 M KOH had been
added. Raising the concentration beyond 3.4 M further inꢀ
creased the proportion of 3 to 4 by reducing its propensity for
protonation. When a strong base was added (2 equiv. tBuOK)
to a solution of 4 under aprotic conditions, full conversion to 3
was observed.
Mechanistic proposal for complex 1a
Crucial for this dehydrogenation process are the CꢀH cleavꢀ
age and the transfer of hydride from methanol, gemdiol or
formate to the Ru center. Several scenarios can be envisaged
for these elementary steps (shown for methanol in Scheme
The complexity observed in the hydride multiplet of 3 in
THFꢀd , could also be removed through the addition of strong
8
base. When BuLi was added to 3, the amine in the pincer ligꢀ
and was deprotonated to give the trans Ruꢀdihydride amidate
1
2). Previously, we had proposed a direct outerꢀsphere addiꢀ
9,19
tion of methanol onto 2, Scheme 3 and Scheme 12 (path A).
Based on our current findings, we now believe that this pathꢀ
way is unlikely to occur: not only have we shown that 4 is rapꢀ
idly formed from 2 in the presence of MeOH, but also that the
addition of formic acid to 2 exclusively gives Ruꢀformate
complex 6 without sign of dihydride 3 formation. Moreover, it
was not possible to computationally locate a transition state
for the direct process A and there is also no reasonable raꢀ
tionale for the dependency of the rate on the base concentraꢀ
tion.
-
complex (3 ) (See SI 8.4). Disappearance of the backbone NH
peak was accompanied by simplification of the hydride multiꢀ
plet that was slightly shifted upꢀfield; all features consistent
-
-
with formation of 3 . Protonation of 3 with water directly led
back to 3. Based on kinetic studies, Ruꢀamidate complexes
have been postulated in ketone hydrogenations using Noyoriꢀ
49
type catalysts. The increased hydride nucleophilicity in the
corresponding anionic catalysts, which were detected at low
temperature, promotes hydrogenation of the less reactive amꢀ
50
ides and imides. Thus, the evidence presented here suggests
Taking into account the intermediacy of 4, an alternative
pathway B could involve a nonꢀtraditional βꢀhydride eliminaꢀ
tion pathway. However, not only were we unable to locate a
transition state, it is unprecedented for this type of complex
and again does not explain the role of base. Hence, we considꢀ
ered pathways that are more likely to occur in the high pH enꢀ
vironment necessary for optimal catalytic activity. Deprotonaꢀ
tion of 4 may occur at the methoxide ꢀCꢀH (pathway C) or the
backbone NꢀH. Pathway C leads directly to anionic complex
the high pH of the catalytic conditions will initiate backbone
NH deprotonation. This is consistent with the slight discrepanꢀ
cy observed in the chemical shifts between the reaction condiꢀ
tions and the protonated material. By DFT we calculated the
energetics of NH backbone deprotonation in the resting states
and found the deprotonation to be barrierꢀless and extremely
exergonic, Scheme 11. The pK of the backbone NꢀH in 4 and
a
3
was calculated (PBE0ꢀNL/def2ꢀTZVPP) to be 9.09 and 8.24,
51
respectively. With a steadyꢀstate pH of between 10ꢀ13, this
clearly indicates that the catalyst is largely deprotonated.
Observations from these in situ experiments thus suggest
the existence of two potential wells in the multiꢀstep process,
8
, as no intermediate could be located due to rapid formaldeꢀ
hyde dissociation. However, backbone NꢀH deprotonation
-
leading to 4 was found to be much more exergonic than CꢀH
ꢀ
1
-
-
-
-
deprotonation (ꢀ23.19 vs. 13.23 kcal mol ), thus disfavoring
pathway C. The high concentration of KOH (pH >10) also
means there will be a substantial proportion of methoxide preꢀ
sent that directly forms 4 from 2. 4 may initiate CꢀH cleavage
of the bound methoxide with ensuing formaldehyde loss
(pathway D). However, this reaction (4 = 8 + CH O) is highly
endergonic (33.45 kcal mol ) and no transition state for the βꢀ
hydrideꢀtype elimination could be located. Moreover, the proꢀ
ton shift from nitrogen to Ru that generates resting state 3
from 8 would require overcoming a barrier of 6.67 kcal mol .
with deprotonated monohydrides 4 , 5 or 6 and dihydride 3
as catalytic resting states. Only Ruꢀformate complex 6 was
recovered from the reaction mixture, which reꢀconfirmed its
relative stability to 4. When the dehydrogenation of aqueous
-
-
1
3
13
methanol was carried out using CH OH, Cꢀ6 was isolated,
3
ꢀ
13
2
where the C label was only incorporated into the formate
group and was not detected in the coordinated CO (See SI
8.5).
ꢀ
1
-
ꢀ
1
ꢀ
1
Scheme 11. DFT calculated energy (kcal mol ) of anionic comꢀ
plex formation. P = PiꢀPr2.
Additional efforts to locate a more viable reaction pathway
focused on the frontier orbital interaction between 2 and
methoxide, gemꢀdiolate and formate (See SI 14). This analysis
demonstrates that both the negatively charged O atom and the
HꢀC atom can coordinate to the Ru center of complex 2
12b,40i
20
(pathway E), similar to that proposed with Fe,
and Ir,
and that the two isomers exist in a dynamic equilibrium. The
Oꢀcoordination of methoxide to 2 generating 4 was found to
−
9
ACS Paragon Plus Environment

Page 11 of 18
Journal of the American Chemical Society
ꢀ
1
−
be exergonic by 23.19 kcal mol , whereas the Hꢀcoordinated
For the Oꢀcoordination of formate to 2 leading to 6 , the
−
ꢀ1
ꢀ1
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
isomer 4H is less stable by 8.82 kcal mol , Scheme 13. The
barrier for CꢀH cleavage and formation of Ruꢀdihydride 3 and
formaldehyde is only 2.58 kcal mol with an overall effective
barrier of 11.40 kcal mol . In attempts to locate lower energy
pathways, we considered cyclic transition states with bridging
solvent molecules for pathways A, B, D and E. However, we
were unable to locate such structures. Nevertheless, pathway E
represents an energetically viable pathway that is consistent
with the experimental results.
