Ring-closing metathesis reactions on azinium salts: Straightforward access to quinolizinium cations and their dihydro derivatives

Article abstract of DOI:10.1021/jo900292b

(Chemical Equation Presented) The ring-closing metathesis reaction of 1-butenyl-2-vinylpyridinium salts and 2-butenyl-1-vinylpyridinium salts using Grubbs second generation and Hoveyda-Grubbs catalysts proved to be an efficient approach to 3,4-dihydro- and 1,2-dihydroquinolizinium salts and the corresponding quinolizinium derivatives by an improved thermal oxidation in the presence of Pd/C without solvent. A comparative study showed that the quinolizinium system was obtained in better yields through the 3,4-dihydroquinolizinium route, thus allowing the synthesis of quinolizinium derivatives or improvements in the yields of some examples reported previously.

Full text of DOI:10.1021/jo900292b

Ring-Closing Metathesis Reactions on Azinium Salts:
Straightforward Access to Quinolizinium Cations and Their Dihydro
Derivatives
Ana Nun˜ez, Beatriz Abarca, Ana M. Cuadro,* Julio Alvarez-Builla, and Juan J. Vaquero*
Departamento de Qu´ımica Orga´nica, UniVersidad de Alcala´, 28871-Alcala´ de Henares, Madrid, Spain
juanjose.Vaquero@uah.es
ReceiVed February 10, 2009
The ring-closing metathesis reaction of 1-butenyl-2-vinylpyridinium salts and 2-butenyl-1-vinylpyridinium
salts using Grubbs second generation and Hoveyda-Grubbs catalysts proved to be an efficient approach
to 3,4-dihydro- and 1,2-dihydroquinolizinium salts and the corresponding quinolizinium derivatives by
an improved thermal oxidation in the presence of Pd/C without solvent. A comparative study showed
that the quinolizinium system was obtained in better yields through the 3,4-dihydroquinolizinium route,
thus allowing the synthesis of quinolizinium derivatives or improvements in the yields of some examples
reported previously.
Introduction
Azinium and azolium salts are representative models of het-
eroaromatic cations. Some of these systems have been known since
the emergence of heterocyclic chemistry as they have a wide range
of applications that include (inter alia) a variety of biological
activities,1 their use as reagents for organic synthesis,2 their role
as viologens3 or in chromophores that have fluorescent,4 nonlinear
optical (NLO)5 and DNA intercalating6 properties, and more
(1) (a) Scriven, E. F. V. In ComprehensiVe Heterocyclic Chemistry; Katritzky,
A. R., Rees, C. W., Eds.; Pergamon: Oxford, 1984; Vol. 1, pp 165-314. (b)
Comins, D. L.; Joseph, S. P. In ComprehensiVe Heterocyclic Chemistry II;
Katrizky, A. R., Rees, C. W., Scriven, E. F. V., Eds.; Pergamon: Oxford, 1996;
Vol. 5, pp 245-300.
FIGURE 1. Types of heteroaromatic cations.
(2) See: Encyclopedia of Reagents for Organic Synthesis (e-EROS); John
Wiley & Sons, Ltd.: New York, 2006.
recently their use as ionic liquids.7 In addition, these salts are easily
obtained by simple alkylation of an azine or an azole.
Less well-known is the other possible class of heteroaromatic
cations represented by the quinolizinium salts8 (Figure 1). The
main feature of these azonia cations is the presence of a
(3) Sliwa, W.; Bachowska, B.; Zelichowicz, N. Heterocycles 1991, 32, 2241.
(4) (a) Haugland, R. Handbook of Fluorescent Probes and Research
Chemicals, 8th ed.; Molecular Probes, Inc.: Eugene, OR, 2001. (b) Applied
Fluorescence in Chemistry, Biology and Medicine; Rettig, W., Strehmel, B.,
Schrader, S., Seifert, H., Eds.; Springer: New York, 1999.
(5) (a) Marder, S. R.; Perry, J. W.; Yakymyshyn, C. P. Chem. Mater. 1994,
6, 1137. (b) Coradin, T.; Nakatani, K.; Ledoux, I.; Zyss, J.; Cle´ment, R. J. Mater.
Chem. 1997, 7, 853. (c) Coe, B. J.; Harris, J. A.; Brunschwig, B. S.; Gar´ın, J.;
Ordun˜a, J. J. Am. Chem. Soc. 2005, 127, 3284. (d) Facchetti, A.; Abbotto, A.;
Beverina, L.; van der Boom, M. E.; Dutta, P.; Evmenenko, G.; Marks, T. J.;
Pagani, G. A. Chem. Mater. 2002, 14, 4996. (e) Mata, J. A.; Uriel, S.; Llusar,
R.; Peris, E. Organometallics 2000, 19, 3797.
(7) (a) Seddon, K. R. Nat. Mater. 2003, 2, 363. (b) Welton, T. Chem. ReV.
1999, 99, 2071.
(8) (a) Bradsher, C. K. In ComprehensiVe Heterocyclic Chemistry; Katritzky,
A. R.; Rees, C. W., Eds.; Pergamon: Oxford, 1984; Vol. 2, p 525. (b) Avendan˜o,
C.; Mene´ndez, C. In ComprehensiVe Heterocyclic Chemistry II; Katrizky, A. R.,
Rees, C. W., Scriven, E. F. V., Eds.; Pergamon: Oxford, 1996; Vol. 8, pp 507-
562. (c) Ihmels, H. Sci. Synth. 2005, 15, 907–945. (d) Ihmels, H.; Faulhaber,
K.; Vedaldi, D.; Dall’Acqua, F.; Viola, G. Photochem. Photobiol. 2005, 81, 1107.
(6) (a) Gago, F. Methods Enzymol. 1998, 14, 277. (b) Wilson, W. D.; Jones,
R. In Intercalation Chemistry; Whittingham, M. S., Jacobson, A. J., Eds.;
Academic Press: New York, 1981.
4166 J. Org. Chem. 2009, 74, 4166–4176
10.1021/jo900292b CCC: $40.75 2009 American Chemical Society
Published on Web 04/30/2009

Ring-Closing Metathesis Reactions on Azinium Salts
FIGURE 2. Quinolizinium system in natural alkaloids.
quaternary bridgehead nitrogen, which is responsible for the
cationic nature of quinolizinium, the simplest cationic hetero-
cyclic system belonging to this type and many other related
cations.
Moreover, quinolizinium ylides are well-known conjugated
heterocyclic mesomeric betaines that are present in different
families of alkaloids such as those based on the indolo[2,3-
a]quinolizinium system represented by sempervirine (3), fla-
vopereirine (4), neooxygambirtannine (5), and afrocurarine (6)
inter alia18 (Figure 2).
The lack of general methodologies to obtain derivatives of
quinolizinium-type cations and our interest in this field led us
to study two new approaches to these azonia cations. One of
these approaches involved palladium-catalyzed cross-coupling
methods,19 and the other concerned metathesis strategies to build
up the quinolizinium core. We report here our full results20 from
a comparative study of ring-closing metathesis (RCM) ap-
proaches to access the dihydro- and quinolizinium systems and
discuss the scope of the best strategy to obtain quinolizinium
derivatives.
Although the synthesis of quinolizinium-type cations is not
as easy as the preparation of azinium- and azolium-type cations,
the potential applications of the former systems appear to be
very similar to those found for the latter class, and in some
cases, new applications or improved behavior have been
reported. As an example, results from our laboratory show that
catiophores based on the quinolizinium system have much more
interesting NLO properties than those based on azinium cations9
(pyridinium, quinolinium, and isoquinolinium), and very prom-
ising antitumor activity has also been found in some DNA
intercalator tetracyclic cationic compounds based on aza-
quinolizinium cations10 targeting MET kinase and Bcl-XL.
