A Four-Step Synthesis of (±)-γ-Lycorane via Pd0-Catalyzed Double C(sp2)-H/C(sp3)-H Arylation

Article abstract of DOI:10.1021/acs.orglett.7b03909

An expedient synthesis of lycorine alkaloids is reported using a palladium(0)-catalyzed double C-X/C-H arylation as the key step. The selectivity of this reaction was controlled through the judicious choice of the two halogen atoms, and its generality was

Full text of DOI:10.1021/acs.orglett.7b03909

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
A Four-Step Synthesis of ( )-γ-Lycorane via Pd0‑Catalyzed Double
C(sp2)−H/C(sp3)−H Arylation
Ronan Rocaboy, David Dailler, and Olivier Baudoin*
University of Basel, Department of Chemistry, St. Johanns-Ring 19, CH-4056 Basel, Switzerland
S
* Supporting Information
ABSTRACT: An expedient synthesis of lycorine alkaloids is
reported using a palladium(0)-catalyzed double C−X/C−H
arylation as the key step. The selectivity of this reaction was
controlled through the judicious choice of the two halogen
atoms, and its generality was demonstrated through the
construction of various substituted pyrrolophenanthridinones.
A selective arene hydrogenation allowed for the completion of the synthesis of ( )-γ-lycorane in just four steps from
commercially available precursors.
arise from this strategy: (1) the generation of the strained B−
ycorines constitute an abundant subclass of fused
Lpolycyclic Amaryllidaceae alkaloids which possess diverse C−D ring system, potentially affecting the efficiency of the
interesting biological properties1 and have been the subject of
second C−H arylation step; (2) the site selectivity of the C−H
numerous synthetic studies.2 As shown with the structure of
activation steps, potentially leading to isomeric mixtures. These
issues notwithstanding, compounds 6 would be readily
accessible from commercially available aniline and benzyl
bromide precursors, thereby providing a straightforward and
scalable access to lycorines. Herein, we report the development
of this domino C−H arylation strategy and its application to
the short and efficient synthesis of lycorine alkaloids.
hippadine (1a), pratosine (1b), assoanine (2), lycorine (4), and
its degradation product γ-lycorane (3),3 the B, C, and D rings
of lycorines occur in diverse oxidation states (Figure 1).
At the onset of this work, we looked for the most appropriate
bis-halogenated precursor to obtain compound 5a selectively
and efficiently (Scheme 1, Table 1). The double C−X/C−H
arylation was first conducted with compound 6a bearing the
two bromine atoms on the same ring. Tricyclohexylphosphine
was chosen as the ligand, because it was previously employed in
both individual C(sp2)−H8d and C(sp3)−H9b arylations. The
well-defined Pd(PCy3)2 complex, which was found to provide
superior yields to in situ generated catalysts in previous strain-
generating C(sp3)−H activation reactions,10 was employed as
the catalyst, and catalytic PivOK/K2CO3, as the basic system.11
Under these conditions, the double C−Br/C−H arylation took
place, but isomer 5b, arising from the electronically favored
activation of the C(sp2)−Hb bond (A → B) instead of the more
sterically accessible C(sp2)−Ha bond (Scheme 1 top), was
isolated as the sole C−H arylation product, consistent with
initial observations from Harayama and co-workers (entry 1).12
To solve this site-selectivity issue, we examined the reaction of
isomeric dibromide 6b, bearing bromine atoms on rings A and
C (entry 2, Scheme 1 bottom).
Figure 1. Examples of lycorine alkaloids and retrosynthetic analysis.
