Cycloaddition products of 3-oxido-1-phenylpyridinium and 1-cyanoacenaphthylene

Article abstract of DOI:10.1039/a709091i

Cycloaddition of 3-oxido-1-phenylpyridinium 8a to 1-cyanoacenaphthylene 9a affords three isomeric adducts (13, 14 and 15). Structures for these compounds resulted from a comparative NMR study, as well as an X-ray crystallographic determination of isomer 15. Attempts to eliminate HCN from compounds 13-15 resulted only in cycloreversion to 1-cyanoacenaphthylene.

Full text of DOI:10.1039/a709091i

View Article Online / Journal Homepage / Table of Contents for this issue
Cycloaddition products of 3-oxido-1-phenylpyridinium and
1-cyanoacenaphthylene
Yvette A. Jackson,*,† Lillian M. Rogers, Robin D. Rogers and Michael P. Cava
The University of Alabama, Department of Chemistry, Box 870336, Tuscaloosa,
AL 35487-0336, USA
Cycloaddition of 3-oxido-1-phenylpyridinium 8a to 1-cyanoacenaphthylene 9a affords three isomeric
adducts (13, 14 and 15). Structures for these compounds resulted from a comparative NMR study,
as well as an X-ray crystallographic determination of isomer 15. Attempts to eliminate HCN from
compounds 13–15 resulted only in cycloreversion to 1-cyanoacenaphthylene.
a potential pathway to compound 7, the simple carbocyclic
analog of these alkaloids.
Introduction
Imerubrine 11 and grandirubrine 22—isolated in 1975 and
1980, respectively, from the tropical American genus Abuta
(Menispermaceae)—were for many years the only known
members of a class of naturally occurring tropoloisoquinoline
alkaloids.
The reaction of 3-oxido-1-phenylpyridinium 8a (prepared in
situ from reaction of 3-hydroxy-1-phenylpyridinium chloride
8 and triethylamine), with 2π-electron dipolarophiles to form
8-azabicyclo[3.2.1]oct-3-en-2-ones as cycloadducts, has been
studied in depth by Katritzky and his co-workers (Scheme 1).6
Recently, other members of the class, namely isoimerubrine
3, pareirubrines A and B (4 and 5) and pareitropone 63 have
been isolated. Banwell and his group4 have reported the total
synthesis of compounds 1 and 2, and subsequently, Boger and
X
6
O–
O
OH
8
N
1
5
+
+
N
+
N
Ph
O
base
–
N
Cl–
3
OMe
OMe
OMe
Ph
Ph
Ph
MeO
MeO
MeO
MeO
MeO
MeO
8
8a
8b
N
N
N
Scheme 1
Various dipolarophiles (e.g. styrene, acrylonitrile, diethyl
maleate and phenylacetylene) and pyridinium salts have been
employed. Hofmann degradation of the quaternized cycload-
ducts usually produces tropones.7
O
OH
OMe
OMe
O
O
One of our approaches to the acenaphthenotropone system
was based on the electronically directed dipolar addition of
the pyridinium dipole 8a to the nitrile 9a (Scheme 2). We
1
3
2
OMe
MeO
MeO
MeO
MeO
MeO
MeO
N
N
N
+
8a
NC
H
OMe
N
N
Ph
O
Ph
O
O
O
X
OH
OH
O
9a: X = CN
b: X = OTf
c: X = Br
11
10
6
4
5
7
+ PhNO
N
+
–O
Ph
O
O
7
12
Scheme 2
Takahashi5 have synthesized compounds 1, 2 and 3. These are
the only total syntheses of the tropoloisoquinoline alkaloids
that have so far been reported. In our efforts to develop an
alternate route to the tropoloisoquinoline system, we explored
anticipated that treatment of the resultant enone 10 with a
strong protic base would result in elimination of hydrogen cyan-
ide to form 11. The compound 11 could then be oxidized to an
unstable N-oxide 12, which could collapse to nitrosobenzene
and the desired tropone skeleton. There are many cases of
analogous aromatization via oxidation of a bridged nitrogen
† Permanent address: Department of Chemistry, The University of the
West Indies, Kingston 7, Jamaica, West Indies.
