Pd-catalyzed bis-cyclization/dimerization reactions of ω-aminovinyl halides

Article abstract of DOI:10.1021/ol401563z

Palladium is shown to catalyze the dimerization and cyclization of vinyl halides to generate pyrrolidine and piperidine dimers connected by a trans-ethylene bridge. The reaction tolerates a variety of N-alkyl substituents, including adamantyl. This remarkable dimerization reaction generates the skeleton of the alkaloid hyalbidone in a single step. A crossover experiment with a vinyl halide and a vinyl bromide is consistent with a Michael-type addition to a vinylpalladium cation to generate a Pd(0) alkylidene intermediate.

Full text of DOI:10.1021/ol401563z

ORGANIC
LETTERS
2013
Vol. 15, No. 14
3694–3697
Pd-Catalyzed Bis-cyclization/Dimerization
Reactions of ω‑Aminovinyl Halides
Avinash Khanna, Ilandari Dewage Udara Anulal Premachandra, Paul D. Sung, and
David L. Van Vranken*
Departmentof Chemistry, University of California at Irvine, Irvine,
California 92697-2025, United States
david.vv@uci.edu
Received June 3, 2013
ABSTRACT
Palladium is shown to catalyze the dimerization and cyclization of vinyl halides to generate pyrrolidine and piperidine dimers connected by a trans-ethylene
bridge. The reaction tolerates a variety of N-alkyl substituents, including adamantyl. This remarkable dimerization reaction generates the skeleton of the
alkaloid hyalbidone in a single step. A crossover experiment with a vinyl halide and a vinyl bromide is consistent with a Michael-type addition to a
vinylpalladium cation to generate a Pd(0) alkylidene intermediate.
We recently reported an intramolecular carbenylative
amination reaction that employed N-tosylhydrazones as
carbene precursors to generate 2-substituted pyrrolidines 2.1
Scheme 1. Dimerization as a Competing Side Reaction in
Carbenylative Amination
During optimization of the carbenylative amination, we
noted that dimerization of the vinyl halide was compet-
ing with the desired carbenylation reaction to form bis-
pyrrolidine 3 (Scheme 1). When the N-tosylhydrazone
was omitted from the reaction, the bis-pyrrolidine dimer 3
was formed in 67% yield. Intrigued by this unprecedented
dimerization, we set out to explore the scope and potential
applications of the reaction.
First we sought to identify which components of the
heavily optimizedcarbenenylative insertion werenecessary
for the dimerization reaction and to improve the yield
of the dimer (Table 1). Among the catalyst precursors
lamine additive proved to be dispensable (entries 18 and 20).
that we examined (Ph3P)4Pd proved to be more efficient
Substitution of lithium tert-butoxide with silver salts as
bases led to a dramatic acceleration of the reaction rate,
reducing the half-life to <1 h (entries 12ꢀ15). The triethy-
Ultimately, the best yields with silver phosphate were
obtained at 55 °C (entries 18 and 20).
Other experiments not shown in Table 1 revealed that
than Pd2dba3 CHCl3 or Pd(II) precatalysts (entries 1ꢀ5).
3
Reducing the amount of lithium tert-butoxide to just
1.1 equiv was beneficial (entries 1, 6, and 7), but surpris-
the added water was essential for good yields and that
ingly, lithium hydroxide was less efficient than lithium
other solvents were less efficient. Surprisingly, doubling
the amount of Pd catalyst to 10 mol % reduced the yield of
dimer 3.
tert-butoxide, even though water is a cosolvent. Other
metal alkoxides were less efficient (entries 7, 10, and 11).
The optimized dimerization reaction generates only the
trans alkene. The configuration of the double bond in 3
(1) Khanna, A.; Maung, C.; Johnson, K. R.; Luong, T. T.; Van
Vranken, D. L. Org. Lett. 2012, 14, 3233–3235.
r
10.1021/ol401563z
Published on Web 07/01/2013
2013 American Chemical Society

