Electronic Fine-tuning of Ruthenium(II) Transfer Hydrogenation Catalysts with Ethanol as the Hydrogen Source

Article abstract of DOI:10.1002/cctc.202001370

A series of ruthenium(II) complexes bearing tridentate N,N’-diallyl-2,6-di(5-butylpyrazol-3-yl)pyridine ligands was synthesized and characterized. Introduction of substituents in the 4-position of the pyrazole rings tune the electron density at the ruthen

Full text of DOI:10.1002/cctc.202001370

The European Society Journal for Catalysis
Supported by
Accepted Article
Title: Electronic fine-tuning of ruthenium(II) transfer hydrogenation
catalysts working with ethanol as the hydrogen source
Authors: Pascal Weingart, Yu Sun, and Werner R. Thiel
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.202001370
Link to VoR: https://doi.org/10.1002/cctc.202001370
01/2020

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
Electronic fine-tuning of ruthenium(II) transfer hydrogenation
catalysts working with ethanol as the hydrogen source
Pascal Weingart,a Yu Suna and Werner R. Thiel*a
[a]
Dr. Pascal Weingart, Dr. Yu Sun, Prof. Dr. Werner R. Thiel
Fachbereich Chemie
Technische Universität Kaiserslautern
Erwin-Schrödinger-Str. 54, 67663 Kaiserslautern, Germany
E-mail: thiel@chemie.uni-kl.de
Supporting information for this article is given via a link at the end of the document.
Abstract: A series of ruthenium(II) complexes bearing tridentate N,N’-
diallyl-2,6-di(5-butylpyrazol-3-yl)pyridine ligands was synthesized and
characterized. Introduction of substituents in the 4-position of the
pyrazole rings tune the electron density at the ruthenium centre, which
was proved by correlation of the 31P NMR chemical shifts with the p
parameters of the Hammett equation. The structural elucidation of a
phosphine-free ruthenium(II) complex proves that one of the allyl side-
chains undergoes chelating coordination to the ruthenium site to
realise a 18 VE centre. This compound is the starting point for
complexes of the type (N,N,N)Ru(L)(Cl)2 bearing ligands L other than
triphenylphosphine. The ruthenium(II) complexes were investigated
for their activity in the transfer hydrogenation with ethanol as the
hydrogen source. Here the logarithms of the measured turn-over
frequencies (TOF) correlate with the p parameters of the Hammett
equation in terms of a linear free-energy relationship.
organic compounds such as formate or alcohols. Therefore it is of
particular interest for lab-scale experiments. Since the first public-
ation of Henbest and co-workers from 1967,[1] large progress in
terms of catalyst optimization has been made especially con-
cerning stereoselective transformations.[3]
[2]
While isopropanol is a common hydrogen source for transfer hy-
drogenation reactions,[4] primary alcohols are not frequently used
as hydrogen source in catalytic transfer hydrogenation reactions.
This is due to that fact that primary alcohols are oxidized to alde-
hydes, which is thermodynamically less favored than the form-
ation of ketones from secondary alcohols. After an initial public-
ation by Grützmacher and co-workers in 2008 [5] only a few further
reports on this topic have appeared.[6] We entered this field after
having developed a highly reactive ruthenium(II) catalyst of the
type A (Scheme 1) for the transfer hydrogenation of ketones with
isopropanol as the hydrogen source.[7] The same catalyst is also
very active in the presence of ethanol as long as the oxidation
product acetaldehyde is removed continuously from the reaction
mixture.[8]
Introduction
The properties of chemical compounds follow their structure. Ac-
cordingly, the quantitative description of structure–reactivity rela-
tionships is of fundamental importance in chemistry. This is in
particular true for homogeneous catalysis, where catalyst optimiz-
ation should be more than “trial and error”. Since transition metal
catalyzed reactions consist of a sequence of individual steps,
drastic structural changes of the ligands bound to a catalytically
active transition metal center can in the best case result in a much
better catalyst or in the worst case lead to a complete loss of
catalytic activity. However, neither of these two cases provides
really clear mechanistic insights, since the barriers connecting the
individual steps change in an unpredictable way with major geo-
metric changes, as do solvation effects. It can therefore be highly
useful to carry out ligand modifications in small steps. Further-
more, it makes sense to consider steric and electronic effects
independently of each other. In the present study we focus on the
electronic influence of 2,6-di(pyrazol-3-yl)pyridine-type ligands on
the catalytic activity of ruthenium(II) complexes used in the trans-
fer hydrogenation with ethanol as the hydrogen source.
Scheme 1. General structure of the ruthenium(II) transfer hydrogenation catal-
ysts bearing a 2,6-di(pyrazol-3-yl)pyridine-type ligand.
In terms of sustainable chemistry, the replacement of isopropanol
by ethanol that can simply be obtained by alcoholic fermentation
from renewable resources is desirable. On the other hand, we
know that the catalysts we use here are highly reactive with iso-
propanol as the hydrogen source, which makes it difficult to con-
tinuously follow the reaction progress by means of GC analysis.
For these reasons, we have chosen to take ethanol as the
hydrogen source in this study.
In the present study we focus on subtle changes of the ligand en-
vironment and their influence on the catalytic activity of the corre-
sponding ruthenium(II) complexes in the transfer hydrogenation
with ethanol as the hydrogen source. Accordingly, the aim of this
Catalytic transfer hydrogenation of ketones and aldehydes prov-
ides access to a broad variety of alcohols. It allows to circumvent
the use of gaseous dihydrogen in high pressure equipment since
the hydrogen atoms to be transferred originate from cheap
1
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
study is to correlate the activity of this class of catalysts with the
substitution pattern of the ligand. This strategy should finally lead
to a knowledge-based optimization of the ligand structure. Pyraz-
ole-based ligands are ideal tools for such a study, since the
pyrazole backbone allows introducing electron-withdrawing
and -donating substituents without changing the steric properties
of the ligand. In the past we had already used chelating ligands
bearing one or two pyrazole rings for catalyst synthesis with a
special focus on getting insight in the structure-activity relation-
ships of transition metal catalysts in other reactions.[9] It can be
mentioned at this point that pyrazole-based ligands can also
easily be fine-tuned with regard to their steric properties.[ 10 ]
Halcrow and co-workers have achieved this by systematically
varying the substituents attached to the N1 atoms of the pyrazole
moieties at the 2,6-di(pyrazol-3-yl)pyridine structure (N,N,N-
ligand) and studying the consequences of such variations on the
spin-state of the corresponding dicationic iron(II) complexes
[Fe(N,N,N)2]2+.[10a]
lower field. In order to work out the influence of the allylic side
chains on the catalyst formation and on the catalytic activity, 2,6-
di(pyrazol-3-yl)pyridine 2b’ bearing propyl instead of propenyl
side chains was synthesized by treating 1b with iodopropane
under the conditions worked out for the allylation reaction.
