Catalyst Deactivation During Rhodium Complex-Catalyzed Propargylic C?H Activation

Article abstract of DOI:10.1002/chem.202102219

Detailed mechanistic investigations on our previously reported synthesis of branched allylic esters by the rhodium complex-catalyzed propargylic C?H activation have been carried out. Based on initial mechanistic studies, we present herein more detailed investigations of the reaction mechanism. For this, various analytical (NMR, X-ray crystal structure analysis, Raman) and kinetic methods were used to characterize the formation of intermediates under the reaction conditions. The knowledge obtained by this was used to further optimize the previous conditions and generate a more active catalytic system.

Full text of DOI:10.1002/chem.202102219

Full Paper
doi.org/10.1002/chem.202102219
Chemistry—A European Journal
Catalyst Deactivation During Rhodium Complex-Catalyzed
Propargylic CÀ H Activation
Saskia Möller,[a] Nora Jannsen,[a] Julia Rüger,[a] Hans-Joachim Drexler,[a] Moritz Horstmann,[a]
Felix Bauer,[b] Bernhard Breit,*[b] and Detlef Heller*[a]
This paper is dedicated to Professor Barry Trost for the outstanding importance of his work in the field of homogeneous catalysis.
Abstract: Detailed mechanistic investigations on our previ-
ously reported synthesis of branched allylic esters by the
rhodium complex-catalyzed propargylic CÀ H activation have
been carried out. Based on initial mechanistic studies, we
present herein more detailed investigations of the reaction
mechanism. For this, various analytical (NMR, X-ray crystal
structure analysis, Raman) and kinetic methods were used to
characterize the formation of intermediates under the reac-
tion conditions. The knowledge obtained by this was used to
further optimize the previous conditions and generate a more
active catalytic system.
Introduction
The synthesis of allylic products from easily available starting
materials is an important research topic in modern organic
synthesis.[1] The allylic moiety enables further functionalizations
such as cyclopropanation, epoxidation or hydroboration,
among others.[1a] Besides, the generation of prochiral branched
allylic products is a valuable tool in asymmetric synthesis.[1a]
Significant progress in this research field was achieved by
transition metal catalysis, especially by allylic substitution[2] and
allylic CÀ H oxidation.[3] However, these methods have intrinsic
disadvantages, such as the presence of a leaving group in the
substrate or the use of stoichiometric amounts of oxidant. For
the synthesis of linear allylic products some more atom
economic methods mostly based on palladium catalysis were
first published by Trost and Yamamoto.[4] In a seminal work,
Breit et al. reported the first example of the branched selective
synthesis of allylic esters by addition of carboxylic acids to
terminal alkynes applying rhodium catalysis (Scheme 1).[5] This
rhodium complex catalyzed propargylic CÀ H activation is
Scheme 1. Rhodium complex-catalyzed formation of allylic esters.
superior to competing methods due to its intrinsic redox
neutrality and atom economy.[5] This reaction is characterized
by high yields and excellent regio- (up to >98:2 branched to
Markovnikov ester) and enantio- (up to 93% ee) selectivities.[5,6]
Beyond carboxylic acids,[5,6,7] the potential of this reaction could
also be applied to sulfonylhydrazines, pyrazoles and ketones.[8]
Apart from this, the variation of the alkyne enables
a
considerable increase in the range of products.[9] First evidences
of the complexity of this reaction were published by the groups
of Breit and Heller.[5,6,7,10]
Preliminary directional studies on the reaction kinetics of
this rhodium complex catalyzed propargylic CÀ H activation
were performed by in situ IR spectroscopy.[10a] Using the initial
rate method, the reaction was found to show zero-order
dependence in substrates and first-order behavior in catalyst.
Both product formation and substrate consumption can there-
fore be expected to be constant over time. However, this does
not correspond to the in situ IR measurements, since a decrease
in reaction rate was detected with increasing conversion.[10a]
These obvious contradictions could not be explained so far and
are part of more detailed mechanistic and kinetic investigations
in the context of the present work. In addition to extensive
NMR studies on isolated elementary steps of the published
catalytic cycle,[10a] isolable intermediates were characterized by
X-ray crystal structure analysis. The combination of these
findings provides the basis for further optimizations of the
catalytic propargylic CÀ H activation and should allow to extend
applications of this reaction.
[a] Dr. S. Möller, N. Jannsen, Dr. J. Rüger, Dr. H.-J. Drexler, M. Horstmann,
Prof. Dr. D. Heller
Leibniz-Institut für Katalyse e.V.