The Oꢀcoordination of gemꢀdiolate 7 to 2 was found to be
exergonic by 6.04 kcal mol , with the Hꢀcoordinated isomer
H more stable by 3.47 kcal mol , Scheme 13. The barrier
for the CꢀH bond dissociation leading to formic acid and 3 is
only 0.58 kcal mol . When considering a hydrogen bonding
interaction between the OH group and the ligand backbone N
atom for 7H , this step becomes barrierꢀless; directly forming
formate and 3 in a highly exergonic (26.11 kcal mol ) reaction
reaction is exergonic by 7.15 kcal mol , with the Hꢀ
coordinated isomer 6H less stable by 2.25 kcal mol . The
barrier for the CꢀH bond dissociation leading to CO and 3 is
13.09 kcal mol , with an overall effective barrier of 15.34 kcal
mol . The potential energy surface shown in Figure 10 implies
that there should be two main resting states, 4 and 3 , which
are detected spectroscopically, vide supra. 6 was also obꢀ
served ( H and P NMR) throughout the reaction. As 6 faces
-
−
ꢀ1
ꢀ
1
ꢀ
2
ꢀ
1
ꢀ1
ꢀ
1
-
-
-
1
31
-
ꢀ
1
a CꢀH cleavage barrier of 15.34 kcal mol , it is a reasonable
-
species to detect in situ.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ
1
Our computed barrier for the direct and methanol promotꢀ
ed reaction of 2 + H = 3 is 18.76 and 13.71 kcal mol , respecꢀ
tively, and is exergonic by 2.31 kcal mol . While Yang calcuꢀ
lated the corresponding barrier as 29.5 and 21.8 kcal mol ,
respectively, with an endergonic reaction of 11.9 kcal mol ,
Lei and coꢀworkers showed the barrier to be 27.1 and
21.0 kcal mol , respectively and endergonic by 7.6 kcal mol
−
ꢀ1
ꢀ1
7
2
−
ꢀ1
ꢀ1
ꢀ1
ꢀ1
−
ꢀ1
ꢀ1
ꢀ
1
19
−
−
−
.
Our experimental observations, such as the reversibility
between 3 and 2 under H atmosphere and the facile formation
[
2+ OCH OH = 3 + HCO ]. The fact that 7H is predicted to
2 2
-
be more stable than 7 , combined with low activation barriers
for its further reaction, suggest this step should be rapid and
highly favorable. This is inꢀline with experimental observaꢀ
2
of 3 from 2, support the computed exergonic property of the
reaction. For the first dehydrogenation step that leads to forꢀ
maldehyde, the obtained effective barriers are largely in
agreement with those of Yang and Lei and coꢀworkers. Howꢀ
ever, for the dehydrogenation of hydroxymethanolate, higher
barriers were calculated from Yang and Lei and coꢀworkers
that do not agree with our experimental and theoretical results
0.58 vs 14.3 and 5.2 kcal mol , respectively). For the last deꢀ
hydrogenation step, we found an effective barrier of 13.54
kcal mol , while Lei and coꢀworkers reported one of about 31
kcal mol , and Yang of 23.4 kcal mol . Yang found the Oꢀ
coordinated formate (6 ) to be more stable than the Hꢀ
coordinated (6H ) by 7.9 kcal mol , which is larger than our
value of 2.25 kcal mol . The closer agreement between the
experimental results and calculations provides confidence in
our computational models and methods.
tions, as the gemꢀdiolate (or CH O) has never been detected
spectroscopically in situ, nor has any complex containing it,
2
-
-
i.e., 7 or 7H . Formaldehyde was indeed tested for (Merck
MColortest) under the reaction conditions and returned a negaꢀ
tive result (See SI 9.1), while a reaction solution containing
added formaldehyde returned a positive test (>1.5 mg/L). The
possibility of rapid, uncatalyzed, baseꢀinduced formaldehyde
decomposition was considered as a possible rationale for the
negative result. However, negligible volumes of gas were
formed when formaldehyde was added to a standard reaction
in the absence of precursor 1a (See SI 9.2). Thus, we are conꢀ
fident the DFT calculations are correct in modelling this cataꢀ
lytic step to be very rapid indeed.
ꢀ
1
(
ꢀ1
5
2
ꢀ1
ꢀ1
-
-
ꢀ1
ꢀ
1
Scheme 12. Possible pathways for the key step involving CꢀH bond cleavage and Ruꢀdihydride formation. P = PiꢀPr . Energies are
given in kcal mol .
2
ꢀ
1
10
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ
1
Scheme 13. DFT calculated energies (kcal mol ) for pathway E. P = PiꢀPr .
2
11
ACS Paragon Plus Environment

Page 13 of 18
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Potential energy surface for the key step in pathway E involving CꢀH bond cleavage and Ruꢀdihydride formation. Energies are
ꢀ1
-
given in kcal mol . Protonations of 3 to 3 in each step are with CH OH, HOCH OH and HCOOH, respectively.
3
2
Based on the experimental and theoretical findings, a cataꢀ
lytic cycle for methanol dehydrogenation can now be proꢀ
posed, Scheme 14, that includes a justification for the highly
basic conditions. Specifically, we can summarize three major
roles for KOH:
Scheme 14. The proposed catalytic cycle for aqueous methanol
reforming. P = iꢀPr .
2
ꢀ
Firstly, it is required to dehydrochlorinate the precataꢀ
lyst 1a
ꢀ
Secondly, formation of the key Hꢀcoordinated intermeꢀ
-
-
-
diates (4H , 7H and 6H ) is found to readily occur from
-
-
-
deprotonated Oꢀcoordinated intermediates 4 , 7 and 6
Lastly, the key dehydrogenation step that produces Ruꢀ
ꢀ
-
-
dihydride (3 ) species from the monohydride species (4
,
-
-
7 and 6 ) is promoted by base. By sequestration of
formaldehyde, formic acid and CO byꢀproducts the
2
high base concentration renders these steps thermodyꢀ
namically feasible and drives the reaction forward.
-
Following the formation of 3 , we propose protonation at
nitrogen to occur, followed by a MeOHꢀassisted H eliminaꢀ
2
tion. Depending on the nature of the acid, direct protonation
may also be a possibility (grey pathway, Scheme 14). Interestꢀ
ingly, the innerꢀsphere CꢀH cleavage step does not involve ligꢀ
and participation any more than acting as a strongly donating
anionic ligand, a finding that might be critical in the developꢀ
ment of new active catalyst systems.
Mechanistic proposal with Ru-complex Me-1a
-
-
As shown in Scheme 14, Nꢀdeprotonated complexes 4 , 6 ,
-
and 7 are key intermediates in the catalytic cycle. We also
demonstrated that Me-1a is a proficient catalyst for aqueous
basic MeOH dehydrogenation, vide supra. Although its cataꢀ
lytic activity is significantly lower, Me-1a can catalyze all
three MeOH dehydrogenation steps, as both formate and carꢀ
bonate were detected after allowing the reaction to reach high
conversion. Obviously, these observations cannot be rationalꢀ
ized on the basis of our proposed mechanism. Thus, stoichioꢀ
metric studies and theoretical calculations were performed usꢀ
ing Me-1a.