Introduction
Ligand synthesis
Pyrazoles are accessible in a few steps by condensation, ring-
closure and substitution reactions which allows broad variations
of their donor properties. The benefit of this strategy becomes
even more evident for multidentate ligands where the parallel
setup of pyrazole donor sites with high-yield conversions reduces
efforts for purification and product isolation. For this study, we
chose four 2,6-di(pyrazol-3-yl)pyridines equipped with different
substituents in the 4-positions of the pyrazole sites for variation of
the ligand’s electronic impact on the catalytic activity of rut-
henium(II) catalysts in transfer hydrogenation. The N-allyl groups
and the butyl group in the 5-positions of the pyrazole rings were
kept constant. Scheme 2 summarizes all relevant details of the
ligand synthesis.
The assembly of the pyrazole rings proceeds via a Claisen-con-
densation starting from pyridine-2,6-dicarboxylicacid diethylester
and 2-hexanone, that gives an intermediate, pyridine-centered te-
traketone.[11] Treating this compound with methyliodide in acetone
in the presence of Na2CO3 allows for the introduction of methyl
groups in the 2-position of the 1,3-diketones. Treatment of the two
intermediate tetraketones with hydrazinehydrate in ethanol solu-
tion leads to the formation of the 2,6-di(5-butylpyrazol-3-yl)pyrid-
ines 1a and 1b in about 45% yield (from pyridine-2,6-dicarboxylic-
acid diethylester). Since pyrazoles behave like electron-rich
aromatic compounds, electrophilic aromatic substitution using
either bromine or a mixture of fuming nitric and concentrated
sulfuric acid makes accessible the derivatives 1c (R = Br, 91%
yield) and 1d (R = NO2, 66% yield) from compound 1b. With the
exception of the nitro compound 1d, the 1H and 13C NMR spectra
of these tridentate ligands show broad resonances due to
tautomeric shifts of the N-H protons. Deprotonation of the acidic
N-H sites with NaH in tetrahydrofurane followed by alkylation with
allylic bromide leads to the 2,6-(1-allyl-5-butyl-pyrazol-3-yl)pyrid-
ines 2a-d (55-75% yield). The ongoing decrease of electron
density in the pyrazole ring caused by the substituents in the
Scheme 2. Synthesis of the allyl-substituted2,6-di(pyrazol-3-yl)pyridine ligands
2a-d and 2b’.
Ruthenium(II) complexes
According to a procedure we published some years ago,[7] treat-
ment of (PPh3)3RuCl2 with ligands 2a-d provides the N,N,N-co-
ordinated ruthenium(II) complexes 3a-d in excellent yields (90-
95%). In contrast, the propyl-functionalized ligand 2b’ does not
react with the ruthenium(II) precursor to give 3b’, although this
substituent has steric and electronic properties similar to the allyl
group. We explain this behaviour by the formation of the first bond
between of one of the allylic groups of 2a-d and the 16 VE
1
series 2a-d clearly effects the H NMR spectra of these com-
pounds: Both the resonances of the pyridine protons and of the
protons of the methylene units (allyl, butyl) that are directly
attached to the pyrazole rings become more and more shifted to
2
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
ruthenium precursor (PPh3)3RuCl2 allowing a stepwise dissocia-
tion of phosphine ligands from the ruthenium centre and thus a
stepwise association of the nitrogen atoms (Scheme 3). In a prior
study we could prove by means of NMR spectroscopy that N,N’-
dialkylated 2,6-di(pyrazol-3-yl)pyridine ligands preferentially
adopt, as shown in Schemes 2 and 3, an all-trans conformation in
solution due to weak C-H∙∙∙N interactions.[12] This hinders the
primary coordination of the bulky ruthenium precursor to one of
the pyrazole or pyridine nitrogen atoms. It has to be mentioned at
this point that olefin coordination to ruthenium(II) will be discussed
later on in more detail.
Scheme 4 Synthesis of the ruthenium(II) complex 3b’ by reduction of the inter-
mediate ruthenium(III) complex 4.
Recrystallization of complexes 3a-d and 3b’ from solutions in
dichloromethane by applying slow diffusion of diethylether
resulted in analytically pure samples. However, only 3a and 3c
provided single crystals suitable for X-ray structure analysis
(Figure 1).
3a
P1
N1
N5
N4
Ru1
N2
N3
Cl2
Cl1
3c
Scheme 3. Formation of the ruthenium(II) catalysts 3a-d via ruthenium-olefin
coordination.
Br2
P1
Br1
N1
Ru1
N5
N3
N4
In order to prove whether hindered complexation and not
chemical instability is the reason why the propyl functionalized
ruthenium(II) complex 3b’ was not accessible, we applied an
alternative route, which we had developed some years ago but
had not yet published.[13] Coordinatively unsaturated RuCl3 can
easily be introduced into the N,N,N-binding pocket of N,N-dialk-
ylated 2,6-di(pyrazol-3-yl)pyridine ligands such as 2b’ (Scheme
4). Stirring a solution of the resulting ruthenium(III) complex 4
(91% yield) in the presence of one equiv. of PPh3 over sodium
amalgam provides compound 3b’ in good yields (71%), which
proves that the complexation process is indeed hindered when 3b’
is treated with (PPh3)3RuCl2.
N2
Cl2
Cl1
Figure 1. Molecular structures of compounds 3a and 3c in the solid state. The
ellipsoids are at the 50% level. Selected bond lengths (Å) and angles (°) are
listed. 3a: one molecule of water that co-crystallized by forming hydrogen bonds
to both chlorido ligands Cl1 and Cl2 and the hydrogen atoms are omitted for
clarity; Ru1-Cl1 2.4714(6), Ru1-Cl2 2.4642(6), Ru1-P1 2.2879(6), Ru1-N1
1.987(2), Ru1-N2 2.082(2), Ru1-N4 2.096(2), Cl1-Ru1-Cl2 90.83(2), Cl1-Ru1-
P1 178.34(2), Cl1-Ru1-N1 87.53(5), Cl1-Ru1-N2 87.59(5), Cl1-Ru1-N4 86.81(5),
Cl2-Ru1-P1 90.83(2), Cl2-Ru1-N1 178.06(6), Cl2-Ru1-N2 103.24(5), Cl2-Ru1-
N4 101.19(5), P1-Ru1-N1 90.82(5), P1-Ru1-N2 92.17(5), P1-Ru1-N4 92.73(5),
N1-Ru1-N2 77.74(7), N1-Ru1-N4 77.70(7), N2-Ru1-N4 155.00(7). 3c: both
butyl chains are disordered, the hydrogen atoms are omitted for clarity; Ru1-Cl1
3
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
2.450(1), Ru1-Cl2 2.464 (1), Ru1-P1 2.311(1), Ru1-N1 1.983(3), Ru1-N2
2.080(3), Ru1-N4 2.088(3), Br1-C5 1.879(4), Br2-C17 1.885(3), Cl1-Ru1-Cl2
87.94(3), Cl1-Ru1-P1 178.04(3), Cl1-Ru1-N1 89.60(9), Cl1-Ru1-N2 88.25(9),
Cl1-Ru1-N4 85.24(9), Cl2-Ru1-P1 91.57(3), Cl2-Ru1-N1 177.18(9), Cl2-Ru1-
N2 103.47(9), Cl2-Ru1-N4 100.63(9), P1-Ru1-N1 90.94(9), P1-Ru1-N2
90.02(9), P1-Ru1-N4 96.72(9), N1-Ru1-N2 77.82(12), N1-Ru1-N4 77.78(12),
N2-Ru1-N4 154.77(12).
driven by the poor solubility of NaCl in acetone. By slow diffusion
of diethylether into a chloroform solution, complex 3bI provided
single crystals suitable for X-ray structure analysis (Figure 3).