Similarly, other groups have reported quinolizinium-based
derivatives with interesting DNA-binding properties,11 fluoro-
phores for detecting biomolecules,12 dyes for photovoltaic
cells,13 and new biological activities such as agents for the
treatment of bipolar affective disorder in mammals,14 selective
inhibitors of human TRPM4,15 and anti-inflammatory activity
in cystic fibrosis bronchial cells.16
Results and Discussion
Although the ring-closing metathesis reaction (RCM) has been
extensively used in heterocyclic chemistry and has proven to
be a powerful heterocyclization strategy,21 the number of studies
dealing with RCM on charged substrates is restricted to a few
examples. These examples include two pioneering works on
ammonium salts,22 the synthesis of the macrocycle moiety of
the (R)-(+)-muscopyridine alkaloid using a RCM on a pyri-
dinium hydrochloride23 and the total synthesis of manzamine,
which involved as the key step a RCM reaction on an alkenyl
pyridinium to build up the 13-membered ring of this alkaloid.24
In addition, the quinolizinium system (and its dihydro form)
is present in a variety of natural alkaloids known as berberines
and protoberberines, with berberine itself (1) and coralyne (2)
being the representative examples of this large family of
alkaloids, which are widely spread over nine botanic families.17
(9) Can˜eque, T. Planned dissertation, Universidad de Alcala´.
(10) Vera, P. Planned dissertation, Universidad de Alcala´.
(11) (a) Basili, S.; Bergen, A.; Dall’Acqua, F.; Faccio, A.; Granzhan, A.;
Ihmels, H.; Moro, S.; Viola, G. Biochemistry 2007, 46, 12721. (b) Ihmels, H.;
Otto, D.; Dall’Acqua, F.; Faccio, A.; Moro, S.; Viola, G. J. Org. Chem. 2006,
71, 8401.
(12) (a) Granzhan, A.; Ihmels, H.; Viola, G. J. Am. Chem. Soc. 2007, 129,
1254. (b) Granzhan, A.; Ihmels, H. Org. Lett. 2005, 7, 5119.
(13) Ahn, K.-S.; Choi, J.-M.; Kang, M.-S.; Shin, B.-C.; Park, J.-W.; Kwon,
M.-S.; Moon, S.-J.; Lee, J.-K.; Lee, J.-W.; Lee, W.-S. Eur. Patent Appl. EP
1755178 A2 20070221, 2007.
(14) Giralt, E.; Tarrago, M. T. PCT Int. Appl. WO 2008077978 A1 20080703,
2008.
(15) Grand, T.; Demion, M.; Norez, C.; Mettey, Y.; Launay, P.; Becq, F.;
Bois, P.; Guinamard, R. Br. J. Pharmacol. 2008, 153, 1697.
(16) Dechecchi, M. C.; Nicolis, E.; Bezzerri, V.; Vella, A.; Colombatti, M.;
Assael, B. M.; Mettey, Y.; Borgatti, M.; Mancini, I.; Gambari, R.; Becq, F.;
Cabrini, G. Am. J. Respir. Cell Mol. Biol. 2007, 36, 615.
(17) (a) Kametani, T.; Ihara, M.; Honda, T. Heterocycles 1976, 4, 483. Kondo,
Y. Heterocycles 1976, 4, 197.
(18) Schmidt, A. AdV. Heterocycl. Chem. 2003, 85, 67.
(19) (a) Barch´ın, B. M.; Valenciano, J.; Cuadro, A. M.; Alvarez-Builla, J.;
Vaquero, J. J. Org. Lett. 1999, 1, 545. (b) Garc´ıa-Cuadrado, D.; Cuadro, A. M.;
Alvarez-Builla, J.; Vaquero, J. J. Synlett 2000, 11, 1904. (c) Garcia-Cuadrado,
D.; Cuadro, A. M.; Barchin, B. M.; Nunez, A.; Can˜eque, T.; Alvarez-Builla, J.;
Vaquero, J. J. J. Org. Chem. 2006, 71, 7989. (d) Garcia-Cuadrado, D.; Cuadro,
A. M.; Alvarez-Builla, J.; Sancho, U.; Castan˜o, O.; Vaquero, J. J. Org. Lett.
2006, 8, 5955.
(20) For preliminary communication, see: Nun˜ez, A.; Cuadro, A. M.; Alvarez-
Builla, J.; Vaquero, J. J. Org. Lett. 2004, 6, 4125.
(21) Handbook of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH: Weinheim,
Germany, 2003. (b) Deiters, A.; Martin, S. F. Chem. ReV. 2004, 104, 2199. (c)
Poulsen, C. S.; Madsen, R. Synthesis 2003, 1. (d) Trnka, T. M.; Grubbs, R. H.
Acc. Chem. Res. 2001, 34, 18.
J. Org. Chem. Vol. 74, No. 11, 2009 4167

Nun˜ez et al.
SCHEME 1. Retrosynthetic Strategies for the Quinolizinium System
SCHEME 2. Synthesis of 3,4-Dihydroquinolizinium Triflate 10a by a RCM Reaction
We envisaged that the quinolizinium system could be obtained
from three different dihydro derivatives, which in turn could
be obtained by three different disconnections (a, b, and c in
Scheme 1) involving ꢀ, γ, and δ bonds with respect to the
nitrogen atom of the pyridinium substrate.
the migration of the double bond proved fruitless, and we
therefore explored the synthesis of 13 by an alternative route
starting from a pyridinium salt functionalized at the C2 position
with the propenyl substituent. Our choice was 2-bromo-1-
methylpyridinium iodide as this is the simplest model to study
the feasibility of the introduction of the allylic substituent at
the C2 position. Unfortunately, all attempts to prepare derivative
16 were unsuccessful using either propenyltributyl stannane
under Stille coupling conditions or allyl trifluoroborate under
Suzuki coupling conditions. The most promising result of these
initial studies was achieved when 14 was alkylated with butenyl
triflate26 to yield, as expected, the corresponding diene 17.
Furthermore, we were able to prove that this diene is a suitable
substrate for a RCM reaction to afford the 3,4-dihydroquino-
lizinium triflate (10a) in moderate yield (55%). This result
strongly supported our strategy based on disconnection c since
it proved that a 1-(3′-butenyl)-2-vinylpyridinum salt can be an
appropriate substrate for the desired RCM reaction.
Initially, strategy b seemed to be the most suitable because
the other two approaches would involve the use of an electron-
poor diene which, in principle, seemed less appropriate to
participate in the metathesis reactions. Furthermore, the dienic
precursor 13 appeared to be more easily available than the
corresponding dienic substrates 9 and 11. The synthesis of 13
was initially attempted from 2-propenylpyridine (14), the
synthesis of which is described25 from the reaction of pyridinium
N-oxide under the conditions shown in Scheme 2. When 14
was subjected to alkylation with either allyl bromide or allyl
iodide, the expected pyridinium diene 13 was not obtained under
a variety of different conditions, with the isomeric diene 15
being isolated in low yields (10-19%). Several attempts to avoid
Clearly, the next step was the preparation of the dienic
substrate 11a in order to test it in the RCM reaction. In this
case, the commercially available 2-vinylpyridine was chosen
as starting material, which was easily alkylated with 3-butenyl
triflate to yield 11a in 82% yield (Scheme 3). It was found that
11a was a good substrate for the RCM reaction under different
(22) (a) Nguyen, S. T.; Grubbs, R. H. J. Am. Chem. Soc. 1993, 115, 9856.
