Towards the rational design of ylide-substituted phosphines for gold(i)-catalysis: From inactive to ppm-level catalysis

Article abstract of DOI:10.1039/d1sc00105a

The implementation of gold catalysis into large-scale processes suffers from the fact that most reactions still require high catalyst loadings to achieve efficient catalysis thus making upscaling impractical. Here, we report systematic studies on the impa

Full text of DOI:10.1039/d1sc00105a

Showcasing research from Professor Gessner’s laboratory
at Ruhr-University Bochum, Germany.
As featured in:
Towards the rational design of ylide-substituted phosphines
I
for gold()-catalysis: from inactive to ppm-level catalysis
Systematic studies on the impact of different aryl substituents
in the backbone of ylide-substituted phosphines reveal that
steric bulk leads to a boost of the catalytic activity in gold
catalysis.
See Viktoria H. Gessner et al.,
Chem. Sci., 2021, 12, 4329.
rsc.li/chemical-science
Registered charity number: 207890

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
Towards the rational design of ylide-substituted
phosphines for gold(I)-catalysis: from inactive to
ppm-level catalysis†
Cite this: Chem. Sci., 2021, 12, 4329
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
Jens Handelmann, Chatla Naga Babu, Henning Steinert,
Thorsten Scherpf, Alexander Kroll and Viktoria H. Gessner
Christopher Schwarz,
*
The implementation of gold catalysis into large-scale processes suffers from the fact that most reactions
still require high catalyst loadings to achieve efficient catalysis thus making upscaling impractical. Here,
we report systematic studies on the impact of the substituent in the backbone of ylide-substituted
phosphines (YPhos) on the catalytic activity in the hydroamination of alkynes, which allowed us to
increase the catalyst performance by orders of magnitude. While electronic changes of the ligand
properties by introduction of aryl groups with electron-withdrawing or electron-donating groups had
surprisingly little impact on the activity of the gold complexes, the use of bulky aryl groups with ortho-
substituents led to a remarkable boost in the catalyst activity. However, this catalyst improvement is not
a result of an increased steric demand of the ligand towards the metal center, but due to steric
protection of the reactive ylidic carbon centre in the ligand backbone. The gold complex of the thus
designed mesityl-substituted YPhos ligand YMesPCy2, which is readily accessible in one step from
a simple phosphonium salt, exhibited a high catalyst stability and allowed for turnover numbers up to
20 000 in the hydroamination of a series of different alkynes and amines. Furthermore, the catalyst was
also active in more challenging reactions including enyne cyclisation and the formation of 1,2-
dihydroquinolines.
Received 7th January 2021
Accepted 28th January 2021
DOI: 10.1039/d1sc00105a
rsc.li/chemical-science
of more than 0.5 mol% of catalyst to achieve high conversions,
thus making large-scale applications impractical. In the past
Introduction
years, several groups have addressed this issue. Thus, a number
of ligands have been reported which enable conversions at
remarkably low loadings, in some cases also on ppm-level. For
example, Zhang and coworkers described the design of an
electron-rich biarylphosphine ligand, which due to additional
substitution at the biaryl moiety enabled a ligand-directed
nucleophilic attack of the alkyne and thus turnover numbers
of 99 000 for the addition of acids to alkynes.5,6 A similar
strategy of improving the catalyst activity by modication of the
catalytic pocket was recently applied to N-heterocyclic carbenes
(NHCs)7 as well as further dialkylbiaryl phosphines.8 In case of
hydroamination reactions, the best results however, were – to
the best of our knowledge – reported by Lavallo and coworkers
using an anionic phosphine and NHC with carborate substitu-
ents, giving rise to TONs close to 100 000.9 Drawback of these
highly sophisticated ligands however, is that they are oen
rather complicated to synthesize thus considerably contributing
to the total costs of the catalyst.
Phosphine ligands are a privileged class of ligands in homog-
enous catalysis.1 Their continuous development and tailoring
has led to many important advances and decisively contributed
to the development of homogenous catalysis into a mature eld
in chemistry. Despite the advances made in the past years, the
design of new ligands is still needed to improve existing reac-
tion protocols and to enable new transformations. Notable
examples for the importance of the design of phosphine ligands
for the improvement of the catalyst performance are achieve-
ments made in hydrogenations,2 C–C and C–X couplings,3 as
well as gold(I) catalysed reactions.4
In gold catalysis, ligand design is particularly important for
the generation of highly productive catalysts that can operate at
low catalysts loadings. However, despite the costs of the metal,
most catalytic transformations with gold still require loadings
Faculty of Chemistry and Biochemistry, Chair of Inorganic Chemistry II, Ruhr
¨
University Bochum, Universitatsstr. 150, 44801 Bochum, Germany. E-mail: viktoria.
gessner@rub.de
Recently, we reported on the class of ylide-substituted
phosphines (YPhos) as easy-to-synthesize and highly efficient
ligands in catalysis.10 Due to the electron-donation from the
ylide to the phosphorus centre YPhos ligands are in general
strong donor ligands with donor strengths comparable to those
† Electronic supplementary information (ESI) available: Experimental procedures
and characterisation data for all new compounds, NMR spectra of the isolated
compounds and crystallographic details. CCDC 1983138–1983147. For ESI and
crystallographic data in CIF or other electronic format see DOI:
10.1039/d1sc00105a
© 2021 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci., 2021, 12, 4329–4337 | 4329

View Article Online
Chemical Science
Edge Article
thus allow for a detailed study of the different factors inu-
encing the catalytic ability of the gold complexes.
