Strategic Approach on N-Oxides in Gold Catalysis – A Case Study

Article abstract of DOI:10.1002/adsc.201801007

An extensive kinetic study of selected key reactions of (oxidative) gold catalysis concentrates on the decrease of the catalytic activity due to inhibition of the gold(I) catalyst caused by pyridine derivatives that are obtained as by-products if N-oxides are applied as oxygen donors. The choice of the examined pyridine derivatives and their corresponding N-oxides has been made regardless of their commercial availability; particular attention has been paid to the practical benefit which up to now has been neglected in most of the reaction screenings. The test reactions were monitored by GC and ¹H NMR spectroscopy. The received reaction constants provide information concerning a correlation between the electronic structure of the heterocycle and the catalytic activity. Based on the collected kinetic data, it was possible to develop a basic set of three N-oxides which have to be taken into account in further oxidative gold(I)-catalyzed reactions. (Figure presented.).

Full text of DOI:10.1002/adsc.201801007

Title: Strategic Approach on N-Oxides in Gold Catalysis – A Case
Study
Authors: Jasmin Schießl, Philipp Stein, Judith Stirn, Kirsten Emler,
Matthias Rudolph, Frank Rominger, and A. Stephen K.
Hashmi
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: Adv. Synth. Catal. 10.1002/adsc.201801007
Link to VoR: http://dx.doi.org/10.1002/adsc.201801007

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
VERY IMPORTANT PUBLICATION
Strategic Approach on N-Oxides in Gold Catalysis – A Case Study
Jasmin Schießl,a Philipp M. Stein,a Judith Stirn,a Kirsten Emler,a Matthias Rudolph,a
Frank Rominger,a A. Stephen K. Hashmia,*
a Organisch-Chemisches Institut, Heidelberg University, Im Neuenheimer Feld 270,
69120 Heidelberg, Germany.
E-mail: hashmi@hashmi.de
Abstract
An extensive kinetic study of selected key reactions of (oxidative) gold catalysis
concentrates on the decrease of the catalytic activity due to inhibition of the gold(I)
catalyst caused by pyridine derivatives that are obtained as by-products if N-oxides
are applied as oxygen donors. The choice of the examined pyridine derivatives and
their corresponding N-oxides has been made regardless of their commercial
availability; particular attention has been paid to the practical benefit which up to now
has been neglected in most of the reaction screenings. The test reactions were
monitored by GC and 1H NMR spectroscopy. The received reaction constants provide
information concerning a correlation between the electronic structure of the
heterocycle and the catalytic activity. Based on the collected kinetic data, it was
possible to develop a basic set of three N-oxides which have to be taken into account
in further oxidative gold(I)-catalyzed reactions.
Keywords
N-oxides; alkynes; kinetic studies; reaction optimization; catalyst inhibition; gold
catalysis
Introduction
The impact of homogeneous gold catalysis increased over the last decades.[1] Since
2010,[2] gold-catalyzed oxidative cyclizations based on -oxo-carbenoids generated by
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
the attack of heteroaromatic N-oxides onto -activated alkynes became one of the
moist fruitful reaction classes in the field of gold catalysis.[3] Triggered by this reactivity
pattern demanding transformations such as C-H insertions, cyclopropanations and
ylide formations were enabled which provide an attractive replacement of α-diazo
carbonyl compounds that were formerly used as precursors for -oxo-carbenoids.[4]
The quest for finding a suitable N-oxide for an oxidative gold-catalyzed transformation
is often not easy as the pyridine by-products can lead to catalyst deactivation at higher
conversions of the starting materials.[5] Interestingly despite this task, the reported
choices of N-oxides merely depend on its commercial availability, despite the usually
simple accessibility of N-oxides. This becomes obvious by an analysis of publications
concerning experimental studies on this topic published from 2010 to 2018 (46
publications at 11.06.2018, Scifinder, keyword: gold catalysis N-oxide; only
experimental studies are included). In sum, 32 different N-oxides were investigated.
Overall, the average number of screened N-oxides is 4.2 per publication. 50 % of these
publications deal with the oxidants like pyridine N-oxide, 3,5-dichloropyridine N-oxide
or 8-methylquinoline N-oxide. 2-Bromopyridine N-oxide, 2,6-dibromopyridine N-oxide,
2,6-dichloropyridine N-oxide, 4-methylpyridine N-oxide or 8-isopropylquinoline N-oxide
is only used in 20 % of the reports. One or more of the remaining 24 N-oxides were
investigated in less than 10 % of these publications. This analysis shows that the
choice of N-oxide is mostly based on its commercial availability and not on its practical
benefit. Already published experimental data[3c,6] and mechanistic studies[7] report that
the efficiency of the formation of the -oxo gold carbenoid on basis of N-oxides is
significantly affected by the formation of unavoidable by-products that can deactivate
the active gold(I) species. Regarding this matter, Gagosz et al. postulated a series of
potential equilibria which affect the activated gold species.[5] Especially, the
deactivation of catalyst caused by the coordination of the pyridine N-oxide and the
corresponding pyridine is significant. Without any doubt, a systematic investigation of
the N-oxide influence on oxidative gold catalysis is needed. This encouraged us to
shed light on the kinetics of these transformations as well as the significance of not yet
or rarely applied N-oxides in homogeneous gold catalysis; regardless of their
commercial availability. It should be noted that acid-assisted gold(I)-catalyzed
reactions are not part of this analysis, however, of a different and still ongoing study.
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
Results and Discussion
As an entry point, the dependency of the inhibition of the gold(I) catalyst on different
pyridine derivatives was examined. To accomplish this, various pyridine gold(I)
complexes and one N-methylmorpholine gold(I) complex were initially isolated.