The ruthenium(II) centres in 3a and 3c are coordinated by the
tridentate N,N,N-donor, two chlorido ligands and a triphenyl-
phosphine ligand in a distorted octahedral geometry. Within the
limits of the standard deviations, there are no significant differen-
ces between the two complexes in terms of bond lengths and
angles. This is in particular true for the distances Ru-N2 and Ru-
N4 where the influence of the substituents at the pyrazole rings
should be most dominant.
In addition to the X-ray structure analyses, 1H NMR spectroscopy
confirms the coordination geometry around the ruthenium(II)
centres in compounds 3a-d and 3b’. Due to the cis-orientation of
the two chlorido ligands, the protons of the methylene groups at
the butyl and the allyl side chains are diastereotopic, a feature that
will be discussed below in more detail. The 31P{1H} NMR chemical
shifts reflect the change of the electronic impact of the pyrazole
rings caused by their substituent. The chemical shifts  (3a: 43.4,
3b: 42.7, 3c: 41.3, 3d: 37.5 ppm) correlate well with the p
parameters (para-substituent constants) of the Hammett equation
(Figure 2).[14] Compound 3b’ shows a 31P NMR resonance at 41.5
ppm. Hammett originally developed the equation to quantify the
influence of substituents at benzene derivatives on a series of
reactions of these compounds. However, the corresponding con-
stants can also be taken to correlate spectroscopic data. In 2014
Nishihara and co-workers for example reported a correlation of
31P{1H} NMR chemical shifts and ligand substituents.[15] They had
investigated platinum(II) complexes bearing para-substituted
phenylacetylenide ligands and chelating aminophosphines, which
were obtained by oxidative addition of the corresponding phenyl-
(trimethylsilyl)acetylenes to platinum(0) precursors. In addition
correlations with the p parameters were found for the coupling
constants 1JPt-P and for the rate constants of the oxidative addition.
Scheme 5. Synthesis of the ruthenium(II) complexes 3bBr and 3bI by halo-
genide exchange.
P1
N1
Ru1
N5
N4
N3
N2
I2
I1
Figure 3. Molecular structures of compound 3bI in the solid state. One molecule
of chloroform that co-crystallized and the hydrogen atoms are omitted for clarity.
The ellipsoids are at the 50% level. Selected bond lengths (Å) and angles (°)
are listed: I1-Ru1 2.7922(5), I2-Ru1 2.7913(6), Ru1-P1 2.2876(7), Ru1-N1
2.008(2), Ru1-N2 2.123(2), Ru1-N4 2.141(2), I1-Ru1-I2 87.31(2), I1-Ru1-P1
175.74(2), I1-Ru1-N1 91.83(6), I1-Ru1-N2 88.60(6), I1-Ru1-N4 88.79(6), I2-
Ru1-P1 89.10(2), I2-Ru1-N1 179.10(7), I2-Ru1-N2 103.41(6), I2-Ru1-N4
102.68(6), P1-Ru1-N1 91.77(7), P1-Ru1-N2 90.00(6), P1-Ru1-N4 94.24(6), N1-
Ru1-N2 76.84(9), N1-Ru1-N4 77.02(9), N2-Ru1-N4 153.62(9).
Based on the solid state structures of compounds 3a, 3c and 3bI
the influence of the halogenido ligands on the structural para-
meters at the ruthenium(II) centre can be discussed. By switching
from chloride to iodide all Ru1-N distances get slightly longer
(0.02-0.05 Å), while the Ru1-P1 distance stays almost unaffected.
The angle N2-Ru1-N4 gets slightly smaller by about 2 ° while the
angle X1-Ru1-N1 (X = Cl, I; 2-4 °) increases by about 2-4 °. All
this can be explained by a somewhat larger steric repulsion of the
iodido ligands - compared to the chlorido ligands - on the nitrogen
donors in the cis-positions. The 31P{1H} NMR resonances of the
halogenide exchanged complexes are observed at 44.9 (3bBr)
and 45.0 ppm (3bI) and thus slightly shifted to lower field than for
3b (42.7 ppm).
Figure 2. Correlation of the 31P NMR chemical shifts  of compounds 3a-d and
the p parameters of the Hammett equation.
Another ligand exchange under the premise of obtaining neutral
complexes is formally possible by exchanging the triphenyl-
phosphine ligand for other neutral two-electron donors. As
recently reported, treatment of 3b with carbon monoxide at
elevated temperatures results in the formation of a ruthenium(II)
Treatment of the dichlorido complex 3b with NaBr and NaI in
acetone makes accessible the dibromido and diiodido complexes
3bBr and 3bI (Scheme 5, yields: 90 and 94%), respectively.
Similar to a Finkelstein reaction,[16] the halogenide exchange is
4
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
carbonyl complex with the chlorido ligands in cis-position to each
other.[7] Thus, this carbonyl complex shows a coordination envir-
onment similar to 3b.
To avoid contamination with triphenylphosphine in the products,
an alternative access to ligand exchanged ruthenium complexes
transition metal complex bearing a pyrazole group with a chelating
olefinic side chain. In 1976 Fukushima et al. published the solid
state structure of a dimeric copper(I) chloride complex, wherein
the copper(I) sites are coordinated in a chelating manner by 1-
allyl-3,5-dimethylpyrazole.[18] In addition, a few similar coordin-
ation geometries are known from complexes with pyridine as the
N-donor and rhodium(I) or ruthenium(II) as the metal centres.[19]
of the type 3b was worked out: Treatment of the ruthenium(II)
[ 17 ]
precursor [(6-cymene)RuCl2]2
with ligand 2b in CH2Cl2
1
resulted - under cleavage of the ruthenium-cymene coordination
- in the formation of the stable complex 5 in almost quantitative
yield as shown in Scheme 6. Herein, the C=C-double bond of one
of the allyl side chains completes the octahedral coordination at
the ruthenium(II) centre. The molecular structure of compound 5
was unambiguously determined by means of an X-ray structure
analysis (Figure 4).
Due to the low symmetry of compound 5, its H and 13C NMR
1
spectra are rather complicated. The H NMR resonances of the
two pyrazole hydrogen atoms appear e.g. as two well separated
singlets at about 6.7 ppm. As expected, the 13C NMR resonances
of the two sp2-hybridized carbon atoms of the coordinating allyl
group are shifted strongly to higher field (92.9, 65.9) compared to
those of the non-coordinating centres (132.1, 118.8 ppm).
