Interconvertions between δ-lactam and δ-lactone derivatives initiated by unique transannular interactions of the rigid cyclohexane boat structure in pentacycloundecane

Article abstract of DOI:10.1021/jo0357723

The pentacycloundecane (PCU) cage structure resembles a perfect boat conformation, and for the first time unique lactam/lactone interconversions on the flagpole carbons of a cyclohexane boat structure are reported. The syntheses of a novel dihydroxy-PCU-δ-lactone and two novel N-substituted PCU-δ-lactams are reported. Hydrolysis of some of the PCU-δ-lactam compounds produced δ-lactones, and reaction of the lactones with ammonia or primary amines again produced δ-lactams. Reaction mechanisms to account for the unusual interconversion reactions induced by transannular interactions are proposed.

Full text of DOI:10.1021/jo0357723

A literature search revealed a vast number of publica-
tions about the separate synthesis of lactam and lactone
derivatives with a cyclohexane boat basis, but very little
is known about the interconversion between these lactam
and lactone derivatives. A number of studies focused on
the transannular interactions of cyclohexane boat struc-
tures as well as transition states involving a boat con-
formation.5,6 The cyclohexane boat structure is a transi-
tion state.7-9 Therefore, performing reactions on the two
flagpole carbons is therefore difficult to envisage, except
if this conformation is conveniently rigidified as in the
pentacyloundecane skeleton or in alternative struc-
tures.10
In ter con ver tion s betw een δ-La cta m a n d
δ-La cton e Der iva tives In itia ted by Un iqu e
Tr a n sa n n u la r In ter a ction s of th e Rigid
Cycloh exa n e Boa t Str u ctu r e in
P en ta cyclou n d eca n e
Hendrik G. Kruger,† Frans J . C. Martins,‡ and
Agatha M. Viljoen*,‡
School of Chemistry, University of Kwazulu-Natal,
Durban 4001, South Africa, and Department of Chemistry,
Potchefstroom University, Potchefstroom 2520, South Africa
The effect of transannular interactions on the penta-
cycloundecane skeleton has been throroughly studied.1,4,11
However, the interconversions between the PCU-lactam
and lactone derivatives have not been investigated.
Certain lactams can be hydrolyzed to their correspond-
ing amino acids,12 and our initial aim was to convert the
different lactam compounds (2-4) to their corresponding
amino acids. Substituent groups on the ring13-15 as well
as ring strain16-18 play an important role on the ease of
hydrolysis. An amino group on the ring carbon atom
adjacent to the nitrogen atom in the â-lactam ring greatly
increases the rate of hydrolysis.16 These type of â-amino
lactams are all moisture sensitive. Hydrolysis of lactam
compounds does not always produce the corresponding
amino acid,19,20 as was the case with the cyano lactam
3.4 The cyano lactam was selectively hydrolyzed to three
different novel lactone compounds (5-7).4
cheamv@puknet.puk.ac.za
Received December 4, 2003
Abstr a ct: The pentacycloundecane (PCU) cage structure
resembles a perfect boat conformation, and for the first time
unique lactam/lactone interconversions on the flagpole
carbons of a cyclohexane boat structure are reported. The
syntheses of a novel dihydroxy-PCU-δ-lactone and two novel
N-substituted PCU-δ-lactams are reported. Hydrolysis of
some of the PCU-δ-lactam compounds produced δ-lactones,
and reaction of the lactones with ammonia or primary
amines again produced δ-lactams. Reaction mechanisms to
account for the unusual interconversion reactions induced
by transannular interactions are proposed.
As part of a program to investigate the synthesis and
chemistry of amino acids with cage structures, the dione
1 was utilized as a substrate in Strecker reactions. The
dione (1) is easily obtained from the Diels-Alder adduct
of cyclopentadiene and p-benzoquinone by intramolecular
photocyclization.1,2 As treatment of the dione with Streck-
er reagents normally leads to cyanohydrin and/or amino
nitrile products,3 the one-pot conversion of the dione to
the three novel δ-lactams 2-4 is quite unique.3,4 The role
of the rigid cyclohexane boat structure of the pentacy-
cloundecane skeleton (see alternative view of the dione
1 below) with respect to transannular interactions was
highlighted in the proposed mechanisms.3,4
The cyanolactam 3 is very susceptible to ring cleavage
in acidic media and is hydrolyzed to the corresponding
(5) For studies about the interactions of substituents on cyclohexane
boat structures: (a) Levisalles, J . Bull. Soc. Chim. Fr. 1960, 551. (b)
Stolow, R. D. J . Am. Chem. Soc. 1961, 83, 2592. (c) Stolow, R. D. J .
Am. Chem. Soc. 1964, 86, 2170. (d) Stolow, R. D.; McDonagh, P. M.;
Bonaventura, M. M. J . Am. Chem. Soc. 1964, 86, 2165. (e) Camps, P.;
Iglesias, C. Tetrahedron Lett. 1985, 26 (44), 5463. (f) Fitjer, L.;
Scheuermann, H.-J .; Klages, U.; Wehle, D.; Stepenson, D. S.; Binsch,
G. Chem. Ber. 1986, 119, 1144.
(6) For Cope rearrangements with transition states involving a
cyclohexane boat structure, see: (a) Hoffmann, R.; Woodward, R. B.
J . Am. Chem. Soc. 1965, 87, 4389. (b) Fukui, K.; Fujimoto, H.
Tetrahedron 1966, 251. (c) Gajewski, J . J .; J imenez, J . L. J . Am. Chem.
Soc. 1986, 108, 468.
(7) Dunitz, J . D. J . Chem. Ed. 1970, 47, 488.
(8) Leventis, N.; Hanna, S. B.; Sotiriou-Leventis, C. J . Chem. Ed.
1997, 74 (7), 813.
(9) Sauers, R. R. J . Chem. Ed. 2000, 77 (3), 332.
(10) Wiberg, K. B.; Matturro, M.; Adams, R. J . Am. Chem. Soc. 1981,
1600.
† University of Kwazulu-Natal.
‡ Potchefstroom University.
(1) Cookson, R. C.; Crundwell, E.; Hill, R. R.; Hudec, J . J . Chem.
Soc. 1964, 3062.
(2) Marchand, A. P.; Allen, R. W. J . Org. Chem. 1974, 39 (11), 1596.
(3) Furniss, B. S.; Hannaford, A. J .; Smith, P. W. G.; Tatchell, A. R.
Vogel’s Textbook of Practical Organic Chemistry, 5th ed.; Longman
Scientific & Technical: New York, 1989.
(11) For transannular interactions on the PCU cyclohexane boat
structure, see: (a) Sasaki, T.; Eguchi, S.; Kiriyami, T. Hiroaki, O.
Tetrahedron 1974, 30, 2707. (b) Barborak, J . C.; Khoury, D.; Maier,
W. F.; Schleyer, P. v. R.; Smith, E. C.; Smith, W. F.; Wyrick, C. J .
Org. Chem. 1979, 44, 4761. (c) Martins, F. J . C.; Viljoen, A. M.; Coetzee,
M.; Fourie L.; Wessels, P. L. Tetrahedron 1991, 47(44), 9215.
(12) Fryth, P. W.; Waller, C. W.; Hutchings B. L.; Williams, J . H. J .
Am. Chem. Soc. 1958, 80, 2736.
(4) Martins, F. J . C.; Viljoen, A. M.; Kruger, H. G.; J oubert, J . A.;
Tetrahedron 1993, 49, 9573.
10.1021/jo0357723 CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/17/2004
J . Org. Chem. 2004, 69, 4863-4866
4863