(b) Kirkland, T. A.; Lynn, D. M.; Grubbs, R. H. J. Org. Chem. 1998, 63, 9904.
(c) Fu¨rstner, A.; Grabowski, J.; Lehmannn, C. W. J. Org. Chem. 1999, 64, 8275.
(d) Fu¨rstner, A.; Thiel, O. R.; Schanz, H.-J.; Nolan, S. P. J. Org. Chem. 2000,
65, 2204.
(23) Fu¨rstner, A.; Leitner, A. Angew. Chem., Int. Ed. 2003, 42, 308.
(24) Magnier, E.; Langlois, Y. Tetrahedron Lett. 1998, 39, 837.
(25) Al-Arnaout, A.; Courtois, G.; Miginiac, C. L. J. Organomet. Chem. 1987,
333, 139.
(26) Dobbs, A. P.; Jones, K.; Veal, K. T. Tetrahedron Lett. 1997, 38, 5383.
4168 J. Org. Chem. Vol. 74, No. 11, 2009

Ring-Closing Metathesis Reactions on Azinium Salts
TABLE 1. Optimization of Reaction Conditions for the Oxidation
of 10a (NR: No Reaction)
SCHEME 3. Improved Synthesis of 10a by a RCM
Reaction
entry
conditions
Pd/C, EtOH reflux
NBS, Et3N
SeO2
chloranil
O2/Pt, AcOH
O2/ Pt, PhNO2
Pd/C, n-BuOH reflux
DDQ/toluene 24 h, 110 °C
DDQ/DMF, 21 h, 120 °C
results (yield %)a
15b
1
2
3
4
5
6
7
8
9
6c
3c
10c
23c
26c
34c
NR
35
conditions using first generation Grubbs catalyst (G-I).27 The
best results (83% yield of isolated product) were achieved in
dichloromethane, at room temperature, for 1.5 h and using 5
mol % of the catalyst. The metathesis reaction of 11a in the
presence of the second generation Grubbs catalyst (G-II) gave
the same yield (83%). It is worth noting that this route to 3,4-
dihydroquinolizinium triflate (10a), based on the RCM reaction
as the key step, affords the desired product in 68% overall yield
while the previously reported three-step synthesis of 10a starting
from 2-methylpyridine gave this dihydro derivative as the
bromide in 32% overall yield.28
10 Pd/C, AcOH, 24 h, 80 °C
11 Pd/C, AcOH, 14 days, 117 °C
12 10% Pd/C, 160 °C, MW, 100 w, 5 min 11a:7a:21a 0.04:0.8:0.8
13 10% Pd/C, 200 °C, MW, 100 w, 5 min 11a:7a:21a 0:1:1
14 30% Pd/C, 200 °C, MW, 100 w, 5 min 11a:7a:21a 0:1:1
NR
66
15 30% Pd/C, 200 °C, 2 h
79
a Yields refer to 7a as isolated products. b Yield reported in ref 29.
c Yields reported in ref 30.
SCHEME 4. Synthesis of the 1,2-Dihydroquinolizinium
Triflate 8a by a RCM Reaction
The conversion of 10a into quinolizinium 7a was first
reported by Boekelheide and Gall in their pioneering synthesis
of the quinolizinium heterocycle, and this oxidation was the
main drawback of this synthesis since the yield was only 15%
when using Pd/C as catalyst in refluxing ethanol.28 An improved
synthesis of 7a was later reported by the same authors, with
the oxidation step improved to give 34% yield using P/C in
refluxing n-butanol.29 Other oxidants such as NBS, chloranil,
selenium dioxide, and oxygen in the presence of platinum either
in acetic acid or nitrobenzene gave the fully oxidized cation in
yields ranging from 3 to 26%.
In order to improve the reported yields of this oxidation, we
also studied the transformation of 10a into 7a in the hope of
finding synthetically more useful conditions for this conversion.
A summary of the conditions and yields for the oxidation of
10a to 7a is given in Table 1. We tested the oxidation using
DDQ in toluene (100 °C) and DMF (120 °C), but only the latter
conditions afforded the oxidation product, albeit in only 35%
yield after 21 h (Table 1, entries 8 and 9). The use of the reported
conditions (Pd/C, HOAc) for the oxidation of some 1,4-
dihydropyridines bearing electron-withdrawing substituents30 led
to a considerable improvement in the yield to 66%, although
the reaction required 14 days to reach completion (Table 1, entry
11). We thought that this disadvantage could by circumvented
by carrying out the reaction under microwave irradiation due
to the well-known efficacy of this technique to reduce reaction
times significantly.31 Attempts to carry out the reaction in a
microwave reactor in the absence of solvent, heating at 160-200
°C at 100 W, afforded variable mixtures of 10a, 7a, and the
1,2,3,4-tetrahydroquinolizinium 21a or the two latter compounds
if the starting material was completely consumed (Table 1,
entries 12-14). Several attempts to avoid the formation of 21a
in the presence of hydrogen acceptors proved unsuccessful either
in the microwave reactor or in a domestic microwave oven with
the reaction carried out in an open vessel. Finally, it was found
that under conventional heating at 200 °C and increasing the
amount of catalyst to 40% the quinolizinium triflate was
obtained in 79% yield in only 2 h (Table 1, entry 15).
Although it seemed feasible to extend the synthetic procedure
developed for 7a to different quinolizinium derivatives, we
decided to explore the approach involving the disconnection of
the ꢀ bond (strategy a in Scheme 1) in order to compare the
two approaches before proceeding to study the scope of these
reactions. The desired diene 9a was synthesized in three steps
from 2-methylpyridine as shown in Scheme 4. The dehydro-
halogenation reaction conditions employed to generate 9a from
20 were those reported by Katritzky32 for the formation of
N-vinyl derivatives. Under these conditions, the diene 9a was
obtained in moderate yield (64%), and this could not be
improved by changing the base (NaHCO3, Na2CO3, and Cs2CO3)
and/or the solvent (MeCN, EtOH/MeOH, CH2Cl2/H2O, MeCN/
i-PrOH). Under the same conditions as used for the RCM
reaction of 11a, diene 9a afforded 8a in moderate yield (52%).
Attempts to improve this yield using the second generation
Grubbs catalyst (G-II) and Hoveyda-Grubbs catalyst (H-G,
see below in Table 6) were unsuccessful. It is noteworthy that
(27) Schwab, P.; France Ziller, J. W.; Grubbs, R. H. Angew. Chem., Int. Ed.
Engl. 1995, 34, 2039.
(28) Boekelheide, V.; Gall, W. G. J. Am. Chem. Soc. 1954, 76, 1832.
(29) Boekelheide, V.; Ross, J. M. J. Am. Chem. Soc. 1955, 77, 5691.
(30) Nakamichi, N.; Kawashita, Y.; Hayashi, M. Org. Lett. 2002, 4, 3955.
(31) (a) Kappe, C. O.; Stadler, A. MicrowaVes in Organic and Medicinal
Chemistry; Wiley-VCH: Weinheim, Germany, 2005. (b) MicrowaVe Assisted
Organic Synthesis; Tierney, J. P., Lidstrom, P., Eds.; Blackwell Publishing:
Oxford, UK, 2005.
(32) Katritzky, A. R.; Rubio-Teresa, O. J. Org. Chem. 1983, 48, 4017.
J. Org. Chem. Vol. 74, No. 11, 2009 4169

Nun˜ez et al.
SCHEME 5. Synthesis of 3-Substituted 2-Vinylpyridines
TABLE 2. Synthesis of 2-Vinylpyridines 18
18g,h
TABLE 3. Synthesis of Pyridinium Dienes 11b-h (NR: No
Reaction)
a Yields refer to 18 as isolated products. b X
trifurylphosphine.