J. Chem. Soc., Perkin Trans. 1, 1998
1865

View Article Online
Table 1 Typical coupling constants for protons of cycloadducts
and, in fact, we have found this reaction to be very useful in
some of our own previous work.8
Protons
J/Hz
Protons
J/Hz
1,3
0–1.5
0–1.0
7.5–8.5
9.5–10.0
5.0–6.0
0.0
6-exo, 6-endo
6-endo, 7-endo
6-endo, 7-exo
6-exo, 7-exo
6-exo, 7-endo
7-endo, 7-exo
13.0
9.5–10.0
4.0
9.0–11.0
6.0–7.0
14.0–14.5
Results and discussion
1,7-endo
1,7-exo
3,4
4,5
5,6-endo
5,6-exo
Acenaphthylene-1-carbonitrile 9a was obtained from readily
available acenaphthylene in 50% yield, by the bromination–
dehydrobromination and halide exchange sequence outlined in
Scheme 3.9,10
5.8–7.0
Ph
Br2, CCl4
DBU
CuCN
8 N
9a
O
1
O
_
7
6
+
Br
Br
Br
N
5
3
9c
Ph
4
X
Scheme 3
17
X
We also obtained 9a by conversion of acenaphthenone to
the unsaturated triflate 9b using N-phenyltrifluoromethane-
sulfonamide11 followed by treatment with LiCN according to
the method of Piers (Scheme 4).12 In our hands, the triflate was
Scheme 5
R
O
N
_
+
N
X
1
O
7
KCN, Pd(PPh3)4
1. LDA
R
9a
5
3
2. Ph-N(Tf)2,
20 h, rt
6
18-crown-6, ∆
4
TfO
O
X
18: R = 5-phenyl-1,2,4-triazin-3-yl, X = CN
19: R = 2,4-dinitrophenyl, X = CN
20: R = Me, X = CO2Me
9b
A
Scheme 4
21: R = -N=CHPh, X = CN
obtained in 73% yield and conversion to the nitrile required
use of KCN, 18-crown-6 and catalytic amounts of tetrakis-
(triphenylphosphine)palladium. The overall yield from ace-
naphthenone using this pathway, however, was only 38%.
The dipolarophile 9a, on refluxing with 3-hydroxy-1-
phenylpyridinium chloride (1.07 molar equivs.) and triethyl-
amine (1.6 molar equivs.) for eighteen hours, produced the
corresponding cycloadducts 13, 14 and 15 as a mixture in the
ratio 1:8:3. In our hands, none of the compound 16 was
isolated.
shown in A, indicating that activation of the dipolarophile by
electron-withdrawing groups is not a requirement for reaction.
Compounds 13 and 14 are the endo- and exo-adducts, respect-
ively, of the cycloaddition according to the normal mode of
reaction shown in Scheme 5. Compound 15 is the exo-adduct of
the reaction in the less common mode shown as A.
Katritzky and his group compared the proton NMR spectra
of exo- and endo-isomers of many of these cycloadducts.6a,6c,14
Table 1 above shows the characteristic coupling constants
observed.
The proton NMR data for our new adducts are shown in
Table 2. Structural assignments were confirmed by double-
resonance experiments. We were able to confirm the stereo-
chemistry of our products by comparison of coupling con-
stants, and by decoupling experiments. The structure of 15 was
unambiguously established by single crystal X-ray determin-
ation as the exo-adduct of cycloaddition in the less common
mode A (Fig. 1). The data immediately disqualifies adduct 16,
since, on referring to the values listed in Table 1, H-5 here
should be a double doublet, coupled to H-6-exo and to H-4
with J values of ~6–7 and ~5–6 Hz respectively. None of the
products showed this feature. The endo-adduct 13 had the
resonance due to H-1 as its distinguishing feature. This showed
up as a double doublet with coupling due to H-3 (J value ~1
Hz) and to H-7-exo (J value of ~8 Hz).