As a consequence, fused biaryls 5 would constitute appealing
retrons allowing access to a variety of lycorine congeners in a
divergent fashion.4 Based on our previous work,5 we envisioned
disconnecting compounds 5 simultaneously at bonds a and b
by means of a Pd0-catalyzed double C−X/C−H arylation,6,7
hence leading to bis-halogenated precursors 6, wherein the
halogen atoms are located either on ring A and C or both on
ring C. The corresponding individual C(sp2)−X/C(sp2)−H
and C(sp2)−X/C(sp3)−H arylations are known,8,9 but to the
best of our knowledge they have not been combined in a
domino process. However, two main issues were expected to
Under the same conditions, the desired product 5a was
isolated, albeit in low yield and with byproducts arising from
competitive arylation at C−Hb and proto-debromination. A
simple solution was found by replacing the bromine atom on
Received: December 15, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b03909
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Mechanistic and Selectivity Considerations for the Double C−X/C−H Arylation
Table 1. Optimization of Substrate and Reaction Conditions
Scheme 2. Scope and Limitations of the Double C−X/C−H
Arylation
yield
entry reactant
conditions
product
(%)
a
1
6a
Pd(PCy3)2 (10 mol %), PivOK (30 mol %),
K2CO3 (4 equiv), mesitylene, 140 °C,
16 h
5b
60
b
b
b
2
3
4
6b
6c
6c
same as above
same as above
5a
5a
5a
22
37
53
Pd(OAc)2 (10 mol %), PCy3 (20 mol %),
PivOH (30 mol %), Cs2CO3 (2 equiv),
mesitylene, 140 °C, 16 h
b
5
6c
Pd(PCy3)2 (10 mol %), PivOH (30 mol %),
5a
90
a
Cs2CO3 (2 equiv), mesitylene, 140 °C,
16 h
(92)
a
b
Yield of the isolated product. NMR yield using trichloroethylene as
internal standard.
ring C with a chlorine atom (6c). Indeed, in this case the
oxidative addition of the C−Br bond to Pd0 should occur faster
than that of the C−Cl bond, to give intermediate C. Upon
activation of the most reactive C(sp2)−Hd bond vs the less
reactive C(sp3)−Hc bonds, ring B would be formed with the
correct regiochemistry via palladacycle D. Then C−Cl oxidative
addition, enabled by the electron-rich PCy3 ligand, would
furnish complex E, followed by activation of a C−Hc bond to
give palladacycle F, which would allow for the construction of
ring D upon reductive elimination. Accordingly, compound 6c
furnished the desired product 5a in higher yield (37%), and
importantly byproducts arising from competitive C−Hb or C−
Hc arylations were not detected (entry 3). Gratifyingly, tuning
the reaction conditions (see the Supporting Information for
details), and in particular replacing potassium with cesium
carbonate (2 equiv), allowed isolation of product 5a in
excellent (92%) yield (entry 5). Although in situ catalyst
generation such as Pd(OAc)2/PCy3 also could be employed
(entry 4), it provided a markedly lower yield than the well-
defined Pd(PCy3)2 complex. When the reaction was stopped at
incomplete conversion (15 min), the monocyclized product 5c,
arising from C(sp2)−Br/C(sp2)−H arylation, was isolated in
68% yield along with 5a, thereby validating the hypothesized
order of events.
a
From the dichloride precursor.
substituents possessing different electronic properties were well
tolerated on rings A and C (5d−l). As shown above, the
regioselectivity control should be good as far as the oxidative
addition to Pd0 occurs first on ring A. Hence, problems were
Next, we studied the scope of the double C−X/C−H
arylation (Scheme 2). We were pleased to find that a variety of
B
DOI: 10.1021/acs.orglett.7b03909
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
expected to arise upon increase of the electron density on ring
A and decrease on ring C, i.e. for substituents slowing down the
oxidative addition on the former and accelerating it on the
latter. This seemed to be the case for compounds 5i−j, for
which a lower yield was indeed obtained, although we were
unable to clearly identify isomeric products in the reaction
mixture. In all other cases good to excellent yields of the desired
polycyclic products were obtained. Substituents in α position to
the amide (R3) were also well tolerated (5m−o), consistent
with previous results on the individual C(sp3)−H arylation
reaction,9 despite the fact that a weaker base was employed in
the current work (PivOK, vs t-BuOK, or LiHMDS in previous
work).
