Coumarin-enaminoester adducts: Structure corrections (X-ray) and some novel transformations. Synthesis of annulated tricyclic 2-pyridones

Article abstract of DOI:10.1055/s-2004-829081

By means of X-ray crystallographic analysis the correct structure of all three types of stable coumarin-enaminoester adducts 6,10 and 11 and the corresponding spontaneous multistep heterocyclizations of ANRORC-type are reported. The initially formed adduc

Full text of DOI:10.1055/s-2004-829081

1584
LETTER
Coumarin-enaminoester Adducts: Structure Corrections (X-ray) and Some
Novel Transformations. Synthesis of Annulated Tricyclic 2-Pyridones
C
oumarin-enam
y
inoester
A
dd
u
ucts bomir D. Raev,a Wolfgang Frey,b Ivo C. Ivanov*a
a
Faculty of Pharmacy, Medical University of Sofia, Dunav 2, BG-1000 Sofia, Bulgaria
Fax +359(2)9879874; E-mail: ivo43bg@yahoo.co.uk
b
Institut für Organische Chemie der Universität Stuttgart, Pfaffenwaldring 55, 70569 Stuttgart, Germany
E-mail: wolfgang.frey@po.uni-stuttgart.de
Received 19 March 2004
report now the true structure of the coumarin-enami-
noester adducts according to the X-ray data of some stable
end products3 of each type (Figure 1, Figure 2, Figure 3)
as well as some of their simple chemical transformations.
Abstract: By means of X-ray crystallographic analysis the correct
structure of all three types of stable coumarin-enaminoester adducts
6, 10 and 11 and the corresponding spontaneous multistep heterocy-
clizations of ANRORC-type are reported. The initially formed ad-
ducts 3–5 undergo lactone ring opening as well as, in many cases, All of these new results show that we made the wrong
spontaneous transannular intramolecular Michael addition, to give
assumption many years ago.
6, 10 and 11. A classic Michael addition of ethyl malonamate 12 to
3-substituted coumarins 1 followed by a lactone ring opening gives
13 as final product.
Key words: enamines, N-heterocycles, pyridines, pyrans, Michael
addition, X-ray analysis
In a series of previously published papers1a–d the Michael
addition of enaminoesters 2a–c to coumarin (1a), 6-nitro-
coumarin (1f) and their derivatives (1b–e,g–i) with elec-
tron-withdrawing 3-substituents (EWG) was described.
On the basis of their spectral properties (IR, 1H NMR, 13C
NMR, MS) the structures of types 3,4 and 5 (Scheme 1),
or their tautomeric forms, were assumed at that time.
Some of these structures, unfortunately incorrect, also
appeared in Houben-Weyl.2
Figure 1 X-ray structure of 6a at the 50% probability level
Since there has been hesitation concerning the interpreta-
tion of some 1H NMR resonance signals (for example NH
or OH) the assigned structures 3–5 have always remained
doubtful. The following facts played the most important
role for those assignments: There was no absorption band
for any free phenolic OH group in the IR spectra of any of
the adducts which were chloroform-soluble. Instead, there
was a peak in the MS (EI) spectra with m/z equal to the
molecular mass of the corresponding starting enamine 2.
It was possible to interpret this defragmentation as a re-
verse elimination of the 4-substituent from 3–5 to give the
corresponding starting coumarins 1a–i whose molecular
masses were often detectable in the mass spectra (EI).
Hence, we believed that the coumarin skeleton remained
unchanged in the adducts. In some degradation trials we
were even able to isolate the starting coumarin, for in-
stance after heating the adduct with trifluoroacetic acid.
Meanwhile, we succeeded in growing crystals of some of
the coumarin-enaminoester adducts so that they became
suitable for X-ray crystallographic analysis. We wish to
Figure 2 X-ray structure of 10e4 at the 50% probability level
(H-atoms are omitted for greater clarity)
SYNLETT 2004, No. 9, pp 1584–1588
2
2.
0
7.