Noteworthy, only a few known pyridine gold(I) complexes are published to date.[8] For
an effective synthesis, one equivalent of the gold(I) complex Ph3PAuNTf2 ([Au] =
0.034 mmol/ml) was treated with one equivalent of a pyridine derivative in DCM at
25 °C for 3 h (Scheme 1). With respect to the pyridine derivatives, representatives of
commonly used N-oxides and pyridines which are characterized by varying electronic
or steric effects were chosen. A precise ratio of the reactants turned out to be crucial
as column chromatography was impossible due to the decomposition of the complex.
After purification by recrystallization, the gold(I) complexes 1a-r were obtained in
excellent yields (85 – 100 %). However, the use of dihalogenated pyridines such as
2,5-/ 2,6-/ 3,5-dibromopyridine and 3,5-dichloropyridine did not afford the desired
pyridine gold(I) complex, only the starting material was regained. These pyridines
might be too electron-deficient to form a stable coordinative bond to the gold(I) center,
-
so the coordination of the counter anion NTf2 is favored. Interestingly, Shi et al.
isolated a pyridine gold(I) complex based on Ph3PAuSbF6 and 3,5-dibromopyridine.[8c]
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
Scheme 1. Synthesis of gold(I) complexes coordinated by N-heterocycles
For some of the unknown structures, single crystal X-ray structures analyses were
conducted (Figure 1); in Table 1 the structural parameters are summarized.[9] The
gold(I) complexes differ only slightly in the selected Au–P and Au–N bond lengths and
the solid state molecular structures of the gold(I) complexes 1k, 1l, 1n and 1r allow no
conclusions concerning their behavior in solution.
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
Figure 1. Solid state molecular structures of the gold(I) complexes 1k, 1l, 1n, 1r.
Table 1. Selected bond lengths and angles of the gold(I) complexes.
Gold(I) complex
Au–P [Å]
2.235 (10)
2.234 (6)
2.231 (11)
2.240 (9)
Au–N [Å]
2.082 (3)
2.104 (19)
2.116 (4)
2.127 (3)
P–Au–N [°]
175.27 (10)
174.38 (6)
169.51 (13)
177.70 (8)
1k
1l
1n
1r
Hereinafter, the catalytic activity of the isolated pyridine gold(I) complexes and
accompanying the deactivation of the activated gold(I) species Ph3PAuNTf2 was
examined on the basis of the literature-known cyclopropanation of styrene by a
progargyl pivalate as a test reaction.[10] One advantage of the cyclopropanation
reaction is certainly the remarkable rapid reaction time if Ph3PAuNTf2 is used as
catalyst. In case of 5 mol% Ph3PAuNTf2, a complete conversion was obtained in less
than 15 seconds; using 0.5 mol% in 4.5 minutes. A potentially significantly decreased
rate constant using a pyridine gold(I) complex instead of Ph3PAuNTf2 still allows the
examination of the reaction by means of common analysis methods. The kinetic
investigations observed by GC were implemented by a standardized procedure (see
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
the Supporting Information for further details). Therefore, one equivalent of propargyl
pivalate was treated with four equivalents of freshly distilled styrene and 5 mol% of the
nitrogen coordinated gold(I) complex using n-dodecane as internal standard in DCM
at 25 °C (Scheme 2). The reaction mixture was stirred with a speed of 400 rpm. To
stop the reaction at a certain point of time, the separation of the catalyst was achieved
by filtration through a short column of silica gel and the samples were analyzed by GC.
Scheme 2. Gold(I)-catalyzed cyclopropanation of styrene with a propargyl pivalate.
The method of internal standard was used to improve the precision of quantitative
analysis.[11] The integration method[12] confirmed the initially assumed second order
reaction. As a result of graphic correlation, the rate constant was determined by
standard kinetic methods. The calculated error is a result of error propagation of
random faults caused by sample preparation. Systematic errors arising from the
influence on the measurement object by instrument were excluded. The coefficient of
determination R2, a criterion to qualify a linear relation, expresses the goodness-of-fit
of a regression if the coefficient is close to one. The results of the kinetic measurements
are represented in Table 2. By way of illustration, the nitrogen coordinated gold(I)
complexes are sorted by descending catalytic activity. Obviously, there are significant
differences concerning the catalytic activity of various pyridine gold(I) complexes. In
order to allow a better comparison, the relative rate constant krel of the unsubstituted
pyridine gold(I) complex 1a was normalized to krel = 1.
Table 2. Rate constants of the gold(I)-catalyzed cyclopropanation.[a]
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
[a]
Reaction conditions: propargyl pivalate (150.9 μmol, 70.8 μmol/ml), styrene
(603.7 μmol), n-dodecane (150.9 μmol), pyridine derivative (5 mol%; if in situ),
gold(I) catalyst (5 mol%), DCM at T = 25 °C, observed by GC monitoring for up
to 5 h.
In the first instance, the comparability of isolated and in situ generated pyridine gold(I)
complexes was examined. As an example serves the reaction of the pre-prepared
gold(I) complex 1n (Table 2; entry 19) and the reaction of Ph3PAuNTf2 in combination
with 8-bromoquinoline in a ratio of 1:1 (Table 2; entry 18) due to the extended reaction
time up to several hours (Figure 2). Obviously, both procedures lead to a comparable
reaction progress. Besides, the calculated rate constants k are consistent within the
error tolerances. The kinetic investigation using the heterocycle 2-bromopyridine
confirm this observation (Table 2; entries 7 and 9). It is evident, therefore, that the
reaction behavior is independent of the usage of isolated or in situ generated gold(I)
complexes.
Pivalate (in situ)
Cyclopropane (in situ)
Pivalate (isolated)
0,08
Cyclopropane (isolated)
0,07
0,06
0,05
0,04
0,03
0,02
0,01
0,00
0
50
100
150
t [min]
Figure 2. Direct comparison of the cyclopropanation of styrene with a propargyl
pivalate using the isolated and the in situ generated 8-bromoquinoline gold(I) complex.
