Deactivation of cationic CuI and AuI catalysts for A3 coupling by CH2Cl2: Mechanistic implications of the formation of neutral CuI and AuI chlorides

Article abstract of DOI:10.1002/anie.201403973

Care should be exercised when using CH₂Cl₂ as a solvent for reactions in which amines are a reagent, since undesirable deactivation of cationic copper(I) and gold(I) catalysts to form the corresponding inactive neutral chloride complexes [LMCl] (M=Cu or Au) can occur as a result of the generation of hydrogen chloride in the medium. Cu^I and Au^I deactivation has been proved for the Mannich three-component coupling reaction. A series of Cu^I and Au^I complexes with potential mechanistic implications were isolated and characterized by X-ray crystallography.

Full text of DOI:10.1002/anie.201403973

Angewandte
Chemie
DOI: 10.1002/anie.201403973
Homogeneous Catalysis
Deactivation of Cationic CuI and AuI Catalysts for A3 Coupling by
CH2Cl2: Mechanistic Implications of the Formation of Neutral CuI and
AuI Chlorides**
Abdessamad Grirrane,* Eleuterio ꢀlvarez, Hermenegildo Garcꢁa,* and Avelino Corma*
Abstract: Care should be exercised when using CH2Cl2 as
a solvent for reactions in which amines are a reagent, since
undesirable deactivation of cationic copper(I) and gold(I)
catalysts to form the corresponding inactive neutral chloride
complexes [LMCl] (M = Cu or Au) can occur as a result of the
generation of hydrogen chloride in the medium. CuI and AuI
deactivation has been proved for the Mannich three-compo-
nent coupling reaction. A series of CuI and AuI complexes with
potential mechanistic implications were isolated and charac-
terized by X-ray crystallography.
complexes in combination with the isolation of potential
intermediate complexes has not been reported. Also, no
information is available on the reaction of these intermediate
complexes with A3 reagents of this Mannich reaction in
toluene or CH2Cl2. The isolation of metal(I) complexes can
shed light on the active or inactive gold(I) and copper(I)
species involved in the catalytic process.
In the initial stage of our study, we selected [Cu-
(MeCN)4]+ [PF6]À (1)[7] as a precursor for the non-aqueous
synthesis of cationic CuI dialkylbiarylphosphane complexes.
The reaction of 1 with 2-di-tert-butylphosphanylbiphenyl (2)
in CH2Cl2 at 458C for 15 h led to the formation of the cationic
copper(I) complex 3 (Scheme 1) in 85% yield (see exper-
imental details in the Supporting Information). When a stoi-
chiometric (1:2) mixture of the synthesized complex 3 and
aniline (4) was stirred for 15 h at room temperature in
CH2Cl2, a new cationic amine–copper(I) complex 5 was
formed in 86% yield (see the Supporting Information) by the
exchange of acetonitrile by aniline (Scheme 1).
Complexes 3 and 5 were fully characterized on the basis of
combustion elemental analysis and spectroscopic data (see
the experimental details as well as Figures S1–S6 and Table S1
for 3 and Figures S7–S12 and Table S2 for 5 in the Supporting
Information). Crystals of complex 3 were obtained at À308C
from solutions in CH2Cl2, on top of which a layer of n-hexane
was carefully added. The addition of toluene (2 mL) to the
filtered solution of complex 5 in CH2Cl2 (2 mL) and
subsequent slow evaporation at ambient temperature enabled
the formation of colorless crystals of 5. Single crystals of both
complexes 3 and 5 were air-stable and of suitable quality for
an X-ray crystallographic diffraction study.
T
here is much current interest in delineating the similarities
and differences between gold(I) catalysts[1] and isoelectronic
metal ions, in particular available and affordable copper(I)
complexes.[2] It is important to establish whether CuI can
exhibit similar catalytic activity to that of AuI in certain
reactions. Herein, we compare the catalytic activity of bulky
2-di-tert-butylphosphanylbiphenyl (Buchwald phosphane)[3]
CuI complexes with respect to AuI analogues for the Mannich
three-component (A3) coupling and show that CuI complexes
are far more active than AuI complexes. We also demonstrate
why CH2Cl2 is not a suitable solvent for this process on the
basis of mechanistic differences derived from the preference
of AuI to coordinate to the alkyne and of CuI to coordinate
with amines. In this study, a series of CuI and AuI complexes
with potential mechanistic implications were isolated and
characterized by X-ray crystallography.