2
0
0
4
Advanced online publication: 29.06.2004
DOI: 10.1055/s-2004-829081; Art ID: G08504ST
© Georg Thieme Verlag Stuttgart · New York

LETTER
Coumarin-enaminoester Adducts
1585
noates react with their a-carbon to give initially the inter-
mediate adducts 3 and 4, respectively, the aminocrotonate
2c adds predominantly with its g-carbon atom.1a–c It
turned out that the initially formed 1:1 C-adducts of types
3–5 are thermodynamically unstable and undergo in-
tramolecular rearrangements spontaneously through an
ANRORC mechanism to form the 2-pyridones 6–8, of
which 6 and 8 are isolatable. Surprisingly, the intermedi-
ates 7 and 8 cyclize via further intramolecular Michael ad-
dition of the phenolic hydroxyl group to the conjugated
double bond and, as an end result, the corresponding sta-
ble tricycles 10 and 11 are built. The 4-(o-hydroxyphe-
nyl)-2-pyridones 6a,f can be purposely cyclized on
heating, or catalytically, to give benzopyrano[3,4-c]py-
ridines 9a,f whose correct structures have already been re-
ported.1b,c
Figure 3 X-ray structure of 11a4 at the 50% probability level
Addition of three types of 3-aminopropenoates 2a–c was
carried out in our studies (Scheme 1; Table 1). While the
3,3-diamino (2a) and the 3-amino-3-ethoxy (2b) prope-
Compounds with structures similar to 10 or 11, containing
a carbon, oxygen or sulfur 1,3-transannular bridge, were
Scheme 1 Three types of non-isolatable, initially formed, coumarin-enaminoester adducts 3, 4 and 5. They undergo further intramolecular
rearrangements to the stable end products 6, 10 and 11. Compounds 9a,f are obtainable from 6a,f on heating.1a,c
Synlett 2004, No. 9, 1584–1588 © Thieme Stuttgart · New York

1586
L. D. Raev et al.
LETTER
Table 1 Reactants, Corrected Structures (6, 8, 10, 11), Former Structures (3, 4, 5, 8) and Some Experimental Data for the Coumarin-enami-
noester Adducts
Reactants
Product,
correct structure
Product,
Yield,1a–c (%)
Mp,1a–c (°C)
(conditions)4
former structurea
1a + 2a (EtOH, r.t., 5 days)
1b + 2a (EtOH, reflux, 5 h)
1c + 2a (EtOH, r.t., 24 h)
1e + 2a (EtOAc, r.t., 5 h)
1f + 2a (EtOH, r.t., 4 d)
6a (X-ray, Figure 1)
3a (4a1a,2; 31b; 41d)
3a (4a1a,2; 31b; 41d)
65
44
47
49
84
53
58
59
40
67
49
49
8
190–193
192–195
122–125
153–156
Dec.>200
Dec.>200
Dec.>170
155–158
153–155
112–115
146–148
153–155
169–172
167–171
130–133
139–140
153–156
143–145
132–135
6a (X-ray, Figure 1)
6c
3c (4c1a,2
)
6e
3e (4d1a,2
3f (5a1c)
3f (5a1c)
3h (5c1c)
3i (5d1c)
8a (121b)
4c (3c1a)
4d (3e1a)
4e (3d1a)
4f (3c1c)
4f (3c1c)
4h (3e1c)
4i (3d1c)
5a (111b)
)
6f
1g + 2a (EtOH, reflux, 10 h)
1h + 2a (EtOH, r.t., 7 d)
6f
6h
1i + 2a (EtOH, r.t., 14 d)
1b + 2c (EtOH, reflux, 30 h)b
1c + 2b (EtOH, r.t., 48 h)
1d + 2b (EtOH, reflux, 20 h)
1e + 2b (EtOH, reflux, 20 h)
1f + 2b (EtOH, reflux, 35 h)
1g + 2b (EtOH, reflux, 8 h)
1h + 2b (EtOH, r.t., 5 d)
1i + 2b (EtOH, r.t., 2 d)
6i
8a (X-ray)3
10c
10d
10e4 (X-ray, Figure 2)
10f
10f
56
74
84
25
50
8
10h
10i
1b + 2c (EtOH, reflux, 30 h)b
1c + 2c (EtOH, reflux, 52 h)
1i + 2c (EtOH, reflux, 32 h)
11a4 (X-ray, Figure 3)
11c
11i
5c (8 1a
5i (6b1c)
)
a Formula numbers as assigned in the corresponding original paper1a–d are given in parentheses.