amino group. However, the N-adamantyl substrate re-
quired a slightly longer reaction time (3 h) to generate 3e.
The bis-piperidines are formed less efficiently than bis-
pyrrolidines under the optimized reaction conditions
(Scheme 4). However the dimerization of 4 directly gen-
erates the skeleton of the bis-piperidine alkaloid hyalbi-
done in a single step. The natural product hyalbidone was
isolated as mixture of meso and d/l isomers from the roots
of H. albus.2 It is not clear whether the mixture of stereo-
isomers occurs in Nature or was the result of epimerization
during isolation.
Table 1. Optimization of the Bis-cyclization/Dimerization of
Vinyl Iodide 1 to Generate Dimer 3
Scheme 3. The Bis-cyclization/Dimerization Reaction Tolerates
a Variety of N-Alkyl Groups
a Temperature dependence studies were carried out in 2-methyl-
tetrahydrofuran.
was established rigorously as E by the 15 Hz coupling
constants in the 13C satellites for the olefinic protons
(Scheme 2). With alkoxide bases, dimer (E)-3 was present
as a mixture of two different stereoisomers in an ∼95:5
ratio, differing in configuration at the stereogenic centers.
With silver phosphate a single stereoisomer was obtained.
The stereochemistry of the dimer 3 was rigorously estab-
lished as meso by a crystal structure (Scheme 2). Of note,
meso dimers of this kind cannot be efficiently made
through olefin metathesis.
The dimeric products accessed by our reaction bear a
remarkable resemblence to the dimeric products noted
by Balme et al. in the Pd-catalyzed reactions of terminal
ω-sulfonylalkyne 6 (Scheme 5). The products observed by
Balme3 were rationalized to arise from carbopalladation of
a terminal alkyne by a hydridopalladium species leading to
Pd(0) alkylidene intermediate 7.4
Scheme 4. One-Step Synthesis of the Hyalbidone Skeleton
Scheme 2. NMR and X-ray Crystallography Establish
the Configuration of the Double Bond and the Relative
Configuration
With optimized conditions for stereoselective formation
of the dimer 3 we next assessed the tolerance of the reaction
to variations of the N-alkyl substituent (Scheme 3). The
reaction was chemoselective, tolerating terminal olefins
and styryl groups to give 3a and 3b in 56% and 48% yield,
respectively. The reaction tolerated electron-rich furans
to give dimer 3c in good yield. Surprisingly, cylopentyl and
even 1-adamantyl substituents were well tolerated on the
(2) (a) Sauerwein, M.; Ishimaru, K.; Shimomura, K. Phytochemistry
1991, 30, 2977–2978. (b) Sauerwein, M.; Shimomura, K.; Wink, M.
Bioforum 1994, 17, 462–463.
ꢀ
(3) (a) Monteiro, N.; Gore, J.; Van Hemelryck, B.; Balme, G. Synlett
1994, 447–449. (b) Monteiro, N.; Balme, G. J. Org. Chem. 2000, 65,
3223–3226.
(4) Hoye, T. R.; Dinsmore, C. J.; Johnson, D. S.; Korkowski, P. F.
J. Org. Chem. 1990, 55, 4518–4520.
Org. Lett., Vol. 15, No. 14, 2013
3695