A3 coupling[4] is a versatile process that finds broad
application in organic synthesis for the straightforward
preparation of high-value propargylamines.[5] There are
precedents for the ability of gold and copper catalysts to
promote this type of Mannich coupling.[6] However, up to
now, a comparison of the activity of cationic LCuI and LAuI
The solid-state structure of
3
(Scheme 1)—the
hexafluorophosphate salt of the cationic 2-di-tert-
butylphosphanylbiphenyl acetonitrile CuI complex (see
Table S1 and Figure S1)—is closely related to the reported
tetrafluoroborate CuI analogue, but with a differently sub-
stituted biarylphosphine ligand and a different counterion.[8]
The selected distances Cu–N, Cu–P, and Cu–Cipso and the
shortest distance between Cu1 and the plane of the com-
plexed arene (C9) for complex 3 are around 1.870(3),
2.1789(8), 2.78, and 2.640(3) ꢀ, respectively. The distances
Cu–N, Cu–P, Cu–Cipso and the shortest distance between Cu1
and the plane of the complexed arene (C13) for complex 5
(see Table S2 and Figure S7) are 1.964(2), 2.1821(7), 2.786,
and 2.577(3) ꢀ, which are shorter than the corresponding
values for Au–N, Au–P, Au–Cipso, and the shortest distance
between Au1 and the plane of the complexed arene (C13) in
the recently reported analogous cationic amine–AuI–L com-
[*] Dr. A. Grirrane, Prof. Dr. H. Garcꢀa, Prof. Dr. A. Corma
Instituto Universitario de Tecnologꢀa Quꢀmica CSIC-UPV
Universidad Politꢁcnica de Valencia
Av. de los Naranjos s/n, 46022 Valencia (Spain)
E-mail: hgarcia@qim.upv.es
acorma@itq.upv.es
Dr. E. ꢂlvarez
Instituto de Investigaciones Quꢀmicas CSIC-US
Departamento de Quꢀmica Inorgꢃnica
Av. Amꢁrico Vespucio 49, 41092 Sevilla (Spain)
[**] Financial support by the Spanish Ministry of the Economy and
Competitiveness (Severo Ochoa and CTQ2012-32315) and the
Generalidad Valenciana (Prometeo 2012-014) is gratefully
acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201403973.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
plex,[9] thus showing the influence of the metal on the metal–
arene interaction.
The copper(I) complexes 3 and 5 were also characterized
by NMR spectroscopy and ESIMS (see the Supporting
Information). 1H, 13C, 31P, and 19F NMR spectroscopy of
solutions in CD2Cl2 provided evidence that the starting
complex 1 was completely converted into 3 and subsequently
into 5. Thus, 31P NMR spectroscopy showed the disappear-
ance of the peaks corresponding to the initial free phosphine 2
The copper(I) complex 5 can also be synthesized directly
from a stoichiometric (1:1:1.4) mixture of 1, 2, and 4
(Scheme 1; see the Supporting Information). A mixture of
these three components was stirred for 24 h at room temper-
ature in CH2Cl2. Filtering of the resulting solution, followed
by the subsequent addition of toluene and slow evaporation at
ambient temperature, afforded colorless crystals of complex 5
in 80% yield and high quality for crystallographic studies.
