Synthesis and ring cyclization-expansion-contraction reactions of some new 2,2-disubstituted indan-1,3-diones and related compounds

Article abstract of DOI:10.1071/CH05169

We have found that the hydrochloride of 2-phenyl-2-[2-(2-piperidyl)ethyl]- 4,5,6,7-tetrahydroindan-1,3-dione 1 possesses marked analgesic activity (100% inhibition referenced to codeine) and report, as part of an extensive synthetic program, the synthesis of 38 new and structurally related compounds. Selective catalytic hydrogenation of the pyridine ring of 2-phenyl-2-[2-(2-pyridyl)ethyl]- indan- 1,3-dione 2 yields the nine-membered nitrogen-containing heterocycle 6 by a novel ring cyclization-expansion reaction. The structural and functional group parameters required for this novel ring-expansion reaction have been extensively and thoroughly investigated through the synthesis of a series of structurally related compounds; principally by modification, substitution, and replacement of the various functionality contained within 2. In addition, we report the synthesis of a series of new 2-methy1-2-(ω-N-phthalimidoalkyl)- indan-1,3-diones 41, 45, and 53, two of which, like the parent 2-phenyl substituted indan-1,3-dione 2, also undergo a novel ring cyclization-expansion reaction to yield eight- and nine-membered nitrogen-containing rings. However, in these cases, further transannular reactions occur to produce the new 5,5- and 5,6-ring-fused nitrogen-containing heterocycles 44, 48 and 51, 52. Hydrazinolysis of the third, 2-methy1-2-(4-N-phthalimidobutyl)-indan- 1,3-dione yields the new azepine-containing ring structure 56 by direct cyclization. Furthermore, some interesting and unexpected chemical properties of the final compounds, which include selective and non-selective pyridine-ring hydrogenations and a few unexpected side reactions, are described. CSIRO 2006.

Full text of DOI:10.1071/CH05169

Full Paper
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2006, 59, 59–74
Synthesis and Ring Cyclization–Expansion–Contraction Reactions
of Some New 2,2-Disubstituted Indan-1,3-diones
and Related Compounds
Craig J. RoxburghA,B,D and Lee BantingC
A Department of Chemistry, University College, University of London, London WC1H OAJ, UK.
B Current address: Institute of Naval Medicine, Alverstoke, Gosport PO12 2DL, UK.
C Department of Chemistry and Pharmacology, University of Portsmouth, Portsmouth PO1 2DT, UK.
D Corresponding author. Email: sic@inm.mod.uk
We have found that the hydrochloride of 2-phenyl-2-[2-(2-piperidyl)ethyl]-4,5,6,7-tetrahydroindan-1,3-dione 1 pos-
sesses marked analgesic activity (100% inhibition referenced to codeine) and report, as part of an extensive synthetic
program, the synthesis of 38 new and structurally related compounds.
Selective catalytic hydrogenation of the pyridine ring of 2-phenyl-2-[2-(2-pyridyl)ethyl]-indan-1,3-dione 2 yields
the nine-membered nitrogen-containing heterocycle 6 by a novel ring cyclization–expansion reaction.The structural
and functional group parameters required for this novel ring-expansion reaction have been extensively and thor-
oughly investigated through the synthesis of a series of structurally related compounds; principally by modification,
substitution, and replacement of the various functionality contained within 2.
In addition, we report the synthesis of a series of new 2-methyl-2-(ω-N-phthalimidoalkyl)-indan-1,3-diones
41, 45, and 53, two of which, like the parent 2-phenyl substituted indan-1,3-dione 2, also undergo a novel ring
cyclization–expansion reaction to yield eight- and nine-membered nitrogen-containing rings. However, in these
cases, further transannular reactions occur to produce the new 5,5- and 5,6-ring-fused nitrogen-containing hetero-
cycles 44, 48 and 51, 52. Hydrazinolysis of the third, 2-methyl-2-(4-N-phthalimidobutyl)-indan-1,3-dione yields
the new azepine-containing ring structure 56 by direct cyclization.
Furthermore, some interesting and unexpected chemical properties of the final compounds, which include
selective and non-selective pyridine-ring hydrogenations and a few unexpected side reactions, are described.
Manuscript received: 8 July 2005.
Final version: 21 December 2005.