Due to the strained ring, the coordination of the allyl group can be
easily broken by a series of 2e- donors (compounds 3b-A to 3b-
H, Scheme 7, yields: 58-95%). Interestingly, the position trans to
the pyridine ring is preferred by most of the donor molecules we
have tested. Only bulky phosphines such as triphenylphosphine,
triphenylphosphite and tricyclohexylphosphine are found bonded
in the trans-position to a chlorido ligand.
Scheme 6. Synthesis of the ruthenium(II) complex 5.
Cl2
N1
N2
N4
Ru1
N5
N3
C11
C12
Cl1
C13
Figure 4. Molecular structures of compound 5 in the solid state. A part of the
hydrogen atoms is omitted for clarity, one of the butyl side chains is disordered.
The ellipsoids are at the 50% level. Selected bond lengths (Å) and angles (°)
are listed: Ru1-Cl1 2.4159(8), Ru1-Cl2 2.3919(8), Ru1-N1 2.044(3), Ru1-N2
1.952(2), Ru1-N4 2.154(2), Ru1-C12 2.204(3), Ru1-C13 2.245(3), C12-C13
1.415(5), C24-C25 1.293(7), Cl1-Ru1-Cl2 176.00(3), Cl1-Ru1-N1 90.75(7), Cl1-
Ru1-N2 97.33(8), Cl1-Ru1-N4 86.07(7), Cl1-Ru1-C12 84.01(8), Cl1-Ru1-C13
94.00(9), Cl2-Ru1-N1 88.60(7), Cl2-Ru1-N2 86.32(8), Cl2-Ru1-N4 89.94(7),
Cl2-Ru1-C12 98.67(8), Cl2-Ru1-C13 86.46(9), N1-Ru1-N2 75.27(10), N1-Ru1-
N4 76.07(10), N1-Ru1-C12 146.76(11), N1-Ru1-C13 174.46(12), N2-Ru1-N4
151.17(11), N2-Ru1-C12 72.92(11), N2-Ru1-C13 106.90(13), N4-Ru1-C12 1
The chelating coordination of the olefin completes the 18 VE shell
of the ruthenium(II) centre. However, it also leads to a drastic dist-
ortion of the octahedral coordination sphere around the ruthen-
ium(II) site. Most obvious is the strong shortening of the Ru1-N1
(2.044(3) Å) bond in comparison to Ru1-N4 (2.154(2) Å). The
bond Ru1-N2 (1.952(2) Å) is by about 0.05 Å shorter than in the
complexes discussed above. The allyl chain is too short to allow
coordination in the ideal trans-position to N1, which makes the
Ru1-C12 (2.204(3) Å) shorter than Ru1-C13 (2.245(3) Å). The two
strongly different bond lengths of the C=C double bonds (C12-
C13: 1.415(5), C24-C25: 1.293(7) Å) illustrate well the influence
of the metal coordination on the length of an olefinic bond. A
further remarkable structural effect of the chelating coordination
is the angle N2-N3-C11 (111.0(2)°), which is much smaller than
the expected value for a relaxed situation (126°). Compound 5 is
up to now only the second example of a structurally characterized
Scheme 7. Ligand exchange reactions starting from complex 5.
Three of these complexes could be obtained as single crystals
which were suitable for an X-ray structure analysis. Figure 5
shows the molecular structures and summarizes characteristic
bond parameters.
5
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
becomes evident, that in solution, the allyl chains will orient in a
way that allows it to avoid steric repulsion with the monodentate
donor. Thus, one of the allyl methylene protons will direct towards
one of the chlorido ligands, which causes a large shift of its
resonance to a lower field, while the other one will direct towards
the butyl chain.
Catalytic transformations with the 18 VE ruthenium(II) compounds
of the type 3 require the dissociation of either the phosphine or
one of the chlorido ligands. From a prior study we know, that an
excess of phosphine slows down the reaction rate in the transfer
hydrogenation of acetophenone catalysed by compound 3b in
isopropanol (used as solvent and hydrogen source).[7] This
speaks for a dissociation of the phosphine in equilibrium as the
entrance step into the catalytic process. It therefore also seemed
reasonable to synthesize cationic ruthenium(II) species by ab-
straction of one or two chlorido ligands and to investigate their
performance in the transfer hydrogenation.
Treatment of compound 3b with one equivalent of silver(I) hexa-
fluorophosphate in dichloromethane resulted in the formation of
the monocationic complex 6a (Scheme 8). Reacting 6a with a
second equivalent of silver(I) acetate gave the acetato complex
6b, while treatment of 6a with an second equivalent of silver(I)
hexafluorophosphate led to the dicationic compound 6c.
Figure 5 Molecular structures of compounds 3b-C, 3b-E and 3B-F in the solid
state. Co-crystallized solvent molecules, disordered side chains and hydrogen
atoms are omitted for clarity. The ellipsoids are at the 50% level. Selected bond
lengths (Å) and angles (°) are listed: 3b-B Ru1-Cl1 2.4154(8), Ru1-Cl2
2.4234(8), Ru1-P1 2.3200(8), Ru1-N1 2.043(2), Ru1-N2 2.096(2), Ru1-N4
2.095(2), Cl1-Ru1-Cl2 178.62(3), Cl1-Ru1-P1 96.10(3), Cl1-Ru1-N1 98.00(7),
Cl1-Ru1-N2 88.21(6), Cl1-Ru1-N4 89.20(6), Cl2-Ru1-P1 82.62(3), Cl2-Ru1-N1
83.30(7), Cl2-Ru1-N2 91.65(6), Cl2-Ru1-N4 91.57(6), P1-Ru1-N1 165.73(7),
P1-Ru1-N2 106.53(6), P1-Ru1-N4 101.51(7), N1-Ru1-N2 76.21(8), N1-Ru1-N4
76.53(8), N2-Ru1-N4 151.96(9); 3b-D Ru1-Cl1 2.4246(9), Ru1-Cl2 2.4071(9),
Ru1-P1 2.2557(8), Ru1-N1 2.064(3), Ru1-N2 2.121(3), Ru1-N4 2.111(2), Cl1-
Ru1-Cl2 173.85(3), Cl1-Ru1-P1 98.99(3), Cl1-Ru1-N1 91.31(7), Cl1-Ru1-N2
87.67(7), Cl1-Ru1-N4 87.48(7), Cl2-Ru1-P1 87.13(3), Cl2-Ru1-N1 82.56(7),
Cl2-Ru1-N2 91.42(7), Cl2-Ru1-N4 90.47(7), P1-Ru1-N1 169.63(7), P1-Ru1-N2
105.01(7), P1-Ru1-N4 103.18(8), N1-Ru1-N2 76.47(10), N1-Ru1-N4 75.90(10),
N2-Ru1-N4 151.80(10); 3b-E Ru1-Cl1 2.4177(5), Ru1-Cl2 2.4194(5), Ru1-N1
1.9829(15), Ru1-N2 2.0831(15), Ru1-N4 2.0787(15), Ru1-N6 2.1671(16), Cl1-
Ru1-Cl2 176.376(17), Cl1-Ru1-N1 88.06(5), Cl1-Ru1-N2 88.98(4), Cl1-Ru1-N4
90.25(4), Cl1-Ru1-N6 92.33(5), Cl2-Ru1-N1 88.35(5), Cl2-Ru1-N2 89.77(4),
Cl2-Ru1-N4 89.49(4), Cl2-Ru1-N6 91.27(5), N1-Ru1-N2 77.49(6), N1-Ru1-N4
78.18(6), N1-Ru1-N6 178.02(6), N2-Ru1-N6 104.46(6), N2-Ru1-N4 155.68(6),
N4-Ru1-N6 99.87(6).