In the NMR spectrum of 15, H-1 appears as a broadened
singlet, with fine splitting due to H-3. This was confirmed by
irradiation at δ 6.22. H-3 appears as a double doublet coupled
with H-4 and H-1, and H-4 appears as part of a two hydrogen
multiplet in the aromatic region, which signal is distinguished
on irradiation at δ 6.22 and 4.86. H-5 appears as a doublet
(coupling with H-4) and H-6 as a singlet at δ 4.69. As shown in
Table 2, the NMR spectra of compounds 14 and 15 are very
similar. With single crystal X-ray establishing the structure of
15, the structure of 14 was thus confirmed, by elimination, to be
the exo-adduct of cycloaddition according to the normal mode
(Scheme 5).
Ph
H
Ph
H
N
N
H
O
H
O
H
7
1
H
5
NC
3
6
H
H
NC
H
H
13
14
Ph
H
Ph
H
N
N
NC
O
H
O
H
NC
H
H
H
H
H
H
15
16
For addition at the 2,6-position of the dipole, the usual
mode of reaction is as shown in Scheme 5, to produce the
6-substituted adduct 17. In their study of peri-, site-, regio-
and stereo-selectivity of these reactions,13 Katritzky and his
group indicate that 6-substituted regioisomers 17 are favored
for monosubstituted ethylenes. Only very few instances of
addition to produce 7-substituted compounds of the type 15
have been found. These instances include formation of com-
pounds 18–216a,14–16 and arise by reaction in the manner
1866
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
Table 2 Proton NMR spectral data for cycloadducts 13–15, 22 and 23
Chemical shifts (δ) (multiplicity, J/Hz)
Compound
H-1a
H-3
H-4
H-5
H-6
H-7
13
14
15
22
23
5.10 (dd, ~1, 7)
4.54 (s)
4.73 (bs)
4.97m
4.57 (d, ~1)
5.28 (dd, ~1, 9.5)
6.31 (dd, 1, 10)
6.22 (dd, 1, 10)
5.27 (dd, 1.5, 9)
6.05 (dd, 1, 9)
6.32 (dd, 5, 9.5)
5.38 (d, 5)
5.08 (d, 5)
4.86 (d, 5)
5.18 (dd, 5, 7)
4.86 (d, 5)
—
—
5.32 (d, 7)
4.55 (s)
—
4.97m
4.21 (d, 7)
7.57 (m with arom. H’s)
7.48 (m with arom. H’s)
6.49 (dd, 5, 9)
4.69 (s)
4.87 (t, ~7)
4.34 (d, 7)
7.44 (m with arom. H’s)
a Numbering of protons (see 13) is non-systematic. Used to coincide with positions described in Table 1, which we use for comparison.
chemically,18 and by electron impact.19 The lack of molecular
ions in the mass spectra of adducts 13–15 (obtained via
electron-impact techniques), and the appearance of base peaks
at m/z 177 (corresponding to Mϩ for 1-cyanoacenaphthylene
9a), provides further indication of the facility with which the
undesired cycloreversion takes place.
We also attempted the cycloaddition reaction using
acenaphthylen-1-yl trifluoromethanesulfonate 9b, 1-bromo-
acenaphthylene 9c and acenaphthylene, as the dipolarophiles.
Compounds 9b and 9c both proved unreactive, whereas
acenaphthylene produced a 17% yield of the cycloadduct as
a 1:1 mixture of two isomers (22 and 23) after 23 hours
at reflux. Starting material (82%) was recovered from the
reaction.
Ph
H
Ph
H
N
N
H
O
H
O
H
H
H
H
H
H
H
H
23
22
Again, examination of the NMR data of the adducts led
to the establishment of their structures. Double-resonance
experiments were used to confirm the assignments of protons.
Note the shielding of H-3 and H-4 in compound 22 as com-
pared with 23 (δ 5.27 and 6.49 as against δ 6.05 and 7.44),
and the signal for H-5 appearing as a double doublet with J = 5
and 7.2 Hz (coupled with H-4 and H-6 respectively), which was
ready evidence of 22 being the endo-adduct. (From Table 1,
J5,6-endo = 0.) All other signals were in accordance with these
assignments.