Albert-Einstein-Str. 29a, 18059 Rostock (Germany)
E-mail: detlef.heller@catalysis.de
[b] Dr. F. Bauer, Prof. Dr. B. Breit
Institut für Organische Chemie
Albert-Ludwigs-Universität Freiburg
Albertstr. 21, 79104 Freiburg (Germany)
E-mail: bernhard.breit@chemie.uni-freiburg.de
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/chem.202102219
© 2021 The Authors. Chemistry - A European Journal published by Wiley-
VCH GmbH. This is an open access article under the terms of the Creative
Commons Attribution Non-Commercial NoDerivs License, which permits use
and distribution in any medium, provided the original work is properly cited,
the use is non-commercial and no modifications or adaptations are made.
Chem. Eur. J. 2021, 27, 1–9
1
© 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202102219
Chemistry—A European Journal
Results and Discussion
species was also followed by 31P NMR spectroscopy, the
evaluation of these data should provide an explanation (Fig-
ure S4). Despite the large number of signals, clear trends can be
observed. For example, during the reaction a broad doublet of
an unknown species disappears at 16.2 ppm and another
doublet at 26.3 ppm grows in intensity. The signals of the σ-
allyl complex (4 in blue), which corresponds to the “resting
state”[14] complex as active catalyst,[10a] can be clearly assigned.
When crystals of the σ-allyl complex are dissolved in 1,2-
dichloroethane (Figure S5), signals of another species appear in
the 31P NMR spectrum (dd, δ=34.9 ppm; dd, δ=30.6 ppm).
These can be assigned to a π-allyl complex (5 in yellow) in
equilibrium with the σ-allyl complex 4.[10a] The 31P NMR spectra
confirm that no phosphorus-containing rhodium complex
remains constant over the reaction time, which supports the
above-made assumption of a non-constant catalyst concentra-
tion (Figure 2). Furthermore, a very prominent set of signals
(Figure S4, marked in red, dd, δ=35.6 ppm; dd, δ=32.8 ppm) is
present almost quantitatively at the end of the reaction. If the
integrals of the individual signal sets are related to each other,
time-dependent data proportional to the concentration of the
rhodium species can be visualized.
In the context of the present work, in situ operando NMR
investigations of rhodium complex catalyzed propargylic CÀ H
activation were performed for the first time. When monitoring
the reaction of benzoic acid and 1-octyne as model compounds
°
at T=70 C using the in situ system [Rh(COD)(μ2-Cl)]2/DPEPhos
as the catalyst by multinuclear NMR spectroscopy for t=
870 min (14.5 h) complex 1H NMR spectra are obtained (Fig-
ure S1). It is remarkable that in addition to the esters described
so far (branched allylic ester 1 and Markovnikov ester 2),
another product is obviously formed, which most likely
possesses vinylic CÀ H atoms according to its NMR signals (ddq,
δ=4.94 ppm; ddq, δ=5.07 ppm). It is known from the literature
that rhodium and other transition metal complexes catalyze the
isomerization of alkynes[11] and allenes[12] into the corresponding
dienes. A reference measurement confirms the hypothesis that
the light gray marked signals in the range of 5.11–4.91 ppm are
caused by 1,3-octadiene 3, which is generated due to alkyne or
allene isomerization during the catalytic cycle[13] (for NMR data
see Supporting Information Figure S2). In a test experiment
using 1,3-octadiene as substrate under similar conditions
°
(2.5 mol% catalyst, 70 C, 16 h see Supporting Information
Figure S3) no conversion was detected. Due to this we assume
the generation of 3 by isomerization to be irreversible.
At the beginning of the reaction approx. 60% of the σ-allyl
complex 4 (“resting state” complex, blue) and 20% of the π-allyl
complex 5 (active catalyst, yellow) are present. The time
dependence of the complex 6 (marked red), which is present
almost quantitatively at the end of the reaction, possibly
indicates the formation of a catalytically inactive species.
Numerical rank matrix analyses are often used to determine
the number of linearly independent partial reactions. However,
this problem can also be solved graphically, using so-called
concentration diagrams.[15] This not only determines the number
of linearly independent partial reactions, but also provides
information about the stoichiometric coefficients of the
reaction.[15] The plot shown in Figure 3 shows an approximately
linear relationship between the concentrations of the products
(Markovnikov ester 2 plotted against the branched allylic ester
1). This indicates that 2 is not formed from 1 (or vice versa) but
A plot of the corresponding concentrations of the substrates
and products as a function of time gives the concentration-time
diagram shown in Figure 1. The generated 1,3-octadiene 3 is
formed in approximately the same concentration as the
Markovnikov ester 2.