b A mixture of products 8a and 11a was obtained from 1b + 2c; separated by column chromatography on silica gel.1b
reported earlier by Claremon et al.5 However, their syn- 13C NMR spectra of the products 6a and 6f, which differ
thesis has been accomplished in a completely different only by the 6-nitro group, are consistent with each other.6,7
way. The authors did not mention whether or not the
transannular nucleophilic addition of the phenolic OH
group occurred spontaneously, as it takes place in 7 or 8
(Scheme 1).
The adducts 6h and 10c could be converted by acidic hy-
drolysis into glutarimide derivatives 13h and 13c, respec-
tively (Scheme 2), whose constitution was confirmed by
means of X-ray analyses. Similarly, acidic hydrolysis of
Most spectral data for the stable products 6a,c,e,f,h,i, 8a, 6f and 10h,i lead to the glutarimides 13f,h,i, which were
10c–f,h,i and 11a,c,i were already published in our previ- independently synthesized by simple Michael addition of
ous communications,1a–c which also describe their de- ethyl malonamate 12 to the coumarins 1c,g,h,i in the
tailed preparation procedures.4 Some experimental data presence of triethylamine.1a,c The initially formed Michael
together with the references for these compounds are sum- adducts 14 (not isolatable) surprisingly undergo spontane-
marized here in Table 1. The structures of the remaining ous intramolecular cyclization to yield the stable end
adducts, which could not be analyzed by means of X-ray products 13 (the intermediate 14g after decarboxylation).
measurements, have been deduced solely from the com- In our previous papers,1a,c the assumed structure 14 was
patibility of their spectral characteristics, above all, from incorrect because we never realized that the amide NH2
their NMR spectra. That is, all compounds of the 6-series group, being a very weak nucleophile, would attack and
1
(Table 1) show analogous H NMR spectra (outside the open the lactone ring. Obviously, the (o-hydroxyphenyl)-
region of the aromatic protons). The same applies to the glutarimide derivatives of type 13 possess greater stability
1
10-series and to the 11-series. For instance, the H- and than 14.
Synlett 2004, No. 9, 1584–1588 © Thieme Stuttgart · New York

LETTER
Coumarin-enaminoester Adducts
1587
three compounds 15a, 16a and 17a, was established by X-
ray analysis,3 which allowed to unambiguously determine
the chemical shifts of the corresponding signals in the 1H
NMR spectra of all five acetyl derivatives.6,7 In our previ-
ous papers1b,d the structure of the O-acetyl derivative 15a
was erroneously ascribed to the N-acetyl compound 16a.
In conclusion, we would like to emphasize that all multi-
step intramolecular transformations of the initial adducts
3–5 into 6,8,10,11 occur spontaneously, apparently under
thermodynamic control. In order to explain the presence
in the MS (EI) spectra of peaks with m/z equal to the mo-
lecular mass of the starting coumarin 1a–i, full reversibil-
ity of these processes (Scheme 1) must be assumed. This
reversibility has also been confirmed by the conversion of
6a,10c into the starting coumarins 1a,c by refluxing with
diluted p-toluenesulfonic or trifluoroacetic acid in anhy-
drous acetonitrile.8
Nonetheless, the Michael addition of enamino esters to
electrophilic coumarins (1a–i) seems to be a useful route
to the tricyclic structures of the types 9–11.
Scheme 2 Preparation of glutarimide derivatives 13.
Acknowledgment
Acknowledgments are due to Dr. Maria Spasova (Memorial Sloan-
Kettering Cancer Center, New York, USA) for recording and dis-
cussing the NMR spectra.
References
(1) (a) Ivanov, I. C.; Raev, L. D. Synth. Commun. 1986, 16,
1679. (b) Ivanov, I. C.; Raev, L. D. Liebigs Ann. Chem.
1987, 1107. (c) Ivanov, I. C.; Raev, L. D. Arch. Pharm.