Fig.
1
Molecular structure of 15 showing the crystallographic
numbering system used
The proton NMR spectrum of the endo-adduct 13 (acenaph-
thylene ring down) is different from that of the exo-adduct 14 in
characteristic ways.
Experimental
a) In 13, H-1 appears as a double doublet with J values of 7
and 1.7 Hz being due to coupling with H-7-exo and H-3,
respectively, whereas in 14, H-1 appears as a singlet.
b) In 13, H-7 appears as a doublet coupled to H-1, whereas
in 14, H-7 appears as a singlet.
c) Signals corresponding to H-3 and H-4 appear at δ 5.28
and 6.32 respectively in the spectrum of compound 13. In
the spectra of 14 and 15, the signals for these protons
appear at δ ~6.31 and ~7.57, respectively. This is no doubt
due to the fact that in the case of 13, the protons H-3 and
H-4 reside within the shielding zone of the acenaphthyl-
ene ring system.
General
All mps are uncorrected. NMR spectra (Bruker 360 MHz
spectrometer) were determined in CDCl3 solution and the
resonances are reported in δ units downfield from TMS.
Elemental analyses were carried out by Atlantic Microlabs,
Atlanta, Georgia.
Acenaphthylene-1-carbonitrile 9a
A mixture of 1-bromoacenaphthylene (25 g, 108 mmol) and
CuCN (17.5 g, 195 mmol) in N,N-dimethylacetamide (105 cm3)
was heated at reflux for 2 h. The mixture was diluted with
CH2Cl2 (~800 cm3) whereupon CuBr precipitated. This was
filtered, and the filter cake washed with CH2Cl2. The CH2Cl2
solution was then washed with water (3 × 150 cm3), dried
(Na2SO4) and concentrated to give a maroon-red liquid which
was purified by flash chromatography (SiO2, hexane–CH2Cl2
5:1) to give the desired nitrile as a yellow solid (12.02 g, 63%),
mp 51–52 ЊC (lit.,9 mp 53 ЊC).
Unfortunately, our attempts at removal of HCN from
these adducts using potassium tert-butoxide or methoxide,
were unsuccessful. Heating the cycloadduct in triethylamine
resulted only in partial (35%) conversion to the nitrile 9a, evi-
dence of thermally induced retro-1,3-dipolar cycloaddition.6d,17
1,3-Dipolar additions are also known to be reversible photo-
J. Chem. Soc., Perkin Trans. 1, 1998
1867

View Article Online
Cycloadducts 13, 14 and 15
was concentrated and chromatographed (SiO2, hexane) to yield
A mixture of the nitrile 9a (12.02 g, 67.91 mmol) and 3-
hydroxy-1-phenylpyridinium chloride 8 (15 g, 72.46 mmol) in
acetonitrile (280 cm3) was heated at reflux with stirring, in an
atmosphere of nitrogen, for 30 min. Triethylamine (15 cm3,
108.7 mmol) was then added to the mixture and heating con-
tinued at reflux for a further 18 h after which the solvent was
removed in vacuo. The resultant solid was dissolved in CH2Cl2,
washed with water, and the CH2Cl2 solution dried (Na2SO4)
and concentrated to yield a very dark oil. Flash chroma-
tography (SiO2, hexane–CH2Cl2 2:1) afforded the starting
nitrile 9a (7.84 g, 65%). Gradually increasing the polarity of the
solvent system to hexane–CH2Cl2 1:1, produced a mixture of
compounds 13 and 14 (3.75 g), and a second band containing
compound 15 as major product.
acenaphthylene (0.82 g, 82%). Increasing the polarity of the
solvent to hexane–EtOAc 3:1 produced the cycloadducts 22
and 23 as a mixture, a dark brown oil (0.37 g).
Further chromatography of the oil (SiO2, hexane–EtOAc
10:1) yielded both compounds 22 and 23 as yellow crystals.