À
at 18.11 ppm (singlet) and to PF6 of starting complex 1 at
À144.43 ppm (septuplet), and the appearance of new peaks at
32.49 (singlet) and À144.36 ppm (septuplet) for 3 or 34.60
(singlet) and À144.13 ppm (septuplet) for 5. Similarly,
1H NMR spectroscopy showed new signals for the methyl
groups of complex 3 at 1.27 and 1.22 ppm instead of those
corresponding to the initial phosphine ligand 2 at 1.11 and
1.07 ppm. ESIMS of a solution obtained by dissolving
complex 3 in CH2Cl2/CH3CN (1:2) showed intense positive
MS peaks at 361.1 and 759.4 Da (see Figure S6), which are
À
+
À
ꢀ
ꢀ
attributable to [3ÀPF6 ÀN CMe] and [(3)2À(PF6 )2À(N
CMe)2+ClÀ]+ (see fragmentation of this peak in Figure S6),
respectively. Complex 5 dissolved in CH2Cl2/CH3CN (1:2)
showed mainly positive MS peaks at 454.2, 361.1, and
759.4 Da, which are attributable to [C26H34CuF6NP2
(5)ÀPF6 ]+
and
the
cationic
CuI
complexes
À
[5ÀPF6ÀÀaniline]+ and [C52H68Cu2F12N2P4 ((5)2)À(PF6)À
2
À(aniline)2+ClÀ]+, respectively (see Figure S12 for details).
With the aim of gaining understanding of the stability of
CuI coordinated to the bulky ligand 2 in complexes 3 and 5, we
performed the Mannich A3 coupling of phenylacetylene,
pyrrolidine, and aqueous formaldehyde with 3 and 5 as
catalysts (Table 1). The catalytic activity of the related gold
complexes [LAu(NCMe)][SbF6] (6)[10] and [LAu(aniline)]-
Table 1: A3 coupling of phenylacetylene, pyrrolidine, and aqueous form-
aldehyde with CuI catalyst complexes 3, 5, 13, 14, and 15 and AuI catalyst
complexes 6, 7, and 10.[a]
Entry
Catalyst
t [h]
Conversion [%][b]
Yield [%][b]
1
2
3
4
5
6
7
8
9
3
5
6
0.2
0.2
12
24
0.2
24
0.2
24
0.2
0.2
100
100
100
100
4
90
2
96
100
100
99
98
95
94
trace
81
trace
87
99
99
7
10
10
13
13
14
15
10
[a] Reaction conditions: phenylacetylene (0.25 mmol), pyrrolidine
(0.35 mmol), aqueous formaldehyde (0.7 mmol), catalyst (6 mol%),
toluene (1 mL). [b] The conversion and yield were determined by
1H NMR spectroscopy and GC of the crude reaction mixture.
Scheme 1. Top: Synthesis of cationic copper(I) complexes 3 and 5.
Bottom: ORTEP views of 3 and 5 (single-crystal X-ray data and the
crystal packing of 3 and 5 are given in Tables S1 and S2 and Figures S1
and S7).
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
[SbF6] (7)[9] was also screened for the sake of comparison (see
Table 1) and with the goal of shedding light on the real active
species involved in the catalytic process for gold(I) and
copper(I). In these four catalytic experiments (Table 1), the
expected product 1-(3-phenylprop-2-ynyl)pyrrolidine (8)[11]
was obtained in good yield, but the reaction took place in
only a few minutes with CuI complexes 3 and 5 (Table 1,
entries 1 and 2), whereas a longer reaction time was required
for the cationic AuI complexes 6 and 7 (Table 1, entries 3 and
4). This result reflects the higher activity of cationic CuI
complexes 3 and 5 versus AuI analogues 6[10] and 7.[9]
The propargylamine 8 was isolated as a light-yellow oil
and characterized by 1H and 13C NMR spectroscopy and GC–
MS (see Figures S13, S14, and S15). The hydrochloride salt of
8 (8H+ ClÀ)[12] (Scheme 2) formed in the A3 reaction in
CH2Cl2 with [LAu(NCMe)][SbF6] (6)[10] as the catalyst was
An alternative catalytic experiment was carried out in
which [LAu(NCMe)][SbF6] (6; 12 mol%) was added to
a mixture of pyrrolidine and phenylacetylene in CD2Cl2 in
the absence of formaldehyde at 708C. After 1 h, the mixture
was analyzed by 31P NMR spectroscopy. Peaks in the spec-
trum at 64.59 (91%) and 60.10 ppm (9%) corresponding to 9
and 10, respectively, replaced the peak at 57.53 ppm for the
starting precatalyst [LAu(NCMe)][SbF6] (6) (see Figure S25).