SCHEME 1. Con ver sion of th e Cya n o La cta m 3 to th e Dih yd r oxyla cton e 10
novel lactone 5 at room temperature.4 Under different
reaction conditions the other novel lactones 6 and 7 were
obtained.4
intermediate cation required for ease of hydrolysis. The
mechanism proposed for the hydrolysis of 3 in acidic
media to produce the lactone 5 was reported previously.4
The δ-lactone 921 with a cyclohexane boat skeleton as
the basis was reported in 1929. cis-4-Hydroxycyclohex-
anecarboxylic acid 8 is converted to the corresponding
cyclohexane boat δ-lactone 9, while the trans form of 8
does not lactonize.5,22
Alkaline hydrolysis of the cyano hydroxylactam 3
unexpectedly led to cyano group substitution. Treatment
of 3 with 30% sodium hydroxide under reflux conditions
produced the novel dihydroxylactone 10. A possible
explanation for the formation of the lactone 10 from
alkaline catalyzed hydrolysis of the lactam 3 is provided
in Scheme 1.
Nucleophilic attack from hydroxyl anions on the car-
bonyl carbon atom of the lactam 3, to displace the cyanide
group in 11, is probably energetically a more likely
process than nitrile group hydrolysis. The highly reactive
imine group in 12 is easily hydrolyzed to the keto acid
13 which spontaneously converts to the lactone 10. The
infrared spectrum, FAB mass spectrum, and NMR data
all support the structure of 10. The product tests negative
for nitrogen. The NMR data of 10 is provided in the
Experimental Section. A diacetate was obtained when the
lactone 10 was acetylated, confirming the presence of two
hydroxyl groups.
It is expected that the lactams 2 and 4 should follow
the same reaction route with alkaline hydrolysis at reflux
temperature. Treatment of the amino hydroxylactam 4
with 30% sodium hydroxide under reflux conditions
produced the expected conversion to the dihydroxylactone
10. However, a similar conversion of the dihydroxylactam
2 to 10 could not be achieved even after the same
treatment for 48 h. Treatment of the amino hydroxylac-
tam 4 under elevated temperature and pressure (170 °C
in a pressurized reaction vessel) in the presence of water
and a catalytic amount of either acetic acid or diethy-
lamine produced the hydrate of 2.26 When amino hy-
droxylactam 4 is refluxed in concentrated hydrochloric
acid for 48h the dihydroxylactone 10 was obtained in
The corresponding lactam of 9 was also reported,12,23
but interconversion between these lactam and lactone
derivatives have not been investigated.
In view of the discussion above, it was decided to
investigate the hydrolysis of the polycyclic δ-lactam
derivatives 2-4 as well as the interconversion of the
lactam and lactone derivatives. These interconversions
on a cyclohexane boat structure had not been reported
before.
The dihydroxylactam 2 could not be converted to the
corresponding lactone in acidic media. Similarly, the
amino hydroxylactam 4 is surprisingly stable under
acidic reaction conditions. The obvious difference between
the lactams 2 and 4 with respect to the cyano lactam 3
is the ability of the cyano group to stabilize4,24,25 an
(13) Sheenan J . C.; Bose, A. K. J . Am. Chem. Soc. 1951, 73, 1761.
(14) Holley R. W.; Holley, A. D. J . Am. Chem. Soc. 1949, 71, 2124.
(15) Holley R. W.; Holley, A. D. J . Am. Chem. Soc. 1950, 72, 2771.
(16) Manhas, M. S.; Bose, A. K. In Beta-Lactams: Natural and
Synthetic; Wiley-Interscience: New York, 1971; pp 61-63.
(17) Koch, T. U.S. 2,453,234, 1948.
(18) Koch, T. Ger. 812,076, 1951.
(19) Perelman, M.; Mizak, S. A. J . Am. Chem. Soc. 1962, 84, 4988.
(20) Opitz, G.; Koch, J . Angew. Chem., Int. Ed. Engl. 1963, 2, 152
(21) Balasˇ, F.; Sˇrol, L. Collect. Czech. Chem. Commun. 1929, I, 658.
(22) Hardegger, E.; Plattner, Pl. A.; Blank, F. Helv. Chim. Acta 1944,
27, 793.
(23) Aubry, A.; Protas, J . Acta Crystallogr. 1973, B29, 2576.
(24) Dixon, D. A.; Charlier, P. A.; Gassman, P. G. J . Am. Chem. Soc.
1980, 102, 3957.
(25) Paddon-Row, M. N.; Santiago, C.; Houk, K. N. J . Am. Chem.
Soc. 1980, 102, 6561.
(26) The dihydroxylactam 2 crystalizes with 1 equiv of water, being
hydrogen bonded to the lactam. Kruger, H. G.; Martins, F. J . C.;
Viljoen, A. M.; Boeyens, J . C. A.; Cook, L. M.; Levendis, D. C. Acta
Crystallogr. 1996, B52, 838.
4864 J . Org. Chem., Vol. 69, No. 14, 2004