) OTf; TFP )
this is the first route that allows the synthesis of the parent 1,2-
dihydroquinolizinium system in a synthetically useful overall
yield (18%), and this enabled the first full characterization of
this system as the previously reported synthesis of the 1,2-
dihydroquinolizinium picrate gave a very poor yield (overall
yield not given).33 Finally, the oxidation of 8a under the
optimized conditions used for 10a gave 7a in almost quantitative
yield.
A simple comparison between these two strategies for the
synthesis of the quinolizinium system shows the advantage of
the approach based on the disconnection of the δ bond (strategy
c) (Scheme 1) since it affords the quinolizinium triflate 7a in a
54% overall yield, while strategy a, based on the ꢀ bond
disconnection, gave 7a in only 18% overall yield.34 Conse-
quently, our next step was to study the scope of the RCM
reaction based on strategy c to produce differently substituted
3,4-dihydroquinolizinium derivatives by using appropriate dienic
pyridinium salts as substrates for the RCM reaction. The
simplest approach to substituted 3,4-dihydroquinolizinium salts
was to use a variety of substituted 2-vinylpyridines as starting
materials to obtain the appropriate cationic dienes. As a result,
our first goal was the synthesis of these intermediates since most
of these pyridine derivatives are currently unknown. It was found
that various vinylpyridines 18b-f could be prepared in moderate
or good yields by a Stille reaction between tributylvinyl stannane
and the corresponding 2-bromo- or 2-trifluoromethanesulfony-
loxypyridine under the optimized conditions for each substrate
(see Table 2). In addition, two more substrates 18g,h were
obtained from 18c and 18b, respectively, under the conditions
shown in Scheme 5. It is worth noting that neither 18g nor 18h
could be obtained from the corresponding substituted 2-bromo-
or 2-trifluoromethanesulfonyloxypyridine under the Stille cou-
pling conditions employed for the preparation of 18b-f.
a Yields refer to 11 as isolated products. b Inseparable mixture of 18c,
11c, and the O-alkylation product.
The N-alkylation of 18b-h was carried out with 3-butenyl-
triflate under the same conditions used for the synthesis of the
diene 11a. The results are summarized in Table 3 and show
that the reaction failed with 18f and 18g while the attempted
N-alkylation of 18c under different conditions gave, in all cases,
inseparable mixtures of the desired diene along with the
O-alkylation product (main component) and the starting material.
For the rest of the vinylpyridines, the N-alkylation reaction
afforded the expected dienes in moderate yields (41-67%). The
failure of the reaction with 18f,g could be due to the electron-
withdrawing effect of the substituents on the pyridine ring. On
(33) Glover, E. E.; Jones, G. J. Chem. Soc. 1959, 1686.
(34) The best synthesis of the quinolizinium bromide was reported in also
54% overall yield by Miyadera, T.; Iwai, I. Chem. Pharm. Bull. 1964, 12, 1338.
4170 J. Org. Chem. Vol. 74, No. 11, 2009

Ring-Closing Metathesis Reactions on Azinium Salts
SCHEME 6. Synthesis of Ethenylpyridines 18i-k
the other hand, the moderate yields obtained in all cases seem
to be more related to the low stability of the alkylating agent.
In fact, this species had to be generated and used in situ in order
to minimize its decomposition, which takes place under moder-
ate heating or even at room temperature with prolonged reaction
times.
To investigate further the potential of the RCM reaction as a
general method for the synthesis of substituted 3,4-dihydro-
quinolizinium salts and eventually the corresponding quinoliz-
inium derivatives after oxidation, we undertook the synthesis
of a series of dienic quinolizinium derivatives 11i-s bearing
substituents in the vinyl and/or butenyl moieties. The synthesis
of some ethenyl pyridines is shown in Scheme 6. Two synthetic
strategies are outlined for 2-(2′-propenyl)pyridine (18k). The
upper route has been reported previously,35 but we found that
a better yield could be obtained by using the lower route, which
involves a Wittig reaction on 2-acetylpyridine with the ap-
propriate ylide. This pyridine derivative was also the starting
compound for the synthesis of the silylenol ether derivative 18j.
The styryl pyridine derivative 18i was obtained in moderate
yield by a Suzuki coupling reaction between 2-bromopyridine
and the 2-(phenylethynyl)boronic acid. On the other hand, two
commercially available 3-substituted 3-butenoles (3-bromo- and
3-methyl-3-butenol) were used for the synthesis of the corre-
sponding triflates. In addition, 3-bromo-3-butenol was also
employed as the starting material for the preparation of some
3-aryl-3-butenoles by a Suzuki coupling reaction according to
literature procedures.36 A literature procedure was also followed
for the preparation of 3-iodo-3-butenol from 3-butynol.37
The synthesis of the series of dienic pyridinium derivatives
with a variety of substitution patterns was carried out using the
method described above for 11a-h. The results of the N-
alkylation reactions of the 2-propenylpyridines with the 3-sub-
stituted butenyl triflates are summarized in Table 4.
for the diene 11k (bearing a methyl substituent), which afforded
the corresponding 1-methyl-3,4-dihydroquinolizinium triflate
(10k) in 91% yield. However, only a moderate yield was
obtained in the RCM reaction of 11i (61%), and in the case of
11j, only partial conversion was observed along with the
formation of polymerization products, which resulted in only
23% yield of the corresponding 3,4-dihydroquinolizinium
derivative 10j. On heating the reaction mixture under reflux
under higher dilution conditions (0.01 M), the yield increased
to 40%. In an attempt to improve this yield, the Hoveyda-Grubbs
catalyst (H-G) was also tested due to its known potential in
metathesis reactions with electron-deficient olefins38 and its
higher thermal stability.39 The reaction performed in CH2Cl2
under refluxing conditions in the presence of 5 mol % of H-G
gave 60% yield of the RCM product, and polymerization
products were not formed, although some starting diene still
remained in the reaction mixture. Next, using the best catalyst,
H-G, the reaction temperature and catalyst loading were varied.
It was found that, in 1,2-dichloroethane at 65 °C, full conversion
of the diene 11j was observed after 3 h and product 10j could
be isolated in 78% yield. The use of 2 mol % of H-G again led
to incomplete conversion of the diene after the same reaction
time.
The yields obtained for dienes 11i-s are similar to or lower
than those obtained from vinylpyridines 18b-h in their reactions
with 3-butenyl triflate. The lower yields seem to be related to
the reduced stability of these substituted dienes, which are prone
to decompose at room temperature. It is noteworthy that the
instability of diene 11p precluded its isolation and characteriza-
tion. On the other hand, the higher temperature needed to
complete some N-alkylations, such as those detailed in entries
2 and 11, favored the decomposition and led to decreases in
the isolated yields.
Before submitting the pyridinium dienes 11 to the ring-closing
metathesis reaction, we further studied the process using the
dienes 11b and 11h as model systems which bear an electron-
donating and an electron-withdrawing group in the pyridine ring,
respectively. In both cases, the second generation Grubbs
catalyst (G-II) proved to be more efficient than G-I as it gave
better yields of the corresponding 3,4-dihydroquinolizinium
derivatives (11b, 80 vs 40%; 11h, 85 vs 78%). Consequently,
catalyst G-II was initially chosen to carry out the RCM reaction,
and very good yields were obtained from dienes 11b,d,e,h, as
shown in Table 5.
It is noteworthy that these optimized conditions for the
synthesis of 10j did not lead to the expected improvement in
the yield of the dihydroquinolizinium 10i. For this reason, the
RCM reaction on substituted dienes on the butenyl moiety and
on disubstituted dienes was, in all cases, tested with both the
G-II and H-G catalysts. The results are shown in Table 6, and
these represent the optimal conditions for the metathesis reaction
in each case.