After a reaction time of 5 h, complex 9 (64.59 ppm) was
a minor component of the mixture (6%) with respect to
complex 10 (60.10 ppm), which predominated (94%; see
Figure S26). Finally, after a reaction time of 6 h, the only gold
species present was complex 10 (60.10 ppm; see Figure S27).
When [LAu(NCMe)][SbF6] (6) was treated at room
temperature with pyrrolidine in an NMR spectroscopic
experiment in CD2Cl2 as the solvent (Scheme 3, top), the
Scheme 2. Formation of the intermediate cationic chloride-bridged
digold(I) complex [(LAu)2(m-Cl)][SbF6] (9), the neutral gold(I) complex
[LAuCl] (10), and the chloride salt 8H+ ClÀ of propargylamine 8, in the
absence of formaldehyde. Conditions for the isolation of complexes 9
and 10 are provided in the Supporting Information. X-ray single-crystal
data and ORTEP views of 8H+ ClÀ, 9, and 10 are given in Tables S3–S5
and Figures S16, S18, and S22.
Scheme 3. Top: Synthesis of the cationic gold(I) complex
[LAu(pyrrolidine)][SbF6] (11) in CD2Cl2 as the solvent. Bottom: Forma-
tion of the fluxional cationic bridged phenylacetylene digold(I) complex
[(s,p)(LAu)2(m-phenylacetylene)][SbF6] (12) as the final Au complex.
Details and conditions for the isolation and characterization of
complexes 11 and 12 are provided in the Supporting Information.
characterized by combustion analysis, X-ray crystallography
(see Table S3 and Figure S16), and ESIMS of a solution
obtained by dissolving 8H+ ClÀ in CH2Cl2/methanol (1:1;
a positive MS peak was observed at 186.1 amu, which
corresponds to the expected mass for 8H+; see Figure S17).
Besides the reaction product, cationic [(LAu)2(m-Cl)][SbF6]
(9)[13] formed as an intermediate complex and neutral
[LAuCl] (10)[10] formed as the final gold chloride complex
(Scheme 2) during the catalytic reaction in CH2Cl2 as the
solvent in the absence or presence of formaldehyde were
detected and fully characterized (see the Supporting Infor-
mation). The X-ray crystal structures of the isolated com-
plexes 9 and 10 were similar to (in the case 9) and a polymorph
of (in the case of 10) those previously reported for these
complexes[10,13] (see Figures S18 and S22 and Tables S4 and
S5). 1H and 31P NMR spectra of 9 and 10 in CD2Cl2 were also
recorded (see Figures S19, S20, S23, and S24). The ESI mass
spectrum of isolated [(LAu)2(m-Cl)][SbF6] (9) dissolved in
CH2Cl2/methanol showed a positive MS peak at 1025.3 amu
corresponding to the expected mass for the chloride-bridged
complex (LAu)2(m-Cl)]+, whereas negative MS peaks
observed at 234.6 and 236.6 amu correspond to the anionic
counteranion [SbF6]À (see Figure S21).
complex [LAu(pyrrolidine)][SbF6] (11) could be isolated and
characterized on the basis of analytical and spectroscopic data
(see experimental details and Figures S28 and S29 in the
Supporting Information) before its transformation into the
final, inactive, neutral species [LAuCl] (10; Scheme 2).
Complex 11 was obtained by quick evaporation of the
reaction solution under reduced pressure, followed by several
washes with pentane. The combustion elemental analysis of
the solid is in accordance with the percentages expected for
the cationic gold(I) complex 11. Finally, when the reagents of
the A3 reaction were exposed to stoichiometric amounts of
[LAu(NCMe)][SbF6] (6) in toluene as the solvent (Scheme 3,
bottom), the fluxional, cationic, bridged phenylacetylene
digold(I) complex [(s,p)(LAu)2(m-phenylacetylene)][SbF6]
(12) corresponding to the final stable gold product previously
reported by us[14] was isolated. The solid-state structure of 12
was confirmed by single-crystal X-ray diffraction of crystals of
complex 12 obtained by slow evaporation at room temper-
ature of filtered solutions in toluene/CH2Cl2. The structure
obtained was a polymorph of our previously reported
structure,[14] but in this case the asymmetric unit contains
two equivalent units of 12 accompanied by one toluene
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
molecule as a molecule of crystallization (see Figure S30 and
Table S6).
plex [{LCu(m-Cl)}2] (13; see Figures S31–S35 and Table S7),
which can also be synthesized by a direct route (see the
Supporting Information). The ORTEP structure of 13 is
shown in Scheme 4.