Scheme 8. The formation of cationic complexes from compound 3b. Scheme
Caption.
The positioning of the monodentate donor is clearly confirmed by
the 1H NMR spectra of the complexes. As already discussed for
the triphenylphosphine derivative 3b, the protons of the methyl-
ene groups of the allyl and butyl chains are diastereotopic, when
the neutral monodentate donor coordinates in the cis-position to
the pyridine site of the tridentate ligands. In particular the chemical
shifts of the allyl methylene protons become largely different in
such a situation. From the structures shown in Figures 1 and 3 it
According to the NMR data of compound 6b, there is no -coord-
ination of any of the two allylic side chains, since the resonances
of the olefinic groups are found in the expected regions for non-
coordinated olefins (1H NMR: 5.7-4.8; 13C NMR: 131.6, 117.3
ppm). As expected for a cationic ruthenium(II) complex, the 31P
NMR resonance is shifted to lower field (: 50.2 ppm) compared
6
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
to compound 3b. The IR spectrum of 6b showing the absorptions
of an asymmetric (as: 1550 cm-1) and a symmetric C-O stretching
vibration (sy: 1435 cm-1) gave the first hint for the presence of a
2-coordinating acetato ligand, which was finally proved by a
single crystal X-ray structure analysis (Figure 6).
with isopropanol are too fast to allow a precise determination of
turn-over frequencies (for more details see below). It is important
to mention at this point that the transfer hydrogenation of ketones
is an equilibrium reaction. If the catalyst is stable during the
course of the reaction, high yields can generally be expected
because the concentration of alcohol (in this case ethanol) in the
reaction mixture is high, since it is used not only as the source of
hydrogen but also as the solvent. Table 1 summarises the results.
P1
-
PF6
N1
N2
Ru1
N3
N4
N5
Scheme 9. Catalytic transfer hydrogenation of acetophenone with ethanol as
the hydrogen source.
O1
O2
Table 1. Ruthenium(II catalysed transfer hydrogenation of acetophenone,a
X denotes the anions according to Scheme 5.
Figure 6. Molecular structures of compound 6b in the solid state. Hydrogen
atoms and the disordering of the PF6- anion are omitted for clarity. The ellipsoids
are at the 50% level. Selected bond lengths (Å) and angles (°) are listed: Ru1-
P1 2.2702(7), Ru1-O1 2.1658(18), Ru1-O2 2.1847(17), Ru1-N1 1.990(2), Ru1-
N2 2.095(2), Ru1-N4 2.075(2), O1-C26 1.273(3), O2-C26 1.266(3), P1-Ru1-O1
105.51(5), P1-Ru1-O2 165.87(5), P1-Ru1-N1 92.03(6), P1-Ru1-N2 93.23(6),
P1-Ru1-N4 92.30(6), O1-Ru1-O2 60.36(7), O1-Ru1-N1 162.46(7), O1-Ru1-N2
100.27(8), O1-Ru1-N4 101.36(7), O2-Ru1-N1 102.10(7), O2-Ru1-N2 90.09(7),
O2-Ru1-N4 90.38(8), N1-Ru1-N2 78.02(9), N1-Ru1-N4 77.77(8), N2-Ru1-N4
155.32(8), Ru1-O1-C26 90.63(15), Ru1-O2-C26 89.97(15), O1-C26-O2
119.0(2).
Yield (%) after t (min)
TOF
Entry
Catalyst
(h-1)
15
90
81
34 b
8
30
97
94
74
16
45
98
98
86
20
60
99
99
93
24
1
2
3
4
3a (R = Me, X = Cl)
3b (R = H, X = Cl)
3c (R = Br, X = Cl)
3d (R = NO2, X = Cl)
609
432
157
19
3b’ (R = H, X = Cl, R’ =
C3H7)
5
78
87
92
95
422
The structural assignment of the composition of complexes 6a
and 6c was carried out on the basis of ESI-MS data and elemental
analysis. For 6a the presence of one chlorido ligand was proven
by ESI-MS (m/z: found 802.16; calcd. 802.24 for
6
3bBr (R = H, X = Br)
3bI (R = H, X = I)
71
61
71
38
32
85
75
82
44
39
90
82
87
50
43
94
85
91
50
46
331
266
354
175
136
7
[C43H48ClN5PRu]+;
found
504.13,
calcd.
504.17
for
[C43H48ClN5PRu]+), while the cationic species observed for 6c
(m/z: found 766.18, calcd. 766.26 for [C43H47N5PRu]+; found
383.59, calcd. 383.63 for [C43H48N5PRu]2+) do not contain chloride
anymore. We believe that as found for compound 5, one of the
allylic side chains undergoes coordination to the ruthenium(II)
centre in both compounds, since there is no hint for coordination
of PF6- from mass spectrometry. However, the solubility of 6a and
6c in non-coordinating solvents is too low to obtain reliable NMR
data. By dissolving them in DMSO-d6, the solvent undergoes
coordination to the ruthenium(II) site and cleaves the ruthenium-
olefin bond. The 1H NMR spectra of 6a and 6c recorded in DMSO-
d6 are therefore largely comparable to those of the other six-
coordinated ruthenium(II) complexes discussed above. Neverthe-
less, the difference in charge of the two compounds is clearly
reflected by their 31P NMR data (6a: 26.4; 6c: 46.6 ppm).
8
6a (R = H, X c = Cl, PF6)
6b (R = H, X c = OAc, PF6)
6c (R = H, X c = PF6)
9
10
a
Reaction conditions: 1 mmol of acetophenone, 100 μL of tetradecane
(internal GC standard), 7.5 mol-% of KOH in 10 mL of EtOH, 1.5 mol-% of
the catalyst, 40 °C, continuous N2-flow of
3
mL min-1, conversions
b
c
determined by GC; 10 min; here X denotes coordinating as well as not
coordinating anions.
In addition to the complexes specified in Table 1 the ruthenium(II)
compounds 3b-A to 3b-H were also investigated for activity in the
transfer hydrogenation of acetophenone in the presence of
ethanol. None of them showed activity! We explain this observ-
ation by a stronger binding of their two-electron ligands to the
ruthenium site. Most of these ligands are bound in the trans-
position to the pyridine donor of the tridentate nitrogen ligand, only
rather bulky ligands (in compounds 3b-A and 3b-B) prefer
coordination trans to the chlorido ligand. A mechanistic explan-
ation must be found in particular for the observation that the olefin
complex 5 is inactive for the transfer hydrogenation. We have
added a sub-chapter at the end of the manuscripts, which
summarizes all mechanistic conclusions.