SCHEME 2. Con ver sion of th e Cya n o La cton e 10 to N-Su bstitu ted La cta m Der iva tives
reasonable yield. The reaction proceeded much faster in
a medium-pressure reaction vessel containing 10% hy-
drochloric acid.
carbon atom resonances in the 13C NMR spectrum (δC
90, 99 and δC 79, 95 are registered as double signals).
It is expected that ammonolysis of the dihydroxylac-
tone 10 should proceed via the same route as proposed
for the ammonolysis of the cyano hydroxylactone 5 (see
Scheme 2). Surprisingly, treatment of 10 (1 g) with 25%
ammonia (10 mL) led to the formation of the amino
hydroxylactam 4 instead of the expected dihydroxylactam
2. When the reaction was repeated with 10 times more
ammonia (100 mL), the dihydroxylactam 2 was obtained.
[Note that only the dihydroxylactam 2 was obtained
when the cyano hydroxylactone 5 (1 g) was treated with
25% ammonia (10 or 100 mL)]. However, treatment of
the dihydroxylactone 10 with ethylamine or benzylamine
produced the expected N-substituted δ-lactams 17 and
18, respectively, in a similar way as for the ammonolysis
of the cyano hydroxylactone 5.
The formation of the amino hydroxylactam 4 upon
treatment of the dihydroxylactone 10 (1 g) with 25%
ammonia (10 mL) can probably be attributed to the
different solubility of the intermediates 15 and 19 (see
Scheme 3). Note that both the reactant 10 and products
2 and 4 are relatively insoluble in aqueous media. In
general structures with a carbonyl group (15) should be
less soluble in aqueous solution compared to its corre-
sponding imine (19). A possible explanation could be that
when the reaction is performed in more diluted conditions
(1 g of 10 in 100 mL of ammonia), enough of the
intermediate 15 is dissolved to yield the kinetic product
2. When the reaction is performed in more concentrated
conditions (1 g of 10 in 10 mL of ammonia), the inter-
mediate 15 presumably does not dissolve sufficiently and
is converted to a more soluble form, namely the corre-
sponding imine 19. Instantaneous transannular cycliza-
tion of 19 to the amino lactam 4 follows.
Ammonolysis of lactones can yield lactams.27-29 As
anticipated, the dihydroxylactam 2 was obtained upon
treatment of the cyano hydroxylactone 5 (1 g) with 25%
ammonia (10 or 100 mL). Nucleophilic attack of ammonia
on the carbonyl carbon atom of 5 will probably lead to
carbon-oxygen bond cleavage with simultaneous nitrile
group elimination. A possible mechanism is suggested in
Scheme 2. Attack of the electron pair on the nearby
nitrogen atom of the carbonyl carbon atom in 15 should
lead to the formation of the dihydroxylactam 2.
Similar conversions of the cyano hydroxylactone 5 to
lactam derivatives were achieved by treatment of 5 with
ethylamine or benzylamine, which afforded the novel
N-substituted lactam compounds 17 and 18, respectively.
The structures of the lactams 17 and 18 were elucidated
1
from 13C and H NMR studies and the data is provided
in the Experimental Section.
Confirmative evidence for the structure of 17 was
obtained from a 13C NMR spectrum recorded in (CD3)2-
SO, which was treated with two drops of a mixture of
60% D2O and 40% H2O to impose partial exchange of
protons. Double signals were observed for both quater-
nary carbon signals indicating that both bear groups with
only one deuterium-exchangeable proton. MS and IR
spectra also support the proposed structure of 17.
The absorption bands in the infrared spectrum of the
N-benzyl-substituted lactam 18 are very similar to that
of 17. The EI mass spectrum of 18 exhibits a molecular
ion at m/z 309, which corresponds with a molecular
formula of C19H19NO3. As in the case of 17 partial
exchange of protons influenced only the quaternary
(27) March, J . In Advanced Organic Chemistry, 3rd ed.; J ohn Wiley
and Sons: New York, 1985; 375.
(28) Torok, D. S.; Ziffer, H.; Meshnick, S. R.; Pan, X.; Ager, A. J .
Med. Chem. 1995, 38, 5045.
(29) El Sayed, K. A.; Orabi, K. Y.; Dunbar, D. C.; Hamann, M. T.;
Avery, M. A.; Sabnis, Y. A.; Mossa, J . S.; El-Feraly, F. S. Tetrahedron
2002, 58, 3699.
Initially, it was not clear whether the dihydroxylactam
2 could serve as intermediate in the formation of the
amino lactam 4 (see Scheme 3).
Treatment of the dihydroxylactam 2 with 25% am-
monia at room temperature for 24 h only produced
J . Org. Chem, Vol. 69, No. 14, 2004 4865