The cationic chloride-bridged digold(I) complex [(LAu)2-
(m-Cl)][SbF6] (9)[13] was recently obtained by incomplete
chloride removal from the complex [LAuCl] (10). The
formation of these compounds 8H+ ClÀ, 9, and 10 indicates
that CH2Cl2 is not an innocent solvent under the conditions of
the A3 reaction, but provides chloride ions that can be bonded
to the metal ion.[15] The formation of these chloride species in
the presence of AuI is also possible even at room temperature
instead of 508C. There is some precedent for the facile
activation of CH2Cl2 in the presence of transition metals, such
as Zn or Co,[16] Rh,[17] and Mg/TiCl4/THF,[18] but as far as we
know, the instability of CH2Cl2 was not detected previously
with 2-di-tert-butylphosphanylbiphenyl AuI or CuI complexes
as catalysts. We tested the catalytic activity of this neutral
gold(I) complex [LAuCl] (10; Table 1, entries 5 and 6) for the
Mannich A3 coupling and found that it was considerably less
active than the initial precatalyst CuI complexes 3 and 5 and
the AuI complexes 6 and 7. It is well-known[1b,c,g,19] that Ag+
salts are required as promoters to activate 10 as precatalyst by
chloride removal, forming insoluble AgCl precipitates, thus
producing supposedly more active cationic Au+. In the
present case, when CH2Cl2 was used as the solvent, the
isolation of chloride-containing AuI complexes 9 and 10
should correspond to the partial and total inactivation,
respectively, of the catalyst. According to the previous
rationalization, CH2Cl2 is not the most appropriate solvent
for AuI catalysis in spite of the ample use of this solvent in
gold catalysis.
The catalytic activity of this neutral copper(I) complex 13
was also tested for the Mannich A3 coupling (Table 1,
entries 7 and 8). We found that 13 was considerably less
active than the initial CuI precatalyst complexes 3 and 5 and
AuI complexes 6 and 7. Thus, the higher reaction rate for the
A3 coupling in toluene as compared to that in CH2Cl2 is
attributable to the absence of chlorine atoms in the medium
to deactivate the transition-metal catalyst. Chloride can also
be formed by the reaction of CH2Cl2 or CD2Cl2 with
pyrrolidine. The resulting HCl or DCl and iminium ions
were identified on the basis of the corresponding positive MS
peaks at 84.1 for [C5H10N]+ and 86.1 Da for [C5H8D2N]+ (see
Figures S36 and S37), and also by their consecutive reaction
with phenylacetylene to give the normal or deuterated
propargylamine 8 (C13H15N) or [D2]8 (C13H13D2N), which
were identified on the basis of the corresponding positive MS
peaks observed by GC–MS at 185.2 or 187.2 Da (see
Figures S15 and S38).
One important conclusion from Table 1 is that the CuI
complexes were considerably more active than the AuI
complexes. To understand this higher activity of CuI, we
attempted to isolate complexes of CuI and AuI with the
various reagents of the A3 coupling in a series of experiments.
