Copper(I)-Catalyzed Ketone, Amine, and Alkyne Coupling for the Synthesis of 2-Alkynylpyrrolidines and -piperidines

Article abstract of DOI:10.1021/acs.orglett.6b02127

A Cu(I)-catalyzed coupling of a ω-chloro ketone, a primary amine, and an alkyne is described. This protocol allows for the synthesis of α-quaternary carbons in 2-alkynyl-substituted N-heterocycles. The key step is the in situ generation of a cyclic ketiminium species, which has enhanced reactivity for alkynylation compared to acyclic ketiminium species.

Full text of DOI:10.1021/acs.orglett.6b02127

Letter
pubs.acs.org/OrgLett
Copper(I)-Catalyzed Ketone, Amine, and Alkyne Coupling for the
Synthesis of 2‑Alkynylpyrrolidines and -piperidines
Wim E. Van Beek, Joren Van Stappen, Philippe Franck, and Kourosch Abbaspour Tehrani*
Organic Synthesis, Department of Chemistry, University of Antwerp, Groenenborgerlaan 171, B-2020 Antwerp, Belgium
S
* Supporting Information
ABSTRACT: A Cu(I)-catalyzed coupling of a ω-chloro ketone,
a primary amine, and an alkyne is described. This protocol
allows for the synthesis of α-quaternary carbons in 2-alkynyl-
substituted N-heterocycles. The key step is the in situ generation
of a cyclic ketiminium species, which has enhanced reactivity for
alkynylation compared to acyclic ketiminium species.
ropargyl amines are an interesting class of molecules. This is
underlined by the overwhelming attention that the reaction
Scheme 1. Overview of KA2 reactions until present
P
providing these molecules, the A3 reaction, has received in the
past decades.1a Other traditional methods for the synthesis of
propargyl amines include amination of propargylic triflates,
propargylic phosphates, propargylic esters, oxyphosphonium
salts, and propargylic halides.1b−f Not only does the propargyl
amine moiety appear in natural products and therapeutics such as
rasagiline,2a selegiline,2b and efavirenz,2c but propargyl amines
also serve as useful intermediates in the synthesis of allyl amines,
pyrroles, pyrrolidines, pyrazoles,3a oxazoles,3b 2-imidazolones,3c
and so on. Ever since the first report of the A3 reaction by Dyatkin
et al. in 1998,4 this reaction has been well established, and only
few challenges are left. One of these challenges is the substitution
of the aldehyde component by a less reactive ketone. This leads
to propargylamines bearing quaternary carbon centers, and a first
synthesis was reported by Van der Eycken et al. in 2010, who
abbreviated this coupling as a KA2 coupling (ketone, amine,
alkyne).5 With CuI as a catalyst and under microwave irradiation
aryl/alkyl alkynes, benzylic amines and cyclohexanone deriva-
tives were coupled efficiently (Scheme 1, A). Acyclic ketones
could not be used due to their lower reactivity compared to
cyclohexanone. In 2011, Chan et al. proved that secondary
amines can also be coupled using a AuBr3 catalyst (Scheme 1,
B).6 Next to cyclic ketones, few acyclic aliphatic ketones could
also be coupled, but aromatic ketones gave no KA2 reaction.
Later, in 2012, Larsen et al. established a dual catalytic system
using 50 mol % of Ti(OEt)4 and 5 mol % of CuCl2 (Scheme 1,
C).7 With this system, acyclic aliphatic ketones could be coupled
with primary/secondary amines and alkynes. Aromatic ketones,
however, remained unreactive. In an attempt to expand the scope
of the KA2 coupling, Ma et al. found a new system consisting of
CuBr and 4 Å MS for the synthesis of propargyl amines from
acyclic and aromatic ketones (Scheme 1, D).8 However, the
scope is limited to the use of secondary amines. In 2016, Ma et al.
published new reaction conditions consisting of a CuBr2/sodium
ascorbate and Ti(OEt)4 as the catalytic system and used this
combination for the coupling of a broad scope of aromatic
ketones (Scheme 1, E).9 Again, primary amines cannot be used in
this coupling.