In a typical catalytic experiment under the standard reacꢀ
tion conditions (MeOH:H O (9:1), KOH 8M) only one species
2
1
2
was observed ( H NMR: δ = ꢀ6.2 (t, J = 17.4 Hz), ꢀ6.7 (t,
JHP = 20.8 Hz); P{ H} NMR: δ = 87.5 (s)) that we suspected
HP
2
31
1
was Ruꢀdihydride Me-3 (See SI 10.1). To confirm the identity
of this species, Me-3 was independently prepared, Scheme 15
21d
(See SI 10.2). The compound was obtained as a mixture of
12
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 18
two isomers in a 98:2 ratio, which, based on NMR data, were
undefined species to be either trans Me-3 hydrogen bonded to
MeOH or a cationic Ruꢀdihydrogen complex. However, even
with 50 equivalents of MeOH, no H was observed to evolve
1
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
identified as trans Me-3 ( H NMR (tolueneꢀd ): δ = ꢀ 6.02 (tm,
8
3
3
31
1
JHP = 20.0 Hz), ꢀ5.75 (tm, J = 18.5 Hz); P{ H} NMR
2
HP
(tolueneꢀd ): δ = 89.59 (s)) and one of the two possible cis
(See SI 10.4). Indeed, theory predicts that protonation of Ruꢀ
dihydride Me-3 by MeOH to afford the corresponding rutheꢀ
nium methoxide Me-4 (or Me-4’) is endergonic by 8.02 (or
8
1
isomers, cis Me-3 and cis Me-3' ( H NMR (tolueneꢀd ): δ = ꢀ
15.77 (td, JHP = 18.6 Hz, J = 5.2 Hz), ꢀ7.13 (td, JHP
21.7 Hz, JHH = 5.2 Hz); P{ H} NMR (tolueneꢀd ): δ =
8
=
HH
1
ꢀ
1
3
1
6
.94 kcal mol ), Figure 11.
8
The key dehydrogenation step is the reformation of Me-3
86.78 (s)) that depend on the relative orientation of methyl
and hydride (‘ = cis). The spectroscopic data indicate the presꢀ
ence of two equivalent phosphorus donors cis orientated to the
hydrides. The hydrides in trans Me-3 have very close chemiꢀ
cal shifts, whereas the two hydrides in the cis isomers are very
different, due to the differing trans influence being exerted.
The NMR data recorded under the reaction conditions are
most similar to that of trans Me-3, and thus we propose this to
be the resting state of the catalytic cycle.
from Me-4. One possibility would be to proceed via βꢀhydride
ꢀ
elimination, where the coordinatively saturated 18ꢀe species
54,55,56
requires temporary decoordination at a cis site.
However,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
this seems highly unlikely considering the stability of the pinꢀ
cer and CO ligands. Indeed, no evidence for this could be
gained from NMR. A dissociative βꢀhydride abstraction is able
to circumvent this problem. It should be accelerated in a polar
20,27d,57
medium by promoting alkoxide dissociation,
however,
DFT was not able to find a suitable low energy pathway and
this would not fully explain the experimental data.
Under neutral conditions, a similar mechanism for 4 to 3 in
the unmethylated catalyst, via Hꢀcoordination of substrate to
the Ruꢀcenter, provided a viable route for CꢀH cleavage and
hydride delivery, Figure 11. The overall dehydrogenation of
Scheme 15. Preparation of transꢀ and cis Me-3. P = PiꢀPr2.
methanol into formaldehyde and H is calculated to be enderꢀ
2
ꢀ
1
gonic by 13.72 kcal mol and the corresponding effective barꢀ
ꢀ
1
Scheme 16. Hydridic H/D exchange in Me-3 upon reaction with
rier is 22.40 kcal mol . In contrast, the dehydrogenation of
methandiol (to HCOOH and H ) and formic acid (to CO and
CH OD. P = PiꢀPr .
3
2
2
2
H ) are predicted to be exergonic processes by ꢀ6.84 and ꢀ
2
ꢀ
1
1
0.68 kcal mol , respectively, Figure 11. Protonation of Me-3
by methanediol resulting in the formation of Me-7 or Me-7’
ꢀ1
is, however, endergonic by 5.52 or 4.22 kcal mol , respectiveꢀ
ly. The effective barrier for this second step is calculated to be
ꢀ1
1
3.62 kcal mol , although the Hꢀcoordinated ruthenium methꢀ
oxide species Me-7H could not be located. Protonation of Me-
by formic acid to Me-6 and Me-6’ was calculated to be exꢀ
3
Protonation of Me-3 by alcohol is an important step in the
ꢀ
1
53
ergonic by 7.77 and 8.63 kcal mol , respectively, Figure 11.
The observation that one equivalent of formic acid is sufficient
catalytic cycle. An excess (2.8 equiv.) of CH OD was added
3
to a sample of Me-3 in C D (See SI 10.3). Two new species
6
6
to promote H liberation and quantitative conversion of Me-3
were formed whose spectroscopic data are consistent with
2
2
1
to Me-6 is inꢀline with this theory. Me-6 was prepared as a
monodeuterated trans[ H]-Me-3 ( H NMR (C D ): δ = ꢀ 5.64
6
6
1
3
2
1
mixture of two isomers (major: 81%, H NMR (tolueneꢀd ): δ
(
t, J = 18.1 Hz)) and trans [ H]-Me-3’ ( H NMR (C D ):
8
HP
6
6
31
1
3
= ꢀ17.26 (t, JHP = 19.4 Hz); P{ H} NMR (tolueneꢀd ): δ =
70.56 (s); minor: 19%, H NMR (tolueneꢀd ): δ = ꢀ17.16 (t, JHP
= 18.7 Hz); P{ H} δ = 72.95 (s)) depending on the relative
orientation of NMe and formate (See SI 10.5). Indeed, the calꢀ
δ = ꢀ 5.96 (t, J = 20.0 Hz)) that arise from H/D exchange
8
HP
1
between Me-3 and CH OD, Scheme 16. Depending on the relꢀ
8
3
31
1
ative orientation of NMe and D, the two monodeuterated isoꢀ
mers were formed in equimolar amounts, indicating that the
rates of exchange are equal no matter the orientation of NMe.