Figure 2. DFT-optimized structure of 5a (ωB97X-D/6-31G**)
showing the HOMO (left) and selected bond angles (right).
The application of this method to the synthesis of lycorine
alkaloids is depicted in Scheme 3. The C−H arylation precursor
We examined various heterogeneous and homogeneous
catalysts and reaction conditions, and selected representative
examples are displayed in Table 2. Whereas PtO2 in AcOH was
a
Scheme 3. Synthesis of Lycorine Alkaloids
Table 2. Selective Hydrogenation of
Pyrrolophenanthridinone 5a
P(H2)
product(s)
d
c
entry
1
catalyst
PtO2 (10 mol %)
solvent
(bar)
(% yield)
AcOH
HFIP
50
50
−
e
2
Pd/C (10 mol %)
10 (41),
11 (5)
e
3
Pd/C (20 mol %)
HFIP
50
10 (55),
11 (17)
f
4
5
Rh/C (30 mol %)
Rh/C (30 mol %)
HFIP
HFIP
6
10 (66) (62)
10
10 (30),
12 (60)
a
Reagents and conditions: (a) 8 (1 equiv), NaH, then 7 (1.1 equiv),
6
7
[Rh(COD)Cl]2/CAAC·HCl/
t-BuOK (20 mol %)
TFE
TFE
50
10 (64)
THF, reflux; (b) Pd(PCy3)2 (10 mol %), PivOH (30 mol %), Cs2CO3
(2 equiv), mesitylene, 140 °C; (c) H2 (6 bar), Rh/C, HFIP, 25 °C; (d)
LiAlH4, THF, reflux.
[Rh(COD)Cl]2/SIMes·HCl/
t-BuOK (20 mol %)
50
10 (73)
a
b
Relative configuration determined by NOESY. Presumed major
c
diastereoisomer based on the configuration of 10 and 11. c = 0.06 M,
6c was obtained in one step from commercially available
substrates 7−8. The double C−H arylation described in Table
1 was scaled up to give 1.4 g of pyrrolophenanthridinone 5a
(81% yield). Then, we considered directly converting 5a to γ-
lycorane 3 through selective arene hydrogenation. We
recognized that achieving selectivity for ring C over ring A
might be particularly challenging because, as shown with the
calculated HOMO of 5a (Figure 2, left), both aromatic rings
bear similar electron density and should therefore react at
comparable rates.13 However, ring C is also more strained than
ring A thanks to the adjacent ring fusions with rings B and D, as
shown with the more distorted bond angles in ring C (Figure 2,
right). As a consequence, ring C might undergo selective
hydrogenation by virtue of strain release.14 Such a selective
arene hydrogenation is rare, but was already reported on the
reduced form of 5ai.e. amine instead of amidein the
context of γ-lycorane semisynthesis using PtO2 as the catalyst in
AcOH,3 albeit only with low yields (8−24%).
d
V = 1.8 mL. NMR yield using trichloroethylene as internal standard.
e
f
Performed at 50 °C instead of 25 °C. Yield of the isolated product.
HFIP = hexafluoroisopropanol, TFE = 2,2,2-trifluoroethanol.
found to be inactive (entry 1), Pd/C in HFIP under forcing
conditions provided 10 as the major product (entries 2−3), but
together with compound 11 arising from the competitive
hydrogenation of ring A. Of note, both hydrogenated products
10−11 were obtained as cis,cis diastereoisomers (d.r. > 95:5).
Gratifyingly, using the more active Rh/C catalyst under milder
conditions (room temperature, 6 bar H2) allowed suppression
of the undesired isomer 11 (entry 4).15 At a higher hydrogen
C
DOI: 10.1021/acs.orglett.7b03909
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
A.; Ratmanova, N. K.; Novoselov, A. M.; Belov, D. S.; Seregina, I. F.;
Kurkin, A. V. Chem. - Eur. J. 2016, 22, 7262. (e) Monaco, A.; Szulc, B.