From these reported methods, it can be concluded that Cu(I)
and Cu(II) salts are the preferred catalysts for the KA2 coupling.
Not surprisingly, efforts were made to synthesize Cu-based
heterogeneous catalysts, as these catalysts are easy to recover and
can be used multiple times. Some examples of such catalysts, used
for simple KA2 couplings of cyclic ketones and secondary amines,
are, for example, Cu2O on TiO2 nanoparticles,10a Cu2O on nano-
ZnO particles,10b a Cu2O on nano-CuFe2O4 system which is
magnetically recoverable,10c a Cu(II)−hydromagnesite nano-
material,10d which was also used in the first decarboxylative KA2
coupling, and a polystyrene-supported N-phenylpiperazine−
Cu(II) complex.10e
KA2 couplings generally work better with secondary amines,
generating in situ ketiminium species A. Since we would like to
expand the scope by using primary amines and to make an entry
toward interesting alkynyl-substituted N-heterocycles D, it was
Received: July 19, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b02127
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Table 1. Evaluation of Different Catalysts for the KA2
Coupling of 5-Chloropentan-2-one, n-Propylamine, And
Phenylacetylene
reasoned that adding a ω-halo atom to the ketone might result in
a cyclic ketiminium species C by substitution of the ω-halo atom
by the in situ generated imine B (Scheme 2). This is supported by
previous work11 from our group that shows that ω-chlorinated
aldimines upon standing become solid due to the formation of
iminium chlorides.
Scheme 2. New Approach toward 2-Alkynyl N-Heterocycles
a
b
entry
catalyst (mol %)
AuPPh3Cl (20)
yield of 4a (%)
1
18
46
77
87
83
87
50
65
99
42
80
2
AgOTf (20)
3
CuCl (20)
4
CuCl2 (20)
5
Cu(OTf)2 (20)
Cu2O (10)
6
7
CuO (20)
8
Cu2O (5)
9
Cu2O (25)
10
11
CuFe2O4 (20)
copper(I) phenylacetylide (20)
a
All reactions were carried out with 1 mmol of 5-chloropentan-2-one,
3.5 mmol of n-propylamine, 1.2 mmol of phenylacetylene, catalyst, and
b
1
2 mL of acetonitrile. Determined from the H NMR spectrum with
1,3,5-trimethoxybenzene as internal standard.
(42%, Table 1, entry 10), and is comparable to CuO (50%, Table
1, entry 7). This yield could probably be improved if Cu2O is
impregnated on the CuFe2O4 nanoparticles as reported
previously.10c When the presumable actual catalyst was used,
i.e., copper phenylacetylide instead of a precatalyst, the yield was
slightly lower (80%, Table 1, entry 11) probably due to difficult
breaking of the copper phenylacetylide polymeric structure.14
The advantage of using copper phenylacetylide is that for
largescale applications this catalyst can easily be recovered by
filtration after basic workup of the reaction. For screening
purposes, however, this catalyst will not be used as the catalyst
should be prepared separately for every other alkyne.
The reaction was then further optimized with regard to the
solvent. Solvents such as EtOAc, THF, 2-MeTHF, TBME,
DMSO, DMF, DCM, EtOH, toluene, and dioxane were also well
tolerated, giving yields ranging from 56 to 99% (see Table S2).
Water as a solvent still gave a yield of 12%, most likely because of
a disfavored ketone−iminium equilibrium, as water is a
byproduct of the reaction. This observation allows the use of
nondried solvents. Finally, the optimized reaction conditions for
the formation of 2-methyl-2-(phenylethynyl)pyrrolidines 4 were
established as follows: 1 equiv of 5-chloropentan-2-one, 3.5 equiv
of n-propylamine, 1.2 equiv of phenylacetylene, and 0.2 equiv of
Cu2O in acetonitrile for 24 h at 60 °C.