This is in contrast to unmethylated 3, where, through hydrogen
bonding, NꢀH orientation was found to be highly influential on
ꢀ
1
culated free energy difference of 0.86 kcal mol corresponds
to a ratio of 81 to 19 in favour of Me-6', in perfect agreement
with experiment. Heating Me-6 for 3 hours at 90 °C triggered
partial (42%) conversion to Me-3 (cisꢀ 19% and trans – 81%)
and gas evolution, experimentally validating the lower barrier
4
6
ꢀ
the process. Coordinatively saturated 18e Me-3 must protoꢀ
nate to form a transient and undetected Ruꢀdihydrogen comꢀ
plex. Indeed, in the presence of added methanol the hydride
ꢀ1
(about 8 kcal mol ) found for this process (See SI 10.6). Our
1
signals of trans Me-3 in benzene are both shifted upꢀfield ( H
calculations suggest that methanol dehydrogenation to formalꢀ
dehyde is the least facile of the three steps when catalysed by
Me-1a preꢀcatalyst. This is in contrast to the process promoted
by 1a, in which formic acid dehydrogenation is the least facile
step.
31
NMR: ꢁδ = ca. ꢀ0.10, P NMR: ꢁδ = ca. ꢀ0.25) (See SI 10.4).
Such changes are typical of dihydrogen bond formation and
result from the fast equilibrium between the dihydrogenꢀ
45
bonded complex and free trans Me-3.
As MeOH is only a weak acid, a large excess of alcohol is
necessary to shift the equilibrium towards protonation of the
The base is clearly necessary to promote the dehydrogenaꢀ
tion with Me-1a, as no reaction is observed in its absence. As
5
3
metal hydride and to eliminate H2. Attempts to displace the
well as sequestering the formaldehyde, formic acid and CO
2
equilibrium towards H elimination by heating to 90 °C were
byꢀproducts and providing a thermodynamic driving force,
KOH is likely to be involved in the keyꢀstep. Indeed, Me-3
was readily generated from Me-1a after treatment with KOMe
in toluene (See SI 10.7).
2
not effective. Along with the observed H/D exchange process,
two new species were observed when increasing quantities of
MeOH were added to Me-3 in tolueneꢀd . We propose these
8
13
ACS Paragon Plus Environment

Page 15 of 18
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ1
Figure 11. Potential energies (kcal mol ) for the key CꢀH bond cleavage step and Ruꢀdihydride formation under neutral and basic condiꢀ
tions. Only showing energies for the trans hydride/methyl monohydride isomer. P = iꢀPr2.
F
In addition, the reaction of cationic Me-1-BAr (obtained
from the reaction of Me-1a with Na[BAr ] (Ar = 3,5ꢀ
C H (CF ) ) ( H NMR (THFꢀd ): δ = ꢀ21.6 (bs); P{ H} NMR
served after the addition of 200 eq. of KOH (See SI 8.3 and
10.8). This enhanced stability towards protonation is one reaꢀ
son why the rates of methanol dehydrogenation employing
Me-1a are slower than with 1a. Thus, at high pH, the reaction
is attenuated by the stability of Me-3 towards protonation and
can explain the dropꢀoff in rate that is observed, Figure 5. The
stabilization of 3 by high pH is much less pronounced than
Me-3, and consequently this counteracting decrease in rate is
present but much less pronounced.
4
F
4
F
1
31
1
6
3
3 2
8
(
THFꢀd ): δ = 67.1 (bs)) with 5 equivalents of MeOH in THFꢀ
8
1
d only afforded a weakly coordinated cationic adduct ( H
8
31
1
NMR (THFꢀd ): δ = ꢀ20.5 (bs); P{ H} NMR (THFꢀd ): δ =
8
8
6
8.1 (bs)) that does not undergo any CꢀH cleavage or hydride
delivery. Only after the addition of KOMe did Me-3 rapidly
form. Thus, Me-3 is formed more readily in the presence of
methoxide and explains the increase in rate observed at lower
base concentrations, Figure 5.
To further rationalize these observations, we considered
possible promotional roles of base in the keyꢀstep and calcuꢀ
lated a route more favorable than the neutral pathway, Figure
11. Thus, starting from Me-4, deprotonation of the bound ꢀ
OCH group was shown to result in the formation of Me-4 ,
where the coordination switches from RuꢀOCH to RuꢀCH O.
Me-3 is generated following rapid CH O dissociation. Ruꢀ
OCH (Me-4) deprotonation and CH O dissociation are enderꢀ
gonic by 1.13 and 1.87 kcal mol , respectively. The formed
CH O can then be consumed by base into hydroxymethanoꢀ
late. In a similar manner, deprotonation of the CꢀH bond of the
bound ꢀOꢀCH ꢀOH group within Me-7 shifts the coordination
from Oꢀ to Cꢀ, which results in formation of Me-7 and subseꢀ
quent dissociation of HCOOH to generate Me-3 . RuꢀOCH ꢀ
OH deprotonation and HCOOH dissociation is exergonic by
CONCLUSIONS
In conclusion, we have characterized the reactivity of cataꢀ
lyst 1a and Me-1a in aqueous methanol dehydrogenation and
proposed mechanisms based on our spectroscopic, experiꢀ
mental and theoretical investigations. At constant temperature,
rates of reaction increased with increasing KOH concentraꢀ
tions for catalyst 1a. A range of Ru complexes that are possiꢀ
ble catalytic intermediates were independently prepared, isoꢀ
lated and characterized by spectroscopy and Xꢀray crystallogꢀ
raphy, and their reactivity examined under aprotic conditions.
Ruꢀamido 2 was highly reactive with MeOH, formic acid and
water and provided mono and dihydride Ru complexes. These
hydride complexes were found to be in an apparent equilibriꢀ
um that could be perturbed by base, to dihydride 3, or acid, to
monohydride 4. Under catalytic conditions, the resting states
-
3
2
2
-
2
3
2
ꢀ
1
2
2
-
-
2
-
-
were shown to be the NꢀH deprotonated 3 and 4 complexes.
In addition to experimentally disproving a number of other
pathways, DFT rationalized a full mechanistic cycle involving
these anionic species in the key step. Thus, it was demonstratꢀ
ed that the ligand does not play a cooperative role in the innerꢀ
sphere CꢀH cleavage step.
Preꢀcatalyst Me-1a was found to be active, albeit less than
1a. A number of possible methylated intermediates were preꢀ
pared and their reactivity investigated. Me-3 was found to be
more stable than 3 to protonation, which accounts for the lowꢀ
er rates observed under the standard reaction conditions. The
rate of dehydrogenation increased with added KOH up to 5 M,
ꢀ
1
1
.35 and 11.71 kcal mol , respectively. Starting from Me-6,
an analogous route was located, wherein deprotonation of the
Ru–OCHO group accompanied Oꢀ to Cꢀ coordination exꢀ
change and is exergonic by 15.87 kcal mol . Interestingly,
CO dissociation was found to be endergonic by 10.26 kcal
ꢀ
1
2
ꢀ
1
mol , indicating very strong CO coordination.