We do not observe alkyne intermediates in our dimer-
ization reactions. To test whether our reaction was occur-
ring through a syn elimination to form a terminal alkyne
intermediate, we synthesized the terminal alkyne 10 and
subjected it to the bis-cyclization/dimerization reaction
conditions. No dimer was formed, and the alkyne starting
material 10 was recovered. Thus, the bis-cyclization/
dimerization reaction reported in this work is complemen-
tary to, yet distinct from, the desulfonylative dimerization
of alkynes reported by Balme.
The resulting Pd(0) carbene b could then undergo oxida-
tive addition, much like Pd(0) complexes with N-hetero-
cyclic carbene ligands, to generate vinylpalladium carbene
c. Migratoryinsertionwouldgenerateanη1-allylpalladium
intermediate d which could undergo an allylic alkylation
through the η3-allylpalladium intermediate e.
Alternatively, the first CꢀN bond might arise through
an intramolecular aminopalladation reaction of f or i
to generate R-iodopalladium carbenoid intermediate g.
R-Elimination of the halide would produce the same
palladium carbene intermediate c that was invoked
for the Michael-type addition mechanism. Palladium car-
benoid g is also set up for reductive elimination to generate
an allyl iodide h that could undergo substitution through
an SN20 reaction or a Pd-catalyzed allylic alkylation.
The starting materials in this bis-cyclization reaction bear
a striking resemblence to N-Boc and N-aryl substrates used
in the Wolfe reaction and related processes.6 However, both
Balme7 and Wolfe8 have independently noted the failure
N-alkylamines (and specifically N-benzylpent-4-en-1-amine)
to engage in intramolecular syn aminopalladation reactions,
presumably due to the difficulty in forming the amidopalla-
dium intermediate i (X = H). Given the lack of precedence
for Wolfe reactions of N-alkylamines, we are reluctant to
invoke syn aminopalladation. Complexes such as f would
be well-suited for an anti aminopalladation, analogous to a
Wacker reaction, leading to the palladium carbenoid g.9
Given the potential for vicinal amino groups to promote
substitution with double inversion,10 the stereospecificity of
the aminopalladation cannot be used to rationalize formation
of the meso bis-pyrrolidine 3 from either diastereomer of
intermediate g.
Scheme 5. The Balme Dimerization of Terminal Alkynes Is
Distinct from Reactions of Vinyl Halides
The Balme system exploited a C-centered nucleophile to
forge the cyclodimer 8. To test the potential for C-nucleo-
philes in our cyclodimerization, substrate 11 was synthesized
and subjected to the optimized conditions from Scheme 3.
None of the desired dimer was formed even when water was
omitted to prevent hydrolysis of the malonate ester. Using
slightly modified conditions, the hindered bis-cyclopentane
12 was formed in a modest 5% yield, representing a single
turnover (Scheme 6).
Scheme 7. A Variety of Mechanistic Pathways Leading to Dimer
Scheme 6. Bis-cyclization/Dimerization with a Carbon
Nucleophile
The beneficial effect of silver is consistent with the in-
volvement of a vinylpalladium(II) cation (Scheme 7).
Inspiredby the palladium carbenesproposedfor theBalme
reaction, we hypothesized that the ω-amino group on
intermediate a might be poised to add to the vinylpalla-
dium cation in a process resembling a Michael addition.5
(5) Tsuji et al. have speculated that Michael-type addition of an
enolate to an allenylpalladium species might produce a palladium
carbene. Tsuji, J.; Watanabe, H.; Minami, I.; Shimizu, I. J. Am. Chem.
Soc. 1985, 107, 2196–2198.
~
(6) Minatti, A.; Muniz, K. Chem. Soc. Rev. 2007, 36, 1142–1152.
(7) Fournet, G.; Balme, G.; Gore, J. Tetrahedron 1990, 46, 7763–
7774.
3696
Org. Lett., Vol. 15, No. 14, 2013