To investigate the scope of the reaction, first the primary
amine component was varied. With n-propylamine, an isolated
yield of 91% was obtained for the product 4a, while the more
sterically hindered isopropylamine also was tolerated and gave a
yield of 87%. The reaction of 5-chloropentan-2-one with tert-
butylamine did not yield any reaction product 4c at 60 °C. By
raising the temperature to 90 °C, product 4c was isolated in 29%
yield. This lower yield can be explained by the difficult formation
of the ketimine and ring-closing reaction11 and by the low boiling
point of t-BuNH2 (46 °C). Amines, with readily removable
protecting groups such as allylamine or benzylamine, furnished
yields of 77% and 75%, respectively. Molecule 4f shows that the
primary amine reacts exclusively, even in the presence of a
The 2-alkynyl N-heterocycles D have been synthesized
previously via an intramolecular hydroamination of aminoalkyl
acetylenes followed by alkynylation.12a,b However, these
approaches require the prior synthesis of the aminoalkynes,
involving a two-step synthesis from either alkynyl alcohols or
from alkynoic acids. In this work, we report a straightforward
entry to 2-alkynyl-substituted N-heterocycles, starting from
commercially available starting materials via a one-pot, three-
component synthesis.
As a starting point for the optimization of the reaction under
investigation commercially available 5-chloropentan-2-one (1),
n-propylamine (2) and phenylacetylene (3) were reacted in the
presence of various metal catalysts in acetonitrile to give 2-
methyl-2-(phenylethynyl)-1-propylpyrrolidine (4a). An excess
of 3.5 equiv of n-propylamine was used, as n-propylamine also
acts as a base to trap the equivalent of HCl that is formed and
because of its volatility. A small excess of 1.2 equiv of
phenylacetylene was used because in all cases traces of the
Glaser homocoupling product were observed. Classical A3
catalysts such as Zn, Fe, Al, Ni, and Sc salts were not able to
catalyze this transformation (see Table S1). Remarkably, even
In(OTf)3, which worked well for chlorinated imines, was not able
to catalyze this reaction.11,13 AuPPh3Cl (Table 1, entry 1) and
AgOTf (Table 1, entry 2) gave low yields of 18 and 46% of 2-
methyl-2-(phenylethynyl)-1-propylpyrrolidine. CuCl (77%,
Table 1, entry 3) gave a lower yield than CuCl2 (87%, Table 1,
entry 4), while when the counteranion was changed to triflate
(83%, Table 1, entry 5) no improvement of the yield was
observed. Changing to Cu2O (87%, Table 1, entry 6) was equally
good, while CuO showed less catalytic activity (50%, Table 1,
entry 7). When the catalyst loading was lowered to 5 mol %, only
65% yield of 2-methyl-2-(phenylethynyl)-1-propylpyrrolidine
(4a) was obtained (Table 1, entry 8), while by increasing the
catalyst loading to 25 mol % the yield became nearly quantitative
(Table 1, entry 9). The coupling reaction proved also feasible
with magnetically recoverable CuFe2O4, albeit in lower yield
B
DOI: 10.1021/acs.orglett.6b02127
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 3. Scope of the Evaluated KA2 Reaction
secondary amine and hydroxy function in a reasonable 75% yield.
(S)-Methylbenzylamine afforded a 7/3 mixture of diastereomers
in 83% yield. It is known that α-amino esters are also good
coupling partners for KA2 couplings.15 Therefore, glycine ethyl
ester was used as amine component, and the product 4h was
obtained in 42% yield. By using the more stable glycine ethyl
ester hydrochloride in the presence of 2 equiv of triethylamine,
the yield increased to 81%. Other α-aminoesters, derived from
natural α-amino acids such as L-valine methyl ester hydrochloride
and L-phenylalanine t-Bu ester hydrochloride, gave under the
same reaction conditions (2 equiv of NEt3) 48% of a 4/1
diastereomeric mixture and a 63% yield of a 7/3 diastereomeric
mixture, respectively.
Next, different alkynes were evaluated. In general, alkyl-
substituted alkynes reacted well in this conversion, resulting in
yields of 40−64%. 1-Pentyne as a coupling partner resulted in a
low yield of 44%, probably due to its low boiling point (40 °C).