In a first set of experiments differently-substituted pyridine derivatives bearing one or
more electron-withdrawing substituents were converted (Table 2; entries 1 – 10, 13 –
19). Unexceptionally, a significantly increased catalytic activity was obtained compared
to the gold(I) complexes of the unsubstituted pyridine (Table 2; entry 23) and quinoline
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
(Table 2; entry 22). Especially, the addition of pentafluoropyridine yielded an
extraordinarily high catalytic activity (Table 2; entry 1). A complete conversion was
already observed after a reaction time of 30 seconds, however, the reaction is still more
than two times slower than the reaction using Ph3PAuNTf2 without additives. An
enhanced catalytic activity correlates with the amount of halide substituents (Table 2;
entries 1 – 3, 5 – 7). Additionally, the catalytic activity of halogenated pyridine
derivatives declines with a lower electronegativity of the halide (Table 2; entries 14 –
17). Furthermore, the substitution pattern of the pyridine derivatives affects the
catalytic activity. In case of bromopyridines as well as dibromopyridines, the 2-
substituted or rather 2,6-disubstituted pyridine derivatives (Table 2; entries 2, 9) show
a higher rate constant than the 3-substituted or 3,5-disubstituted isomers (Table 2;
entries 6, 15). As expected, the catalytic activity of the 2,5-dibromopyridine derivative
(Table 2; entry 3) is decreased compared to the 2,6-dibrominated analogue and
increased in comparison to the 3,5-dibrominated isomer. Significantly reduced rate
constants were obtained by using pyridine derivatives including exclusively electron-
donating substituents (Table 2; entries 24 – 29, 31, 33 – 35). In case of electron-
donating substituents the catalytic activity is up to ten times lower than using the
unsubstituted analogues. Pyridine derivatives with increased σ-donor properties
caused by an enhanced electron density of the heterocycle form a stable coordinative
bond to the gold(I) center. As a consequence, the dissociation of the pyridine gold(I)
complex occurs in low probability. This becomes particularly obvious in case of 2,6-di-
tert-butyl-4-methylpyridine as additive (Table 2; entry 34), here, the reaction rate is
more than three times lower than using the unsubstituted pyridine. Interestingly, the
steric effects are not as pronounced as the electronic effects. A comparison of the
additives
2,6-di-tert-butyl-4-methylpyridine,
2,6-di-tert-butylpyridine
and
2-
methylpyridine clearly demonstrates this (Table 2; entries 27, 28, 34). In case of
electron-donating substituents, the substituent pattern is not that pronounced.
Independent of the position, the rate constants of the methyl-substituted pyridine
derivatives coincide within the error tolerances (Table 2, entries 28, 29, 31). The
quinoline gold(I) complexes are characterized by a similar trend as the pyridine
derivatives. The catalytic activity decreases with an enhanced electron density of the
nitrogen atom. The different reactivity of the unsubstituted pyridine derivative’s gold(I)
complex is caused by varying basicity (Table 2; entries 22, 23, 30, 32). Due to the peri
steric effect,[13] a higher basicity is expected for acridine, quinoline and isoquinoline in
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
comparison to pyridine. The basicity order is acridine > isoquinoline > pyridine >
quinoline. Obviously, pyridine derivatives with an increased basicity show a lower
reactivity. The notably declined catalytic activity of the acridine derivative confirms this
hypothesis (Table 2; entry 32). In comparison to pyridine derivatives, diazines have a
lower electron density due to a further electron withdrawing nitrogen atom. As a result,
an increased catalytic activity was observed by using diazine gold(I) complexes (Table
2; entry 11, 12, 20). N-Methylmorpholine was investigated as an exceptional aliphatic
heterocycle (Table 2; entry 36). Within a reaction time of two days, only a low
conversion was observed.
Nevertheless, in case of oxidative gold catalysis using N-oxides up to one equivalent
of the corresponding pyridine derivative exists in the reaction mixture at late reaction
states. Therefore, in a second set of experiments, the influence of the pyridine
derivative concentration was examined (Table 3). Due to its relatively high reaction
rate in case of 5 mol% pyridine derivative, 2-bromopyridine in combination with the
gold catalyst Ph3PAuNTf2 was chosen for further investigations. As expected, the
catalytic activity of the gold(I) complex significantly decreases in combination with an
enhanced amount of 2-bromopyridine (Table 3; Figure 3). The graphic correlation of
the rate constants depending on the equivalents of deployed 2-bromopyridine shows
an interesting effect (Figure 4). An exponential dependence using various ratios of
Ph3PAuNTf2 to pyridine derivative becomes obvious. This is caused by a slight
equilibrium shift towards the pyridine gold(I) complex with increasing concentration of
2-bromopyridine.
Table 3. Influence of various 2-bromopyridine concentrations on the cyclopropanation.
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
0,08
0,07
0,06
0,05
0,04
0,03
0,02
0,01
0,00
0.05 equiv
0.10 equiv
0.15 equiv
0.20 equiv
0.25 equiv
0.50 equiv
1.00 equiv
0
10
20
30
40
50
60
70
80
90
100
t [min]
Figure 3. Decrease of the propargyl pivalate depending on the 2-bromopyridine
concentration.
Figure 4. Exponential dependence of the rate constant.
In sum, the kinetic investigation of the cyclopropanation of styrene with a propargylic
pivalate demonstrates that the electronic structure of the pyridine additive significantly
affects the catalytic activity, whereas the steric effect is less pronounced. Moreover,
the pyridine concentration is of considerable interest. These observations are crucial
for the application of pyridine derivatives as organic bases and as starting materials in
homogeneous gold catalysis. Due to insertion of defined substituents, it might be
possible to predict and ultimately increase the reactivity.