Evidently, no linear dependency for substrate decrease or
product increase can be observed, as it would be expected
from the previous kinetic investigations (zero order depend-
ence) using in situ IR measurements.[10a] However, if the
concentration of the active catalyst decreases during the
reaction, for example due to the formation of a catalytically
inactive rhodium complex, the decrease in the rate of product
formation in the reaction progress, as shown in Figure 1, can be
easily explained. As the formation of the corresponding catalyst
Figure 2. Ratio of rhodium complexes obtained from 31P NMR spectra
(Figure S4) as a function of time; the sum of all signals remains constant over
the course of the reaction within the limits of measurement accuracy.
Figure 1. Concentrations of substrates and products obtained from 1H NMR
spectra (Figure S1) as a function of time.
Chem. Eur. J. 2021, 27, 1–9
www.chemeurj.org
2
© 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202102219
Chemistry—A European Journal
that both products are the result of a parallel reaction, which is
in line with the mechanistic proposal made earlier.[10a]
In contrast to the analysis of the concentration data for the
products, the plot of the percentage of active catalyst species (4
and 5) as a function of the probably inactive species 6 gives a
straight line which shows that the species primarily present at
the end of the reaction is formed from the mixture of σ- and π-
allyl complex in the sense of a subsequent reaction (Figure 4).
The 31P NMR spectrum of the addition of benzoic acid to 1-
°
octyne catalyzed by [Rh(COD)(μ2-Cl)]2/DPEPhos (in situ) at 70 C
in 1,2-dichloroethane almost exclusively contains two signal
sets after 15 h of reaction time (Figure 5). Precipitation of
rhodium complexes from this reaction mixture is described in
the synthetic procedures (Supporting Information). By manual
sorting under inert atmosphere, the separation of crystals
suitable for X-ray crystal structure analysis was possible (Fig-
ure S6).
Figure 3. Concentration diagram according to Mauser;[15] plot of the
concentration of the Markovnikov ester 2 as a function of the concentration
of the branched allylic ester 1 determined from 1H NMR data (Figure S1).
The connectivity found in the molecular structure (Figure 6)
is in favor of the assumed formation of the already discussed σ-
allyl complex 4[10a] (Figure S7, right). It shows the rhodium(III)
center in octahedral coordination geometry surrounded by the
diphosphine, chloride, a benzoate ligand and a vinyl ligand that
originates from 1-octyne. However, bond lengths of the alkyl
chain point to a different position of the CÀ C double bond
(Figure S7) and indicate the presence of the σ-vinyl complex
with a terminal double bond instead of the σ-allyl species.
Further details on the characterization of this compound by
multinuclear NMR experiments can be found in the Supporting
Information (Figures S7–S14).
To further distinguish between the σ-vinyl and σ-allyl
complex, Raman measurements were carried out (Figure 7)
allowing for an unequivocal assignment of the position of
carbon-carbon multiple bonds. The Raman bands were assigned
by quantum mechanical spectra simulations (Supporting In-
formation) and reference measurements of the DPEPhos ligand
and the benzoic acid (Figure S15 and S16). The characteristic
stretching vibration of the CÀ C double bond in the σ-allyl
complex 4 appears at 1655 cmÀ 1 while in the σ-vinyl complex 6
Figure 4. Concentration diagram according to Mauser;[15] plot of the
percentage of active catalyst species (σ- and π-allyl complex, 4 and 5) as a
function of the percentage of catalytically inactive species 6 determined
from the 31P NMR data (Figure S4).[16]
Figure 5. 31P NMR spectra of the catalyzed addition of [benzoic acid]-
=0.47 mol/L and [1-octyne]=0.61 mol/L using the in situ generated precata-
°
lyst [[Rh(DPEPhos)(μ2-Cl)]2] 7=0.023 mol/L in 1,2-dichloroethane-d4 at 70 C
after 900 minutes; σ-allyl complex (4, *); unknown species which may be
catalytically inactive (6, *).
Figure 6. Molecular structure of the σ-vinyl complex 6. Hydrogen atoms are
omitted for clarity (ORTEP, 30% probability ellipsoids).