When 1-hexyne with a boiling point of 71−72 °C was used, the
yield increased to 55%. When 4-chlorobut-1-yne was used,
product 4m was isolated in 53% yield. Trace amounts of 2-(but-
3-en-1-yn-1-yl)-2-methyl-1-propylpyrrolidine and 4-(2-methyl-
1-propylpyrrolidin-2-yl)-N-propylbut-3-yn-1-amine were ob-
served, resulting from the conjugated elimination of HCl and
the substitution of chlorine by n-propylamine in 2-(4-chlorobut-
1-yn-1-yl)-2-methyl-1-propylpyrrolidine (4m). The use of cyclo-
hexylacetylene results in the formation of product 4n in 64%.
Next, a range of aromatic alkynes were evaluated. Electron-
donating substituents (4-t-Bu, 2-Me, 4-Cl, 2-MeO, 3-MeO, 4-
MeO) were tolerated well, resulting in yields of 60−98%. 4-
Ethynyl-N,N-dimethylaniline leads to a very high yield of 95%.
The reaction of 1 with 2-ethynylaniline also led to the expected
alkynylpyrrolidine 4v in 86% yield. Under these reaction
conditions, no additional intramolecular hydroamination,
leading to a 2-indolopyrrolidine as previously reported under
comparable conditions, was observed.16 Next to electron-
donating groups, electron-withdrawing groups like 4-nitro and
4-trifluoromethyl gave reasonable yields of 65 and 70%. In a
special case, the 2-ethynylpyridine could also be coupled, albeit in
a moderate yield of 45%.
Next, different ketones were screened. As stated above,
aromatic ketones are difficult coupling partners in KA2 reactions.
Under our conditions, the commercially available 4-chlorobutyr-
ophenone could be easily coupled with n-propylamine and
phenylacetylene leading to 2-phenyl-2-(phenylethynyl)-1-pro-
pylpyrrolidine (4z) in 63% yield. In addition, substituted 4-
chlorobutyrophenones could be coupled, and the corresponding
pyrrolidines were isolated in 52−86% yield. In general, the yields
are lower than for the aliphatic ketones, probably because of
increased steric hindrance in aromatic ketiminium intermediates
(Scheme 3). When 1,7-dichloroheptan-4-one was used, solely
pyrrolidine 4ae, which contains a cyclopropyl ring, was obtained
in toluene. Besides the commercially available ketones tested
above, a wide variety of chloroketones can be accessed by a
plethora of literature methods.17 For example, bicyclic product
4af, containing an ester moiety, was prepared from ethyl 1-(2-
chloroethyl)-2-oxocyclohexanecarboxylate. Lastly, the formation
of other nitrogen heterocycles was evaluated. To our delight, the
formation of piperidine 4ag from the commercially available 6-
chlorohexan-2-one was possible in 88% yield. The formation of
4ah from 5-chloro-1-phenylpentan-1-one did not work at 60 °C
and required a higher temperature of 90 °C, leading to a
comparable yield of 88%. In addition, product 4ai, which
contains a Bn group, was obtained at elevated temperature (90
a
Unless otherwise stated, 3.5 equiv of amine was used. 10 mmol scale.
b
c
d
90 °C. 2 equiv of amine was used. 1 equiv of amine was used; 1
e
equiv of NEt3 was added. 1 equiv of aminoester·HCl was used, and 2
equiv of NEt3 was added. Prepared from 1,7-dichloroheptan-4-one in
toluene. Prepared from ethyl 1-(2-chloroethyl)-2-oxocyclohexanecar-
boxylate.
f
g
°C) in 61% yield. Unfortunately, the reaction of 4-chlorobutan-2-
one or 7-chloroheptan-2-one with phenylacetylene and Cu2O
under the developed conditions did not lead to any formation of
the corresponding 2-alkynylazetidines and -azepanes. In order to
show the robustness of the developed method and to prove that
the reaction could be scaled up, a gram-scale experiment was
conducted. On a 10 mmol scale, the product 4a could be isolated
in 78% yield.