HO
Introduction
O
O
N
2-Substituted indan-1,3-diones are of pharmacological inter-
est and importance since these are known to possess
anti-inflammatory,[1] hypermetabolic,[2] anticoagulant,[3]
uricosuric,[4] analgetic,[5] rodenticidal,[6] antibacterial,[7]
bronchodilator,[8] and parasiticidal[9] properties. In the light
of the above, and the fact that substituted indan-1,3-diones
that also possess various nitrogen-containing functionalities
have been relatively unexplored, led us to embark on an exten-
sivesyntheticprogramto investigate thechemistry andpoten-
tial pharmacological activity of these new structures. Further,
during the course of this project we synthesized and iso-
lated the hydrochloride of 2-phenyl-2-[2-(2-piperidyl)ethyl]-
4,5,6,7-tetrahydroindan-1,3-dione 1 (Fig. 1), which was
found to possess marked analgesic activity (100% refer-
enced to codeine).[10] This marked analgesic activity added
further interest to the project, and focussed some of our
synthetic endeavours towards related structures. (Note: we
have/had no control over the decision of which compounds
Ph
H
N
HCl
O
Codeine
1
MeO
Fig. 1. The structures of compound 1 and codeine.
werepharmacologicallytested, oroverthelimitedresultswith
which we were furnished.)
In addition, during the course of this project, we dis-
covered that selective catalytic hydrogenation of the pyri-
dine ring of 2-phenyl-2-[2-(2-pyridyl)ethyl]-indan-1,3-dione
2 yielded, via a ‘one-step’ novel ring cyclization–expansion
reaction, a nine-membered nitrogen-containing heterocyle
6 in a relatively excellent 60% overall yield.[11] Further to
the above mentioned biological activity of indan-1,3-diones,
medium-sized nitrogen-containing rings, related in structure
to 6, are of importance since they are reported to exhibit
© CSIRO 2006
10.1071/CH05169
0004-9425/06/010059

60
C. J. Roxburgh and L. Banting
O
synthesized target compounds: In particular, the expected
selective catalytic hydrogenation of the pyridine portion of
the quinoline ring of the indan-1,3-dione structure 15 was
obtained. However, under identical hydrogenation condi-
tions, the corresponding 1,2,3,4-tetrahydroindan-1,3-dione
compound 17 afforded the unexpected selective hydrogena-
tion of the benzene portion of the quinoline ring.
H
N
R1
R2
(A)
(A)
(CH2)n
(B)
O
(a)
(b)
O
R1
Some further synthetic modifications of the final ring
structures were investigated, which produced en route some
new chemical entities. Specifically, it was found that the
isoindole 30 produced, on treatment with a hydride reagent,
the new 5,6-fused nitrogen-bridgehead heterocycle 33, the
stereochemistry of which has been deduced. In addition, the
nine-membered ring system 6, when treated with sodium
borohydride, underwent an unexpected hydride-induced
transannular-contraction reaction (a lactam-to-lactone inter-
change) to afford the new lactone-containing compound 57.
NR2R3
(CH2)n
O
Fig. 2. Overview of the structural modifications detailed here.
sedative,[12,13] muscle relaxant,[12,13] anticonvulsant,[12,13]
and marketed psychosedative[14–16] and tranquillizing[14–16]
properties. The known biological activity of these structures
was, therefore, an additional driving force to thoroughly
investigate the scope of this potentially very useful, and
relatively simple, ‘one-pot’ route to medium-sized rings.
Furthermore, medium-sized rings are classically difficult
to prepare, usually requiring multiple steps, and invariably
associated with poor yields. Consequently, we extensively
investigated the generality and applicability of this ring-
expansion reaction to other structurally related compounds.
This was achieved by dissection, modification, substitution,
and replacement of the functionality associated with the
parent structure 2, and subsequent investigation of the reac-
tivity of this introduced functionality, to see whether this
affectedthecourseandoutcomeoftheabove-mentionedring-
expansion-reaction sequence. The detailed structural modi-
fications made are described in detail under the appropriate
sectionsthroughoutthispaper;however, abriefoverviewwith
specific reference to the general structures, depicted in Fig. 2,
is given here. With reference to Fig. 2: (a) The degree of sat-
uration contained within the indan-1,3-dione nucleus, –(A)–,
was varied in order to investigate the effects of introducing
partial and full saturation; (b) variations in the length of the
alkyl linking chain, –(CH2)n–, for n = 1 and 2, were made,
and an additional structure (n = 1) that possesses a 3-pyridyl
nucleus, were investigated; (c) replacement of the pyridine
ring (B) with a quinoline, with both fully unsaturated and
partially saturated indan-1,3-dione nuclei, –(A)–, was inves-
tigated; (d) substitution of the pyridine ring with an alkyl
group (R2 = 5-ethyl) was achieved and investigated; (e) the
size of the ring-containing the 1,3-dione functionality (A)
was varied in order investigate the effects that a change in
ring strain induces; and lastly (f) the substitution of R1 = Ph
for an alkyl grouping (R1 = Me) was investigated. With refer-
ence to Fig. 2b: (a) variation of the length of the alkyl linking
chain –(CH2)n–, for n = 2 and 3 and R1 = Ph, possessing
both a partially saturated indan-1,3-dione nucleus, –(A)–, and
terminal N-phthalimidoalkyl (–NR2R3) groups, was investi-
gated, and finally (b) an investigation of the variation of the
length of the alkyl linking chain –(CH2)n–, for n = 2, 3, and
4 with R1 = Me, possessing both an indan-1,3-dione nucleus
and terminal N-phthalimidoalkyl (–NR2R3), was made.