Results and discussion
Ligand synthesis and properties
We started our studies by preparing a series of YPhos ligands of
type YArPCy2 (Scheme 1). In this ligand structure, electron-
Fig. 1 (Left) Gold complexes based on ylide-substituted phosphines
(YPhos) and comparison of different YPhos ligands and their activity in
catalysis.
donating (YpOMe, 2) and electron-withdrawing (YpCF PCy2, 3) as
3
well as sterically demanding groups (YoTolPCy2, 4) were
compared with the parent phenyl compound YPhPCy2 (1). All
ligands were prepared in one step from simple phosphonium
salts C (prepared from PPh3 and the corresponding benzyl
halides)10 using a slight excess of base (KH) and half an equiv.
Cy2PCl. Thus, ligands 1–4 were obtained as yellow solids in
moderate to good yields of 42–68%. The YPhos ligands are
characterized by two sets of doublets in the 31P{1H} NMR
spectrum. While these signals appear in a similar region for the
ligands 1–3 (Table 1), they are clearly shied for the ortho-tolyl-
substituted compound. Likewise, the 2JPP coupling constant of 4
(2JPP ¼ 160.4 Hz) considerably differs from those of the other
ligands (2JPP ¼ 182.4–185.0 Hz) thus indicating differences in
the geometry of 4 compared to the other ligands. This is
conrmed by the crystal structures of the compounds (Fig. 2b).
The most striking difference between 4 and 1–3 concerns the
orientation of the aryl group relative to the P–C–P plane. While
the ligands 1–3 exhibit small P–C–C–C torsion angles between
8.0 and 21.7ꢀ and thus an almost co-planar arrangement of the
P–C–P and the aryl moiety, the aryl group in 4 is almost
perpendicularly arranged (P–C–C–C ¼ 75.5(3)ꢀ). Hence, no
delocalization of the carbanionic charge is possible in case of
the tolyl system. This also results in the shortest P1–C1 bond in
4 for all ligands due to the strongest electrostatic interactions in
the ylide linkage.
of N-heterocyclic carbenes (NHC). Accordingly, high activities
were observed with YPhos-ligated complexes in Au-catalysed
hydroaminations10a,11 as well as Pd-catalysed coupling12 and a-
arylation reactions.13 In case of gold-catalysed hydroaminations
we observed that YPhos ligands such as A (Fig. 1) with electron-
withdrawing groups in the backbone (–SO2Tol or –CN) and only
moderate donor strengths gave the most active catalysts, while
the most electron-rich donors (e.g. B) showed no activity at all.
This is surprising since the rate-limiting protodeauration step
of the hydroamination is facilitated by electron-rich ligands.14
Hence, strong donors ligands typically showed the highest
activities.5–9,15
To understand the inactivity of the most electron-rich YPhos
ligands and to develop principles for future ligand design, we
turned our attention towards ligands with aryl substituents in
the ylide-backbone. This decision was based on two consider-
ations: (i) the ylidic carbon atom in the electron-rich ligands is
highly reactive and thus prone to side-reactions (protonation/
auration). Stabilization of the carbanionic charge might be
crucial and thus aryl groups should be suitable substituents due
to the possible charge delocalization into the aromatic moiety.
(ii) Aryl groups can be tuned electronically and sterically and
To determine the steric properties of the ligands the buried
volume (%Vbur) was calculated based on the structures of the
isolated LAuCl complexes,16 which were synthesized from the
free ligands and (THT)AuCl (Scheme 1) and isolated as off-white
solids (Fig. 3). All four ligands exhibit similar spatial properties
and approx. cover half of the sphere around the metal atom. %
Vbur varies only slightly within the ligand series thus adopting
values between 47.0 and 49.9% (Table 2). The same holds true
for the calculated Tolman cone angles,17 thus conrming the
bulkiness of the ligands, but also indicating that catalytic
differences due to varying spatial protection of the metal by the
ligands should be minimal.
Scheme 1 Preparation of the YPhos ligands 1–4.
Table 1 31P{1H} NMR spectroscopic and crystallographic data for ligands 1–4
PhPCy2 (1) pOMePCy2 (2)
Y
Y
Y
pCF PCy2 (3)
YoTolPCy2 (4)
YMesPCy2 (5)
3
dP(PPh3) [ppm]
dP(PCy2) [ppm]
2JPP [Hz]
19.8
18.8
21.2
13.9
9.9
7.2
170.3
1.700(2)
111.9(1)
87.2(1)
ꢁ5.3
ꢁ5.0
ꢁ5.1
ꢁ1.3
185.0
182.4
184.3
160.4
˚
P1–C1 [A]
1.721(2)
115.1(1)
21.7(1)
1.721(3)
113.2(2)
8.0(4)
1.726(2)
112.5(1)
17.2(2)
1.696(2)
113.4(1)
75.5(3)
P1–C1–P2 [ꢀ]
P1–C1–C2–C3 [ꢀ]
4330 | Chem. Sci., 2021, 12, 4329–4337
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Edge Article
Chemical Science
contrasts with the free ligands where only the ortho-tolyl ligand
4 showed this behaviour. Presumably, this rotation of the aryl
group is enforced by steric effects. Upon coordination of gold
the cyclohexyl groups adjust their orientation relative to each
other and to the metal, thus forcing the aryl unit out of the
P–C–P plane. This can nicely be seen by comparison of the
structures of the free ligands (Fig. 2) and their gold complexes
(Fig. 3). The steric pressure on the aryl substituent is further
amplied by the widening of the P–C–P angle upon metal
coordination. As such, P–C–P angles larger than 120ꢀ are found
in the gold complexes, while they range between 112.5(1) and
115.1(1)ꢀ in the free ligands. Since the rearrangement of the
ligand structure already occurs upon coordination of the small,
linear AuCl fragment, it can be assumed that this will also be
the case for any other metal fragment binding to the YPhos
ligands. Hence, no p-delocalization of the carbanionic charge
into the aryl substituents should be possible in the metal
complexes, so that only inductive effects should play a role in
the ligand donor strength. Accordingly, the Tolman electronic
parameters (TEP) of all ligands are in the same range and reect
the different inductive effects of the substituents at the aryl
Fig. 2 Molecular structures of ligands 3 and 4. Ellipsoids are drawn at
the 50% probability level. Structures of the other ligands are given in
the ESI,† important parameters are listed in Table 1.