To gain insight into the reaction mechanism, as proposed in
Scheme 2, a series of control experiments was set up. First, the
reaction of 5-chloropentan-2-one with dipropylamine and
phenylacetylene was evaluated. At 85 °C, a complex mixture of
substitution product 5-(dipropylamino)pentan-2-one (5), the
hydroamination product N-(1-phenylvinyl)-N-propylpropan-1-
amine (6), and the subsequent alkynylation product 2,4-
diphenyl-N,N-dipropylbut-3-yn-2-amine (7) could be distin-
1
guished in the H NMR spectrum of the reaction mixture
(Scheme 4, route a). When the nonchlorinated 2-pentanone was
C
DOI: 10.1021/acs.orglett.6b02127
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Optimization data, experimental procedures, character-
ization of new compounds, and spectral data (PDF)
Scheme 4. Control Experiments
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: kourosch.abbaspourtehrani@uantwerpen.be.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was financed by the Agency for Innovation by Science
and Technology (IWT-Flanders), the University of Antwerp
(BOF), and the Hercules Foundation. We thank Heidi Seykens
(Organic Synthesis, University of Antwerp) for HRMS measure-
ments.
REFERENCES
■
(1) (a) Peshkov, V. A.; Pereshivko, O. P.; Van der Eycken, E. V. Chem.
Soc. Rev. 2012, 41, 3790. (b) Czernecki, S.; Valer
́
y, J.-M. J. Carbohydr.
Chem. 1990, 9, 767. (c) Imada, Y.; Yuasa, M.; Nakamura, I.; Murahashi,
S.-I. J. Org. Chem. 1994, 59, 2282. (d) Castro, B.; Selve, C. Bull. Soc.
Chim. Fr. 1971, 4368. (e) Hennion, G. F.; Hanzel, R. S. J. Am. Chem. Soc.
1960, 82, 4908. (f) Kopka, I. E.; Fataftah, A.; Rathke, M. W. J. Org. Chem.
1980, 45, 4616.
(2) (a) Yu, P. H.; Davis, B. A.; Boulton, A. A. J. Med. Chem. 1992, 35,
3705. (b) Tsai, S.-C.; Yuan, R.-Y. J. Exp. Clin. Med. 2012, 4, 209.
(c) Jiang, B.; Si, Y. G. Angew. Chem., Int. Ed. 2004, 43, 216.
(3) (a) Chen, J.; Properzi, R.; Uccello, D. P.; Young, J. A.; Dushin, R.
G.; Starr, J. T. Org. Lett. 2014, 16, 4146. (b) Nilsson, B. M.; Hacksell, U.
J. Heterocycl. Chem. 1989, 26, 269. (c) Ranjan, A.; Yerande, R.;
Wakchaure, P. B.; Yerande, S. G.; Dethe, D. H. Org. Lett. 2014, 16, 5788.
(4) Dyatkin, A. B.; Rivero, R. A. Tetrahedron Lett. 1998, 39, 3647.
(5) Pereshivko, O. P.; Peshkov, V. A.; Van der Eycken, E. V. Org. Lett.
2010, 12, 2638.
(6) Cheng, M.; Zhang, Q.; Hu, X.-Y.; Li, B.-G.; Ji, J.-X.; Chan, A. S. C.
Adv. Synth. Catal. 2011, 353, 1274.
(7) Pierce, C. J.; Nguyen, M.; Larsen, C. H. Angew. Chem., Int. Ed. 2012,
51, 12289.
(8) Tang, X.; Kuang, J.; Ma, S. Chem. Commun. 2013, 49, 8976.
(9) Cai, Y.; Tang, X.; Ma, S. Chem. - Eur. J. 2016, 22, 2266.
(10) (a) Albaladejo, M. J.; Alonso, F.; Moglie, Y.; Yus, M. Eur. J. Org.
Chem. 2012, 3093. (b) Hosseini-Sarvari, M.; Moeini, F. New J. Chem.
2014, 38, 624. (c) Nemati, F.; Elhampour, A.; Farrokhi, H.; Bagheri
Natanzi, M. Catal. Commun. 2015, 66, 15. (d) Chinna Rajesh, U.; Gulati,
U.; Rawat, D. S. ACS Sustainable Chem. Eng. 2016, 4, 3409.