During the course of this synthetic program we noted
some interesting reactivity and chemical properties of the
Results and Discussion
Synthesis of 2-Pyridyl and 2-Piperidyl-Containing
Structures Possessing Tetrahydro- and
Hexahydroindan-1,3-dione Nuclei. Selective
Ene-1,4-dione and Pyridine Ring Reductions
Since we reported that the selectively catalytically reduced
piperidyl variant of 2 underwent a novel ring cyclization–
expansion reaction to a nine-membered ring (Scheme 1),
we were interested to investigate whether this reaction
sequence could be reproduced utilizing compounds pos-
sessing slightly different chemical functionality. Initially,
we were interested in synthesizing compounds that con-
tained small structural variations, i.e., the partially saturated
9 and corresponding fully saturated indan-containing sys-
tem 12 (see Scheme 2). In addition, the marked analgesic
activity of 1 offered a further reason for us to synthesize
and explore structurally similar variants as possible candi-
dates for pharmacological screening. Compound 7[17] readily
underwent a Michael-type addition to 2-vinyl pyridine, in
an ethanol reflux, to yield 2-phenyl-2-[2-(2-pyridyl)ethyl]-
4,5,6,7-tetrahydroindan-1,3-dione 9 in an excellent 84%
yield.
Selective catalytic hydrogenation of the pyridine portion
of this compound, in the presence of two equivalents of
hydrochloric acid, gave the hydrochloride of 2-phenyl-2-
[2-(2-piperidyl)ethyl]-4,5,6,7-tetrahydroindan-1,3-dione 10
(91%). Attempts to prepare the corresponding free piperidyl
compound, by treatment with aqueous base, resulted in
polymerization, presumably by intermolecular Michael-type
additions of the resulting secondary amines to the activated
3a,7a-ene-dione functionality. High dilution basification of
this compound followed by N-acetylation (access acetylchlo-
ride/triethylamine) allowed isolation and characterization by
the N-acetamido derivative 11a/11b (55%). Presumably, the
N-acetylation reaction proceeds more rapidly than the com-
peting intermolecular Michael-type polymerization, as a 55%
yield of the amide 11a/11b is obtained. (1H NMR showed
this to be a 1/1 mixture of its two amide rotamers at room
temperature.)

Synthesis and Reactions of 2,2-Disubstituted Indan-1,3-diones
61
Carbinolamine intermediate
O
O
Oꢀ
O
O
N
N
Ph
H2, PtO2,
Ph
H
H
N
N
EtOH, HCl
Ph
O
ꢀO
O
Ph
2
3
4
5
O
N
H
6
O
H
Ph
Scheme 1.
O
OH
O
Ph
Zn, HOAc, ꢁ
Ph
O
7
8
2-Vinylpyridine,
EtOH, ꢁ
O
O
Ph
3a
Ph
H2, PtO2,
H
N
N
EtOH, HCl
7a
O
O
9
10
(i) Dil NaOH
(ii) CH3COCl, NEt3
Zn, HOAc, ꢁ
O
O
O
H
H
Ph
Ph
N
N
O
O
12
11a
H2, PtO2,
EtOH, HCl
1/1 mixture of rotamers
O
O
H
H
O
Ph
Ph
H
N
N
O
O
13
11b
Scheme 2.
In view of the instability of the above compound,
we decided to investigate the hexahydroindan-1,3-dione-
containing systems 12 and 13. This was primarily to investi-
gate exactly what functional groups and structural parameters
were required in order for the novel ring cyclization–
expansion reaction, typified by the conversion of 2 → 6,
to progress. It would also produce further analogues for
pharmacological screening.