1995, 328, 53. (d) Raev, L. D.; Voinova, E.; Ivanov, I. C.;
Popov, D. Pharmazie 1990, 45(9), 696.
(2) Kantlehner, W. In Houben-Weyl, 4th ed., Part 2, Vol. E 15;
Kropf, H.; Schaumann, E., Eds.; Thieme: Stuttgart, 1993,
2197.
(3) X-ray Crystal Structure Analyses: atomic coordinates,
bond lengths [Å] and angles [deg], anisotropic displacement
parameters, hydrogen coordinates, torsion angles [deg] have
been deposited at Cambridge Crystallographic Data Centre,
12 Union Road, Cambridge CB2 1EZ, UK, under deposition
numbers CCDC 177460 for 6a, CCDC 177461 for 8a,
CCDC 177459 for 10e, CCDC 181931 for 11a, CCDC
233450 for 13c, CCDC 177462 for 13h, CCDC 177463 for
15a, CCDC 177464 for 16a, CCDC 177464 for 17a [fax:
+44 (1223)336033; homepage: http://
Scheme 3 Acetylation of 6a,f
Compounds 6a,f have been additionally characterized by
acetylation (Scheme 3). Appropriate reaction conditions
have been worked out for each of the possible acetyl de-
rivatives 15–17. Thus, 15a and 16a,b were obtained by
treatment of 6a,f with acetic anhydride at room tempera-
ture in the presence of pyridine or acetic acid, respective-
ly. The O,N-diacetyl derivative 17a was prepared by
treatment of 6a with excess of acetic anhydride in pyri-
dine, or by acetylation of each of the monoacetyl deriva-
tives (Scheme 3). Our attempts to achieve selective O-
acetylation of 6f in order to prepare 15b (X = NO2,
Scheme 3) failed and, therefore, compound 17b could be
obtained solely by refluxing 16b with acetic anhydride in
pyridine. Apparently, the nitro group deactivates O-
acetylation under milder conditions. The constitution of
www.ccdc.cam.ac.uk].
(4) (a) Typical Procedure for the Preparation of Coumarin-
enaminoester Adducts (6a,c,e,f,h,i; 8a; 10c–f,h,i; 11a,c,i):
A mixture of 15 mmol of the corresponding coumarin 1a–i
and 15 mmol of an enaminoester 2a–c in 40 mL anhyd EtOH
was stirred (temperature and period of time are given in
Table 1). After cooling, the separated crystals of the
corresponding adduct were filtered, washed with cold EtOH
and recrystallized, usually from EtOH. If a mixture of
products was obtained (TLC monitoring), the solvent was
removed in vacuo and the residue was separated by column
chromatography on siliga gel.
(b) Systematic names: Ethyl 12-benzoyl-9-ethoxy-11-oxo-
8-oxa-10-azatricyclo[7.3.1.02,7]trideca-2 (7),3,5-triene-13-
carboxylate (10e); Ethyl {11-oxo-8-oxa-10-
azatricyclo[7.3.1.02,7]trideca-2,4,6-trien-9-yl}acetate (11a).
Synlett 2004, No. 9, 1584–1588 © Thieme Stuttgart · New York

1588
L. D. Raev et al.
LETTER
(7) Ethyl 2-Amino-4-(2-hydroxy-5-nitrophenyl)-6-oxo-
(5) Claremon, D. A.; Hirshfield, J.; Lumma, P. K.; McClure, D.
E.; Springer, J. P. Synthesis 1986, 144.
1,4,5,6-tetrahydro-3-pyridinecarboxylate (6f)
(unpublished data): 1H NMR (400 MHz, DMSO-d6):
d = 1.04 (m, 3 H, CH3), 2.50 (d, 2J = 16.2 Hz, 1 H, 5-HA),
2.80 (dd, 3J = 7.3 Hz, 2J = 16.2 Hz, 1 H, 5-HB), 3.95 (m, 2 H,
OCH2), 4.40 (d, 3J = 7.3 Hz, 1 H, 4-H), 7.01 (d, 1 H, 3¢-H),
7.75 (s, 1 H, 6¢-H), 8.03 (d, 1 H, 4¢-H.), 6.20–7.80 (br, 2 H,
NH2), 9.80 (s, 1 H, NH or OH), 11.35 (br s, 1 H, OH or NH).