Compound 22 (148 mg, 6.9%), mp 186–188 ЊC; m/z (relative
intensity) 323 (0.5), 171 (9), 152 (100); (HRMS calcd. for
C23H17NO: 323.1310; found: 323.1307).
Compound 23 (165 mg, 7.7%), mp 159–160 ЊC; m/z (relative
intensity) 323 (0.5), 171 (11), 152 (100); (HRMS calcd. for
C23H17NO: 323.1310; found: 323.1302).
References
1 (a) M. P. Cava, K. T. Buck, I. Noguchi, M. Srinivasan, M. G. Rao
and A. I. DaRocha, Tetrahedron, 1975, 31, 1667; (b) J. V Silverton,
C. Kabuto, K. T. Buck and M. P. Cava, J. Am. Chem. Soc., 1977,
99, 6708.
2 M. D. Menachery and M. P. Cava, Heterocycles, 1980, 14, 943.
3 (a) H. Morita, K. Matsumoto, K. Takeya, H. Itokawa and Y. Iitaka,
Chem. Pharm. Bull., 1993, 41, 1418; (b) H. Morita, K. Matsumoto,
K. Takeya, H. Itokawa and Y. Iitaka, Chem. Lett., 1993, 339; (c)
H. Morita, K. Matsumoto, K. Takeya and H. Itokawa, Chem.
Pharm. Bull., 1993, 41, 1478; (d) H. Morita, K. Takeya and
H. Itokawa, Biorg. Med. Chem. Lett., 1995, 5, 597; (e) H. Itokawa,
K. Matsumoto, H. Morita and K. Takeya, Heterocycles, 1994, 37,
1025.
Recrystallization of the first batch (3.75 g) of product from
CH2Cl2–MeOH produced 13 as fluffy yellow needles (0.27 g,
1.1%), mp 178 ЊC; m/z (relative intensity) 177 (100), 171 (10),
150 (26) (Found: C, 82.6; H, 4.6; N, 8.0. Calc. for C24H16N2O:
C, 82.7; H, 4.6; N, 8.0%).
The mother liquor was concentrated and recrystallized from
MeOH to give 14 as a yellow powder (2.01 g, 8.5%), mp 146 ЊC;
m/z (relative intensity) 177 (100), 171 (70), 150 (65) (Found: C,
82.55; H, 4.6; N, 8.0. Calc. for C24H16N2O: C, 82.7; H, 4.6; N,
8.0%).
The second band was further purified by chromatography
(SiO2, hexane–EtOAc 2:1) and compound 15 was obtained as
yellow crystals (0.83 g, 3.5%), mp 168–169 ЊC; m/z (relative
intensity) 177 (100), 171 (60), 150 (80) (Found: C, 82.8; H, 4.6;
N, 8.0. Calc. for C24H16N2O: C, 82.7; H, 4.6; N, 8.0%).
4 (a) M. G. Banwell, E. Hamel, N. K. Ireland and M. F. Mackay,
Heterocycles, 1994, 39, 205; (b) M. G. Banwell and N. K. Ireland,
J. Chem. Soc., Chem. Commun., 1994, 591.
5 D. L. Boger and K. Takahashi, J. Am. Chem. Soc., 1995, 117, 12 452.
6 (a) A. R. Katritzky, A. Boonyarakvanich and N. Dennis, J. Chem.
Soc., Perkin Trans. 1, 1980, 2, 343; (b) N. Dennis, A. R. Katritzky,
T. Matsuo, S. K. Parton and Y. Takeuchi, J. Chem. Soc., Perkin
Trans. 1, 1974, 746; (c) N. Dennis, A. R. Katritzky, S. K. Parton,
Y. Nomura, Y. Takahashi and Y. Takeuchi, J. Chem. Soc., Perkin
Trans. 1, 1976, 2289; (d) J. Banerji, N. Dennis, J. Frank, A. R.
Katritzky and T. Matsuo, J. Chem. Soc., Perkin Trans. 1, 1976, 2334;
(e) N. Dennis, A. R. Katritzky and Y. Takeuchi, Angew. Chem., Int.
Ed. Engl., 1976, 15, 1.