Initial synthesis of the hexahydroindan system 8 was
achieved by a zinc reduction of the internal 3a,7a dou-
ble bond of 7 to yield 2-phenyl-hexahydroindan-1,3-dione
(Scheme 2).[17] Even with the use of a variety of catalysts

62
C. J. Roxburgh and L. Banting
(sodium metal, crushed sodium or potassium hydroxide
pellets), this would not, however, undergo a Michael-type
addition with 2-vinyl pyridine. Since α,β-unsaturated ketones
are reported to be reduced to saturated ketones using zinc
dust in acetic acid,[18] this method was applied to 9 in the
hope of obtaining selective reaction over other present func-
tionalities, in particular, the possible problems associated
with the dimerization of the pyridine nuclei through radical
anion mechanisms, and keto-group reductions. Initial prob-
lems (polymer formation) were encountered in obtaining
the required selectivity over other present functionalities.
However, this was overcome by limiting the time of the reac-
tion to 25 min to produce 12 (60%). Selective catalytic hydro-
genation of the pyridine ring of 12, using Adam’s platinum
transitional carbinolamine intermediate, comparable with 4,
and due to the strong preference for the methylindan-nucleus
to adopt an equatorial conformation to the piperidine ring.
Both these factors make attack of the secondary amine at the
keto-groups energetically and sterically unfavourable, which
precludes the formation of the expected eight-membered
ring.
To produce further variants for pharmacological screen-
ing, and to retain the more rigid methylene linkage, we also
synthesizedthe3-pyridyl-substitutedanalogue60, sinceposi-
tioning, and thus biological availability, of the nitrogen atom
can affect biological availability, and hence activity. This
compound (Accessory Materials) was synthesized using a
similar procedure to that of the 2-pyridyl-substituted ana-
logue. However, yields were poor, probably due to the lower
electrophilicity at the 3-position. In addition, quantities of
a side product, characterized as 14 (Fig. 3) were obtained,
which indicated a Michael-type addition of base (methoxide)
to the internal ene-1,4-dione of the 4,5,6,7-tetrahydroindane
nucleus. Yields of 2-phenyl-2-[2-(3-pyridyl)methyl]4,5,6,7-
tetrahydroindan-1,3-dione 58 were not sufficiently high to
attempt subsequent selective catalytic hydrogenation of the
pyridine ring.
1
oxide, yielded 13 (56%), which was deduced by H NMR
to possess a cis-ring fusion. This system did not undergo
a ring cyclization–expansion reaction to a nine-membered
ring-containing system. We believe this is probably due to a
combination of the differing electronic, steric, and ring-strain
effects,[19] relative to the fully unsaturated indan-1,3-dione
system.
Synthesis of 2-Substituted Tetrahydroindan-1,3-diones
Possessing Methylene Linkages to 2- and 3-Substituted
Pyridine and Piperidine Rings
Synthesis of 2-Substituted Indan-1,3-diones and
2-Substituted Tetrahydroindan-1,3-diones Containing
2-Substituted Quinoline, Tetrahydroquinoline,
2-(ω-N-Phthalimidoalkyl)-1,3-diones, and
5-Ethyl-Substituted Pyridyl Nuclei. Selective
and Non-Selective Catalytic Hydrogenation
of the Pyridine Ring Functionality
We were also interested in investigating a shorter pyri-
dine/piperidine to indan-1,3-dione alkyl-linkage, i.e., the
methylene-containing analogue 58 (see Table 1 and Acces-
sory Materials). This was predominantly for two reasons:
first, a methyl linking group between the 2-piperidyl and
indan nuclei would add structural rigidity to the overall
molecule, which is particularly important when considering
potential drug candidates and their specificity of interaction
with receptor sites in the human body; and second, to inves-
tigate whether the overall structure that possessed the shorter
methyl-linkage would undergo the novel ring cyclization–
expansionreactionsequence, in comparision withstructure 2.
The latter would be of interest, particularly when consider-
ing that the syntheses of medium-sized rings by the more
classical direct-cyclization routes are generally difficult to
achieve, require multiple steps, and are usually accompanied
bypooryields. Itwasalsoofinterestsinceitofferedthepoten-
tial to further extend the novel ring cyclization–expansion
reaction to produce eight-membered medium-sized nitrogen-
containing heterocyclic rings in potentially good overall
yields.