Chem. Eur. J. 2021, 27, 1–9
www.chemeurj.org
3
© 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202102219
Chemistry—A European Journal
Figure 7. Raman spectra of a) the σ-allyl complex 4, C=C stretching vibration
at 1655 cmÀ 1; b) the σ-vinyl complex 6, vinylic C=C stretching vibration at
1608 cmÀ 1
.
it is found bathochromically shifted by 47 cmÀ 1 (Figure 7). The
band at 1608 cmÀ 1 is therefore in the typical range for terminal
carbon-carbon double bonds.[17]
The σ-vinyl complex 6 is obviously responsible for catalyst
deactivation during the reaction, which becomes dominant
with increasing conversion. An NMR scale experiment using 6
as catalyst proves that this species is indeed catalytically
inactive (Figure S17).
To elucidate the mechanism for the formation of the
catalytically inactive σ-vinyl complex 6, the isolable σ-allyl
complex 4 (i.e. the “resting state“ compound) was reacted with
Figure 8. Synthesis of a SiMe3 substituted σ-vinyl complex 9, 31P NMR
spectrum of the reaction of σ-allyl complex 4 with an excess of trimeth-
ylsilylacetylene 8 in 1,2-dichloroethane-d4 and the molecular structure of the
SiMe3 substituted σ-vinyl complex 9. Hydrogen atoms are omitted for clarity
(ORTEP, 30% probability ellipsoids).
°
all reactants at both room temperature as well as at T=70 C
and examined using 31P NMR spectroscopy (Figure S18). These
experiments clearly show that the catalytically inactive σ-vinyl
complex 6 is formed from reaction of the σ-allyl compound 4
°
with 1-octyne at 70 C. To further verify whether this reaction
Scheme 2. Disproportionation of the monohydride benzoate complex 10.
sequence corresponds to an exchange of the alkyl ligand, the
σ-allyl complex 4 was reacted with trimethylsilylacetylene 8. In
fact, it was possible to isolate a SiMe3 substituted σ-vinyl
complex 9, which proves the assumption of an exchange
reaction and excludes for example a simple (thermal) rearrange-
ment (Figure 8).
If the precatalyst [Rh(DPEPhos)(Cl)]2 7 is reacted with an
°
excess of benzoic acid at 70 C, the additional rhodium species
10 was formed (Scheme 2). 10 can be isolated and character-
ized by NMR spectroscopy. It corresponds to a product of the
oxidative addition of benzoic acid to the monomerized
precatalyst 11 and gives a hydride signal at À 15.15 ppm (dt,
Figure 9. 31P NMR spectrum of the conversion of [Rh(DPEPhos)(μ2-Cl)]2 and
°
benzoic acid (in the ratio 1:54) at 70 C in 1,2-dichloroethane-d4.
J
Rh–H =11.8 Hz, JP–H =17.4 Hz) in 1H NMR spectrum. By 1H-31P-
HMBC NMR spectroscopy (Figure S19) the hydride signal could
be clearly assigned to the doublet in the 31P NMR spectrum at
43.03 ppm (Figure 9).
group and two chloride ligands. During the reaction it can be
observed in H NMR (Figure S22) that the free cyclooctadiene
1
If the precatalyst is generated in situ and reacted with
benzoic acid, another benzoate complex 12 (27.86 ppm, d in
31P NMR spectrum) is formed in addition to the hydrido
benzoate complex (Figure S20), presumably by disproportiona-
tion of 10 (Scheme 2).
In this rhodium(III) species the hydride ligand is replaced by
another chlorine atom, forming the dichlorido benzoate
complex 12. The molecule was characterized by NMR and X-ray
crystal structure analysis (Figures S21) and possesses an octahe-
dral rhodium(III) center coordinated with DPEPhos, a benzoate
generated as a result of the in situ production of the precatalyst
reacts to cyclooctene. If 10 is converted with 1-octyne, an
immediate formation of the σ-allyl 4 and σ-vinyl complex 6 can
already be observed at room temperature (Figure S23). The
reaction is too fast for kinetic NMR investigations. However, if
12 is converted with 1-octyne, the reaction is slow enough for
NMR spectroscopic monitoring. This NMR spectroscopic mon-
itoring confirms the formation of the σ-vinyl species 6 from this
benzoate complex 10 (Figure S24). Additional validation of this
Chem. Eur. J. 2021, 27, 1–9
www.chemeurj.org
4
© 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202102219
Chemistry—A European Journal
conversion is provided by the concentration diagram of these
data (Figure 10).
2.9·10À 6 mmol/mL·s but it is also converted into the catalytically
inactive σ-vinyl species in the presence of 1-octyne
6
The reaction from 12 to 6 is likely to take place via benzoate
complex 10. The mechanism for this step is still unclear, but
due to the excess of 1-octyne, a reaction of it seems to be a
likely hydride source for the formation of 10. If the reaction
from 10 to 6 takes place immediately at room temperature,
then it can be assumed that the formation of the monohydride
species 10 from 12 is the slowest step of this catalytic reaction.