2
The activity of Me-1a was shown to increase with increasꢀ
ing base concentration up to 4 M KOH, after which it dropped,
Figure 5. This “bellꢀshaped” behavior indicates that base is
playing at least two competing roles. Me-3 was found to be
much more stable than 3 towards protonation and was obꢀ
14
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 18
but dropped with higher concentrations. KOH is essential to
Ishida, K.; Kobayashi, T.; Sayo, N.; Saito, T. Org. Process Res. Dev.
2012, 16, 166 – 171.
11) (a) RodriguezꢀLugo, R. E.; Trincado, M.; Vogt, M.; Tewes, F.;
SantisoꢀQuinones, G.; Grützmacher, H. Nature Chem. 2013, 5, 342 –
47; (b) Monney, A.; Barsch, E.; Sponholz, P.; Junge, H.; Ludwig, R.;
Beller, M. Chem. Commun. 2014, 50, 707 – 709; (c) Hu P.; Diskinꢀ
Posner Y.; BenꢀDavid Y.; D. Milstein ACS Catal. 2014, 4, 2649 –
2652; (d) Heim, L. E.; Thiel, D.; Gedig, C.; Deska, J.; Prechtl, M. H.
G. Angew. Chem. Int. Ed. 2015, 54, 10308 – 10312.
(12) (a) Alberico, E.; Sponholz, P.; Cordes, C.; Nielsen, M.;
Drexler, H.ꢀJ.; Baumann, W.; Junge, H.; Beller, M. Angew. Chem. Int.
Ed. 2013, 52, 14162 – 14166; (b) Bielinski, E. A.; Förster, M.; Zhang,
Y.; Bernskoetter, W. H.; Hazari, N., Holthausen, M. C. ACS Catal.
2015, 5, 2404 – 2415.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ensure methoxide is present in solution, from which hydride
transfer to Ru proceeds to afford dihydride Me-3. However, at
higher KOH concentrations, the stability of Me-3 is too high
and its protonation rate decreases.
We anticipate these results to be particularly valuable in
the development of new catalysts that can operate at lower
base concentrations and temperatures in this important reacꢀ
tion.
(
3
ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ASSOCIATED CONTENT
Experimental and computational details are supplied as Supportꢀ
ing Information. This material is available free of charge via the
Internet at http://pubs.acs.org.
(
13) Selected references: (a) Gunanathan, C.; Milstein, D. Chem.
Rev. 2014, 114, 12024 – 12087; (b) Annibale, V. T.; Song, D. RSC
Adv. 2013, 3, 11432 – 11449; (c) Luca, O. R.; Crabtree, R. H. Chem.
Soc. Rev. 2013, 42, 1440 – 1459; (d) Zhao, B.; Han, Z.; Ding, K. An-
gew. Chem. Int. Ed. 2013, 52, 4744 – 4788 (e) Schneider, S.; Meiners,
J.; B. Askevold, Eur. J. Inorg. Chem. 2012, 412ꢀ429; (f)
Lyaskovskyy, V.; de Bruin, B. ACS Catal. 2012, 2, 270ꢀ279; (g)
Gunanathan, C.; Milstein, D. Acc. Chem. Res. 2011, 44, 588ꢀ602; (h)
van der Vlugt, J. I. ; Reek, J. N. H. Angew. Chem. Int. Ed. 2009, 48,
AUTHOR INFORMATION
Corresponding Author
* Matthias.beller@catalysis.de
‡
These authors contributed equally.
8
832 – 8846.
14) (a) Noyori, R.; Yamakawa, M.; Hashiguchi, S. J. Org. Chem.
(
ACKNOWLEDGMENT
2001, 66, 7931 – 7944; (b) Yamakawa, M., Ito, H., Noyori, R. J. Am.
Chem. Soc. 2000, 122, 1466ꢀ1478; (c) Khusnutdinova, J. R.; Milstein,
D. Angew. Chem. Int. Ed. 2015, 54, 12236 – 12273, Angew. Chem.
2015, 127, 12406 – 12445.
EA would like to thank the DAAD, Programme n. 50015559, for
a scholarship. AJJL would like to thank the Alexander von Humꢀ
boldt Foundation for generous funding. LKV would like to thank
the VCI for a scholarship (n. 196241). MN would like to thank the
Marie Curie FP7 International Outgoing Fellowship (IOF n.
(15) Li, H.; Hall, M. B. J. Am. Chem. Soc. 2015, 137, 12330 –
1
2342.
(16) (a) Campos, J., Sharninghausen, L. S.; Manas, M. G.; Crabtree,
3
1
27565) and the Danish Council for Independent Research (n. 12ꢀ
31997) for financial support.
R. H. Inorg. Chem. 2015 54, 5079 – 5084.
17) Fujita, K.; Kawahara, R.; Aikawa T.; Yamaguchi, R. Angew.
Chem. Int. Ed. 2015, 54, 9057 – 9060.
18) Handgraaf, J. W., Meijer, E. J. J. Am. Chem. Soc. 2007, 129,
099 – 3103.
19) (a) Yang, X. ACS Catal. 2014, 4, 1129 – 1133; (b) Lei, M.;
Pan, Y.; Ma X. Eur. J. Inorg. Chem. 2015, 794 – 803.
20) Blum, O.; Milstein, D. J. Organomet. Chem. 2000, 593 – 594,
79 –484.
21) (a) Kothandaraman, J.; Goeppert, A.; Czaun, M.; Olah G. A.;
(
REFERENCES
(
(1) Armaroli, N.; Balzani, V. Chem. Asian J. 2011, 6, 768 – 784.
(2) Armaroli, N.; Balzani, V. ChemSusChem 2011, 4, 21 – 36.
3
(
(
3) Bagotsky, V. S. Fuel Cells: Problems and Solutions, 2nd Ed.,
WileyꢀVCH, Weinheim, 2012.
4) (a) Dalebrook, A. F.; Gan, W.; Grasemann, M.; Moreta, S.; Lauꢀ
(
(
4
renczy, G. Chem. Commun. 2013, 49, 8735 – 8751; (b) Eberle, U.;
Felderhoff, M.; Schüth, F. Angew. Chem. Int. Ed. 2009, 48, 6608 –
6630.