These complexes could have implications with regard to the
reaction mechanism. In this context, we were unable to obtain
a phenylacetylene–copper(I) complex by mixing equimolar
amounts of phenylacetylene and the complex [Cu(L)-
(NCMe)][PF6] (3). This negative result contrasts sharply
with the previously reported behavior of [Au(L)(NCMe)]-
[SbF6] (6), which forms isolable, fluxional digold complexes
with phenylacetylene.[14]
Our results, observations, and conclusions were similar
when [LCu(NCMe)][PF6] (3) was used as a precatalyst for the
A3 coupling or as a stoichiometric reagent with pyrrolidine in
CH2Cl2 (Scheme 4). Thus, we were able to isolate and fully
characterize the neutral, dichloride-bridged dicopper(I) com-
In contrast to the failure to isolate the copper acetylide
complex, when [Cu(L)(NCMe)][PF6] (3) was treated with
pyrrolidine in the presence of formaldehyde and with toluene
as the solvent (Scheme 5, top), complex [LCu(pyrrolidine)]-
[PF6] (14) was isolated and fully characterized by analytical
and spectroscopic techniques (see experimental details,
Figures S39–S44, and Table S8 in the Supporting Informa-
tion). Combustion elemental analysis of the cationic cop-
per(I) complex 14 was in accordance with the percentages
expected for its formula. Crystals of complex 14 of sufficient
quality for X-ray crystallography were obtained by recrystal-
lization of the precipitate formed during the reaction of 3 and
pyrrolidine in a mixture of CH2Cl2 and toluene (see Fig-
ure S39 and Table S8). Finally, when the A3 reaction was
performed in the presence of a stoichiometric amount of
[LCu(NCMe)][PF6] (3) in toluene as the solvent (Scheme 5,
middle), the [LCu(propargylamine)][PF6] complex 15, corre-
sponding to the complex of the Cu catalyst with the final
product, was isolated and characterized on the basis of
analytical and spectroscopic data (see experimental details,
Figures S45–S50, and Table S9 in the Supporting Informa-
tion). Combustion elemental analysis of complex 15 was in
accordance with the percentages expected for its formula.
Colorless crystals of complex 15 were obtained by the slow
Scheme 4. Top: Synthesis of the bridged dichloride dicopper(I) com-
plex [{LCu(m-Cl)}2] (13) in CH2Cl2 as the solvent. Bottom: ORTEP view
of the neutral CuI complex 13 (X-ray single-crystal data are given in
Table S7 and Figure S31).
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Scheme 6. Plausible mechanism for the A3 coupling leading to prop-
argylamine 8 with cationic CuI complex 3 as the precatalyst in the
absence of CH2Cl2. Intermediates 14 and 15 were isolated. Intermedi-
ate 16 is proposed.
pyrrolidine would take place spontaneously or be promoted
by Brønsted acids. Following the formation of the copper(I)
propargylamine complex 15 from the plausible key copper
1
intermediate 16 (Scheme 6; see Figures S51 and S52 for H
and 31P NMR spectra of 16), propargylamine 8 would be
released from the coordination sphere of the metal by
replacement with free pyrrolidine present in reaction
medium. According to this mechanistic study, the major
difference between CuI and AuI catalysis is the lack of
formation of a s,p-dicopper–phenylacetylene adduct analo-
gous to the [(s,p)(LAu)2(m-phenylacetylene)][SbF6] complex
(12; see Scheme S1 in the Supporting Information).
Scheme 5. Top: Synthesis of the cationic copper(I) complex
[LCu(pyrrolidine)][PF6] (14) in toluene. Middle: Synthesis of the cation-
ic copper(I) [LCu(propargylamine)][PF6] complex 15 as the final Cu
complex upon mixing stoichiometric amounts of 3, pyrrolidine, and
phenylacetylene. Bottom: ORTEP views of the cationic CuI complexes
14 and 15 (X-ray single-crystal data are given in Tables S8 and S9 and
Figures S39 and S45).
In conclusion, we have shown herein that AuI and CuI
complexes can react with hydrogen chloride generated in situ
by the condensation of CH2Cl2 and pyrrolidine to form AuI
and CuI chloride complexes of low catalytic activity. There-
fore, this solvent is inadequate at least for Mannich reactions.
Furthermore, we found that CuI complexes are considerably
more active than analogous AuI complexes for A3 coupling.
The reaction mechanism also presents remarkable differences
between AuI and CuI catalysis owing to the s and p alkyne
nucleophilicity of gold, in contrast with the preference of
copper for copper(I)–nitrogen coordination. Deuterium
labeling together with the isolation and characterization of
CuI and AuI intermediates has shown high potential to shed
light on the identity and activity of CuI and AuI complexes
formed during the catalytic Mannich reaction.
evaporation at room temperature of a filtered solution of 15
in toluene (see Table S9 and Figure S45). ORTEP structures
of the CuI complexes 14 and 15 are shown in Scheme 5.