The resulting linear relationship with a slope of approximately
(2.5·10À 6 mmol/mL·s). Its stationary concentration during catal-
ysis is thus very small. With these initial rates it can now be
explained why only a small portion of 12 (approx. 4%) could be
detected during monitoring of the catalytic reaction (Figure S3,
grey-blue species). Although the formation of the σ-vinyl
complex 6 is by an order of magnitude slower than the actual
catalytic reaction (Table 1), the formation of the catalytically
inactive species occurs irreversibly. Thus, it can be concluded
that almost all rhodium catalyst is irretrievably deactivated
during propargylic CÀ H activation under the used standard
reaction conditions. Even repeated addition of substrate would
not lead to further product formation.
°
À 45 proves that the catalytically inactive σ-vinyl complex 6 is
formed pseudo directly from the benzoate complex 12
(Scheme 3).
In order to relate this deactivation path to the original
catalysis, initial rates of these individual steps as well as of the
rhodium complex catalyzed propargylic CÀ H activation, i.e. the
total reaction, were determined (Table 1).[18]
This analysis shows that formation of the benzoate complex
12 and the σ-vinyl species 6 occur at approximately equal rates.
The presented results significantly extend the initially
proposed reaction pathway[10a] (Figure S25) and allow for the
refinement of the mechanistic picture (Figure 11). After an
experimentally confirmed monomerization of 7 to 11 of the
precatalyst 7,[19] coordination of 1-octyne occurs, as described
earlier.[10a] However, the results also indicate that the alkyne
complex 14 is not asymmetric, as previously assumed but has a
symmetrical structure and is neutral in nature due to the
coordination of two alkyne molecules.[20] Based on NMR data (d,
δ=23.3 ppm, JRhP =117 Hz) and comparison with literature data
([RhCl(NBD)(DPEPhos)], d, δ=15.5 ppm, JRhP =120 Hz),[21] this
compound is most likely a rhodium(I) complex that possesses a
fivefold coordinated metal center (Figure S26). Furthermore, its
formation is reversible.[22] Subsequent ligand exchange with
benzoic acid and intramolecular alkyne protonation leads via
transition state 15 to the vinyl species 16. Competing dimeriza-
tion of the alkyne from complex 14 is also formally feasible but
could be excluded experimentally.[23] From 16 reductive elimi-
nation explains the formation of the side product, the
Markovnikov ester 2 (Figure S27). β-Hydride elimination leads
via the allene hydrido complex 17 to σ-/π-allyl equilibrium 4
and 5.[10a] Reversible reductive elimination of the branched allyl
ester 1 from π-allyl complex 5 closes the catalytic cycle and
regenerates 11.[10a]
The benzoate complex 12 is formed with
a rate of
Figure 10. Concentration diagram according to Mauser;[15a] plot of the
percentage of the benzoate complex 12 determined from 31P NMR data as a
function of the percentage of the σ-vinyl complex 6.
The formation of the side product 1,3-octadiene 3 could
likely be explained by the hydrometallation of the internal
double bond of the allene ligand in complex 17. The so formed
internal σ-allyl complex 18 could react by a consecutive
reductive elimination to the monohydride benzoate complex
10 and release 3.
Although a reaction of the precatalyst 7 [Rh(DPEPhos)(μ2-
Cl)]2 with benzoic acid at room temperature was excluded in
our earlier work,[10a] new investigations at 70 C, i.e. under
Scheme 3. Reaction of the dichloride benzoate complex 12, forming the σ-
vinyl complex 6.
°
standard reaction conditions, have shown that a parallel
reaction in fact takes place. The benzoate complex 10 from this
conversion and its formation was explored in detail. Thus, in the
presence of an olefin or diolefin (like COD due to in situ
precatalyst generation) the benzoate complex 12 is visible, as a
result of a disproportionation of 10 and hydrogenation of the
respective olefin by 13. The catalyst deactivation associated
with benzoate complexes 10 (or 12) is an important key step.
By coordinating 1-octyne as a vinyl ligand to the benzoate
Table 1. Initial rates of the total reaction as well as the formation of the
benzoate complex 12 and the σ-vinyl species 6. ([catalyst]=0.0023 mol/L,
[alkyne]=0.61 mol/L, [benzoic acid]=0.47 mol/L in 1,2-dichloroethane-d4
°
at 70 C)
reaction sequence
Initial rates [mmol/mL·s]
total reaction (product formation)
formation benzoate complex 12
formation σ-vinyl complex 6
4.1·10À 5
2.9·10À 6
2.5·10À 6
Chem. Eur. J. 2021, 27, 1–9
www.chemeurj.org
5
© 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202102219
Chemistry—A European Journal
Figure 11. Modified reaction mechanism of rhodium complex catalyzed propargylic CÀ H activation based on the results presented in this work.
complex 10, the catalytically inactive σ-vinyl complex 6 is
formed irreversibly.