13C NMR (100.6 MHz, DMSO-d6): d = 14.8 (CH3), 30.5
(C-4), 36.9 (C-5), 58.6 (O-CH2), 75.0 (C-3), 115.9
(Carom.-3¢), 123.1 (Carom.-6¢), 124.7 (Carom.-4¢), 131.7
(Carom.-1¢), 139.8 (Carom.-5¢), 154.6 (C-2), 162.0 (Carom.-2¢),
168.2 (C=O, ester), 170.3 (C=O, lactam).
(6) Ethyl 2-Amino-4-(2-hydroxyphenyl)-6-oxo-1,4,5,6-
tetrahydro-3-pyridinecarboxylate (6a) (unpublished
data): 1H NMR (400 MHz, DMSO-d6): d = 1.02 (t, J = 7.0
Hz, 3 H, CH3), 2.51 (d, 2J = 16.0 Hz, 1 H, 5-HA), 2.80 (dd,
2J = 16.0 Hz, 3J = 7.3 Hz, 1 H, 5-HB), 3.99 (m, 2 H, OCH2),
4.35 (d, 3J = 7.3 Hz, 1 H, 4-H), 6.67 (m, 1 H, 6¢-H), 6.80 (m,
2 H, 3¢-H, 5¢-H), 6.99 (m, 1 H, 4¢-H.), 6.00–8.00 (br, 2 H,
NH2), 9.45 (s, 1 H, NH or OH), 9.59 (s, 1 H, OH or NH). 13
C
NMR (100.6 MHz, DMSO-d6): d = 13.5 (CH3), 28.7 (C-4),
36.1 (C-5), 57.4 (O-CH2), 74.9 (C-3), 114.0 (Carom.), 117.9
(Carom.), 125.5 (Carom.), 126.5 (Carom.), 128.4 (Carom.), 152.4
(C-2¢), 153.4 (C-2.), 167.4 (C=O, ester), 170.0 (C=O,
lactam).
Salient NMR signals: O-Acetyl derivative (15a): mp 170–
172 °C (toluene). 1H NMR (300 MHz, DMSO-d6): d = 2.32
(s, 3 H, OCOCH3), 7.03–7.25 (m, 6 H, 4Harom and NH2), 9.73
(s, 1 H, lactam-NH). N-Acetyl derivative (16a): mp 184–186
°C (EtOH). 1H NMR (300 MHz, DMSO-d6): d = 2.21 (s, 3
H, NCOCH3), 9.61 (br s, 1 H, OH), 10.68 (s, 1 H, amide-
NH), 11.49 (s, 1 H, lactam-NH). O,N-Diacetyl derivative
(17a): mp 180–181 °C (EtOH). 1H NMR (300 MHz, CDCl3):
d = 2.23 (s, 3 H, NCOCH3), 2.34 (s, 3 H, OCOCH3), 10.96
(s, 1 H, amide-NH), 11.98 (s, 1 H, lactam-NH). 13C NMR (75
MHz, CDCl3): d = 21.0 (OCOCH3), 25.3 (NCOCH3).
Salient 1H NMR signals: N-Acetyl derivative (16b): mp
218–220 °C (2-PrOH). 1H NMR (250 MHz, DMSO-d6):
d = 2.23 (s, 3 H, NCOCH3), 10.72 (s, 1 H, amide-NH), 11.40
(s, 1 H, lactam-NH), 11.5 (br s, 1 H, OH). O,N-Diacetyl
derivative (17b): mp 165–166.5 °C (2-PrOH). 1H NMR (250
MHz, CDCl3): d = 2.29 (s, 3 H, NCOCH3), 2.40 (s, 3 H,
OCOCH3), 11.12 (s, 1 H, amide-NH), 12.01 (s, 1 H, lactam-
NH).
(8) Raev, L. D.; Ivanov, I. C. Third National Congress of
Pharmacy (Abstracts), 17-19 October 1996, Sofia –
Panichishte (Bulgaria); poster 11.P5.
Synlett 2004, No. 9, 1584–1588 © Thieme Stuttgart · New York