Subsequently, further kinetic investigations were conducted on the basis of test
reactions with commonly used or easily accessible N-oxides. The N-oxides were
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
synthesized according to literature procedure (see Supporting Information for further
details).[14] However in case of the pyridine derivatives pentafluoropyridine, dimethyl
pyridine-2,5-dicarboxylate, 2,6-di-tert-butyl-pyridine, 2,6-di-tert-butyl-4-methylpyridine,
8-bromoquinoline, common synthetic strategies[15] did not yield the desired product.
Due to this, these N-oxides were excluded from further consideration. Initially, the
literature-known oxazole synthesis introduced by Zhang’s group was investigated as a
test reaction.[16] The oxazole synthesis is not only a key reaction type in homogeneous
gold catalysis, it also includes the application of easily accessible starting materials
and has a suitable reaction time of a few hours. As in the case of the cyclopropanation
reaction, the kinetic investigations were observed by GC using a standardized
procedure to ensure the comparability (see the Supporting Information for further
details). Therefore, one equivalent of phenylacetylene was treated with 1.3 equivalents
of N-oxide, one equivalent of n-dodecane as internal standard and 5 mol% of
Ph3PAuNTf2 in acetonitrile at 60 °C (Scheme 3). The reaction mixture was stirred with
a speed of 400 rpm. Here again, samples of the reaction mixture were taken in time
intervals adjusted to the reaction time. To abort the reaction at a certain point of time,
the separation of the catalyst was gained by filtration through a short column of silica
gel. Afterwards, the samples were analyzed by GC monitoring. The evaluation of the
kinetic data was conducted in accordance with the analysis of the cyclopropanation.
The integration method[12] proved the initially assumed pseudo second order reaction.
Due to the fact that acetonitrile performs as reactant and solvent, the concentration of
acetonitrile can be regarded as constant. The rate constant k* takes the acetonitrile
concentration into account. In Table 4, the results of the kinetic measurements are
shown. The experimental data was sorted by N-oxides with descending rate constants.
An exemplary reaction progress is shown in Figure 5.
Scheme 3. Gold(I)-catalyzed oxazole synthesis.
Table 4. Rate constants of the gold(I)-catalyzed oxazole synthesis.[a]
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
[a]
Reaction conditions: phenylacetylene (0.30 mmol, 0.0967 mmol/ml), N-oxide
(0.39 mmol), n-dodecane (0.30 mmol), Ph3PAuNTf2 (5 mol%), acetonitrile at T
= 60 °C, observed by GC monitoring for up to 5 h.
Phenylacetylene
2-Methyl-5-phenyloxazole
0,10
0,08
0,06
0,04
0,02
0,00
0
50
100
150
200
t [min]
Figure 5. Exemplary oxazole synthesis using 8-methylquinoline N-oxide (Table 4;
entry 2).
The initial kinetic investigation of differently substituted pyridine N-oxides with one or
more electron-withdrawing substituents show up to seven times increased reactivity
than using the unsubstituted analogue (Table 4; entries 1, 3 – 12, 15). Exclusively, the
2,6-dibromopyridine N-oxide performed worse than first expected (Table 4; entry 25).
2,5-dibromopyridine N-oxide achieved the highest reactivity by far (Table 4; entry 1),
however, the application of dihalogenated N-oxides did not lead to a full conversion
after a reaction time of 24 h (Table 4; entries 1, 6, 7). As mentioned in connection with
the cyclopropanation, the enhanced reactivity correlates with the amount of halide
substituents (Table 4; entries 1, 6 – 9). In case of the monohalogenated pyridine N-
oxides, a strong dependence of the electronegativity on the reactivity becomes obvious
(Table 4; entries 3, 4, 9, 11). It clearly demonstrates that substituents with an increased
electronegativity obtain a higher reaction rate. The only exception is shown by using
the dihalogenated pyridine N-oxides (Table 4; entries 6, 7). Here, the rate constants
are nearly consistent within the error tolerances. This is caused by the slight difference
in electronegativity of bromine and chlorine. Additionally, the substitution pattern of the
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
pyridine N-oxide affect the reactivity. Independent of electron-withdrawing (Table;
entries 8, 9) or donating substituents (Table 4; entries 20, 21, 27), substituents in 2-
position gain a higher reaction constant than their 3-substituted isomer; 4-substituted
derivatives show by far the worst reactivity. With the exception of 2,6-dibromopyridine
N-oxide, this observation is also transferable to the dibrominated N-oxides (Table 4;
entries 1, 6). Due to the negative inductive effect of the CF3- and CHF2-substituents,
the corresponding N-oxides achieved a more than two times higher rate constant than
the unsubstituted analogue (Table 4; entries 10, 12). As expected, pyridine N-oxides
including solely electron-donating substituents lead to significantly decreased reaction
rates (Table 4; entries 16, 20 – 22, 27, 28). The introduction of an electron-donating
substituent increases the N–O bond length and considerably enhances the electron
density on the oxygen. Furthermore, N-oxides with electron donating substituents
show a greater complexation ability than N-oxides with electron-withdrawing
substituents. Belova et al. confirm this expectation by DFT calculations.[17] The
Hantzsch ester N-oxide nearly yielded no consumption (Table 4; entry 30). After a
reaction time of 72 h, only 1 % of the phenylacetylene was converted. The application
of substituted quinoline N-oxides showed different trends than the pyridine derivatives.
8-Methylquinoline N-oxide gained a three times higher reaction constant than
unsubstituted quinoline N-oxide (Table 4; entries 2, 14). The reaction using 8-
isopropylquinoline N-oxide is characterized by a decreased reaction rate (Table 4;
entry 13). An outstanding high reactivity was estimated using 8-iodoquinoline N-oxide
(Table 4; entry 5), however, the N-oxide did not meet the expectations. With one
exception, the rate constants of N-oxides of the pyridine derivatives quinoline, pyridine,
isoquinoline and acridine are consistent with the basicity order; the higher the basicity,
the lower the reactivity (Table 4; entries 14, 17, 19, 23). Although a further electron-
withdrawing nitrogen atom decreases the electron density, the diazine N-oxides
perform worse than the N-oxide of pyridine or quinoline (Table 4; entries 18, 24, 26).