When vinyl iodide 1 was subjected to the dimerization
reaction in the presence of norbornadiene an 8:1 mixture
of cyclopropane 13 and dimer 3 was obtained. The forma-
tion of cyclopropanes is consistent with Pd(0) alkylidene
intermediates. We recently showed that vinyl iodide 1 and
related compounds 1aꢀ1e could cyclopropanate norbor-
nadiene (Scheme 8), consistent with palladium-alkylidene
intermediate b in Scheme 7.11,3 Unfortunately, cyclopro-
panation of norbornenes is also consistent with a double
carbopalladation sequence proposed by Catellani et al.
or a carbopalladation/aminopalladation/reductive elimi-
nation sequence.12,13 When styrene was added to the reac-
tion mixture, no cyclopropanation was observed. Thus, it
is unclear whether formation of cyclopropanes from vinyl
iodide 1 and norbornadiene is evidence for a Pd(0) alky-
lidene intermediate or merely symptomatic of the unique
reactivity of norbornenes.
Scheme 9. A Crossover Experiment
be expected to undergo intramolecular aminopalladation
at comparable rates regardless of whether the halide substi-
tuent X was bromide or iodide. The intermediate g (X = Br)
would either generate mixed dimer 15 or, if unreactive, reduce
the catalytic turnover and yield; neither of these results was
observed. In contrast, the exclusive formation of dimer 3b is
consistent with the Michael-type addition/oxidative addition
pathway in which both of the vinyl halides that are incorpo-
rated into the dimer participate through successive oxidative
additions. The ability to access Pd(0) alkylidene intermediates
such as bin Scheme 7 from vinyl halides would offer immense
potential for construction of complex molecules.14
Scheme 8. A Cyclopropanation Experiment
In conclusion we have described the first example of
a bis-cyclization/dimerization reaction of vinyliodides that
generates meso bis-pyrrolidines and bis-piperidines. The
reaction tolerates a range of N-alkyl substituents and was
used to synthesize the skeleton of the alkaloid hyalbidone
in a single step. A crossover experiment is consistent with a
novel Michael-type addition of an amine to a vinylpalla-
dium cation giving rise to a Pd(0) alkylidene intermediate.
In order to gather mechanistic insight we set up a cross-
over experiment, taking advantage of the sluggish oxida-
tive addition of vinyl bromides relative to vinyl iodides
(Scheme 9). Vinyl bromide 14 gives little reaction at 55 °C
over 3 h, conditions under which vinyl iodide 1b generates
dimer 3b in 48% yield. The vinyl bromide can be coaxed to
form dimer 3 in 29% yield at higher temperature and with
an extended reaction time. When 50 mol % of vinyl iodide
1b and 50 mol % of vinyl bromide 14 were subjected to the
reaction, the only dimer that formed was the bis-cinnamy-
lamine dimer 3b, isolated in 49% yield. None of the bis-
benzylamine dimer 3 and none of the mixed dimer 15 were
observed during the reaction or after workup. Some of the
vinyl bromide (35%) starting material was recovered,
whereas the vinyl iodide was completely consumed.
Acknowledgment. We thank Dr. Joe Ziller (UCI) for
X-ray crystallographic analysis of 3. A.K. is supported by a
graduate fellowship from the National Science Foundation.
We thank Professor James S. Nowick (UCI) for suggesting
the use of 13C satellites to assign coupling constants in our
symmetrical dimers.
Based on the result of this crossover experiment, the
aminopalladation pathway in Scheme 7 is untenable because
the intermediate cationic Pd(II) olefin complex f would
Supporting Information Available. Experimental pro-
cedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
3
(8) Beaudoin, M.; Wolfe, J. P. Tetrahedron 2005, 61, 6447–6459.
˚
€
ꢀ
€
(9) Akermark, B.; Backvall, J. E.; Siirala-Hansen, K.; Sjoberg, K.;
Zetterberg, K. Tetrahedron Lett. 1974, 15, 1363–1366.
(10) Kondo, Y.; Suzuki, N.; Takahashi, M.; Kumamoto, T.; Masu,
H.; Ishikawa, T. J. Org. Chem. 2012, 77, 7988–7999.
(11) Khanna, A.; Premachandra, I. D. U. A.; Sung, P. D.; Van
Vranken, D. L. Org. Lett. 2013, 15, 3158–3161.
(14) (a) Schweizer, S.; Song, Z.-Z.; Meyer, F. E.; Parsons, P. J.; de
Meijere, A. Angew. Chem., Int. Ed. 1999, 38, 1452–1454. (b) Fillion, E.;
Taylor, N. J. J. Am. Chem. Soc. 2003, 125, 12700–12701. (c) Busacca,
C. A.; Swestock, J.; Johnson, R. E.; Bailey, T. R.; Musza, L.; Rodger,
C. A. J. Org. Chem. 1994, 59, 7553–7556. (d) Hashmi, A. S. K.; Ruppert,
€
T. L.; Knofel, T.; Bats, J. W. J. Org. Chem. 1997, 62, 7295–7304. (e)
(12) Catellani, M.; Chiusoli, G. P. J. Organomet. Chem. 1984, 275,
257–262.
(13) (a) Catellani, M.; Chiusoli, G. P.; Giroldini, W.; Salerno, G. J.
Organomet. Chem. 1980, 199, C21–C23. (b) Arcadi, A.; Marinelli, F.;
Bernocchi, E.; Cacchi, S.; Ortar, G. J. Organomet. Chem. 1989, 368, 249–
256.
Farina, V.; Hossain, M. A. Tetrahedron Lett. 1996, 37, 6997–7000. (f)
Hashmi, A. S. K.; Choi, J.-H.; Bats, J. W. J. Prakt. Chem. 1999, 341,
342–357.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 14, 2013
3697