Importantly, the catalytic activity of the cationic copper(I)
complexes 14 and 15 for the Mannich A3 coupling (Table 1,
entries 9 and 10) was as high as that of the initial CuI
precatalyst complexes 3 and 5 (Table 1, entries 1 and 2).
According to these experimental results, we propose
a reaction mechanism for the A3 coupling with [Cu(L)-
(NCMe)][PF6] (3) or [Cu(L)(NH2Ph)][PF6] (5) in the absence
of CH2Cl2 (Scheme 6). The main features of this mechanism
are the coordination of CuI precatalysts 3 or 5 with pyrrolidine
to generate the isolated intermediate [Cu(L)(pyrrolidine)]-
[PF6] (14) and the presumed formation of the p-complex
intermediate 16 between copper complex 14 and phenyl-
acetylene. The condensation between formaldehyde and
Experimental Section
See the Supporting Information for experimental details of the
preparation, isolation, and full characterization of the new copper and
gold complexes 3, 5, 11, 13, 14, and 15, as well as propargylamines 8
and 8D2, 8H+ ClÀ, and gold complexes 9, 10, and 12, including NMR,
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
ESIMS, and GC–MS spectra, combustion analysis, and X-ray
crystallography data.
[6] a) S. B. Park, H. Alper, Chem. Commun. 2005, 1315 – 1317;
b) B. R. Buckley, A. N. Khan, H. Heaney, Chem. Eur. J. 2012, 18,
3855 – 3858; c) V. K.-Y. Lo, Y. Liu, M.-K. Wong, C.-M. Che, Org.
Lett. 2006, 8, 1529 – 1532.
[7] S. F. Rach, F. E. Kuhn, Chem. Rev. 2009, 109, 2061 – 2080.
[8] P. Pꢂrez-Galµn, N. Delpont, E. Herrero-Gómez, F. Maseras,
A. M. Echavarren, Chem. Eur. J. 2010, 16, 5324 – 5332.
[9] A. Grirrane, H. Garcia, A. Corma, E. lvarez, Chem. Eur. J.
2013, 19, 12239 – 12244.
[10] E. Herrero-Gómez, C. Nieto-Oberhuber, S. López, J. Benet-
Buchholz, A. M. Echavarren, Angew. Chem. 2006, 118, 5581 –
5585; Angew. Chem. Int. Ed. 2006, 45, 5455 – 5459.
[11] D.-Y. Yu, Y.-G. Zhang, Adv. Synth. Catal. 2011, 353, 163 – 169.
[12] T. Pang, H. Lu, T. Pang, Acta Crystallogr. Sect. E 2009, 65, o2806.
[13] A. Homs, I. Escofet, A. M. Echavarren, Org. Lett. 2013, 15,
5782 – 5785.
X-ray crystallographic data for
3
(CCDC 986314),
5
(CCDC 986315), 8H+ ClÀ (CCDC 986316), 9 (CCDC 986317), 10
(CCDC 986318), 12 (CCDC 986319), 13 (CCDC 986320), 14
(CCDC 986321), and 15 (CCDC 986322) can also be obtained free
of charge from The Cambridge Crystallographic Data Centre via
http//www.ccdc.cam.ac.uk/data request/cif.
Reactions were carried out in a reinforced glass reactor (5 mL
capacity) with a conical bottom. The reactor was equipped with
a manometer to monitor the pressure and an inlet/outlet valve to
enable samples to be taken during the course of the reaction.
Received: April 3, 2014
Published online: && &&, &&&&
[14] A. Grirrane, H. Garcia, A. Corma, E. lvarez, ACS Catal. 2011,
1, 1647 – 1653.
Keywords: coordination chemistry · copper intermediates ·
gold intermediates · homogeneous catalysis · propargylamines
.
[15] a) J. E. Mills, C. A. Maryanoff, R. M. Cosgrove, L. Scott, D. F.
McComsey, Org. Prep. Proced. Int. 1984, 16, 97 – 114; b) H. J.