The aliphatic N-methylmorpholine manifested in penultimate place (Table 4; entry 29).
To confirm the examined trends, the set of N-oxides was investigated by means of
another literature-known key reaction type in gold catalysis, an oxidative ring
expansion. Here, the synthesis of a functionalized tropone derivative serves as a test
reaction.[18] Due to the higher resolution in time of NMR kinetics, the kinetic
investigations were conducted by in situ monitoring with 1H NMR spectroscopy at 25 °C
(see the Supporting Information for further details). The use of a continuous
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
1
measurement enables the recording of 256 H NMR spectra within a period of
maximum twelve hours. The sample composition consists of one equivalent of alkynyl
quinol, 1.3 equivalents of N-oxide and 5 mol% Ph3PAuNTf2 in deuterated DCM
(Scheme 4). Due to the in situ monitoring, the use of an internal standard is not
necessary.
Scheme 4. Gold(I)-catalyzed oxidative ring expansion of an alkynyl quinol.
Analogous to the analysis of the cyclopropanation and the oxazole synthesis, the
evaluation of the kinetic data was implemented. The integration method[12] serves for
the determination of the reaction order. Once more, this procedure confirmed the
initially assumed second order reaction. The results of the kinetic investigations are
summarized in Table 5, whereby the rate constants of the N-oxides are sorted in
descending order. As observed before, there are substantial differences concerning
the reactivity of the N-oxides. An exemplary reaction progress is shown in Figure 6.
Table 5. Rate constants of the gold(I)-catalyzed oxidative ring expansion of an alkynyl
quinol.[a]
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
[a]
Reaction conditions: alkynyl quinol (0.060 mmol, 0.10 mmol/ml), N-oxide
(0.078 mmol), Ph3PAuNTf2 (5 mol%), deuterated DCM at T = 25 °C, observed
by 1H NMR monitoring for up to 12 h.
Alkynyl quinol
Tropone
0,10
0,08
0,06
0,04
0,02
0,00
0
2
4
6
8
10
t [h]
Figure 6. Exemplary oxidative ring expansion using 3,5-dibromopyridine N-oxide
(Table 5; entry 5).
The kinetic investigations clearly show a dependence of the electronic effect caused
by differently substituted pyridine N-oxides on the reactivity. Pyridine N-oxides with one
or more electron withdrawing substituents show more than eighty times higher rate
constants than the unsubstituted analogue (Table 5; entries 3 – 5, 7 – 14, 20). The
decrease of electron density using an enhanced amount of halide substituents leads
to an increased reactivity. Exclusively, the 2,6-dibromopyridine N-oxide performed
worse (Table 5; entry 8). Different monohalogenated pyridine N-oxides yielded the
same rate constants within the error tolerances (Table 5; entries 11 – 14). Here, the
electronegativity is not affecting the reactivity. However, a dependence of the
substitution pattern of the pyridine N-oxides on the rate constants was observed.
Electron-withdrawing substituents in 2-position achieved a higher reaction rate than
the 3-substituted isomer (Table 5; entries 7, 12). Interestingly, in case of electron-
donating substituents a reverse effect was monitored (Table 5; entries 19, 21, 24).
Pyridine N-oxides containing electron donating substituents (Table 5; entries 15, 16,
19, 21, 24, 26) yielded a significantly lower reactivity compared to the N-oxides with
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
electron-withdrawing substituents, however, not necessarily lower than the
unsubstituted pyridine N-oxide. The application of the Hantzsch ester N-oxide did not
afford a reaction (Table 5; entry 30). As observed in case of the oxazole synthesis, the
substituted quinoline N-oxides exhibited different trends than the pyridine derivatives.
8-Isopropylquinoline N-oxide achieved the highest rate constant by far, closely
followed by 8-methylquinoline N-oxide (Table 5; entries 1, 2). The use of 8-
iodoquinoline N-oxide promised an outstanding high reactivity, however, the N-oxide
did not meet the expectations (Table 5; entry 6). The reactivity of the N-oxides of the
pyridine derivatives quinoline, isoquinoline, pyridine and acridine exclusively are in
accordance with the basicity order (Table 5; entries 17, 22, 23, 28). Thereby, the
acridine N-oxide which possesses the highest basicity in this series afforded the lowest
rate constant. The diazine N-oxides obtained a lower reaction rate than the N-oxide of
pyridine and quinoline (Table 5; entries 18, 25, 27). In accordance with the oxazole
synthesis, the aliphatic N-methylmorpholine gained the penultimate place (Table 5;
entry 29).
The kinetic studies of the oxazole synthesis and the oxidative ring expansion show a
definite trend concerning the necessary electronic properties of the N-oxides.
Interestingly, one of the best N-oxides, the 2,5-dibromopyridine N-oxide, was not used
in literature before. Although the synthesis of the 2,5-dibromopyridine N-oxide is as
simple as for every other N-oxide as well. The use of 2,5-dibromopyridine N-oxide in
further oxidative gold(I)-catalyzed reactions shall ensure the transferability of the
results. Thereby, a comparison to already published results was aimed. Therefore, the
general procedure is in accordance to literature to guarantee a comparability of the
obtained results. Hence, the literature-known synthesis of indenone was examined.[19]
One equivalent of 1,5-diyne was treated with 1.5 equivalents of N-oxide and 5 mol%
Ph3PAuNTf2 in deuterated acetonitrile at 60 °C. Hexamethylbenzene was used as
1
internal standard for H NMR-monitoring. The results are shown in Table 6. As
expected the 2,5-dibromopyridine N-oxide significantly improves the reaction. The
highest published yield (58 %) was achieved by 8-ethylquinoline N-oxide after a
reaction time of 17 hours. In case of 2,5-dibromopyridine N-oxide a complete
consumption and additionally a quantitative yield were monitored after even 30 min.