The metathesis reaction involving substitution in the butenyl
diene moiety (dienes 11l-o) afforded the 2-methyl- and
2-phenyl-3,4-dihydroquinoliziniums 10l and 10m in good yields
using the G-II catalyst under the conditions shown in entries 4
and 5. However, neither H-G nor G-II catalysts were able to
The monosubstituted dienes on the ethenyl moieties 11i-k
were next subjected to the RCM reaction using the conditions
detailed in Table 6. These conditions proved to be very efficient
(38) (a) Randl, S.; Gessler, S.; Wakamatsu, H.; Blechert, S. Synlett 2001, 3,
430. (b) Randl, S.; Connon, S. J.; Blechert, S. Chem. Commun. 2001, 1692. (c)
Randl, S.; Connon, S. J.; Blechert, S. Chem. Commun. 2001, 1796.
(39) (a) Garber, S. B.; Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H. J. Am.
Chem. Soc. 2000, 122, 8168. (b) Kingsbury, J. S.; Hoveyda, A. H. J. Am. Chem.
Soc. 2005, 127, 4510.
(35) (a) Bachman, G. B.; Micucci, D. J. Am. Chem. Soc. 1948, 70, 2381. (b)
Adamson, D. W.; Billinghurst, J. W. J. Chem. Soc. 1950, 1039.
(36) Huang, Q.; Fazio, A.; Dai, G.; Campo, M. A.; Larock, R. C. J. Am.
Chem. Soc. 2004, 126, 7460.
(37) Kamiya, N.; Chikami, Y.; Ishii, Y. Synlett 1990, 675.
J. Org. Chem. Vol. 74, No. 11, 2009 4171

Nun˜ez et al.
TABLE 4. Synthesis of Substituted Pyridinium Dienes 11i-s
a Yields refer to 11 as isolated products. b Decomposition.
produce the cyclization reaction of substrates 11n,o, both of
which bear a halogen as a substituent in the butenyl moiety.
This result seems to be consistent with previous reports on
the failure of cross-metathesis reactions using vinyl halides40
or those describing the feasibility of the RCM reaction with
vinyl chlorides but not with vinyl bromides.41 Apparently, the
higher reactivity of the catalyst toward the halogen when
compared to the terminal alkene leads to the formation of an
unreactive Fisher carbene, thus precluding the RCM reaction.
Substitution in both the ethenyl and butenyl moieties required
the use of the H-G catalyst, higher temperatures, and longer
(40) Chatterjee, A. K.; Morgan, J. P.; Scholl, M.; Grubbs, R. H. J. Am. Chem.
Soc. 2000, 122, 3783.
(41) Chao, W.; Weinreb, S. M. Org. Lett. 2003, 5, 2505.
4172 J. Org. Chem. Vol. 74, No. 11, 2009

Ring-Closing Metathesis Reactions on Azinium Salts
TABLE 5. Synthesis of 3,4-Dihydroquinolizinium Salts 10b-h by
RCM Reaction of 11
SCHEME 7. Mechanism for the Formation of
3,4-Dihydroquinolizinium Salts 10 by the RCM Reaction of
Pyridinium Dienes 11
TABLE 6. Synthesis of Salts 10 by RCM Reaction of 11
a Yields refer to 10 as isolated products. b Reaction time: 1 h.
c Reaction time: 2 h.
reaction times to afford the desired 1,2-disubstituted 3,4-
dihydroquinolizinium salts, as exemplified by the formation of
10q (entry 8). However, even harsher conditions were not
sufficient to achieve the cyclization of dienes 11r,s, which were
partially recovered from the reaction mixtures after 24 h at 120
°C in tetrachloroethane (Table 6, entries 9 and 10).
The failure of the RCM reaction on dienes 11r,s along with
the higher temperatures and/or longer reaction times needed to
complete the reaction on dienes 11l and 11m when compared
with dienes bearing the same substituents (Me and Ph) in the
ethenyl moiety (10k and 10i) strongly supports the mechanism
of the reaction as being initiated with the insertion of the Ru
alkylidene into the double bond of the butenyl alkene rather
than into the double bond of the ethenyl moiety because of the
electronic deficiency of the latter. If this is the case, the steric
effect of the substituent on the butenyl moiety will have a crucial
influence on the rate of the reaction (Scheme 7).
The final goal of our study was to transform the substituted
3,4-quinolizinium salts 10 into the corresponding quinolizinium
derivatives 7 in order to demonstrate both the scope of the
oxidation conditions used for the formation of the quinolizinium
7a from 10a and the power of the RCM reaction to access the
quinolizinium derivatives or to improve their synthesis in cases
where they are known. Thus, the synthesized 3,4-dihydroquino-
lizinium derivatives were subjected to the oxidation conditions
explored for the transformation of 10a into 7a. The results
obtained in a comparative study under two different sets of
conditions [reflux in acetic acid or heating at 200 °C in the
presence of a high ratio of Pd/C (10%)] are summarized in Table
a Yields refer to 10 as isolated product.
7. The oxidation of the 3,4-dihydroquinolizinium derivatives
with substituents in the pyridinium ring (10b, 10d, 10e, and
10h) were successful except for derivative 10h, which under
both sets of conditions tested could not be completely trans-
J. Org. Chem. Vol. 74, No. 11, 2009 4173

Nun˜ez et al.
TABLE 7. Synthesis of Quinolizinium Salts 7 by Oxidation of
3,4-Dihydroquinolizinium Salts 10
The oxidation of derivatives with a substituent in the
dihydroquinolizinium ring seemed to be even harder in acetic
acid. For instance, the oxidation of 10k proceeded cleanly to
afford 7k in quantitative yield, but 21 days were required for
completion of the reaction. A slightly lower yield (94%) was
obtained under thermal conditions, but complete oxidation took
only 2 h. For these reasons, derivatives 10i, 10j, 10l, 10m, and
10q were only tested under thermal conditions, and the yields
are shown in Table 7 (entries 6, 7, and 9-11). These conditions
were found to be unsuitable for the oxidation of 10j, which
decomposed under these conditions. On the other hand, 7q
required the longest reaction time to afford the lowest yield of
the corresponding quinolizinium derivative (entry 11). Methyl
(7k,l) and phenyl (7i and 7m) derivatives were obtained in good
yields when the substituent is in the C1 position (entries 6 and
8) and in moderate yields if either the phenyl or the methyl
substituent are in the C2 position (entries 9 and 10).
While most of the 3,4-dihydroquinolizinium derivatives 10
synthesized here are unknown, most of the corresponding quino-
lizinium salts obtained from their oxidation have been reported
previously. It is worth noting the preparation for the first time of
the 1,2-dimethylquinolizinium salt 7q with an overall yield of 10%
(entry 11) and the significant improvement achieved for 1-bromo-,
3-bromo-, and 1-phenylquinolizinium salts 7b42 (29 vs 2%), 7d42
(17 vs 9%), and 7i33 (15 vs 0.04%) (entries 2, 3, and 6,
respectively). By contrast, 3-methyl-, 2-methyl-, and 2-phe-
nylquinolizinium triflates 7e,34 7l,43 and 7m43 were obtained in
lower yields than those previously reported (entries 4, 9, and 10,
respectively). The parent quinolizinium 7a was obtained with the
same yield (54%) as that previously reported38 (entry 1), and
1-methylquinolizinium 7k was achieved in a very similar yield34
(27 vs 26%, entry 8).