Federsel, E. Koenberg, L. Lilljequist, B. M. Swahn, J. Org. Chem.
1990, 55, 2254 – 2256; c) J. E. Mills, C. A. Maryanoff, D. F.
McComsey, R. C. Stanzione, L. Scott, J. Org. Chem. 1987, 52,
1857 – 1859; d) J. Gao, Q.-W. Song, L.-N. He, Z.-Z. Yang, X.-Y.
Dou, Chem. Commun. 2012, 48, 2024 – 2026; e) Y. Zhang, D. Yu,
World Patent WO2012005692A1, Agency for Science Technol-
ogy and Research, Singapore, 2012, p. 19.
[1] a) A. S. K. Hashmi, Chem. Rev. 2007, 107, 3180 – 3211; b) D. J.
Gorin, B. D. Sherry, F. D. Toste, Chem. Rev. 2008, 108, 3351 –
3378; c) A. Fꢁrstner, Chem. Soc. Rev. 2009, 38, 3208 – 3221; d) A.
Corma, A. Leyva-Pꢂrez, M. J. Sabater, Chem. Rev. 2011, 111,
1657 – 1712; e) T. C. Boorman, I. Larrosa, Chem. Soc. Rev. 2011,
40, 1910 – 1925; f) A. S. K. Hashmi, M. Buehrle, Aldrichimica
Acta 2010, 43, 27 – 33; g) E. Jimꢂnez-NfflÇez, A. M. Echavarren,
Chem. Rev. 2008, 108, 3326 – 3350.
[16] R. Pattacini, S. Jie, P. Braunstein, Chem. Commun. 2009, 890 –
892.
[2] a) N. Yoshikai, E. Nakamura, Chem. Rev. 2012, 112, 2339 – 2372;
b) D. S. Surry, S. L. Buchwald, Chem. Sci. 2010, 1, 13 – 31; c) S. V.
Ley, A. W. Thomas, Angew. Chem. 2003, 115, 5558 – 5607;
Angew. Chem. Int. Ed. 2003, 42, 5400 – 5449; d) A. Casitas, X.
Ribas, Chem. Sci. 2013, 4, 2301 – 2318.
[3] D. S. Surry, S. L. Buchwald, Angew. Chem. 2008, 120, 6438 –
6461; Angew. Chem. Int. Ed. 2008, 47, 6338 – 6361.
[4] L. W. Bieber, M. F. Da Silva, Tetrahedron Lett. 2004, 45, 8281 –
8283.
[17] J. J. Brunet, X. Couillens, J. C. Daran, O. Diallo, C. Lepetit, D.
Neibecker, Eur. J. Inorg. Chem. 1998, 349 – 353.
[18] T.-H. Yan, C.-C. Tsai, C.-T. Chien, C.-C. Cho, P.-C. Huang, Org.
Lett. 2004, 6, 4961 – 4963.
[19] a) V. Michelet, P. Y. Toullec, J.-P. GenÞt, Angew. Chem. 2008,
120, 4338 – 4386; Angew. Chem. Int. Ed. 2008, 47, 4268 – 4315;
b) M. Rudolph, A. S. K. Hashmi, Chem. Soc. Rev. 2012, 41,
2448 – 2462.
[5] W.-J. Yoo, L. Zhao, C.-J. Li, Aldrichimica Acta 2011, 44, 43 – 51.
6
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Communications
Homogeneous Catalysis
A. Grirrane,* E. ꢁlvarez, H. Garcꢂa,*
A. Corma*
&&&& — &&&&
Deactivation of Cationic CuI and AuI
Catalysts for A3 Coupling by CH2Cl2:
Mechanistic Implications of the
Formation of Neutral CuI and AuI
Chlorides
Proceed with caution! Care should be
exercised when using CH2Cl2 as a solvent
for reactions in which amines are
a reagent, since undesirable deactivation
of cationic CuI and AuI catalysts to form
the corresponding inactive neutral chlo-
ride complexes [LMCl] (M=Cu or Au)
can occur as a result of the generation of
hydrogen chloride in the medium. This
phenomenon was studied on the basis of
a Mannich three-component coupling
reaction (see scheme).
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7
These are not the final page numbers!