groups. Thus, the donor strength follows the order: YCF PCy2 <
3
Y
PhPCy2 < YpOMePCy2 z YoTolPCy2. YoTolPCy2 and YpOMePCy2
exhibit similar donor strengths and are thus similarly strong
donors than the YPhos ligand B with a methyl group in the
backbone (Fig. 1).
Comparison of the catalytic performance
Fig. 3 Molecular structures of the gold(I) complexes of the ligands 3
and 4.
To examine the impact of the different aryl substituents on the
activity of the complexes in gold(I) catalysed hydroaminations
we chose the reaction of phenylacetylene with aniline as test
reaction, since many other ligands have been tested in this
transformation, thus allowing for a more detailed compar-
ison.9–20,22 The reactions were conducted at mild conditions
It must be noted that all ligands feature a similar geometry in
the gold complexes. As such, the PPh3 moiety always points
towards the gold centre, suggesting that the ligand should be
able to stabilize a cationic gold species by additional arene–gold
interactions. This was also observed for other YPhos gold
complexes10,11 and has become a more general design principle
for ligands in gold catalysis including phosphines and N-
heterocyclic carbenes.7,18
ꢀ
(50 C) with low catalyst loadings of only 0.1 mol% LAuCl and
NaBArF4 for halide abstraction. Surprisingly, only small changes
were observed with the ligands 1–3 (Fig. 4), suggesting that
electronic changes in the aryl group have only little impact on
the catalytic activity. YPhPCy2 and its methoxy analogue gave
even lower yields than simple PPh3 aer 24 h. YpCF PCy2 also
delivered only approx. 40% conversion. However, the almost
linear reaction prole with 3 suggests a higher stability (but low
Interestingly, the aryl substituents in the ylide-backbone of
all LAuCl complexes rotate out of the P–C–P plane. This
3
Table 2 Comparison of the electronic and steric properties of the YPhos ligands 1–4
P1–C1–C2–C3
in L$AuCl [ꢀ]
nRh [cmꢁ1
]
TEPa [cmꢁ1
]
%Vbur
Cone angle [ꢀ]c
b
Ligand
Y
Y
Y
Y
Y
PhPCy2 (1)
pOMePCy2 (2)
1948.9
1947.2
1952.6
1947.5
n.d.d
2052.5
2051.5
2054.6
2051.7
n.d.
48.3
47.0
49.9
48.9
50.4
194.5
197.2
204.1
199.4
194.7
50.4(1)
71.2(1)
36.6(1)
81.7(3)
85.3(1)
pCF PCy2 (3)
3
oTolPCy2 (4)
MesPCy2 (5)
a
TEPs were determined by nCO in the Rh(acac)(CO)(L) complexes using the linear relationship between nCO for Ni(CO)3(L) and Rh(acac)(CO)(L)
b
20
˚
reported in ref. 19. Calculated with the SambVca 2.1 program for the LAuCl complexes; M–P distance ¼ 2.28 A, including H atoms.
c
d
¨
Calculated from the crystal structure according to the method described by Muller and Mingos, see ref. 21. The complex eliminates both CO
molecules and therefore no stretching frequency could be recorded.
© 2021 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci., 2021, 12, 4329–4337 | 4331

View Article Online
Chemical Science
Edge Article
Table
3
Comparison of the arene gold interaction in the gold
complexes of the YPhos ligands 1–5; calculated valuesa, bond lengths
in [A˚]
Ligand
Au–C (ipso/ortho/sum)
r(BCP)
Y
Y
Y
Y
Y
PhPCy2$AuCl
pOMePCy2$AuCl
3.234/3.244/6.478
3.195/3.297/6.492
3.361/3.369/6.730
3.207/3.130/6.337
3.241/3.106/6.347
0.0138
0.0137
0.0113
0.0153
0.0153
pCF PCy2$AuCl
3
oTolPCy2$AuCl
MesPCy2$AuCl
a
Level of theory: PW6B95D3/def2tzvp (MWB60 for Au).
a role in the catalytic activity. Earlier studies have shown that
the P–C–P angle in the ligand structure varies strongly
depending on changes at the metal centre. This exibility is
limited by bulky groups in the ligand backbone thus resulting
in more compact/rigid structures. This ultimately leads to
a closer contact between the phenyl group of the PPh3 moiety
and the gold centre and hence to stronger secondary ligand
metal interactions. Quantum theory of atoms-in-molecules
(QTAIM) calculations conrmed the presence of weak arene–
gold interactions in all complexes by the detection of a bond
critical point (Table 3, see ESI† for details). This complexation
behaviour is similar to well established biaryl phosphines.