Table 6. Gold(I)-catalyzed cyclization of 1,5-diyne.[a]
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
[a]
Reaction conditions: 1,5-diyne (0.140 mmol, 0.292 mmol/ml), N-oxide
(0.210 mmol), hexamethylbenzene (0.016 mmol), Ph3PAuNTf2 (5 mol%),
deuterated acetonitrile at T = 60 °C, observed by 1H NMR monitoring for 17 h.
[b]
Yield after a reaction time of 30 min.
Additionally, the oxidative cyclization of propynyl arene into indan-2-one was
investigated.[5,20] Therefore, one equivalent of 3-phenyl-1-propyne was converted with
1.2 equivalents of N-oxide and 4 mol% Ph3PAuNTf2 in deuterated CHCl3 at 20 °C.
1,4-Dinitrobenzene was used as internal standard for 1H NMR monitoring. In Table 7,
the conversion related to different reaction times is listed in a direct comparison of 2-
(tert-butyl)-6-chloropyridine N-oxide and 2,5-dibromopyridine N-oxide. Although the
reaction comes to a standstill after a reaction time of six hours, a nearly two times
higher conversion was yielded by using 2,5-dibromopyridine N-oxide. 2-(tert-butyl)-6-
chloropyridine N-oxide afforded the best results in the published reaction screening. It
should be noted that further published results were achieved using a biarylphosphonite
gold(I) complex. Due to this, further results are not comparable to this investigation.
Table 7. Gold(I)-catalyzed oxidative cyclization of propynyl arene into indan-2-one.[a]
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
[a]
Reaction conditions: prop-2-yn-1-ylbenzene (0.12 mmol, 0.20 mmol/ml),
N-oxide (0.24 mmol), hexamethylbenzene (0.027 mmol), Ph3PAuNTf2 (4 mol%),
deuterated chloroform at T = 20 °C, observed by 1H NMR -monitoring.
Conclusions
In an initial systematic kinetic study, the inhibition of the gold(I) catalyst by various
pyridine derivatives was examined by means of a cyclopropanation reaction. Related
to this a series of pyridine gold(I) complexes has been synthesized and evaluated by
single crystal X-ray structure analyses. Comparison experiments have proven that the
isolated and in situ generated pyridine gold(I) complexes achieve a similar catalytic
activity. The electronic structure of these complexes significantly affects the catalytic
activity. Pyridine derivatives with increased σ-donor properties caused by an enhanced
electron density of the nitrogen atom of the heterocycle form a more stable coordinate
bond to the Lewis-acidic gold(I) center, finally, the degree of inhibition rises. This trend
became apparent in case of pyridine derivatives with different electronic properties and
various substitution patterns. The influence of the pyridine concentration on the
catalytic activity is noteworthy, an exponential dependence was observed. Especially,
these results are of major importance for the transfer to other applications. The
insertion of defined substituents to selected pyridine derivatives which might be applied
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
for example as organic base or starting material possibly increase the reactivity.
Moreover, a detailed kinetic study of two key reactivity types of oxidative gold catalysis
using yet unapplied or only rarely applied N-oxides, regardless of their commercial
availability, was conducted. For both test reactions, similar reaction trends – apart from
a few exceptions – in accordance with the results of the cyclopropanation were
observed. In all cases, outstanding results were achieved using 2,5-dibromopyridine
N-oxide and 8-methylquinoline N-oxide. The practical benefit of 2,5-dibromopyridine
N-oxide is additionally shown by the exemplary use in further test reactions. This
exploration can be used to categorically exclude N-oxides which were commonly an
integral component of investigations and to consider others. These investigations are
the key for a basic set of N-oxides which should not be neglected in a reaction
optimization screening in homogeneous oxidative gold catalysis. The basic set of at
least three N-oxides should include 2,5-dibromopyridine N-oxide, 2-
trifluoromethylpyridine N-oxide and 8-methylquinoline N-oxide.
Experimental Section
General remarks
Chemicals were, if not noted otherwise, used from the stock of the University of
Heidelberg or were bought from commercial suppliers such as Sigma-Aldrich, Strem
and Carbolution. Deuterated solvents were purchased from Euriso-Top. The
experiments were carried out in standard laboratory glassware.
NMR (Nuclear magnetic resonance) spectra were recorded at room temperature on
Bruker Devices. The following instruments were available to choose from: Avance III-
300 and Avance DRX-300. All chemical shifts δ were given in ppm and coupling
constants J in Hz. 1H and 13C spectra were calibrated using the residual solvent signal
(CDCl3: 7.26 / 77.16 ppm; CD2Cl2: 5.32 / 53.84 ppm).
Gas Chromatography (GC) was processed on HP 58090 SERIES II with a HP 1
column. Nitrogen was used as the carrier gas.
General procedures
General procedure 1 (GP 1) – Synthesis of nitrogen coordinated gold(I)-complexes
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
A quoted amount of precatalyst Ph3PAuNTf2 (1.0 equiv) as well as pyridine derivative
(1.0 equiv) were dissolved in dry DCM and stirred for 2 h at room temperature. The
purity of the substances was taken into account when calculating the required amount
of educts. Afterwards, the reaction mixture was filtered through Celite and the solvent
was removed under reduced pressure. Finally, the residue was recrystallized from
pentane.