(
Prakash G. K. S. J. Am. Chem. Soc. 2016, 138, 778 – 781; (b)
Kothandaraman, J.; Czaun, M.; Goeppert, A.; Haiges, R.; Jones, J.ꢀP.;
May, R. B.; Prakash, G. K. S.; Olah, G. A. ChemSusChem 2015, 8,
(
5) (a) Trincado, M.; Banerjeea, D.; Grützmacher, H. Energy Envi-
ron. Sci., 2014, 7, 2464ꢀ2503; (b) Brethauer, S.; Studer, N. H. Chimia,
015, 69, 572 – 581; (c) Goeppert, A.; Czaun, M.; Jones, J.ꢀP.; Praꢀ
kash, S. G. K.; Olah G. A. Chem. Soc. Rev. 2014, 43, 7995 – 8048.
6) (a) Olah, G. A.; Goeppert, A.; Prakash, S. G. K. Beyond Oil and
1
1
442 – 1451; (c) Choi, J.ꢀH.; Prechtl, M. H. ChemCatChem 2015, 7,
023 – 1028; (d) Choi, J.ꢀH.; Heim, L. E.; Ahrens, M.; Prechtl, M. H.
2
G. Dalton Trans 2014, 43, 17248 – 17254; (e) Marziale, A. N.;
Friedrich, A.; Klopsch, I.; Drees, M.; Celinski, V. R.; Schmedt auf der
Günne J.; Schneider, S. J. Am. Chem. Soc. 2013, 135, 13342 – 13355;
(f) Han, Z.; Rong, L.; Wu, J.; Lei, Z,; Wang Z.; Kuiling Ding, K.
Angew. Chem. Int. Ed. 2012, 51, 13041 – 13045; (g) Spasyuk, D.;
Gusev. D. G. Organometallics 2012, 31, 5239 – 5242; (h) Zhang, L.;
Raffa, G; Nguyen, D. H.; Swesi, Y.; CorbelꢀDemailly, L.; Capet, F.;
Trivelli, X.; Desset, S.; Paul, S.; Paul, J.ꢀF.; Fongarland, P.;
Dumeignil, F.; Gauvin, R. M. J. Catal. 2016, 340, 331 – 343.
(
nd
Gas: The Methanol Economy, 2 . Ed., WileyꢀVCH, Weinheim, 2009;
b) Olah, G. A.; Prakash, S.G. K.; Goeppert A. J. Am. Chem. Soc.
(
2
2
011, 33, 12881 – 12898; (c) Olah, G. A. Angew. Chem. Int. Ed.
013, 52, 104 – 107 and references therein.
(
7) Thermochemical data have been calculated based on the Nist
Chemistry Webbook at http://webbook.nist.gov/chemistry/.
(8) (a) Palo, D. R.; Dagle, R. A.; Holladay, J. D. Chem. Rev. 2007,
107, 3992 – 4021; (b) Sá S.; Silva H.; Brandãoa L.; Sousa J. M.;
Mendes A. Appl. Catal. B: Environmental 2010, 99, 43 – 57.
(
22) (a) Ruibo, X.; Chakraborty, S.; Bellows, S. M.; Yuan, H.;
Cundari, T. R.; Jones, W. D. ACS Catal. 2016, 6, 2127 – 2135; (b)
Zhang, Y; MacIntosh, A. D.; Wong, J. L.; Bielinski, E. A.; Williard,
P. G.; Mercado, B. Q.; Hazari, N.; Bernskoetter, W. H. Chem. Sci.
(
9) Nielsen, M.; Alberico, E.; Baumann, W.; Drexler, H.ꢀJ.; Junge,
H.; Gladiali, S.; Beller, M. Nature 2013, 495, 85 – 89.
10) (a) Bertoli, M.; Choualeb, A.; Lough, A. J.; Moore, B.;
(
2
015, 6, 4291 – 4299; (c) Bornschein, C.; Werkmeister, S.; Wendt,
Spasyuk, D.; Gusev, D. G. Organometallics 2011, 30, 3479 – 3482;
(b) Nielsen, M.; Kammer, A.; Cozzula, D.; Junge, H.; Gladiali, S.;
Beller, M. Angew. Chem. Int. Ed. 2011, 50, 9593 – 9597; (c) Kuriyaꢀ
ma, W.; Matsumoto, T.; Ino, Y.; Ogata, O. (Takasago International
Corporation), PCT Int. Appl. WO/2011/048727 A1, 2011; (d) Kuꢀ
riyama, W.; Matsumoto, T.; Ogata, O.; Ino, Y.; Aoki, K.; Tanaka, S.;
B.; Jiao, H.; Alberico, E.; Baumann, W.; Junge, H.; Junge, K.; Beller,
M. Nature Communications 2014, 5: 4111; (d) Werkmeister, S.;
Junge, K.; Wendt, B.; Alberico, E.; Jiao, H.; Baumann, W.; Junge, H.;
Gallou, F.; Beller, M. Angew. Chem. Int. Ed. 2014, 53, 8722 – 8726;
(
e) Sharninghausen, L. S.; Mercado, B. Q.; Crabtree, R. H.; Hazari N.
Chem. Commun. 2015, 51, 16201 – 16204.
15
ACS Paragon Plus Environment

Page 17 of 18
Journal of the American Chemical Society
(
23) (a) Jing, Y.; Chen, X.; Yang, X.; Organometallics 2015, 34,
Muckerman,J. T.; Manbeck, G. F.; Fujita, E. Chem. Rev. 2015, 115,
12936 – 12973.
(40) CO hydrogenation: (a) Filonenko, G. A.; van Putten R.;
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5716 – 5722; (b) Zhang, G.; Vasudevan, K. V.; Scott, B. L.; Hanson,
S. K. J. Am. Chem. Soc. 2013, 135, 8668 – 8681; (c) Zhang, G.; Hanꢀ
son, S. K. Chem. Commun. 2013, 49, 10151 – 10153; (d) Xu, R.;
Chakraborty, S.; Yuan, H.; Jones, W. D. ACS Catal. 2015, 5,
350−6354; (e) Zhang, G.; Yin, Z.; Tan J. RSC Adv. 2016, 6, 22419 –
2423.