Interestingly, the highest electron density at the bond critical
point was found for YTol, thus suggesting the strongest stabili-
zation in this complex.23
Fig. 4 Comparison of the catalytic efficiency of the gold complexes of
L1–L4 in catalysis. Reaction conditions: aniline (1 eq.), phenylacetylene
(1 eq.), LAuCl : NaBArF ¼ 1 : 1, 50 ꢀC (if not stated otherwise).
Conversion was determined by NMR spectroscopy.
activity) of the catalytically active species, which we attributed to
the better stabilization of the negative charge by the p-CF3C6H4
group. This corroborates with previous ndings which showed
that only YPhos ligands with highly electron-withdrawing
groups (–SO2Tol and –CN) in the ylide-backbone exhibit high
activities in gold catalysis (see above).
To our delight, introduction of the ortho-tolyl moiety resulted
in a surprisingly strong increase of activity. Full conversion to
the imine was already reached aer 4 h reaction time. Fortu-
nately, 4 was also active at RT and gave almost full conversion
aer 5 h. This superior activity of ligand 4 clearly demonstrates
that changes of the steric prole of the aryl group have a larger
impact on the activity than electronic effects (note that 2 and 4
are similarly strong donors according to their TEP, Table 2).
However, it is not the spatial inuence that the ligand passes
onto the metal which is decisive here (note that %Vbur only
marginally changes in whole ligand series), but the steric
Motivated by the excellent performance of 4 we envisioned
that a further steric protection of the ylidic carbon atom could
further improve the catalyst performance. Thus, the mesityl
analogue 5 (Mes ¼ 2,4,6-trimethylphenyl), in which the carba-
nionic centre is protected from both sides of the P–C–P plane,
was targeted.
Ligand optimization
protection of the carbanionic centre in the ylide-backbone. As The synthesis of YMesPCy2 (5) was attempted via the same
suggested by the previously observed inactivity of the most strategy as used for 1–4 (Scheme 2). However, phosphorylation
electron-rich YPhos ligands,10a the ylidic centre might be of the phosphonium salt D turned out to be less straight-
susceptible to protonation reactions or migration of the metal forward. Treatment of D with n-BuLi and Cy2PCl led to an
from the phosphine donor to the ylidic carbon atom, which equilibrium between the ylide and the phosphonium salt 5-H,
might lead to a complete shutdown of the catalytic ability. Steric which due to steric hinderance was largely on the reactant side
protection of the carbanionic carbon centre seems to suppress (Scheme 2). This problem was bypassed by addition of NaBF4 to
these side-reactions and thus improves the catalyst stability. remove the chloride from the equilibrium (precipitation of
This observation represents an important nding for further NaCl) thus preventing the reformation of Cy2PCl. Deprotona-
ligand design. Whereas electronic stabilization of the ylide is tion nally yielded YMesPCy2, which was isolated as yellow solid
effective for stabilizing the catalytically active gold species, it
also limits the tunability of the electronic properties since the
electron-withdrawing group ultimately leads to a decrease of the
ligand donor strength. This can be circumvented by the steric
protection of the ylide moiety, which also allows the generation
of ligands with stronger donor properties.
It is noteworthy that besides the steric protection of the ylidic
carbon atom also the increased rigidity of the ligand structure
caused by the introduction of the ortho-methyl group might play
¼ 2,4,6-trimethylphenyl).
Scheme 2 Preparation of the mesityl-substituted YPhos ligand 5 (Mes
4332 | Chem. Sci., 2021, 12, 4329–4337
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Edge Article
Chemical Science
aryl ring (Fig. 5b). This distortion is necessary to allow the
mesityl group to t into the pocket of the PPh3 and PCy2
moieties. The increased steric bulk in the backbone also causes
a slight increase of the steric demand of the ligand towards the
metal. With %Vbur ¼ 50.4 YMesPCy2 is the bulkiest ligand in the
series of YPhos ligands 1–5 (Table 2). Unfortunately, attempts to
determine the TEP of 5 repeatedly failed. Although treatment of
5 with Rh(acac)CO2 resulted in gas evolution, the IR spectrum
showed no signal in the region expected for the C–O vibration.
This is probably due to the elimination of both CO ligands from
the metal as was also observed for bulky NHCs and other YPhos
ligands.24
PPM-level catalysis
With YMesPCy2$AuCl in hand we next addressed the evaluation
of its catalytic performance. To our delight, the complex per-
formed equally well as the complex with the tolyl-substituted
phosphine 4 at 50 ꢀC with 0.1 mol% catalyst. However, in
contrast to YoTolPCy2 it even kept its high performance when
further decreasing the catalyst loading (Fig. 6, le). While for
example, YoTolPCy2 only delivered approx. 85% conversion aer
24 h with 0.05 mol% (TON ¼ 1610), YMesPCy2 gave full conver-
sion. Hence, the mesityl group in YMesPCy2 leads to the expected
further increase of the catalyst stability and thus to a better
performance at low catalyst loadings. This high activity is also
seen at room temperature (entry 1, Table 4). Even with only
0.1 mol% catalyst full conversion could be reached within 24 h.
The increased robustness of the active YMesPCy2-Au species
allowed a further decrease of the catalyst loading. With
0.005 mol% (50 ppm) high yields of 76% (TON ¼ 15 200,
Table 3) can be reached at 70 ꢀC, while turnover numbers as
Fig. 5 (a) Molecular structures of 5 and its gold complex and (b)
extract of the molecular structure of 5 showing the distortion of the
mesityl substituent.
in 49% yield in a one-pot reaction starting from the phospho-
nium salt D.