In conclusion, the results described above show that RCM is
a viable reaction on N-alkenyl-R-vinylpyridinium systems to
give a variety of 3,4-dihydroquinolizinium derivatives in good
overall yields from readily available starting materials. More-
over, our studies also allowed a significant improvement in the
oxidation of these 3,4-dihydroquinolizinium derivatives to the
corresponding quinolizinium salts under thermal conditions in
the presence of Pd/C. This approach, which is based on a RCM
reaction as the key step for the construction of the fused six-
membered ring on quinolizinium-type systems, can provide
access to biologically relevant cations based on this cationic
heteroaromatic system.
a Yields refer to 7 as isolated products. b Conditions: 40% Pd/C
(10%)/HOAc. c Conditions: 40% Pd/C (10%)/200 °C. d Incomplete
reaction. e Decomposition.
Experimental Section
General Procedure for the Preparation of Dienes 11. A
solution of the corresponding 3-butenol (2.6 mmol) and dry pyridine
(0.205 g, 2.6 mmol) in dry CCl4 (2 mL) was stirred at room temperature
for 5-10 min under argon. The mixture was added dropwise (5-10
min) to a cooled (-10 °C) solution of triflic anhydride (0.733 g, 6
mmol) in dry CCl4 (3 mL). The resulting white solid was filtered off
through sodium sulfate; the solution was added by cannula to a solution
of the vinyl derivative 18 (2 mmol) in dry CCl4 (2 mL), and the reaction
mixture was stirred for 24 h at room temperature or 40 °C. The solvent
was evaporated under reduced pressure, and the residue was purified
by flash chromatography on silica gel (eluent: CH2Cl2/MeOH 9.5:0.5)
or by washing with Et2O.
formed into the desired oxidation product and gave inseparable
mixtures from which the corresponding quinolizinium derivative
could not be isolated (some charged decomposition products
were also formed).
A similar result was obtained for 10d under acidic conditions,
although in this case, the thermal oxidation allowed the
completion of the reaction and the isolation of the 2-bromo-
quinolizinium derivative 7d in 67% yield (Table 7, entry 3).
The quinolizinium derivatives 7b and 7e were both obtained in
good yields under both sets of conditions, although for 7e, very
long times were required to complete the reaction under acidic
conditions (entry 4) whereas a more acceptable time was needed
for the oxidation of the 1-bromo-substituted derivative 10b (12
h vs 14 days, entries 2 and 4).
1-(But-3-enyl)-2-vinylpyridinium triflate (11a). Following the
general procedure, the reaction of 2-vinylpyridine (0.21 g, 2 mmol)
(42) Sanders, G. M.; van Dijk, M.; van der Plas, H. Heterocycles 1981, 15,
213.
(43) Glover, E. E.; Jones, G. J. Chem. Soc. 1958, 3021.
4174 J. Org. Chem. Vol. 74, No. 11, 2009

Ring-Closing Metathesis Reactions on Azinium Salts
and 3-butenyl triflate (0.53 g, 2.6 mmol) afforded 0.507 g (82%)
of 11a as a pale-yellow oil: IR (NaCl) νmax (cm-1) 3087, 1621,
1257, 1512, 1161, 1030, 785; 1H NMR (300 MHz, CDCl3) δ 8.92
(d, 1H, J ) 6.2 Hz), 8.39 (t, 1H, J ) 7.9 Hz), 8.02 (d, 1H, J ) 8.2
Hz), 7.90 (t, 1H, J ) 6.4 Hz), 7.13 (dd, 1H, J ) 11.3, 17.0 Hz),
6.28 (dd, 1H, J ) 1.3, 17.0 Hz), 6.13 (dd, 1H, J ) 1.1, 11.2 Hz),
5.84-5.71 (m, 1H), 5.06 (d, 1H, J ) 10.2 Hz), 4.94 (dd, 1H, J )
1.3, 17.0 Hz), 4.79 (t, 2H, J ) 6.9 Hz), 2.65 (dd, 2H, J ) 6.9, 14.1
Hz); 13C NMR (75 MHz, acetone-d6) δ 153.2, 146.6, 133.2, 130.5,
128.3, 127.6, 127.3, 119.7, 58.2, 34.7; MS (ESI+) m/z (relative
intensity) 160 (M+). Anal. Calcd for C12H14NSO3F3: C, 46.60; H,
4.56; N, 4.53; S, 10.37. Found: C, 46.52; H, 4.79; N, 4.44; S, 10.41.
Ring-Closing Metathesis of Dienes 11. General Procedure
for 3,4-Dihydroquinolizinium 10. The ruthenium catalyst G-II
(5 mol %, 0.01 mmol, 8.5 mg) or H-G (5 mol %, 0.01 mmol, 6.2
mg) in CH2Cl2 or ClCH2CH2Cl (1 mL) was added to a solution of
the corresponding diene 11 (0.2 mmol) in dry CH2Cl2 or
ClCH2CH2Cl (1.5 mL) under an argon atmosphere. The reaction
mixture was stirred at room temperature or under heating for the
appropriate time. The solvent was evaporated under reduced
pressure, and the residue was purified by flash chromatography
(eluent: CH2Cl2/MeOH 9.3:0.7).
3,4-Dihydroquinolizinium Triflate (10a). Following the general
procedure, after stirring for 1.5 h, 46.8 mg (83%) of 10a was obtained
as a gray powder: mp 124-125 °C (acetone/Et2O); IR (NaCl) νmax
(cm-1) 3086, 1646, 1507, 1263, 1150, 1031, 808; 1H NMR (300 MHz,
CDCl3) δ 8.92 (d, 1H, J ) 6.2 Hz), 8.33 (t, 1H, J ) 7.8 Hz), 7.80 (t,
1H, J ) 7.1 Hz), 7.65 (d, 1H, J ) 7.9 Hz), 6.89-6.83 (m, 1H), 6.72
(d, 1H, J ) 9.7 Hz), 4.81 (t, 2H, J ) 7.7 Hz), 2.85 (dd, 2H, J ) 7.7,
12.3 Hz); 13C NMR (75 MHz, acetone-d6) δ 146.7, 146.1, 140.6, 126.6,
126.0, 122.1, 53.7, 22.9; MS (ESI+) m/z (relative intensity) 132 (M+).
Anal. Calcd for C10H10NSO3F3: C, 42.71; H, 3.58; N, 4.98; S, 11.40.
Found: C, 42.40; H, 3.76; N, 4.85; S, 11.34.
87). Anal. Calcd for C16H14NSO3F3: C, 53.78; H, 3.95; N, 3.92; S,
8.97. Found: C, 53.97; H, 3.66; N, 4.01; S, 9.18.
1,2-Dimethyl-3,4-dihydroquinolizinium Triflate (10q). Fol-
lowing the general procedure, after heating for 19 h and purified
by chromatography, 37.3 mg (61%) of 10q was obtained as a gray
needles: mp 124-125 °C; IR (KBr) νmax (cm-1) 3081, 1570, 1512,
1258, 1148, 1032, 793; 1H NMR (300 MHz, acetone-d6) δ 8.89 (d,
1H, J ) 6.3 Hz), 8.57 (td, 1H, J ) 1.0, 8.3 Hz), 8.09 (d, 1H, J )
8.2 Hz), 7.92 (t, 1H, J ) 6.5 Hz), 4.83 (t, 2H, J ) 7.5 Hz), 2.88
(t, 2H, J ) 7.7 Hz), 2.22 (t, 3H, J ) 0.8 Hz), 2.18 (s, 3H); 13C
NMR (75 MHz, acetone-d6) δ 150.3, 146.6, 145.9, 145.1, 124.2,
123.3, 121.7 (c, J ) 320.5 Hz), 121.5, 53.5, 20.0, 13.3; MS (ESI+)
m/z (relative intensity) 160 (M+, 100). Anal. Calcd for
C12H14NSO3F3: C, 46.60; H, 4.56; N, 4.53; S, 10.37. Found: C,
46.49; H, 4.51; N, 4.48; S, 10.22.