In order to minimize the formation of the σ-vinyl complex 6,
now recognized as the reason for catalyst deactivation, the
formation of its precursor, the benzoate complex 10 must
therefore be reduced as much as possible. Due to this, it is
desirable to increase the concentration of alkyne in solution at
constant benzoic acid concentration to facilitate the alkyne
coordination to 11 to form 14. By this, the actual catalytic cycle
is preferred over the competitive formation of the benzoate
complex 10 and the resulting catalytically inactive σ-vinyl
species 6. In fact, it could be shown experimentally that an
increase of concentration of alkyne from 1.2 eq. to 4.7 eq. in the
rhodium complex catalyzed addition of benzoic acid to 1-
Scheme 4. Optimized rhodium complex catalyzed propargylic CÀ H activa-
tion; the catalyst concentration given is based on the amount of used
benzoic acid.
octyne leads to a reduced formation of the catalytically inactive Conclusion
σ-vinyl compound 6 (Figure S28).[24] The results show that a
significantly lower catalyst/substrate ratio is possible and the
catalyst concentration can be reduced from initially 5 mol% to
2.5 mol% with unchanged high selectivity (Scheme 4).
Besides detailed kinetic measurements by using in situ oper-
ando NMR spectroscopic reaction monitoring, the studies
presented include mechanistic investigations by using NMR and
Raman spectroscopy as well as the structural characterization of
intermediates of the catalytic cycle with X-ray crystal structure
analysis. Especially the interpretation of NMR data by using so-
Chem. Eur. J. 2021, 27, 1–9
www.chemeurj.org
6
© 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202102219
Chemistry—A European Journal
Asao, J. Am. Chem. Soc. 1994, 116, 6019–6020; c) B. M. Trost, V. J. Gerusz,
J. Am. Chem. Soc. 1995, 117, 5156–5157; d) Y. Yamamoto, M. Al-Masum,
N. Fujiwara, N. Asao, Tetrahedron Lett. 1995, 36, 2811–2814; e) B. M.
Trost, P.-Y. Michellys, V. J. Gerusz, Angew. Chem. Int. Ed. Engl. 1997, 36,
1750–1753; f) M. Al-Masum, Y. Yamamoto, J. Am. Chem. Soc. 1998, 120,
3809–3810; g) I. Kadota, A. Shibuya, Y. S. Gyoung, Y. Yamamoto, J. Am.
Chem. Soc. 1998, 120, 10262–10263; h) N. T. Patil, N. K. Pahadi, Y.
Yamamoto, Synthesis 2004, 2186–2190; i) N. T. Patil, F. Nawaz Khan, Y.
Yamamoto, Tetrahedron Lett. 2004, 45, 8497–8499; j) N. T. Patil, N. K.
Pahadi, Y. Yamamoto, Can. J. Chem. 2005, 83, 569–573.
called concentration diagrams provided important information
on the parallel formation of the ester products and the
generation of the catalytically inactive species 6. On the basis of
these studies, individual elementary steps of the catalytic
propargylic CÀ H activation reaction by using the precatalyst
[Rh(DPEPhos)(μ2-Cl)]2 7 could be specifically investigated. Thus,
although this method is rarely used, its application to complex
catalytic reactions is a very valuable tool.
The suspected catalyst deactivation could be confirmed
experimentally for the first time by identifying a catalytically
inactive rhodium(III) σ-vinyl species 6, which is formed from a
rhodium(III) benzoate complex 10 at low alkyne concentrations.
After considering these new insights into the reaction pathway
for optimizations of the catalytic protocol, it was possible to
more than double the productivity of the catalyst system in
initial test reactions. Further studies will focus on the optimized
slow addition of benzoic acid for example by syringe pump in
combination with in situ IR-spectroscopy to minimize the
catalyst deactivation.
[5] A. Lumbroso, P. Koschker, N. R. Vautravers, B. Breit, J. Am. Chem. Soc.
2011, 133, 2386–2389.
[6] P. Koschker, M. Kähny, B. Breit, J. Am. Chem. Soc. 2015, 137, 3131–3137.