General procedure 2 (GP 2) – Requirement for GC observed kinetic studies
The internal standard (IS) n-dodecane (5.7 µl, 25.1 µmol, 1 equiv) and the analyte
were dissolved in 1.0 ml DCM in various stochiometric ratios (n-dodecane/analyte, 1:1,
1:0.5, 1:0.25). After analyzing the samples by means of gas chromatography, the
corresponding response factors were calculated according to the following formula.
areaIS ∙ canalyte
RFanalyte
=
areaanalyte ∙ cIS
General procedure 3 (GP 3) – Kinetic studies – Cyclopropanation
2-Methylbut-3-yn-2-yl pivalate (27.6 µl, 150.9 µmol, 1 equiv), styrene (69.2 µl,
603.7 µmol, 4 equiv) and the internal standard n-dodecane (34.3 µl, 150.9 µmol,
1 equiv) were dissolved in 2.0 ml dry DCM and stirred with a speed of 400 rpm. After
addition of the nitrogen coordinated gold(I) complex, samples of the reaction mixture
were taken in time intervals adjusted to the reaction rate. Separation of the catalyst
was achieved by filtration through a short column of silica gel (PE/EA, 1:1). Then, the
samples were analyzed by means of gas chromatography.
Note: In case of in situ activation, the precatalyst Ph3PAuNTf2 and the corresponding
pyridine derivative were dissolved in 2.0 ml dry DCM and stirred with a speed of 400
rpm for 1 h at room temperature. Then, styrene and internal standard, n-dodecane,
were added. After addition of the 2-methylbut-3yn-2-yl pivalate, samples of the reaction
mixture were taken as described in the general procedure (GP 3).
General procedure 4 (GP 4) – Synthesis of N-oxides
The quoted amount of N-derivative (1 equiv) was dissolved in dry DCM. After cooling
to 0 °C, the calculated amount of m-CPBA (1.6 equiv) was added gradually. The
reaction mixture was allowed to warm up to room temperature und stirred overnight.
After complete conversion of the starting material, the solution was quenched with
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
saturated NaHCO3-solution and 2M NaOH-solution. The aqueous layer was separated
and extracted with DCM. The combined organic layers were dried over Na2SO4,
filtrated and concentrated under reduced pressure. The crude product was purified by
column chromatography.
General procedure 5 (GP 5) – Kinetic studies – Gold(I)-catalyzed oxazole synthesis
Phenylacetylene (0.30 mmol, 1 equiv), N-oxide (0.39 mmol, 1.3 equiv) and the internal
standard (0.30 mmol, 1 equiv) were dissolved in 3.0 ml acetonitrile and stirred with a
speed of 400 rpm at 60 °C. After addition of the gold(I) complex Ph3PAuNTf2
(0.015 mmol, 0.05 equiv), samples of the reaction mixture were taken in time intervals
adjusted to the reaction rate. Separation of the catalyst was achieved by filtration
through a short column of silica gel (PE/EA, 1:1). Then, the samples were analyzed by
means of gas chromatography.
General procedure 6 (GP 6) – Kinetic studies – Gold(I)-catalyzed ring expansion
A NMR tube was charged with a solution of 4-hydroxy-4-(phenylethynyl)cyclohexa-2,5-
dien-1-one (0.060 mmol, 1 equiv) and N-oxide (0.078 mmol, 1.3 equiv) in 0.6 ml
D2-dichloromethane. Afterwards, the quoted amount of Ph3PAuNTf2 (0.003 mmol,
0.05 equiv) was added. The reaction was observed at T = 25 °C by 1H NMR kinetic for
12 h. Here, the equipment Avance DRX-300 of Bruker Devices was selected. Due to
often varying relaxation times, the measurement period distinguishes from substance
to substance. Calculations of the measured time interval Δt are based on Equation 1:
Δt = (DS + NS)(AQ + T1)
DS – number of dummy scans NS – number of scans;
(1)
AQ – acquisition time (observation period) T1 – spin-lattice relaxation time constant
References
[1] a) A. S. K. Hashmi, Chem. Rev. 2007, 107, 3180 – 3211; b) A. Arcadi, Chem. Rev.
2008, 108, 3266 – 3325; c) F. Gagosz, Actualite Chimique 2010, 347,12 – 19; d) P.
Garcia, M. Malacria, C. Aubert, V. Gandon, L. Fensterbank, ChemCatChem 2010, 2,
493 – 497; e) M. Rudolph, A. S. K. Hashmi, Chem. Soc. Rev. 2012, 41, 2448 – 2462;
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
f) L. Liu, G. B. Hammond, Chem. Soc. Rev. 2012, 41, 3129 – 3139; g) D. Zhang, X.
Tang, M. Shi, Acc. Chem. Res. 2014, 47, 913 – 924.
[2] a) L. Ye, W. He, L. Zhang, J. Am. Chem. Soc. 2010, 132, 8550 – 8551; b) B. Lu, C.
Li, L. Zhang, J. Am. Chem. Soc. 2010, 132, 14070 – 14072.
[3] a) J. Xiao, X. Li, Angew. Chem. Int. Ed. 2011, 50, 7226 – 7236; Angew. Chem.
2011, 123, 7364 – 7375; b) A. S. K. Hashmi, F. D. Toste, Modern Gold Catalyzed
Synthesis, Wiley-VCH, 1st edn., 2012; c) L. Zhang, Acc. Chem. Res. 2014, 47, 877 –
888; d) L. Ye, L. Zhang, Synthesis of Oxygenated and Nitrogen-Containing
Heterocycles by Gold-Catalyzed Alkyne Oxidation; in: M. Bandini, Au-Catalyzed
Synthesis and Functionalization of Heterocycles, Springer, 1st edn., 2016.
[4] a) M. R. Fructos, T. R. Belderrain, P. de Frémont, N. M. Scott, S. P. Nolan, M. M.
Díaz-Requejo, P. J. Pérez, Angew. Chem. Int. Ed. 2005, 44, 5284 – 5288; Angew.
Chem. 2005, 117, 5418 – 5422; b) A. Prieto, M. R. Fructos, M. M. Díaz-Requejo, P. J.