Y
MesPCy2 features two doublets in the 31P{1H} NMR spec-
trum (9.90 and 7.16 ppm) with a large coupling constant of
170.2 Hz. The crystal structures of 5 and 5$AuCl (Fig. 5) conrm
the increased steric bulk of the ligand: besides the almost ideal
perpendicular orientation of the mesityl group relative to the
P–C–P plane (5: 87.2(1)ꢀ; 5$AuCl: 85.3(1)ꢀ), the mesityl groups
are also extremely distorted, showing a marked deviation from
the ideal planarity of an aromatic substituent. In both struc-
tures, the para-(C1 and C9) and the ortho-methyl groups (C87
and C10) are tilted above and below the plane of the aromatic
ring, respectively, thus resulting in an overall bending of the
ꢀ
high as 20.000 can be obtained at 80 C.
To further evaluate the activity of 5 we compared the
performance of its gold complex with that of gold chloride
complexes of well-established, commercially available ligands.
Fig. 6 (Left) Comparison of the catalytic activity of the gold complexes with 4 and 5 at different catalyst loadings; (Right) comparison of the
activity of the AuCl complexes of 5, IPr and CyJohnPhos at 0.05 and 0.01 mol% catalyst loading. Reaction conditions: aniline (1 eq.), phenyl-
acetylene (1 eq.), gold complex : NaBArF ¼ 1 : 1, 50 ꢀC (if not stated otherwise). Conversion was determined by NMR spectroscopy.
© 2021 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci., 2021, 12, 4329–4337 | 4333

View Article Online
Chemical Science
Edge Article
Table 4 Hydroamination of alkynes with different primary amines with catalysts based on 5a
Entry
Ligand
R1
R2
R3, R4
T [ꢀC]
Time [h]
Mol% cat
Yield [%]
TON
1
2
3
4
5
6
7
8
5
5
5
4
4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
Ph
H
H
H
H
H
Ph, H
Ph, H
Ph, H
Ph, H
Ph, H
Ph, H
Ph, H
Ph, H
p-OMeC6H4, H
p-OMeC6H4, H
p-OMeC6H4, H
p-OMeC6H4, H
p-OMeC6H4, H
p-OMeC6H4, H
o-MeC6H4, H
Ph, H
Ph, H
Ph, H
Mes, H
Ph, Me
25
50
70
50
50
70
70
80
50
70
80
50
70
80
50
80
80
80
50
70
70
70
70
24
24
24
24
24
72
48
72
1
0.1
99 (85)b
97
74
99
85
990
0.05
0.01
0.1
0.05
0.005
0.0025
0.001
0.1
0.01
0.001
0.1
0.01
0.001
0.1
0.2
0.2
0.2
0.1
0.5
0.5
0.7
0.1
1950
7400
990
1610
15 200
18 800
20 000
990
7400
13 043
990
6600
18 256
990
490
490
265
990
138
144
100
990
76
47c
20
99
74
9
Ph
Ph
Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
72
72
1
13 (8)b
99
p-MeOC6H4
p-MeOC6H4
p-MeOC6H4
24
24
24
24
24
24
24
48
48
48
24
66
18 (11)b
99
Ph
n-Bu
Ph
Ph
Ph
Ph
Ph
Ph
Ph
98
98
53
99
69 (65)
72
70 (62)
99
p-MeC6H4, Me
p-MeC6H4, Et
–C2H4OC2H4-morpholine
a
Reaction conditions: 5 mmol acetylene, 5 mmol amine, gold complex : NaBArF ¼ 1 : 1. Yields were determined by NMR spectroscopy. b 20 mmol
50 mmol acetylene, 50 mmol amine, gold complex : NaBArF ¼ 1 : 1. With additional substrate aer 24 h (50 mmol). Yields were determined by
acetylene, 20 mmol amine, gold complex : NaBArF ¼ 1 : 1. Yields were determined by NMR spectroscopy. Values in brackets are isolated yields.
c
NMR spectroscopy. Values in brackets are isolated yields.
We chose the N-heterocyclic carbene IPr (IPr ¼ 1,3-bis(2,6- were achieved for a variety of substrates (Table 3). Even internal
diisopropylphenyl)imidazol-2-ylidene) and the biaryl phos- alkynes (entries 17 and 18), which are in general difficult to
phine CyJohnPhos which have been used in gold catalysis in the couple with other ligands including YPhos ligand A, were
past (Fig. 6, right).8,25 All ligands were tested under the same
reaction conditions, i.e. at 50 ꢀC with 0.05 and 0.01 mol%
catalyst loading. The conversion over time plots clearly conrm
the superior performance of 5 at low catalyst loadings. For
example, while YMesPCy2$AuCl provides approx. 95% yield aer
10 h reaction time with 0.05 mol% catalyst loading, only approx.
65% of product are formed when using the NHC- and biaryl-
phosphine catalysts. Overall, YMesPCy2 provides a highly active
and efficient catalyst for the hydroamination of acetylene. It's
activity is higher than those of simple phosphine ligands and
NHC and even higher than that observed for the sulfonyl-
substituted YPhos ligand A, which is more difficult to synthe-
sise since it requires the formation of a highly reactive meta-
lated intermediate.10 Thus, YMesPCy2 is not only more active but
also easier to synthesise. This also represents a decisive
advantage compared to the most efficient gold(I) catalysts re-
ported for the hydroamination of acetylene with aniline in
literature. A comparison of the ligand structures and the
performance of their gold complexes is shown in Fig. 7.5,7,9,26
Fig. 7 Comparison of the most efficient gold(I) catalysts in the
Next, the performance of 5 in the hydroamination of a series
hydroamination of phenylacetylene with aniline (Dipp
diisopropylphenyl).