Catalysis
Since compound 3b had turned out to be an excellent catalyst for
the transfer hydrogenation of ketones and aldehydes, even with
ethanol as the hydrogen source,[7,8] it seemed reasonable to also
investigate the other above introduced ruthenium(II) complexes
for this transformation, since they are structurally closely related.
We chose acetophenone (Scheme 9) as the model substrate and
took the reaction conditions we recently had published for the
transfer hydrogenation with ethanol. Ethanol was taken as the
hydrogen source instead of isopropanol since the reaction rates
7
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
The catalytic activity strongly depends on the electronic situation
of the ruthenium centre. Fitting the data of Table 1 allows to
calculate turn-over frequencies from the slope of the curves (see
the Supporting Information) at the beginning of the reaction. Just
as the p parameters of the Hammett equation correlate linearly
with the 31P NMR data of catalysts 3a-d (see Figure 2), so do the
turn-over frequencies in a logarithmic scale (Figure 7), which is a
typical case of a linear free-energy relationship as expressed by
the Hammett equation (log(k) = log(k0) +·). The first conclusion
that can be drawn from these results is that 31P NMR spectro-
scopy is well suited to predict the catalytic activity of dichlorido-
triphenylphosphineruthenium(II) complexes with substituted 2,6-
(1-allyl-5-butyl-pyrazol-3-yl)pyridines. As expected, there is little
difference in the activities of complexes 3b and 3b’, which proves,
that in contrast to their importance in the catalyst synthesis the
allylic double bonds do not strongly interfere in the catalytic
process.
gas-phase acidities (G(HX)) of the hydrohalic acids (HX ⇌ H+ +
[22]
X-; HCl: 328.1, HBr: 318.3, HI: 309.3 kcal/mol)
as a measure
for the interaction of the halide with the ruthenium(II) centre. In
our opinion, this is much more accurate than, for example, using
the corresponding pKa values which are strongly solvent
dependent and show large variations.[ 23 ] Figure 8 shows the
corresponding correlation.
Figure 8 Correlation or the turn-over frequencies (TOF) of compounds 3b, 3bBr
and 3bI with the gas-phase acidities (G(HX)) of the corresponding hydrohalic
acids HX.
HCl is the hydrohalic acid with the lowest acidity, thus chloride the
anion with the highest basicity giving the most active catalyst. This
is in general agreement with the findings for the influence of sub-
stituents at the N,N,N-donor ligand discussed above. In the first
instance it nevertheless seems surprising, that substituents that
are located at the N,N,N-donor ligand should have a more potent
influence on the catalytic activity than halido ligands being bound
directly to the ruthenium(II) site. However, the pyrazole donors
transfer electron density directly into the ruthenium(II) dx2-y2 orbital.
In case the hydrido ligand, being formed in the course of the
catalysis by -hydrogen elimination from an ethanolato ligand, is
located in the trans-position to the pyridine donor, the influence of
the equatorial donor sites should be more pronounced than the
influence of the halido ligand in the axial position. DFT calcul-
ations (see the Supporting Information) on a model complex of 3b
bearing methyl groups instead of allyl substituent and hydrogen
atoms instead of the butyl groups suggest that this is indeed the
case: Only -hydrogen elimination in the equatorial position
(Scheme 10, top) is possible. The corresponding transition state
for the axial elimination (Scheme 10, bottom) could not be located.
We explain this by the steric interference of the substituents at the
pyrazole nitrogen atoms with the acetaldehyde that is formed in
the course of the -hydrogen elimination step.
Figure 7. Correlation or the turn-over frequencies (TOF, logarithmic scale) of
compounds 3a-d and the p parameters of the Hammett equation.
The influence of the pyrazole substituents on the TOF values is a
rather dramatic one: The methyl substituted compound 3a is by a
factor of 32 more active than the nitro derivative 3d. Calculation
of the reaction factor  by log(TOF/TOF0) = · gives negative
values (-0.887 - -1.890), which speaks for an accumulation of
positive charge in the rate determining step. This might be the
transfer of a hydrido ligand from the ruthenium(II) centre to the
ketone. We recently found by DFT calculations the hydride
transfer to be the rate determining step for the transfer hydro-
genation for another class of ruthenium(II) catalysts.[20]
The exchange of the chlorido ligands for bromido and iodido
ligands results in a moderate decrease of transfer hydrogenation
activity. At a first glance, this correlates with the positioning of the
halogenides in the spectrochemical series of ligands. However, it
is hard to quantify this effect e.g. by using the ligand field splitting
0 since this will depend not only on the halogenido ligand but in
a complex way also on the metal site and on the other ligands that
are present. Corresponding quantitative 0 data for ruthenium
complexes and all three halogenides are not available in the
literature. In 2009 Ishii et al. published an attempt to quantify the
spectrochemical series for
a broad variety of ligands L
coordinated to octahedral first row transition metal (M) complexes
of the type [LM(H2O)5]x+,[21] but they left out second and third row
transition metals. Under the assumption that the transfer of
electron density from the halido ligands to the ruthenium(II) centre
should mainly occur via -donation, we took the well-established
8
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
allylic groups and these incoming ligands may provide additional
stabilization for this situation, which prevents this type of
ruthenium(II) compounds from being active in transfer hydro-
Scheme 10 Two different routes of -hydrogen abstraction (reverse: aldehyde
insertion) leading to a hydrido ligand in the equatorial respectively axial position.
For clarity the N,N,N-donor ligand is shown in a simplified way, the grey dots
depict the substituents at the nitrogen atoms.
Finally, the catalytic activity of the cationic complexes 6a-c shall
shortly be discussed. For a correct interpretation of the measured
turn-over frequencies it must be taken into account that the
transfer hydrogenation is performed in the presence of a fivefold
excess of KOH relative to the amount of catalyst used. The pre-
sence of cationic compounds in the reaction mixture is therefore
not likely. Compound 6a that still bears one chlorido ligand is able
to almost reach the equilibrium conversion of acetophenone. It
therefore behaves in almost the same way as compound 3b, with
just a slightly worse performance. In contrast the chloride-free
compounds 6b and 6c, although showing good activities in the
beginning of the reaction, do not reach the equilibrium conversion.
The catalytic conversion stops after 30-45 min. This is a clear hint
for catalyst decomposition during the course of the reaction.
genation.
Scheme 11 Mechanistic explanation for the site-selective association of diffe-
rent ligands starting from olefin complex 5.