R.; Rao, Z. X.; Barniol-Xicota, M.; Sehailia, M.; Borges, B. M. A.;
Hilton, S. T. Chem. - Eur. J. 2017, 23, 4750. (f) Doan, B. N. D.; Tan, X.
Y.; Ang, C. M.; Bates, R. W. Synthesis 2017, 49, 4711.
pressure, the completely hydrogenated product 12 was mainly
formed (entry 5). We also examined various homogeneous
catalysts and in particular Rh-carbene complexes as reported by
Zeng and co-workers.16 Similar results to Rh/C were obtained
using two different carbene precursors, but at a higher pressure
(entries 6−7).17
(4) Tsukano, C.; Muto, N.; Enkhtaivan, I.; Takemoto, Y. Chem. -
Asian J. 2014, 9, 2628.
The hydrogenation of 5a was then repeated using Rh/C as
the catalyst and 6 bar of H2 pressure on a 0.5 mmol scale, to
give γ-lycorane in 58% yield in two steps after amide reduction
with LiAlH4 (Scheme 3). Overall, ( )-γ-lycorane was obtained
in only four steps, 47% overall yield, from commercially
available precursors 7−8, which to the best of our knowledge is
the shortest and highest-yielding synthesis to date.18 In addition
to 5a, dimethoxy-substituted compound 5e was converted to γ-
lycorane analogue 9 through the same reductive sequence.
Moreover, pyrrolophenanthridinones 5a and 5e are valuable
platforms for the synthesis of more oxidized lycorine alkaloids
such as hippadine 1a and pratosine 1b, which were synthesized
as previously reported.4
In conclusion, an expedient synthesis of lycorine alkaloids
was achieved using a palladium(0)-catalyzed double C−X/C−
H arylation, the selectivity of which being controlled through
the judicious choice of the two halogen atoms. The generality
of this reaction was demonstrated through the construction of
various substituted pyrrolophenanthridinones. A selective arene
hydrogenation allowed for the completion of the synthesis of
( )-γ-lycorane in just four steps from commercially available
precursors.
(5) Pierre, C.; Baudoin, O. Org. Lett. 2011, 13, 1816.
(6) For selected examples of Pd0-catalyzed double C−X/C−H
functionalization reactions leading to polycyclic compounds: (a) Gon-
́ ́
zalez, J. J.; García, N.; Gomez-Lor, B.; Echavarren, A. M. J. Org. Chem.
1997, 62, 1286. (b) Kamikawa, K.; Takemoto, I.; Takemoto, S.;
Matsuzaka, H. J. Org. Chem. 2007, 72, 7406. (c) Ohno, H.; Iuchi, M.;
Fujii, N.; Tanaka, T. Org. Lett. 2007, 9, 4813. (d) Ohno, H.; Iuchi, M.;
Kojima, N.; Yoshimitsu, T.; Fujii, N.; Tanaka, T. Chem. - Eur. J. 2012,
18, 5352. (e) Huang, R. Y.; Franke, P. T.; Nicolaus, N.; Lautens, M.
Tetrahedron 2013, 69, 4395. (f) Pedroni, J.; Saget, T.; Donets, P. A.;
Cramer, N. Chem. Sci. 2015, 6, 5164.
(7) For reviews: (a) Ohno, H. Asian J. Org. Chem. 2013, 2, 18.
(b) Tsui, G. C.; Lautens, M. In Domino Reactions: Concepts for Efficient
Organic Synthesis; Tietze, L. F., Ed.; Wiley-VCH Verlag GmbH & Co.
KGaA.: 2014; pp 67−103.
(8) (a) Garden, S. J.; Torres, J. C.; Pinto, A. C. J. Braz. Chem. Soc.
2000, 11, 441. (b) Harayama, T.; Sato, T.; Hori, A.; Abe, H.; Takeuchi,
Y. Synlett 2003, 1141. (c) Torres, J. C.; Pinto, A. C.; Garden, S. J.
Tetrahedron 2004, 60, 9889. (d) Campeau, L.-C.; Parisien, M.; Jean,
A.; Fagnou, K. J. Am. Chem. Soc. 2006, 128, 581. (e) Umemoto, H.;
Dohshita, M.; Hamamoto, H.; Miki, Y. Heterocycles 2011, 83, 1111.