2-(But-3-enyl)pyridine (19). To a solution of 2-picoline (0.186
g, 2 mmol) in dry THF (2 mL), under argon atmosphere at -78
°C, was added tert-butyllithium (1.235 mL, 2.1 mmol), and the
reaction mixture was stirred for 1 h. Then, allyl bromide was added
dropwise over the cooled solution, and the reaction mixture was
stirred at -78 °C for 30 min and left to stand overnight at room
temperature. Then, the mixture was extracted with water (5 mL)
and EtOAc (2 × 5 mL), the organic phase dried over MgSO4, and
the solvent evaporated under reduced pressure. Compound 19 was
isolated as a yellow oil by flash chromatography on silica gel using
hexane/EtOAc (8:2) as eluent (0.215 g, 81%): IR (NaCl) νmax (cm-1
)
1
2925, 1591, 1474, 1435; H NMR (300 MHz, acetone-d6) δ 8.47
(d, 1H, J ) 4.8 Hz), 7.64 (td, 1H, J ) 1.8, 7.5 Hz), 7.21 (d, 1H,
J ) 7.7 Hz), 7.14 (t, 1H, J ) 4.9 Hz), 5.93-5.80 (m, 1H),
5.06-4.98 (m, 2H), 2.83 (t, 2H, J ) 7.3 Hz), 2.50-2.42 (m, 2H);
13C NMR (200 MHz, acetone-d6) δ 162.0, 149.9, 138.9, 136.8,
123.4, 121.7, 115.0, 38.1, 34.2; HRMS (DIP/QI) m/z calcd for
C9H12N: 134.0968. Found: 134.0970.
2-(But-3-enyl)-1-(2-chloroethyl)pyridinium Triflate (20). To
a solution of 2-chloroethanol (0.105 g, 1.3 mmol) in dry CCl4 (1
mL), under argon atmosphere, was added dry pyridine (0.103 g,
1.3 mmol), and the reaction mixture was stirred at room temperature
for 5-10 min. Then, this solution was added dropwise (5-10 min)
over a cooled solution (-10 °C) of triflic anhydride (0.327 g, 0.219
mL, 1,3 mmol) in dry CCl4 (1.5 mL). The resulting white solid
formed was filtered off through sodium sulfate, and the solution
was added via cannula to a solution of 2-(but-3-enyl)-pyridine 19
(1 mmol) in dry CCl4 (1.5 mL). The reaction mixture was stirred
at room temperature for 24 h. Removal of the solvent under reduced
pressure and purification by flash chromatography on silica gel using
CH2Cl2/MeOH (9.5:0.5) gave 20 (0.307 g, 89%) as a yellow oil:
IR (NaCl) νmax (cm-1) 3087, 1632, 1512, 1454, 1261, 1158, 1030,
779; 1H NMR (200 MHz, acetone-d6) δ 9.14 (d, 1H, J ) 6.4 Hz),
8.68 (td, 1H, J ) 1.3, 9.1 Hz), 8.24 (d, 1H, J ) 8.0 Hz), 8.14 (t,
1H, J ) 6.4 Hz), 6.03-5.89 (m, 1H), 5.30 (t, 2H, J ) 5.7 Hz),
5.17-5.04 (m, 2H), 4.35 (t, 2H, J ) 5.8 Hz), 3.51 (t, 2H, J ) 7.5
Hz), 2.72-2.64 (m, 2H); 13C NMR (300 MHz, acetone-d6) δ 159.6,
147.4, 147.1, 136.4, 130.3, 126.6, 117.4, 58.8, 43.6, 32.5, 32.4;
MS (ESI+) m/z 196 (M+), 198 (M + 2). Anal. Calcd for
C12H15NClSO3F3: C, 41.68; H, 4.37; N, 4.05; S, 9.27. Found: C,
41.37; H, 4.59; N, 4.31; S, 9.15.
2-(But-3-enyl)-1-vinylpyridinium Triflate (9a). To a solution
of 20 (0.086 g, 0.25 mmol) in EtOH/MeOH (3:1) (5 mL) was added
dropwise at -10 °C for 10 min a solution of aqueous NaOH (10
N, 0.275 mmol, 1.1 equiv). The reaction mixture was stirred for
15 min, neutralized with acetic acid, and the solvent removed under
reduced pressure at 20 °C. The residue was filtered off, and the
solution was concentrated under reduced pressure at 20 °C and
purified by flash chromatography on silica gel using CH2Cl2/MeOH
(9.5:0.5) as eluent. Compound 9a (49.0 mg, 64%) was obtained as
a brown oil: IR (NaCl) νmax (cm-1) 3493, 1627, 1497, 1276, 1158,
1031, 787; 1H NMR (300 MHz, acetone-d6) δ 9.07 (d, 1H, J ) 6.2
Hz), 8.69 (td, 1H, J ) 1.5, 8.0 Hz), 8.19 (d, 1H, J ) 8.0 Hz), 8.13
(t, 1H, J ) 7.5 Hz), 7.89 (c, 1H, J ) 8.2 Hz), 6.22 (dd, 1H, J )
2.2, 14.8 Hz), 6.07 (dd, 1H, J ) 2.0, 8.2 Hz), 5.99-5.85 (m, 1H),
9-Bromo-3,4-dihydroquinolizinium Triflate (10b). Following
the general procedure, after stirring for 1 h, 57.6 mg (80%) of 10b
was obtained as a pale-brown solid: mp 145-147 °C (CH2Cl2/Et2O);
IR (NaCl) νmax (cm-1) 3084, 1632, 1477, 1262, 1149, 1029, 784;
1H NMR (300 MHz, CDCl3) δ 9.03 (d, 1H, J ) 5.8 Hz), 8.86 (d,
1H, J ) 8.2 Hz), 7.94 (t, 1H, J ) 8.1 Hz), 7.27-7.17 (m, 2H),
4.97 (t, 2H, J ) 7.5 Hz), 3.02-2.85 (m, 2H); 13C NMR (75 MHz,
acetone-d6) δ 150.1, 147.2, 146.2, 143.8, 126.3, 121.7, 121.2, 55.4,
22.8; MS (ESI+) m/z (relative intensity) 211 (M+), 213 (M + 2).
Anal. Calcd for C10H9NBrSO3F3: C, 33.35; H, 2.52; N, 3.89. Found:
C, 33.12; H, 2.74; N, 3.51.
1-Methyl-3,4-dihydroquinolizinium Triflate (10k). Following
the general procedure, after stirring for 2 h and purified by
precipitation, 53.9 mg (91%) of 10k was obtained as a white solid:
mp 130-132 °C (CH2Cl2/Et2O); IR (NaCl) νmax (cm-1) 3076, 1620,
1510, 1442, 1262, 1149, 1031, 736; 1H NMR (300 MHz, acetone-
d6) δ 8.99 (d, 1H, J ) 6.2 Hz), 8.64 (t, 1H, J ) 7.9 Hz), 8.13 (d,
1H, J ) 8.2 Hz), 8.02 (t, 1H, J ) 7.1 Hz), 6.77 (t, 1H, J ) 4.6
Hz), 4.88 (t, 2H, J ) 7.5 Hz), 2.85 (dd, 2H, J ) 4.6, 7.6 Hz), 2.27
(s, 3H); 13C NMR (75 MHz, acetone-d6) δ 147.0, 146.4, 136.4,
128.3, 126.2, 124.2, 54.7, 22.8, 18.1; MS (ESI+) m/z (relative
intensity) 146 (M+, 100), 147 (M + 1, 25), 441 (2M + OTf, 14).