At this point it is important to make clear that binding of ethanolate
(or ethanol) to isomer A’ cannot lead to an activation of the sub-
strate, otherwise the olefin complex 5 would be active in transfer
hydrogenation too. An explanation for this behaviour can be as
follows: The binding of a small and hard donor like ethanolate in
trans-position to the pyridine site should not be very strong due
the relatively strong trans-effect of pyridine (we have rather short
N1-Ru distances) and a poor stabilization of this situation by
dispersive interactions with the allylic side chains. At the same
time, the binding of the two chlorido ligands to the ruthenium
centre is strong, since they do not exert a strong trans-effect to
each other. However, for the formation of a ruthenium(II) hydride
species via -hydride elimination from an alcoholato complex, the
dissociation of a chlorido ligand is necessary (species B). The
stability of the binding of the two trans-oriented chlorido ligands to
ruthenium also explains the fact that complexes with donors trans
to pyridine are inactive in transfer hydrogenation.
Mechanistic aspects
We are trying at the end of the manuscript to fit all relevant data
into a mechanistic picture, in the knowledge that mechanisms
must always be adapted as new insights become available. In this
study a large number of structurally rather closely related ruthen-
ium(II) complexes was investigated. In contrast to intuition, some
of these compounds show pronounced differences in catalytic
activity in the transfer hydrogenation while others behave more or
less as expected. A mechanistic explanation must be found in
particular for the inactivity of olefin complex 5. Since complex 5 is
the starting material for the synthesis of the complexes 3b-A -
3b-H.
The situation changes, when a donor with a strong trans-effect
such as PPh3 coordinates to the position cis to the pyridine site
(compound 3b, Scheme 12). This allows dissociation of the trans-
chlorido ligand in equilibrium, leading to the cationic 16 VE inter-
mediate C, which can either isomerize to species C’ with PPh3
and chloride in trans-position to each other or can coordinate fur-
ther PPh3 (species D). The latter reaction explains the reduction
of the reaction rate in the presence of more than one equivalent
of PPh3. Corresponding cationic ruthenium complexes with two
triphenylphosphine ligands in trans position to each other were
obtained with 2,6-di(pyrazol-3-yl)pyridine ligands containing N-H
instead of N-allyl units at the pyrazole rings.[24] Coordination of
ethanolate (or ethanol) to the cationic intermediate C’ allows the
second chlorido ligand (or HCl) to dissociate and to open up a free
coordination site for -hydride elimination.
The best point to start with a mechanistic discussion is therefore
olefin complex 5. The coordination of the olefin ligand must be
reversible. Only dissociation of the olefin allows opening up a
coordination site for the income of another ligand (Scheme 11).
From DFT calculations on the model system mentioned above we
know, that the resulting 16 VE ruthenium(II) complex A’ should be
in equilibrium with isomer A, where the chlorido ligands are in cis-
position. The energy difference between compounds A (less
stable) and A’ should be just a few kJ mol-1. We believe that due
to steric reasons, ligands as bulky as PPh3 cannot coordinate to
the free coordination site of A’ (trans to the pyridine site), which is
shielded by the two allylic side chains. PPh3 can only coordinate
to A and thus in trans-position to a chlorido ligand. In contrast,
smaller ligands should preferentially coordinate to A’ and are
bound strongly in this position. Disperse interactions between the
9
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
correlations with the activities (TOFs) of the catalysts. The TOFs
were measured with acetophenone as the substrate and ethanol
acting as the hydrogen source and the solvent. While there is a
strong dependence of the TOF from the substituents in the pyraz-
ole-4-positions, exchange of chlorido ligands at the ruthenium(II)
complex against bromido or iodido ligands has a minor impact on
the catalyst activity. This can be explained by the location of a
hydrido ligand that is formed during catalyst activation in the
equatorial position, while the remaining halido ligand is located in
an axial position. The TOF data of these ruthenium(II) complexes
correlate with the gas-phase acidities of the corresponding
hydrohalic acids HX. We thus have established a set of experi-
mental (31P shifts) and theoretical (P, gas-phase acidities
(G(HX)) of HX) parameters that allow to predict and to optimize
the activities of a class of highly active transfer hydrogenation
catalysts.
Experimental Section
General information: Purification and drying of solvents was carried out
according to standard methods. Synthesis and catalysis were performed
under an atmosphere of nitrogen although the ruthenium catalysts are not
sensitive towards oxygen and moisture. Synthesis procedures for the
ligands and ruthenium complexes described herein, spectroscopic and X-
ray data as well as the corresponding references are deposited in the
Supporting Information. CCDC 2010583-2010590 contain the supple free
of charge from The Cambridge Crystallographic Data Centre via
Scheme 12 Mechanistic explanation for the site-selective activation of ethan-
olate (or ethanol) occurring at compound 3b or ruthenium(II) complexes with the
same coordination geometry.
Finally, the poorer activity of the cationic ruthenium complexes
6a-c in the transfer hydrogenation has to be explained. It looks
like the presence of two equivalents of chloride anions with
respect to the amount of ruthenium(II) has a stabilizing effect and
prevents the catalyst from degradation. The highest activity of
compound 6a in this series, which carries still one chlorido ligand,
might be considered as a hint in this direction.
Acknowledgements
The authors gratefully acknowledge financial support by the EU-
INTERREG project BIOVAL.
Keywords: transfer hydrogenation • ruthenium • N-donor ligand
• Hammett equation • correlation
Conclusion
By introduction of substituents in the pyrazole-4-positions of N,N’-
diallyl-2,6-di(5-butylpyrazol-3-yl)pyridine ligands, the electron
density at the metal centre of ruthenium(II) transfer hydrogenation
catalysts can be simply fine-tuned. The P parameters of the
Hammett equation and the 31P NMR chemical shift of the triphen-
ylphosphine ligand attached to the ruthenium(II) site deliver exact
References
[1]
[2]
J. Trocha-Grimshaw, H. B. Henbest, Chem. Commun. 1967, 544-544.
a) G. Brieger, T. J. Nestrick, Chem. Rev. 1974, 5, 567-580; b) R. A. W.
Johnstone, A. H. Wilby, I. D. Entwistle, Chem. Rev. 1985, 85, 129-170;
c) W. Baratta, E. Herdtweck, K. Siega, M. Toniutti, P. Rigo, Organometal-
lics 2005, 24, 1660-1669; d) J. F. Bower, I. S. Kim, R. L. Patman, M. J.
Krische, Angew. Chem. Int. Ed. 2009, 48, 34-46; e) C. Gunanathan, D.
Milstein, Chem. Rev. 2014, 114, 12024-12087 f) D. Wang, D. Astruc,
Chem. Rev. 2015, 115, 6621-6686; g) S. Elangovan, C. Topf, S. Fischer,
H. Jiao, A. Spannenberg, W. Baumann, R. Ludwig, K. Junge, M. Beller,
J. Am. Chem. Soc. 2016, 138, 8809-8814; h) G. Zassinovich, G. Mestroni,
S. Gladiali, Chem. Rev. 1992, 92, 1051-1069.
[4]
a) Y. Shvo, D. Czarkie, Y. Rahamim, J. Am. Chem. Soc. 1986, 108, 7400-
7002; b) J. S. M. Samec, J.-E. Bäckvall, P. G. Andersson, P. Brandt,
Chem. Soc. Rev. 2006, 35, 237-248; c) C. Kayan, N. Meric, M. Aydemir,
Y. S. Ocak, A. Baysal, H. Temel, Appl. Organomet. Chem. 2014, 28, 127-
133; d) X.-H. Zhu, L.-H. Cai, C.-X. Wang, Y.-N. Wang, X.-Q. Guo, X.-F.