2
Schulpen E. N.; Hensen E. J. M.; Pidko E. A. ChemCatChem 2014, 6,
1526 – 1530; (b) Filonenko, G. A.; Conley, M. P.; Coperet, C.; Lutz,
M.; Hensen, E. J. M.; Pidko, E. A. ACS Catal. 2013, 3, 2522 – 2526;
(c) Huff, C. A.; Sanford, M. S. ACS Catal. 2013, 3, 2412 – 2416; (d)
Hull, J. F.; Himeda, Y.; Wang, W.ꢀH.; Hashiguchi, B.; Periana, R.;
Szalda, D. J.; Muckerman, J. T.; Fujita, E. Nat. Chem. 2012, 4, 383ꢀ
388; (e) Schmeier, T. J.; Dobereiner, G. E.; Crabtree, R. H.; Hazari,
N. J. Am. Chem. Soc. 2011, 133, 9274 – 9277; (f) Langer, R.; Diskinꢀ
Posner, Y.; Leitus, G.; Shimon, L. J. W.; BenꢀDavid, Y.; Milstein, D.
Angew. Chem. Int. Ed. 2011, 50, 9948 – 9952 (g) Yamashita, M.;
Nozaki, K.; J. Am. Chem. Soc. 2009, 131, 14168 – 14169; HCOOH
dehydrogenation: (h) Bielinski E. A.; Lagaditis P. O.; Zhang Y.;
Mercado B. Q.; Würtele C.; Bernskoetter W. H.; Hazari N.; Schneider
S. J. Am. Chem. Soc. 2014, 136, 10234 – 10237; (i) Zell T.; Butschke
B.; BenꢀDavid Y.; Milstein D. Chem. Eur. J. 2013, 19, 8068 – 8072;
(l) Tanaka R.; Yamashita M.; Chung L. W.; Morokuma K.; Nozaki
K.; Organometallics 2011, 30, 6742 – 6750.
6
2
(24) Vasudevan, K. V.; Scott, B. L.; Hanson S. K. Eur. J. Inorg.
Chem. 2012, 4898 – 4906.
25) GómezꢀGallego, M.; Sierra, M. A. Chem. Rev. 2011, 111, 4857
4963.
26) Reactant molecules, methanol and water, can effectively be
(
–
(
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
removed from the rate equation. The catalyst and base concentration
during the steady state period are also constant, making the kinetics
effectively pseudo zero order, as exemplified by a constant rate of gas
evolution in the shortꢀtoꢀmedium term
(
27) (a) Sandoval, C. A.; Ohkuma, T.; Muñiz, K.; Noyori, J. Am.
Chem. Soc. 2003, 125, 13490ꢀ13503; (b) Rautenstrauch, V.; Hoangꢀ
Cong, X.; Churlaud, R.; AbdurꢀRashid, K.; Morris, R. H. Chem. Eur.
J. 2003, 9, 4954 – 4967; (c) Hamilton, R. J.; Bergens, S. H. J. Am.
Chem. Soc. 2006, 128, 13700 – 13701; (d) Barattta, W.; Siega, K.;
Rigo, P. Chem. Eur. J. 2007, 13, 7479 – 7486; (e) Baratta, W.; Baldꢀ
ino, S.; Calhorda, M. J.; Costa, P. J.; Esposito, G.; Herdtweck, E.;
Magnolia S.; Mealli C.; Messaoudi A.; Mason S. A.; Veiros L. F.
Chem. Eur. J. 2014, 20, 13603 – 13617.
(41) Steiner, T. Angew. Chem. Int. Ed. 2002, 41, 48 – 76.
(42) (a) Crabtree, R. H. Chem. Rev. 2016, 116, 8750 – 8769; (b)
AbdurꢀRashid K., Clapham S. E., Hadzovic A. Harvey J. N., Lough
A. J., Morris, R. H. J. Am. Chem. Soc. 2002, 124, 15104 – 15118; (c)
Field, L. D., Hambley, T. W., Yau, B. C. K. Inorg. Chem. 1994, 33,
2009 – 2017.
(28) The solid state structure of antiꢀ1a has been recently reported
Zhang, L.; Nguyen, D. H.; Raffa, G.; Trivelli, X.; Capet, F.; Desset,
S.; Paul, S.; Dumeignil, F.; Gauvin, R. M. ChemSusChem 2016, 9,
(43) (a) Pitman, C. L.; Brereton, K. R.; Miller, A. J. M. J. Am.
Chem. Soc. 2016, 138, 2252 – 2260; (b) Bullock, R. M.; Appel, A.
M.; Helm, M. L. Chem. Commun. 2014, 50, 3125 – 3143; (c) Jacobꢀ
sen, H.; Berke, H., Recent Advances in Hydride Chemistry Page 89 –
116, Peruzzini, M.; Poli, R., Eds. Elsevier, Amsterdam 2001.
1
413 – 1423.
29) Despite exposure of the solid to high vacuum for a prolonged
(
1
time, the H NMR spectrum of 2 often contained a “tꢀBu” peak from
tꢀBuOH. A strong hydrogen bond between the alcohol, which is too
bulky to add across the RuꢀN amido bond, and 2 may explain these
observations. See Prokopchuk, E. D.; Tsui, B. T. H.; Lough; A. J.,
Morris R. H. Chem. Eur. J. 2014, 20, 16960 – 16968.
1
31
(44) Decoupling from phosphorus in the H{ P} NMR spectrum
resolved the multiplet into two wellꢀresolved doublets, with only the
2
residual JHꢀH coupling remaining for each hydride.
(45) (a) Belkova, N. V.; Epstein, L. M.; Shubina, E. S. Eur. J.
Inorg. Chem. 2010, 3555ꢀ3565; (b) Bakhmutov, V. I. Eur. J. Inorg.
Chem. 2005, 245 – 255; (c) Custelcean, R.; Jackson, J. E. Chem. Rev.
2001, 201, 1963 – 1980.
(46) Friedrich, A.; Drees, M.; auf der Günne, J. S.; Schneider, S. J.
Am. Chem. Soc. 2009, 131, 17552 – 17553.
(
3
30) A related amido complex, bearing a PMe in place of CO has
been reported by Schneider that shows similar structural features:
Friedrich, A.; Drees, M.; Käss, M.; Herdtweck, E.; Schneider, S. In-
org. Chem. 2010, 49, 5482 – 5494.
(
31) Jordan, R. B. Reaction Mechanisms of Inorganic and Organoꢀ
metallic Systems 2007, 3rd Ed., Oxford University Press.
32) The remaining 10% of the total Ru content is due to minor
(47) Käss, M.; Friedrich, A.; Schneider, S. Angew. Chem. Int. Ed.
2009, 48, 905 – 907.
(
1
ꢀ
ꢀ
monohydride species, among which two were more abundant ( H
NMR (THFꢀd
78.6 (s); H NMR (THFꢀd ): δ ꢀ 18.58 (t, JHP = 17.5 Hz) and P{ H}
NMR: δ = 76.9 (s)).