[7] A. Lumbroso, N. Abermil, B. Breit, Chem. Sci. 2012, 3, 789–793.
[8] a) K. Xu, V. Khakyzadeh, T. Bury, B. Breit, J. Am. Chem. Soc. 2014, 136,
16124–16127; b) T. M. Beck, B. Breit, Org. Lett. 2016, 18, 124–127;
c) A. M. Haydl, L. J. Hilpert, B. Breit, Chem. Eur. J. 2016, 22, 6547–6551.
[9] a) T. M. Beck, B. Breit, Eur. J. Org. Chem. 2016, 5839–5844; b) C. Li, C. P.
Grugel, B. Breit, Chem. Commun. 2016, 7, 5840–5843; c) Z. Liu, B. Breit,
Angew. Chem. Int. Ed. 2016, 55, 8440–8443; Angew. Chem. 2016, 128,
8580–8583; d) J. Kuang, S. Parveen, B. Breit, Angew. Chem. Int. Ed. 2017,
56, 8422–8425; Angew. Chem. 2017, 129, 8542–8545; e) D. Berthold, B.
Breit, Org. Lett. 2018, 20, 3, 598–601; f) J. Zheng, B. Breit, Angew. Chem.
Int. Ed. 2018, 58, 3392–3397; Angew. Chem. 2018, 131, 3430–3435.
[10] a) U. Gellrich, A. Meissner, A. Steffani, M. Kähny, H. J. Drexler, D. Heller,
D. A. Plattner, B. Breit, J. Am. Chem. Soc. 2014, 136, 1097–1104; b) P.
Koschker, B. Breit, Acc. Chem. Res. 2016, 49, 1524–1536.
[11] R. Shintani, W. L. Duan, S. Park, T. Hayashi, Chem. Commun. 2006, 3646–
Experimental Section
For experimental details, see the Supporting Information.
3647.
[12] C. M. Ting, Y. L. Hsu, R. S. Liu, Chem. Commun. 2012, 48, 6577–6579.
[13] Test reactions with the precatalyst and the alkyne under the exclusion
of benzoic acid have shown that the dimeric alkyne is formed instead of
the diene. The diene detected during catalysis should therefore result
from a complex of the catalytic cycle.
[14] A “resting state” is typically defined as the coordination compound that
is present in highest concentration under stationary conditions. It
should however be kept in mind that in catalysis that can be described
using Michaelis-Menten type kinetics, the pre-equilibrium between
catalyst and catalyst-substrate complex can be altered by modification
of experimental conditions (such as substrate concentration), which in
consequence changes the resting state during catalysis.
Crystal-structure analysis: Deposition Numbers 2000244 (6),
2000245 (9) and 2000243 (12) contain the supplementary crystallo-
graphic data for this paper. These data are provided free of charge
by the joint Cambridge Crystallographic Data Centre and Fachinfor-
mationszentrum Karlsruhe Access Structures service www.ccdc.ca-
m.ac.uk/structures.
Acknowledgements
[15] a) H. Mauser, Formale Kinetik-Experimentelle Methoden der Physik und der
Chemie, Bertelsmann Universitätsverlag, 1974; b) J. Polster, Talanta
1984, 31, 113–116; c) B. Ursin, H. Kleindienst, Zeitschrift für Naturfor-
schung 1986, 41, 989–997; d) H. Mauser, G. Gauglitz Photokinetics:
Theoretical Fundamentals and Applications. Elsevier, 1998.
[16] Even if the equilibrium between the allyl complexes 4 and 5 is not
adjusted, the sum of both complex species does not change and is
therefore independent of the status of the equilibrium. For this reason,
the sum of the allyl complexes was plotted as a function of the
percentage of catalytically inactive species 6 in the concentration
diagram.
We thank the DFG for the financial support (grant No. HE2890/
7-1 and BR1646/10-2). Open access funding enabled and
organized by Projekt DEAL.
Conflict of Interest
[17] G. Socrates, in Infrared and Raman Characteristic Group Frequencies:
Tables and Charts, Wiley, 2004.
The authors declare no conflict of interest.
[18] Based on the initial part of the concentration-time data (ca. 30–40%
conversion) of the reaction monitoring the formation rate of the esters
1 and 2, the benzoate complex 12 and the σ-vinyl compound 6 have
been determined. For this purpose, the sum of two exponential
functions was chosen as a differentiable substitute function, whereby
this corresponds to the respective initial rate at time t=0 (calculation
example in Figure S30).
Keywords: catalyst deactivation · concentration diagrams · CÀ H
activation · kinetics · rhodium catalysis
[19] A. Mannu, M. Ferro, S. Möller, D. Heller, J. Chem. Res. 2018, 42, 402–404.