¼
2,6-
of alkynes with different amines was tested. Excellent activities
4334 | Chem. Sci., 2021, 12, 4329–4337
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Edge Article
Chemical Science
readily converted to the imine, albeit requiring somewhat at room temperature with only 0.16 mol% catalyst loading.
higher reaction temperatures. Also, secondary amines, could be Thus, 5$AuCl is similarly active in this reaction than a specially
converted into the corresponding enamines and isolated in designed NHC-based catalyst described previously.7a The
high yields (entries 20 to 21). To our delight, also the hydro- formation of 10 thus also conrms the propensity of 5$AuCl to
amination with the aliphatic amine morpholine could be real- serve as (pre)catalyst in more complex C–C bond formation
ized giving the product in near quantitative yield aer 24 h reactions. Lastly, the catalyst also showed high activity in the
(entry 23). In contrast, no activity was seen for n-butylamine and hydration of phenylacetylene with water, thus also demon-
phenylacetylene under the same reaction conditions.
strating the stability of the complex under aqueous conditions
The activity of YMesPCy2$AuCl in the hydroamination with (Scheme 3d).
secondary amines led us to investigate its performance in
reactions with two equiv. or excess of alkyne. Primary amines,
Conclusions
such as aniline underwent double hydroamination to form
bisenamines such as 6 (Scheme 3a). This reaction also pro-
ceeded at low catalyst loadings and mild reaction temperatures
of only 50 ꢀC, thus giving 6a in 52% isolated yield.27 In contrast,
secondary amines were found to react with two equiv. of alkyne
to form 1,2-dihydroquinolines of type 7 via a hydroamination,
C–H activation and C–C bond formation sequence (Scheme 3b).
Such a cyclization reaction has been performed with gold
complexes with N-heterocyclic carbenes and cyclic alkyl(amino)
carbenes (CAAC), however, under more forcing reaction
conditions.28
Next, we tested the activity of 5 in a more complex reaction to
further prove the scope of application of the YPhos-gold
complex. We chose the domino process with 1,6-enyne 8 and
trimethoxy benzene 9 as nucleophile.29,30 To our delight, 5$AuCl
proofed to be highly efficient also in this transformation, thus
giving the cyclized compound 10 in quantitative yields aer 4 h
In conclusion, we developed a highly active, easy-to-synthesise
YPhos-based gold catalyst for hydroamination reactions of
alkynes. Systematic studies on the impact of the electronic and
steric properties of aryl groups in the ligand backbone
impressively demonstrated the importance of steric protection
of the ylidic carbon centre for the activity and stability of the
catalyst. While electronic changes at the aryl substituent only
had a minor impact on the catalytic activity and only gave way to
low conversions under mild reaction conditions, an increase of
the steric bulk by introduction of an ortho-tolyl and mesityl
substituent led to a boost in the catalytic activity by several
orders of magnitude. The gold complex of the mesityl-
substituted ligand YMesPCy2 thus showed high conversions
with only 50 ppm of catalyst loading and reached turnover
numbers of 20 000. Thus, this complex competes with the most
active catalysts based on highly sophisticated phosphine and
NHC complexes reported in literature, albeit its simple molec-
ular design and facile synthesis. Furthermore, YMesPCy2 can
also be applied in further gold catalysed reactions including the
formation of dihydroquinolines via an amination and C–C bond
formation sequence or enyne cyclisation.
Overall, the remarkable increase of activity by backbone
modication emphasizes the importance of controlling ligand
properties also beyond typical electronic and steric properties as
measured by the Tolman electronic parameter or cone angle.
These results demonstrate the importance of ligand design and
will be helpful for further improvements of the ligand struc-
tures also in other catalytic transformations.
Conflicts of interest
The authors have led patent WO2019030304 covering the
YPhos ligands and precatalysts discussed, which is held by
UMICORE and products will be made commercially available
from.
Acknowledgements
This work was supported by RESOLV, funded by the Deutsche
Forschungsgemeinscha (DFG, German Research Foundation)
under Germany's Excellence Strategy – EXC-2033 – Pro-
jektnummer 390677874 and by the European Research Council
(Starting Grant: Ylide Ligands 677749). We also thank UMI-
CORE for nancial support and the donation of chemicals. The
Scheme 3 Application of YMesPCy2$AuCl in bisenamine and 1,2-
dihydroquinoline formation. Yields were determined by NMR spec-
troscopy. Values in brackets correspond to isolated yields.
© 2021 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci., 2021, 12, 4329–4337 | 4335

View Article Online
Chemical Science
Edge Article
authors thank Ilja Rodstein for the synthesis of amine
precursors.
621–626; (c) T. Scherpf, R. Wirth, S. Molitor, K.-S. Feichtner
and V. H. Gessner, Angew. Chem., Int. Ed., 2015, 54, 8542–
8546.
11 C. Schwarz, J. Handelmann, D. M. Baier, A. Ouissa and
V. H. Gessner, Catal.: Sci. Technol., 2019, 9, 6808–6815.
Notes and references
1 (a) C. Torborg and M. Beller, Adv. Synth. Catal., 2009, 351, 12 (a) P. Weber, T. Scherpf, I. Rodstein, D. Lichte, L. T. Scharf,
3027; (b) B. Schlummer and U. Scholz, Adv. Synth. Catal.,
2004, 346, 1599; (c) V. Farina, Adv. Synth. Catal., 2004, 346,
1553; (d) D. J. Ager, A. H. M. de Vries and H. G. de Vries,
Chem. Soc. Rev., 2012, 41, 3340.