(
(48) Due to fast equilibration, the HCO /CO anions gave rise to a
3 3
31
1
13
): δ ꢀ 18.09 (t, JHP = 17.8 Hz) and P{ H} NMR: δ =
single C NMR peak, the chemical shift of which depends on their
relative concentration and therefore on the pH. See: (a) Moret, S.;
Dyson, P. J.; Laurenczy, G. Dalton Trans., 2013, 42, 4353 – 4356; (b)
Mani, F.; Peruzzini, M.; Stoppioni, P. Green Chem. 2006, 8, 995 –
8
1
31
1
8
33) (a) Buckingham, A. D.; Stephens, P. J. J. Chem. Soc. 1964,
1
000.
49) (a) Dub, P. A.; Henson, N. J.; Martin, R. L.; Gordon, J. C. J.
2
747 – 2759; (b) Appleton, T. G.; Clark, H. C.; Manzer, L. E. Coord.
Chem. Rev. 1973, 10, 335 – 422.
34) (a) Koike, T.; Ikariya, T. Organometallics 2005, 24, 724 – 730;
b) Hamilton, R. J.; Bergens, S. H. J. Am. Chem. Soc. 2008, 128,
(
Am. Chem. Soc. 2014, 136, 3505 – 3521; (b) Hartmann, R.; Chen, P.
Angew. Chem. Int. Ed. 2001, 40, 3581 – 3585.
(
(
(
50) John, J. M.; Takebayashi, S.; Dabral, N.; Miskolzie, M.; Berꢀ
gens S. H. J. Am. Chem. Soc. 2013, 135, 8578 – 8584.
51) The proton affinity (PA) of each complex was calculated and
11979 – 11987; (c) Takebayashi, S.; Dabral, N.; Miskolzie, M.;
Bergens S. H. J. Am. Chem. Soc. 2011, 128, 9666 – 9669.
(
(
35) (a) Bryndza, H. E.; Tam, W. Chem. Rev. 1988, 88, 1163 –
188; (b) Fulton, J. R.; Holland, A. W.; Fox, D. J.; Bergman, R. G.
Acc. Chem. Res. 2002, 35, 44 – 56.
36) Traces of protic solvent must provide the required proton for
the formation of 3.
37) The molecular structure of a similar Ru carbonate complex
compared to a calibration curve that was populated with known pK
and calculated PAs.
a
s
1
(52) (a) Kapoor, S.; Barnabas, F. A.; Sauer Jr., M. C.; Meisel, D.;
Jonah, C. D. J. Phys. Chem., 1995, 99, 6857 – 6863; (b) Ashby, E. C.;
Doctorovich, F.; Liotta, C. L.; Neumann, H. M.; Barefield, E. K.;
Konda, A.; Zhang, K.; Hurley J. J. Am. Chem. Soc. 1993, 115, 1171 –
(
(
bearing the diphenyl substituted PNP ligand has been reported. See
Ref. 21b.
(
Morris, R. H. Dalton Trans. 2013, 42, 10214ꢀ10220; (b) Kohl, S. W.;
Weiner, L.; Schwartsburd, L.; Konstantinovski, L.; Shimon, L. J. W.;
BenꢀDavid, Y.; Iron, M. A.; Milstein, D. Science 2009, 324, 74 – 77.
1
173.
53) Besora, M.; Lledós, A.; Maseras, F. Chem. Soc. Rev. 2009, 38,
(
38) For example: (a) Procopchuk, D. E.; Collado, A.; Lough, A. J.;
957 – 966.
(54) Examples of hemilability at the side donor sites of pincer ligꢀ
ands: (a) Canovese, L.; Visentin, F.; Chessa, G.; Uguagliati, P.; Santo,
C.; Bandoli, G.; Maini, L. Organometallics 2003, 22, 3230 – 3238;
(
39) (a) Grasemann, M.; Laurenczy, G. Energy Environ. Sci. 2012,
, 8171 – 8181; (b) Beller, M.; Bornscheuer, U. T. Angew. Chem. Int.
Ed. 2014, 53, 4527 – 4528; (c) Wang, W.ꢀH.;. Himeda, Y.;
(
b) Bassetti, M.; Capone, A.; Salamone, M. Organometallics 2004,
5
2
3, 247 – 252; (c) Zhang, J.; Leitus, G.; BenꢀDavid, Y.; Milstein, D.
16
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 18
J. Am. Chem. Soc. 2005, 127, 10840 – 10841; (d) Choualeb, A.;
Lough, A. J.; Gusev, D. G. Organometallics 2007, 26, 5224 – 5229;
e) Vuzman, D.; Poverenov, E.; Shimon, L. J. W.; DiskinꢀPosner, Y.;
Milstein, D., Organometallics 2008, 27, 2627 – 2634; (f) Fulmer, G.
R.; Kaminsky, W.; Kemp, R. A.; Goldberg, K. I.; Organometallics
dephatt, R. J. Dalton Trans. 2008, 1243 – 1250; (c) Moxham, G. L.;
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
RandellꢀSly, H.; Brayshaw, S. K.; Weller, A. S.; Willis, M. C. Chem.
Eur. J. 2008, 14, 8383 – 8397; (d) Moxham, G. L.; RandellꢀSly, H.
E.; Brayshaw, S. K.; Woodward, R. L.; Weller, A. S.; Willis, M. C.;
Angew. Chem., Int. Ed. 2006, 45, 7618 – 7622.
(56) Nꢀalkylation of amine ligands is known to decrease the stabilꢀ
ity constants of their metal ion complexes: Meyerstein, D. Coord.
Chem. Rev. 1999, 185-186, 141 – 147.
(57) (a) Smythe, N. A.; Grice, K. A.; Williams, B. S.; Goldberg, K.
I. Organometallics 2009, 28, 277 – 288; (b) Fafard, C. M.; Ozerov, O.
V.; Inorg. Chim. Acta 2007, 360, 286 – 292.
(
2
011, 30, 1627 – 1636; (g) Lindner, R.; van den Bosch, B.; Lutz, M.;
Reek, J. N. H.; van der Vlugt, J. I. Organometallics 2011, 30, 499 –
510; (h) Ruddy, A. J.; Mitton, S. J.; McDonald, R.; Turculet, L.
Chem.Commun. 2012, 48, 1159 – 1161.
(
55) Examples of hemilability at the central donor site of the pincer
ligand: (a) Bernhammer, J. C.; Frison, G.; Huynh, H. V. Dalton
Trans. 2014, 43, 8591 – 8594; (b) Zhao, C.; Jennings, M. C.; Pudꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
17
ACS Paragon Plus Environment