Anal. Calcd for C11H12NSO3F3: C, 44.74; H, 4.10; N, 4.74; S, 10.86.
Found: C, 44.56; H, 4.32; N, 4.65; S, 10.58.
2-Phenyl-3,4-dihydroquinolizinium Triflate (10m). Following
the general procedure, after heating for 3 h and purified by
chromatography, 49.3 mg (69%) of 10m was obtained as a green
solid: mp 122-123 °C; IR (NaCl) νmax (cm-1) 2924, 1504, 1257,
1225, 1161, 1029, 774; 1H NMR (300 MHz, acetone-d6) δ 8.97 (d,
1H, J ) 6.2 Hz), 8.57 (td, 1H, J ) 1.1, 7.9 Hz), 8.09 (d, 1H, J )
8.0 Hz), 7.96 (td, 1H, J ) 1.1, 7.5 Hz), 7.88-7.85 (m, 2H),
7.55-7.51 (m, 4H), 5.08 (t, 1H, J ) 7.7 Hz), 3.45 (td, 1H, J )
0.9, 8.0 Hz); 13C NMR (75 MHz, acetone-d6) δ 150.5, 146.4, 145.5,
136.9, 131.5, 129.8, 127.2, 127.0, 125.5, 116.8, 113.1, 54.2, 25.4;
MS (ESI+) m/z (relative intensity) 208 (M+, 100), 209 (M + 1,
J. Org. Chem. Vol. 74, No. 11, 2009 4175

Nun˜ez et al.
5.15-5.02 (m, 2H), 3.37 (t, 2H, J ) 7.5 Hz), 2.66-2.58 (m, 2H);
13C NMR (200 MHz, acetone-d6) δ 147.6, 145.5, 137.2, 136.4,
129.3, 126.8, 121.0, 117.2, 32.8, 31.7; HRMS (ESI+) m/z calcd
for C11H14N: 160.1126. Found: 160.1122.
From 10a: A mixture of 10 (25 mg) and Pd/C 10% (40%, 10
mg) was heated at 200 °C for 2 h (monitored by NMR). The
reaction mixture was worked up as above to give 19.6 mg (79%)
of 7a as yellow crystals: mp 148-150 °C (CH2Cl2/Et2O); IR (NaCl)
1
νmax (cm-1) 3091, 1651, 1264, 1143, 1031, 801; H NMR (300
1,2-Dihydroquinolizinium Triflate (8a). To solution of the diene
9a (0.2 mmol) in dry CH2Cl2 (1.5 mL) was added ruthenium catalyst
G-II (5 mol %) in CH2Cl2 (1 mL) under argon atmosphere. The
reaction mixture was stirred for 3 h at room temperature. Then,
the solvent was evaporate under reduced pressure and the residue
purified by flash chromatography (eluent: CH2Cl2/MeOH 9.5:0.5)
to give 29.0 mg (52%) of 8a as a brown solid: mp 153-155 °C;
IR (KBr) νmax (cm-1) 2434, 1628, 1502, 1264, 1144, 1030, 791;
1H NMR (300 MHz, acetone-d6) δ 8.96 (d, 1H, J ) 6.2 Hz), 8.57
(td, 1H, J ) 1.3, 7.9 Hz), 8.13-8.05 (m, 2H), 7.58 (d, 1H, J ) 7.7
Hz), 6.60-6.58 (m, 1H), 3.52 (t, 2H, J ) 8.2 Hz), 2.69-2.61 (m,
2H); 13C NMR (200 MHz, acetone-d6) δ 146.7, 142.8, 130.6, 129.1,
128.1, 126.9, 26.9, 18.1; HRMS (ESI+) m/z calcd for C9H10N:
132.0813. Found: 132.0806.
Synthesis of Quinolizinium Triflates 7. General Procedure
A: A solution of the corresponding 3,4-dihydroquinolizinium salt 10
(25 mg) in acetic acid (2 mL) and Pd/C 10% (40%,
10 mg) was heated under reflux for the appropriate time. The reaction
mixture was concentrated in vacuo; acetone was added to the residue,
and the mixture was filtered through Celite. The solvent was removed
under reduced pressure, and the quinolizinium salt was purified by
recrystallization.
General Procedure B: A mixture of 10 (25 mg) and Pd/C 10%
(40%, 10 mg) was heated at 200 °C for 0.5-5 h. The quinolizinium
salt was isolated as in procedure A and purified by recrystallization
or by flash column chromatography.
Quinolizinium triflate (7a). From 8a: A mixture of 1,2-
dihydroquinolizinium triflate 8a (25 mg) and Pd/C 10% (40%, 10
mg) was heated at 200 °C for 2 h (monitored by NMR). Then,
acetone was added to the reaction mixture and the resulting
suspension was filtered through Celite. Removal of the solvent under
reduced pressure afforded 7a as a pale-yellow solid that was purified
by flash chromatography on silica gel [CH2Cl2/MeOH (9:1)] or by
recrystallization (23.8 mg, 96%).
MHz, acetone-d6) δ 9.47 (d, 2H, J ) 6.8 Hz), 8.68 (d, 2H, J ) 8.6
Hz), 8.50 (td, 2H, J ) 0.9, 7.1 Hz), 8.20 (t, 2H, J ) 6.9 Hz); 13C
NMR (75 MHz, acetone-d6) δ 143.5, 137.4, 137.3, 126.6, 124.3,
121.5 (c, J ) 319.6 Hz); MS (ESI+) m/z (relative intensity) 130
(M+, 100). Anal. Calcd for C10H8NSO3F3: C, 43.01; H, 2.89; N,
5.02; S, 11.48. Found: C, 43.29; H, 2.54; N, 5.08; S, 11.37.
1,2-Dimethylquinolizinium Triflate (7q). Following the general
procedure, after heating for 24 h, 7q (14.0 mg, 57%) was obtained
as a brown oil: IR (NaCl) νmax (cm-1) 2918, 1639, 1407, 1278,
1
1159, 1031, 638; H NMR (300 MHz, acetone-d6) δ 9.34 (d, 1H,
J ) 6.8 Hz), 9.23 (d, 1H, J ) 7.2 Hz), 8.72 (d, 1H, J ) 8.9 Hz),
8.43 (t, 1H, J ) 7.2 Hz), 8.11-8.02 (m, 2H), 2.82 (s, 3H), 2.76 (s,
3H); 13C NMR (300 MHz, acetone-d6) δ 148.6, 137.9, 137.3, 135.1,
133.7, 127.1, 124.5, 123.4, 55.4, 20.9, 14.8; MS (ESI+) m/z 158
(M+). Anal. Calcd for C12H12NSO3F3: C, 46.75; H, 4.25; N, 4.54;
S, 10.40. Found: C, 46.66; H, 4.30; N, 4.70; S, 10.12.
Acknowledgment. We thank the Spanish Ministerio de
Ciencia e Innovacio´n (projects CTQ2005-011060 and CTQ2008-
04313) for financial support and acknowledge grants from the
Comunidad de Madrid (A.N.) and Universidad de Alcala´ (B.A.).
Note Added after ASAP Publication. Due to a production
error Scheme 1 was incorrect, the correct Scheme 1 was
published ASAP on May 6, 2009.
Supporting Information Available: Experimental proce-
dures and full spectroscopic data and NMR spectra for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
JO900292B
4176 J. Org. Chem. Vol. 74, No. 11, 2009