[20] The doublet as coupling pattern in the 31P NMR spectrum shows that
the molecule is probably symmetrical and therefore does not
correspond to the originally postulated asymmetrical compound [RhCl-
(DPEPhos)(1-octyne)]. In addition, the resulting species cannot be
converted with benzene to an arene complex, so it can be assumed
that it is not cationic. (C. Fischer, R. Thede, H.-J. Drexler, A. König, W.
Baumann, D. Heller, Chem. Eur. J. 2012, 18, 11920–11928). By adding 1-
octyne to the pentacoordinated complex [RhCl(DPEPhos)(NBD)], (A.
Meißner, A. König, H.-J. Drexler, R. Thede, C. Fischer, W. Baumann, D.
Heller, Chem. Eur. J. 2014, 20, 14721–14728) the species sought can be
detected spectroscopically by 31P NMR. It can be concluded that the
signal (22.4 ppm; d, JRhP =117 ppm) can be assigned to the pentacoordi-
nated species [RhCl(DPEPhos)(1-octyne)2].
[1] a) A. Lumbroso, M. L. Cooke, B. Breit, Angew. Chem. Int. Ed., 2013, 52,
1890–1932; Angew. Chem. 2013, 125, 1942–1986; Angew. Chem. Int. Ed.
2013, 52, 1890–1932; b) J. F. Hartwig, L. M. Stanley, Acc. Chem. Res.
2010, 43, 1461–1475.
[2] a) B. M. Trost, D. L. van Vranken, Chem. Rev. 1996, 96, 395–422; b) B. M.
Trost, L. M. Crawley, Chem. Rev. 2003, 103, 2921–2944; c) Z. Lu, S. Ma,
Angew. Chem. Int. Ed. 2008, 47, 258–297; Angew. Chem. 2008, 120, 264–
303.
[3] a) M. S. Chen, M. C. White, J. Am. Chem. Soc. 2004, 126, 1346–1347; b) G.
Liu, S. S. Stahl, J. Am. Chem. Soc. 2007, 129, 6328–6335; c) G. Yin, Y. Wu,
G. Liu, J. Am. Chem. Soc. 2010, 132, 11978–11987.
[4] a) B. M. Trost, W. Brieden, K. H. Baringhaus, Angew. Chem. Int. Ed. Engl.
1992, 31, 1335–1336; Angew. Chem. 1992, 104, 1392–1394; Angew.
Chem. Int. Ed. 1992, 31, 1335–1336; b) Y. Yamamoto, M. Al-Masum, N.
Chem. Eur. J. 2021, 27, 1–9
www.chemeurj.org
7
© 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202102219
Chemistry—A European Journal
[21] S. Möller, H.-J. Drexler, D. Heller, Acta Crystallogr. 2019, C75, 1434–1438.
[22] In order to test the reversibility of the formation of this new compound,
the reaction mixture of precatalyst and 1-octyne was evacuated. If the
reaction to the new alkyne complex is irreversible, no signals from the
precatalyst should be detectable after removal of the alkyne by
vacuum. However, it can be observed that after reducing the alkyne
concentration in solution, more than half of the alkyne complex has
been transformed into other phosphorus-containing rhodium com-
plexes, including the precatalyst.
[24] The remaining 1-octyne due to the stoichiometry of the reaction could
be separated from the products and reused after the catalytic reaction
is completed.
Manuscript received: June 21, 2021
Accepted manuscript online: July 14, 2021
Version of record online: ■■■, ■■■■
[23] The signals of the dimeric alkyne were found only under exclusion of
benzoic acid, but not in the catalytic reaction.
Chem. Eur. J. 2021, 27, 1–9
www.chemeurj.org
8
© 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
Dr. S. Möller, N. Jannsen, Dr. J. Rüger,
Dr. H.-J. Drexler, M. Horstmann, Dr. F.
Bauer, Prof. Dr. B. Breit*, Prof. Dr. D.
Heller*
1 – 9
Catalyst Deactivation During
Rhodium Complex-Catalyzed Propar-
gylic CÀ H Activation
In this study, the complete mechanis-
tic elucidation of rhodium-catalyzed
propargylic CÀ H activation is
activated species has been studied in
detail by NMR and Raman spectro-
scopy as well as X-ray crystallography.
Furthermore, the present work
includes a kinetic study carried out by
in situ operando NMR spectroscopy.
presented. Particular attention is paid
to the deactivation pathway of the
catalytic reaction. The associated de-