(f) Mishra, S.; De, S.; Kakde, B. N.; Dey, D.; Bisai, A. Indian J. Chem.,
Sect. A 2013, 52, 1093.
(9) (a) Shaughnessy, K. H.; Hamann, B. C.; Hartwig, J. F. J. Org.
Chem. 1998, 63, 6546. (b) Lee, S.; Hartwig, J. F. J. Org. Chem. 2001,
66, 3402. (c) Zhang, T. Y.; Zhang, H. Tetrahedron Lett. 2002, 43, 193.
(10) (a) Guyonnet, M.; Baudoin, O. Org. Lett. 2012, 14, 398.
(b) Dailler, D.; Danoun, G.; Baudoin, O. Angew. Chem., Int. Ed. 2015,
54, 4919.
(11) Lafrance, M.; Fagnou, K. J. Am. Chem. Soc. 2006, 128, 16496.
(12) (a) Harayama, T.; Toko, H.; Hori, A.; Miyagoe, T.; Sato, T.;
Nishioka, H.; Abe, H.; Takeuchi, Y. Heterocycles 2003, 61, 513.
(b) Harayama, T.; Kawata, Y.; Nagura, C.; Sato, T.; Miyagoe, T.; Abe,
H.; Takeuchi, Y. Tetrahedron Lett. 2005, 46, 6091.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.7b03909.
Full optimization tables, procedural and spectral data,
computational details (PDF)
(13) Rylander, P. N. Hydrogenation methods; Academic Press:
London, 1985; Chapter 9, pp 117−122.
AUTHOR INFORMATION
■
(14) Valls, N.; Bonjoch, J.; Bosch, J. J. Org. Chem. 1992, 57, 2508.
(15) Maegawa, T.; Akashi, A.; Yaguchi, K.; Iwasaki, Y.; Shigetsura,
M.; Monguchi, Y.; Sajiki, H. Chem. - Eur. J. 2009, 15, 6953.
(16) Wei, Y.; Rao, B.; Cong, X.; Zeng, X. J. Am. Chem. Soc. 2015, 137,
9250.
Corresponding Author
*E-mail: olivier.baudoin@unibas.ch.
ORCID
Olivier Baudoin: 0000-0002-0847-8493
Notes
(17) Of note, we failed to achieve enantioselective hydrogenation
using chiral carbenes.
(18) The previous shortest synthesis of ( )-γ-lycorane includes 6−7
steps, 30.6% overall yield.3e
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was financially supported by the University of Basel.
■
We thank E. Hormann, University of Basel, for the preparation
̈
of the Pd(PCy ) complex, Dr. D. Haussinger, University of
̈
3
2
Basel, for NMR experiments, and S. Mittelheisser and Dr. H.
Nadig, University of Basel, for MS analyses.
REFERENCES
(1) Jin, Z. Nat. Prod. Rep. 2007, 24, 886.
(2) Ghavre, M.; Froese, J.; Pour, M.; Hudlicky, T. Angew. Chem., Int.
■
Ed. 2016, 55, 5642.
(3) Degradation study: (a) Kotera, K. Tetrahedron 1961, 12, 248.
For selected recent syntheses: (b) Conner, E. S.; Crocker, K. E.;
Fernando, R. G.; Fronczek, F. R.; Stanley, G. G.; Ragains, J. R. Org.
Lett. 2013, 15, 5558. (c) Liu, D.; Ai, L.; Li, F.; Zhao, A.; Chen, J.;
Zhang, H.; Liu, J. Org. Biomol. Chem. 2014, 12, 3191. (d) Andreev, I.
D
DOI: 10.1021/acs.orglett.7b03909
Org. Lett. XXXX, XXX, XXX−XXX