2 W. Tang and X. Zhang, Chem. Rev., 2003, 103, 3029–3070.
3 A. Biffis, P. Centomo, A. Del Zotto and M. Zecca, Chem. Rev.,
2018, 118, 2249–2295.
L. J. Gooßen and V. H. Gessner, Angew. Chem., Int. Ed., 2019,
58, 3203–3207; (b) L. Scharf, I. Rodstein, M. Schmidt,
T. Scherpf and V. H. Gessner, ACS Catal., 2020, 10, 999.
13 X.-Q. Hu, D. Lichte, I. Rodstein, P. Weber, A.-K. Seitz,
T. Scherpf, V. H. Gessner and L. J. Gooßen, Org. Lett., 2019,
21, 7558–7562.
14 W. Wang, G. B. Hammond and B. Xu, J. Am. Chem. Soc., 2012,
134, 5697–5705.
4 Selected reviews: (a) A. M. Asiri and A. S. K. Hashmi, Chem.
¨
Soc. Rev., 2016, 45, 4471; (b) A. Furstner, Angew. Chem., Int. 15 (a) E. Deck, H. E. Wagner, J. Paradies and F. Breher, Chem.
Ed., 2018, 57, 4215; (c) R. Dorel and A. M. Echavarren,
Chem. Rev., 2015, 115, 9028–9072; (d) A. S. K. Hashmi,
Angew. Chem., Int. Ed., 2010, 49, 5232–5241.
5 Y. Wang, Z. Wang, L. Yuxue, G. Wu, Z. Cao and L. A. Zhang,
Nat. Commun., 2014, 5, 3470.
Commun., 2019, 55, 5323–5325; (b) T. Witteler,
H. Darmandeh, P. Mehlmann and F. Dielmann,
Organometallics, 2018, 37, 3064–3072; (c) K. D. Hesp and
M. Stradiotto, J. Am. Chem. Soc., 2010, 132, 18026–18029;
(d) N. L. Rotta-Loria, A. J. Chisholm, P. M. MacQueen,
R. McDonald, M. J. Ferguson and M. Stradiotto,
Organometallics, 2017, 36, 2470–2475.
6 (a) Z. Wang, C. Nicolin, C. Hervieu, Y.-F. Wong, G. Zanoni
and L. Zhang, J. Am. Chem. Soc., 2017, 139, 16064; (b)
X. Cheng, Z. Wang, C. D. Quintanilla and L. Zhang, J. Am. 16 L. Falivene, Z. Cao, A. Petta, L. Serra, A. Poater, R. Oliva,
Chem. Soc., 2019, 141, 3787; (c) X. Li, S. Liao, Z. Wang and
L. Zhang, Org. Lett., 2017, 19, 3687.
V. Scaraon and L. Cavallo, Nat. Chem., 2019, 11, 872–879.
17 C. A. Tolman, J. Am. Chem. Soc., 1970, 92, 2956.
¨
7 (a) Y. Tang, I. Benaissa, M. Huynh, L. Vendier, N. Lugan, 18 (a) J. Iglesias-Siguenza, C. Izquierdo, E. Diez, R. Fernandez
´
S. Bastin, P. Belmont, V. Cesar and V. Michelet, Angew.
and J. M. Lassaletta, Dalton Trans., 2016, 45, 10113–10117;
(b) C. Burstein, C. W. Lehmann and F. Glorius,
Tetrahedron, 2005, 61, 6207–6217; (c) H. Tinnermann,
L. D. M. Nicholls, T. Johannsen, C. Wille, C. Golz,
R. Goddard and M. Alcarazo, ACS Catal., 2018, 8, 10457–
10463; (d) L. Shengrong, P. Alessio, C. Xinpeng, M. Xu,
Z. Giuseppe and L. Zhang, Angew. Chem., Int. Ed., 2018, 57,
Chem., Int. Ed., 2019, 58, 7977–7981; (b) I. Varela,
H. Faustino, E. Dıez, J. Iglesias-Siguenza, F. Grande-
Carmona, R. Fernandez, J. M. Lassaletta, J. L. Mascarenas
and F. Lopez, ACS Catal., 2017, 7, 2397–2402; (c) K. Azouzi,
C. Duhayon, I. Benaissa, N. Lugan, Y. Canac, S. Bastin and
V. Cesar, Organometallics, 2018, 37, 4726–4735; (d)
´
¨
´
˜
´
´
´
´ ´
8250–8254; (e) P. Perez-Galan, N. Delpont, E. Herrero-
N. Marion, R. S. Ramon and S. P. Nolan, J. Am. Chem. Soc.,
´
2009, 131, 448–449; (e) P. Klumphu, C. Desfeux, Y. Zhang,
S. Handa, F. Gallou and B. H. Lipshutz, Chem. Sci., 2017, 8,
Gomez, F. Maseras and A. M. Echavarren, Chem.–Eur. J.,
2010, 16, 5324–5332.
´ ˇ
´
6354–6358; (f) S. G. Mahamulkar, I. Cısarova and U. Jahn, 19 L. Chen, P. Ren and B. P. Carrow, J. Am. Chem. Soc., 2016,
Adv. Synth. Catal., 2018, 360, 4215–4224. 138, 6392–6395.
8 (a) D. Malhotra, M. S. Mashuta, G. B. Hammond and B. Xu, 20 A. Poater, B. Cosenza, A. Correa, S. Giudice, F. Ragone,
Angew. Chem., Int. Ed., 2014, 53, 4456; (b) S. L. A. Porta,
X. Cheng, X. Ma, G. Zanoni and L. Zhang, Angew. Chem.,
V. Scarano and L. Cavallo, SambVca, Eur. J. Inorg. Chem.,
2009, 1759–1766.