(e) Perumgani, P. C.; Keesara, S.; Parvathaneni, S.; Mandapati, R. M.
New J. Chem. 2016, 40, 5113.
reacted with n-propylamine and phenylacetylene, besides some
imine 8, no alkynylated product was formed (Scheme 4, route b).
When dipropylamine was used instead of n-propylamine in
combination with nonchlorinated ketone, hydroamination of
phenylacetylene, followed by alkynylation of the intermediate
formed iminium occurred, as described by Larsen et al. (Scheme
4, route c).18a,b However, no incorporation of the ketone
component was seen. This proves that both the primary amine
and the presence of the ω-chlorine atom on the ketone are
necessary for the coupling reaction. Next, ω-chloroimine 10 was
synthesized19 and reacted with phenylacetylene at 90 °C in the
presence of 0.25 equiv of Cu2O and 1 equiv of NEt3. The
corresponding 2-alkynylpiperidine 4aj was isolated in 33% yield,
proving that the proposed reaction mechanism in Scheme 2 is
plausible. Alkynylation reactions of enamines have been reported
to proceed via the corresponding iminium compounds, thus
supporting the proposed reaction mechanism.20 Preliminary
alkylation of the primary amine by the alkyl chloride moiety of 1,
followed by intramolecular iminium formation is more plausible,
since 5-aminopentan-2-one 11 (as its ammonium salt), which
was prepared independently (see the SI) upon reaction with
phenylacetylene, 1 equiv of propylamine and 0.25 equiv of Cu2O
furnished the alkynylpiperidine 4a in 70% yield already at rt after
2 h.
In conclusion, a new, mild Cu(I)-catalyzed synthesis of 2-
alkynylpyrrolidines and -piperidines bearing a quaternary α-
carbon center was reported. Different primary aliphatic amines
can be used in combination with ω-chlorinated ketones. A broad
range of functional groups (Cl, MeO, CF3, NO2, NH2, NMe2) is
well tolerated on the alkyne moiety. Different substituted
aromatic ketones, which traditionally are difficult reaction
partners, could be coupled. Moreover, this synthesis uses simple,
cheap, and commercially available starting materials. The key
step is the in situ generation of a cyclic ketiminium species, which
has an enhanced reactivity toward alkynylation in comparison to
acyclic ketiminium species.
(11) Kushwaha, K.; Malakar, C. C.; Stas, S.; Lemier
Tehrani, K. RSC Adv. 2015, 5, 10139.
̀
e, F.; Abbaspour
(12) (a) Han, J.; Xu, B.; Hammond, G. B. J. Am. Chem. Soc. 2010, 132,
916. (b) Liu, X. Y.; Che, C. M. Angew. Chem., Int. Ed. 2008, 47, 3805.
(13) Malakar, C. C.; Maes, B. U. W.; Abbaspour Tehrani, K. Adv. Synth.
Catal. 2012, 354, 3461.
(14) Buckley, B. R.; Khan, A. N.; Heaney, H. Chem. - Eur. J. 2012, 18,
3855.
(15) Sharma, N.; Sharma, U. K.; Mishra, N. M.; Van der Eycken, E. V.
Adv. Synth. Catal. 2014, 356, 1029.
(16) Takada, H.; Kumagai, N.; Shibasaki, M. Org. Lett. 2015, 17, 4762.
(17) Roman, B. I.; De Kimpe, N.; Stevens, C. V. Chem. Rev. 2010, 110,
5914.
(18) (a) Palchak, Z. L.; Lussier, D. J.; Pierce, C. J.; Yoo, H.; Larsen, C.
H. Adv. Synth. Catal. 2015, 357, 539. (b) Nelson, K.; Larsen, C. H.
Synlett 2014, 25, 2681.
(19) Sulmon, P.; De Kimpe, N.; Schamp, N. Synthesis 1989, 8.
(20) Koradin, C.; Polborn, K.; Knochel, P. Angew. Chem., Int. Ed. 2002,
41, 2535.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.6b02127.
D
DOI: 10.1021/acs.orglett.6b02127
Org. Lett. XXXX, XXX, XXX−XXX