SCHEME 3. Con ver sion of th e Dih yd r oxyla cton e 10 to th e Am in o La cta m 4
starting material on subsequent evaporation of the excess
ammonia. Since the amino hydroxylactam 4 is less solu-
ble than the dihydroxylactam 2 the reaction was repeated
with a minimum amount of ammonia in an effort to
induce precipitation of the desired product 4. Instead the
hydrate26 of the dihydroxylactam 2 was obtained. The
intermediacy of the dihydroxylactam 2 in the conversion
of the dihydroxylactone 10 to 4 was confirmed by forced
dissolution of 2 in ammonia (150 °C for 18 h), which
indeed produced the amino hydroxylactam 4.
formation of the cyano lactam 3 should also occur,
followed by conversion of the latter to the amino lactam
4. This experimental observation, combined with the rate
of hydrolysis of the lactam compounds 3 and 4, implies
that the amino lactam 4 is the thermodynamically
determined product.
Ack n ow led gm en t. The authors thank Dr. Glenn
Maguire, University of KwaZulu-Natal, for his kind
advice and assistance with the presentation of the
results and Dr. L. Fourie and Mr. A. J oubert, University
of Potchefstroom, for the MS and NMR spectra, respec-
tively.
Treatment of the cyano lactam 3 (1 g) with ammonia
(10 or 100 mL) also produced the amino lactam 4. A
similar mechanism as proposed in Scheme 3 accounts for
this transformation of the lactam 3 via intermediates 15
and 19 to the lactam 4.
Su p p or tin g In for m a tion Ava ila ble: All experimental
details. This material is available free of charge via the
Internet at http://pubs.acs.org.
This observation explains the reason why treatment
of the dione 1 with an excess of sodium cyanide and
ammonia produces4 the amino lactam 4 although the
J O0357723
4866 J . Org. Chem., Vol. 69, No. 14, 2004