¨
Int. Ed., 2018, 57, 8250–8254; (c) H. Tinnermann, 21 T. E. Muller and D. M. P. Mingos, Transition Met. Chem.,
L. D. Nicholls, T. Johannsen, C. Wille, C. Golz, R. Goddard
1995, 20, 533–539.
and M. Alcarzo, ACS Catal., 2018, 8, 10457–10463; (d) 22 (a) Z. Wang, C. Nicolin, C. Hervieu, Y.-F. Wong, G. Zanoni
¨
L. Noel-Duchesneau, N. Lugan, G. Lavigne, A. Labande and
and L. Zhang, J. Am. Chem. Soc., 2017, 139, 16064–16067;
(b) L. Noel-Duchesneau, N. Lugan, G. Lavigne, A. Labande
´
¨
V. Cesar, Eur. J. Inorg. Chem., 2015, 1752–1758; (e) L. Noel-
´
Duchesneau, N. Lugan, G. Lavigne, A. Labande and
and V. Cesar, Organometallics, 2014, 33, 5085; (c) E. Deck,
´
V. Cesar, Organometallics, 2014, 33, 5085–5088.
H. E. Wagner, J. Paradies and F. Breher, Chem. Commun.,
2019, 55, 5323–5325; (d) K. D. Hesp and M. Stradiotto, J.
Am. Chem. Soc., 2010, 132, 18026–18029; (e) N. L. Rotta-
Loria, A. J. Chisholm, P. M. MacQueen, R. McDonald,
M. J. Ferguson and M. Stradiotto, Organometallics, 2017,
36, 2470–2475; (f) X. Cheng, Z. Wang, C. D. Quintanilla
and L. Zhang, J. Am. Chem. Soc., 2019, 141, 3787–3791.
9 (a) V. Lavallo, J. H. Wright and F. S. Tham, Angew. Chem., Int.
Ed., 2013, 52, 3172–3176; (b) S. P. Fisher, S. G. McArthur,
V. Tey, S. E. Lee, A. L. Chan, I. Banda, A. Gregory,
K. Berkley, C. Tsay, A. L. Rheingold, G. Guisado-Barrios
and V. Lavallo, J. Am. Chem. Soc., 2020, 142, 251–256.
10 (a) T. Scherpf, C. Schwarz, L. T. Scharf, J.-A. Zur, A. Helbig
and V. H. Gessner, Angew. Chem., Int. Ed., 2018, 57, 12859– 23 Note that in the crystal structure the Au–C interactions partly
12864; (b) C. Schwarz, T. Scherpf, I. Rodstein,
K.-S. Feichtner and V. H. Gessner, ChemistryOpen, 2019, 8,
deviate from those of the calculated structures. This is
presumably due to further intermolecular interactions
4336 | Chem. Sci., 2021, 12, 4329–4337
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Edge Article
Chemical Science
´
´
´
within the crystal thus further demonstrating that these 27 A. Leyva-Perez, J. R. Cabrero-Antonino, A. Cantın and
interactions are weak and can be overcompensated by A. Corma, J. Org. Chem., 2010, 75, 7769–7780.
intermolecular interactions or interactions with the solvent. 28 (a) X.-Y. Liu, P. Ding, J.-S. Huang and C.-M. Che, Org. Lett.,
24 (a) T. Scherpf, I. Rodstein, M. Paaßen and V. H. Gessner,
Inorg. Chem., 2019, 58, 8151–8161; (b) R. Dorta,
E. D. Stevens, N. M. Scott, C. Costabile, L. Cavallo,
C. D. Hoff and S. P. Nolan, J. Am. Chem. Soc., 2005, 127,
2485–2495.
2007, 9, 2645–2648; (b) X. Zeng, G. D. Frey, R. Kinjo,
B. Donnadieu and G. Bertrand, J. Am. Chem. Soc., 2009,
131, 8690–8696; (c) B. Rao, H. Tang, X. Zeng, L. Liu,
M. Melaimi and G. Bertrand, Angew. Chem., Int. Ed., 2015,
54, 14915–14919.
25 (a) V. Kadiyala, P. B. Kumar, S. Balasubramanian and 29 For examples of gold-catalysed enyene cyclisations: (a)
G. V. Karunakar, J. Org. Chem., 2019, 84, 12228–12236; (b)
T. J. O'Connor and F. Dean Toste, ACS Catal., 2018, 8,
5947–5951; (c) F. Sirindil, S. Golling, R. Lamare,
J.-M. Weibel, P. Pale and A. Blanc, Org. Lett., 2019, 21,
8997–9000.
X. Han, P. Retailleau, V. Gandon and A. Voituriez, Chem.
Commun., 2020, 56, 9457–9460; (b) C. Obradors,
D. Leboeuf, J. Aydin and A. M. Echavarren, Org. Lett., 2013,
15, 1576–1579.
ˆ
30 (a) P.-Y. Toullec, E. Genin, L. Leseurre, J.-P. Genet and
26 S. Yazdani, G. P. Junor, J. L. Peltier, M. Gembicky, R. Jazzar,
D. B. Grotjahn and G. Bertrand, ACS Catal., 2020, 10, 5190–
5201.
V. Michelet, Angew. Chem., Int. Ed., 2006, 45, 7427–7430;
(b) C. H. M. Amijs, C. Ferrer and A. M. Echavarren, Chem.
Commun., 2007, 698–700.
© 2021 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci., 2021, 12, 4329–4337 | 4337