Pérez, P. Pérez-Galán, N. Delpont, A. M. Echavarren, Tetrahedron 2009, 65, 1790 –
1793; c) V. V. Pagar, A. M. Jadhav, R.-S. Liu, J. Org. Chem. 2013, 78, 5711 – 5716;
d) V. V. Pagar, R.-S. Liu, Angew. Chem. Int. Ed. 2015, 54, 4923 – 4926; Angew. Chem.
2015, 127, 5005 – 5008.
[5] G. Henrion, T. E. J. Chavas, X. Le Goff, F. Gagosz, Angew. Chem. Int. Ed. 2013,
52, 6277 – 6282; Angew. Chem. 2013, 125, 6397 – 6402.
[6] a) Y. Wang, K. Ji, S. Lan, L. Zhang, Angew. Chem. Int. Ed. 2012, 51, 1915 – 1918;
Angew. Chem. 2012, 124, 1951 – 1954; b) A. S. K. Hashmi, T. Wang, S. Shi, M.
Rudolph, J. Org. Chem. 2012, 77, 7761 – 7767.
[7] J. Schulz, J. Jašík, A. Gray, J. Roithová, Chem. Eur. J. 2016, 22 , 9827 – 9834.
[8] a) S. E. Thwaite, A. Schier, H. Schmidbaur, Inorg. Chim. Acta 2004, 357, 1549 –
1557; b) P. de Frémont, N. Marion, S. P. Nolan, J. Organomet. Chem. 2009, 694, 551
– 560; c) W. Yuan, X. Dong, Y. Wei, M. Shi, Chem. Eur. J. 2012, 18, 10501 – 10505;
d) S. Orbisaglia, B. Jacques, P. Braunstein, D. Hueber, P. Pale, A. Blanc, P. de
Frémont, Organometallics 2013, 32, 4153 – 4164.
[9] 1859090 (1k), 1859091 (1l), 1859092 (1n) and 1859093 (1r) contain the
supplementary crystallographic data for this paper. These data can be obtained free
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[10] a) M. J. Johansson, D. J. Gorin, S. T. Staben, F. D. Toste, J. Am. Chem. Soc.
2005, 127, 18002 – 18003; b) J. Petuškova, H. Bruns, M. Alcarazo, Angew. Chem. Int.
Ed. 2011, 50, 3799–3802; Angew. Chem. 2011, 123, 3883 – 3886.
[11] R. L. Grob, E. F. Barry, Modern Practice of Gas Chromatography, Wiley
Interscience, Hoboken, NJ, 4th edn., 2005.
[12] a) H. G. Elias, Makromoleküle: Chemische Struktur und Synthesen, Wiley-VCH,
6th edn., 1999; b) W. Bechmann, J. Schmidt, Einstieg in die Physikalische Chemie für
Nebenfächler, B. G. Teubner, 1st edn., 2001; c) P. W. Atkins, J. Paula, Physikalische
Chemie, Wiley-VCH, 4th edn., 2006; d) P. L. Houston, Chemical Kinetics and Reaction
Dynamics, Dover Publications, Inc., Minola, New York, 2012.
[13] M. Lõkov, S. Tshepelevitsh, A. Heering, P. G. Plieger, R. Vianello, I. Leito, Eur. J.
Org. Chem. 2017, 4475 – 4489.
[14] a) D. Rong, V. A. Phillips, R. Sánchez Rubio, M. A. Castro, R. T. Wheelhouse,
Tetrahedron Lett. 2008, 49, 6933 – 6935; b) G. Li, C. Jia, K. Sun, Y. Lv, F. Zhao, K.
Zhou, H. Wu, Org. Biomol. Chem. 2015, 13, 3207 – 3210; c) J. Zhao, P. Li, C. Xia, F.
Li, RSC Adv. 2015, 5, 32835 – 32838; d) X. Ma, H. Dang, J. A. Rose, P. Rablen, S. B.
Herzon, J. Am. Chem. Soc. 2017, 139, 5998 – 6007.
[15] a) Y. Wang, K. Ji, S. Lan, L. Zhang, Angew. Chem. Int. Ed. 2012, 51, 1915 – 1918;
Angew. Chem. 2012, 124, 1951 – 1954; b) M. Butler, G. M. Cabrera, J. Mol. Struct.
2013, 1043, 37 – 42; c) H. Hwang, J. Kim, J. Jeong, S. Chang, J. Am. Chem. Soc.
2014, 136, 10770 – 10776; d) C. Zhu, M. Yi, D. Wei, X. Chen, Y. Wu, X. Cui, Org. Lett.
2014, 16, 1840 – 1843.
[16] W. Yang, R. Zhang, F. Yi, M. Cai, J. Org. Chem. 2017, 82, 5204 − 5211.
[17] N. V. Belova, N. I. Giricheva, M. S. Fedorov, Struct. Chem. 2015, DOI:
10.1007/s11224-015-0621-9.
[18] J. Zhao, J. Liu, X. Xie, S. Li, Y. Liu, Org. Lett. 2015, 17, 5926 − 5929.
[19] P. Nösel, S. Moghimi, C. Hendrich, M. Haupt, M. Rudolph, F. Rominger, A. S. K.
Hashmi, Adv. Synth. Catal. 2014, 356, 3755 – 3760.
This article is protected by copyright. All rights reserved.

10.1002/adsc.201801007
Advanced Synthesis & Catalysis
[20] a) C. Khin, A. S. K. Hashmi, F. Rominger, Eur. J. Inorg. Chem. 2010, 1063-1069;
b) L. Huang, M. Rudolph, F. Rominger, A. S. K. Hashmi, Angew. Chem. Int. Ed.
2016, 55, 4808-4813; Angew. Chem. 2016, 128, 4888-4893; c) L. Huang, F.
Rominger, M. Rudolph, A. S. K. Hashmi, Chem. Commun. 2016, 52, 6435-6438.
.
TOC graphic
This article is protected by copyright. All rights reserved.