Hou, J. Mol. Catal. A 2014, 393, 134-141; e) O. Prakash, H. Joshi, K. N.
Sharma, P. L. Gupta, A. K. Singh, Organometallics 2014, 33, 3804-3812;
f) A. K. Sharma, H. Joshi, K. N. Sharma, P. L. Gupta, A. K. Singh,
Organometallics 2014, 33, 3629-3639; g) A. Azua, M. Finn, H. Yi, A. B.
Dantas, A. Voutchkova-Kostal, ACS Sust. Chem. Eng. 2017, 5, 3963-
3972; h) V. H. Mai, S.-H. Lee, G. I. Nikonov, Chem. Select 2017, 2, 7751-
7757.
[3]
a) S. Hashiguchi, A. Fujii, J. Takehara, T. Ikariya, R. Noyori, J. Am. Chem.
Soc. 1995, 117, 7562-7563; b) R. Noyori, S. Hashiguchi, Acc. Chem. Res.
1997, 30, 97-102; c) K.-J. Haack, S. Hashiguchi, A. Fujii, T. Ikariya, R.
Noyori, Angew. Chem. Int. Ed. Engl. 1997, 36, 285-288; d) T. Ikariya, A.
J. Blacker, Acc. Chem. Res. 2007, 40, 12, 1300-1308; e) S. Gladiali, E.
Alberico, Chem. Soc. Rev. 2006, 35, 226-236.
[5]
[6]
T. Zweifel, J.-V. Naubron, T. Büttner, T. Ott, H. Grützmacher, Angew.
Chem. Int. Ed. 2008, 47, 3245-3249.
a) T. Zweifel, D. Scheschkewitz, T. Ott, M. Vogt, H. Grützmacher, Eur. J.
Inorg. Chem. 2009, 36, 5561-5576; b) H. Lundberg, H. Adolfsson, Tetra-
hedron Lett. 2011, 52, 2754-2758; c) W.-P. Liu, M.-L. Yuan, X.-H. Yang,
K. Li, J.-H. Xie, Q.-L. Zhou, Chem. Commun. 2015, 51, 6123-6125; d) N.
10
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
[15] Y. Okuda, Y. Ishiguro, S. Mori, K. Nakajima, Y. Nishihara, Organometal-
lics 2014, 33, 1878-1889.
Castellanos-Blanco, A. Arévalo, J. J. García, Dalton Trans. 2016, 45,
[16] H. Finkelstein, Ber. Dt. Chem. Ges. 1910, 43, 1528-1532.
[17] M. A. Bennett, A. K. Smith, J. Chem. Soc. Dalton Trans. 1974, 233-241.
[18] K. Fukushima, A. Kobayashi, T Miyamoto, Y. Sasaki, Bul. Chem. Soc.
Jpn. 1976, 49, 143-146.
13604-13614; e) Y. Wang, Z. Huang, X. Leng, H. Zhu, G. Liu, Z. Huang,
J. Am. Chem. Soc. 2018, 140, 4417-4429; f) R. A. Farrar-Tobar, B. Woz-
niak, A. Savini, S. Hinze, S. Tin, J. G. de Vries, Angew. Chem. Int. Ed.
2019, 58, 1129-1133.
[19] a) G. Vasapollo, A. Sacco, C. F. Nobile, M. A. Pellinghelli, M. Lanfranchi,
J. Organomet. Chem. 1986, 312, 249-262; b) G. Vasapollo, C. F. Nobile,
M. Latronico, M. Lanfranchi, M. A. Pellinghelli, J. Organomet. Chem.
1987, 336, 429-439.
[7]
L. Taghizadeh Ghoochany, S. Farsadpour, Y. Sun, W. R. Thiel, Eur. J.
Inorg. Chem. 2011, 23, 3431-3437.
[8]
[9]
P. Weingart, W. R. Thiel, ChemCatChem 2018, 10, 4844-4848.
a) W. R. Thiel, J. Eppinger, Chem. Eur. J. 1997, 3, 696-705; b) K. Muller,
A. Schubert, T. Jozak, A. Ahrens-Botzong, V. Schünemann, W. R. Thiel,
ChemCatChem 2011, 3, 887-892; c) K. Muller, Yu Sun, A. Heimermann,
F. Menges, G. Niedner-Schatteburg, C. van Wüllen, W. R. Thiel, Chem.
Eur. J. 2013, 19, 7825-7834.
[20] C. Kerner, J. Lang, M. Gaffga, Y. Sun, G. Niedner-Schatteburg, W. R.
Thiel, ChemPlusChem 2017, 82, 212-224.
[21] T. Ishii, S. Tsuboi, G. Sakane, M. Yamashita, B. K. Breedlove, Dalton
Trans. 2009, 680-687.
[22] a) A. Trummal, L. Lipping, I. Kaljurand, I. A. Koppel, I. Leito, J. Phys.
Chem. A 2016, 120, 3663-3669; b) F. Menges, J. Lang, Y. Nosenko, C.
Kerner, M. Gaffga, L. Taghizadeh Ghoochany, W. R. Thiel, C. Riehn, G.
Niedner-Schatteburg, J. Phys. Chem. A 2017, 121, 4422-4434.
[23] D. D. Perrin, Dissociation Constants of Inorganic Acids and Bases in
Aqueous Solution, Pergamon, Toronto, 1982.
[10] a) T. D. Roberts, M. A. Little, L. J. Kershaw Cook, S. A. Barrett, F. Tuna,
M. A. Halcrow, Polyhedron 2013, 64, 4-12; H. Glas, M. Barz, W. R. Thiel,
J. Organometal. Chem. 2001, 621, 153-157.
[11] D. Zabel, A. Schubert, G. Wolmershäuser, R. L. Jones Jr., W. R. Thiel,
Eur. J. Inorg. Chem. 2008, 3648-3654.
[12] C. Groß, Y. Sun, T. Jost, T. Grimm, M. Klein, G. Niedner-Schatteburg, S.
Becker, W. R. Thiel, Chem. Commun. 2020, 56, 368-371.
[13] S. Schäfer, diploma thesis, TU Kaiserslautern, 2013.
[14] a) L. P. Hammett, J. Am. Chem. Soc. 1937, 59, 96-103; b) S. L. Keenan,
K. P. Peterson, K. Peterson, K. Jacobson, J. Chem. Educ. 2008, 85, 558-
560.
[24] T. Jozak, D. Zabel, A. Schubert, Y. Sun, W. R. Thiel, Eur. J. Inorg. Chem.
2010, 32, 5135-5145.
11
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001370
ChemCatChem
FULL PAPER
Entry for the Table of Contents
Correlation of electronic substituent effects with 31P NMR data and measured TOF values allows a knowledge-based optimization of
ruthenium-based transfer hydrogenation catalysts.
12
This article is protected by copyright. All rights reserved.