7 A. R. Katritzky and Y. Takeuchi, J. Chem. Soc. C, 1971, 878.
8 M. P. Cava, M. Sprecker and W. Hall, J. Am. Chem. Soc., 1974, 96,
1817.
9 J. Nakayama, E. Oshima, A. Ishii and M. Hoshino, J. Org. Chem.,
1983, 48, 60.
10 A. Castellan, G. Dumartin, R. Galante and H. Bouas-Laurent,
Bull. Soc. Chim. Fr., 1976, 217.
11 J. McMurry and W. Scott, Tetrahedron Lett., 1983, 24, 979.
12 E. Piers and F. Fleming, J. Chem. Soc., Chem. Commun., 1989, 756.
13 A. R. Katritzky, N. Dennis, M. Chaillet, C. Larrieu and M. El
Mouhtadi, J. Chem. Soc., Perkin Trans 1, 1978, 408.
14 N. Dennis, B. Ibrahim, A. R. Katritzky, I. G. Taulov and
Y. Takeuchi, J. Chem. Soc., Perkin Trans. 1, 1974, 1883.
15 A. R. Katritzky, J. Banerji, N. Dennis, J. Ellison, G. J. Sabonji
and E. Wurthwein, J. Chem. Soc., Perkin Trans. 1, 1979, 2528.
16 A. R. Katritzky, J. Banerji, A. Boonyarakvanich, A. T. Cutler,
N. Dennis, A. S. Q. Rizvi, G. J. Sabonji and H. Wilde, J. Chem.
Soc., Perkin Trans. 1, 1979, 2, 399.
X-Ray data collection, structure determination and refinement for
15
A yellow single crystal of 15 was mounted on a fiber and
transferred to the Siemens SMART CCD area detector
diffractometer system. The crystal was cooled to Ϫ100 ЊC
during data collection by using a stream of cold nitrogen gas.
Compound 15 crystallizes in the monoclinic space group Cc
with a = 19.351(12), b = 10.069(6), c = 8.966(6) Å, β = 91.91(1)Њ,
V = 1746.2(2) Å3, µ = 0.082 mmϪ1, and ρcalc = 1.325 Mg mϪ3 for
Z = 4. Graphite monochromated Mo-Kα radiation (λ = 0.7107
Å) was utilized to collect 5400 reflections within the θ range
2.11 to 27.88Њ. Of the total reflections collected, 3352 were
unique (Rint = 0.0196) and 3142 were considered observed
[I > 2σ(I)]. The geometrically constrained hydrogen atoms
were placed in calculated positions and allowed to ride on the
bonded atom with B = 1.2 × Ueqv (C). Refinement of non-
hydrogen atoms was carried out with anisotropic temperature
factors. Structure solution and refinement was accomplished
using SHELXTL, Ver. 5 and led to final residuals of
R1 = 0.0350 and wR2 = 0.0776 based on 245 parameters refined
against all data. The final minimum and maximum residuals in
the final difference Fourier map were 0.242 and Ϫ0.150 e ÅϪ3
.
17 R. Huisgen, H. Hauck, R. Grashey and H. Seidl, Chem. Ber., 1968,
101, 2568.
18 K. Burger and J. Fehn, Tetrahedron Lett., 1972, 1263.
19 Y. Nomura, F. Furusaki and Y. Takeuchi, J. Org. Chem., 1972,
37, 502.
Cycloadducts 22 and 23
Acenaphthylene (1 g, 6.6 mmol) and 3-hydroxy-1-phenyl-
pyridinium chloride 8 (1.36 g, 6.6 mmol) in acetonitrile (25
cm3) were heated at reflux, in an atmosphere of nitrogen, for
30 min. Triethylamine (1.1 cm3, 7.92 mmol) was then slowly
added to the mixture, and heating continued at reflux for a
further 23 h. The black solution was concentrated and the
residue diluted with CH2Cl2, and filtered. The CH2Cl2 solution
Paper 7/09091I
Received 18th December 1997
Accepted 13th March 1998
1868
J. Chem. Soc., Perkin Trans. 1, 1998