The synthesis of 58 was achieved via the initial syn-
thesis of 2-picolyl chloride hydrochloride, derived from
2-pyridylcarbinol,[20] and treatment of this with the sodio-
derivative of 2-phenyl-4,5,6,7-tetrahydroindan-1,3-dione.
(Attempts to alkylate with the corresponding picolyl bro-
mides failed. These proved to be too reactive and underwent
total self-quaternization to form a solid red polymer.) Selec-
tive catalytic hydrogenation of the pyridine ring of 58, in the
presence of hydrochloric acid, yielded the hydrochloride of
59(seeAccessoryMaterials). However, productionofthefree
base did not produce the expected eight-membered ring. This
is probably due to the strain of the required 5-membered ring
We were further interested to investigate the generality of
the ring cyclization–expansion sequence, as undergone by
compound 2, by variation of the heterocyclic component,
in particular, by the introduction of a quinoline ring. This
was achieved by a Michael-type addition of 2-vinylquinoline,
preparedbythemethodofKotanandSurnina,[21] to2-phenyl-
indan-1,3-dione (75%) (Scheme 3) to yield the quinoline
adduct 15. Selective catalytic hydrogenation of the pyridine
ring was achieved at room temperature and pressure using
Adam’s platinium oxide as the catalyst, together with a trace
of acid, to yield the selective 1,2,3,4-tetrahydroquinoline-
reduced compound 16 in a 64% yield. Basification generated
the free amine. However, unlike the corresponding piperidyl
structure 2, it failed to undergo the potential ring cyclization–
expansion reaction to a nine-membered ring. We believe that
this is due to two reasons: (1) a reduction in the nucle-
ophilicity of the amine by conjugation of its lone pair with
the fused benzene ring, and (2) the steric bulk associated
O
H
Ph
N
O
MeO
14
Fig. 3.

Synthesis and Reactions of 2,2-Disubstituted Indan-1,3-diones
63
Table 1. 1,3-Dione-containing structures
Structure
Compound
R
Rꢀ
10
9
11a/11b
58
59
60
17
19
21
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
2-(2-Piperidyl)ethyl
2-(2-Pyridyl)ethyl
2-(2-(N-Acetyl)piperidyl)ethyl
2-(2-Pyridyl)methyl
2-(2-Piperidyl)methyl
2-(3-Pyridyl)methyl
2-(2-Quinolyl)ethyl
2-(2N-Ethylphthalimido)
2-(3N-Propylphthalimido)
O
R
Rꢂ
O
O
R
12
13
Ph
Ph
2-(2-Pyridyl)ethyl
2-(2-Piperidyl)ethyl
Rꢂ
O
O
R
35
36
Ph
Ph
2-(2-Pyridyl)ethyl
2-(2-Piperidyl)ethyl
Rꢂ
O
O
R
38
39
Ph
Ph
2-(2-Pyridyl)ethyl
2-(2-Piperidyl)ethyl
Rꢂ
O
2
Ph
Ph
Me
Ph
Ph
Ph
Ph
Ph
Me
Me
Me
2-(2-Pyridyl)ethyl
2-(5-Ethyl-2-pyridyl)ethyl
2-(2-Pyridyl)ethyl
2-(2-Pyridyl)methyl
2-(2-Piperidyl)methyl
24
61
62
63
15
16
29
41
45
53
O
R
2-(2-Quinolyl)ethyl
Rꢂ
2-(2-(1,2,3,4-Tetrahydroquinolyl)ethyl)
2-(3N-Phthalimidopropyl)
2-(2N-Phthalimidoethyl)
2-(3N-Phthalimidopropyl)
2-(4N-Phthalimidobutyl)
O
with the benzo-fused ring system energetically disfavouring
nucleophilic attack at the keto-groups.
selectively hydrogenated over other present functionalities,
in particular, ketone groups. We believe that in the case of
the fully unsaturated indan-containing system 15, the mostly,
overall, planar geometry of the molecule allows it to adhere,
to a greater extent, planar to the surface of the catalyst, thus
allowing the expected selective reduction of the pyridine
nucleus to occur. However, in the case of the tetrahydroindan-
containing variant 17, the molecule is unable to adhere to
the surface of the catalyst as effectively as the more planar
unsaturated indan-nucleus 15 (see Fig. 4), due to the slight
increase in steric bulk associated with the partially saturated
nucleus (half chair conformation of the cyclohexene ring)
and, more importantly, due to its resultant overall confor-
mation. This possibly leads to the pyridine portion of the
quinoline-ring system being unable to adhere to the catalyst
surface as effectively as the benzene portion of the quinoline
ring. The result of this very small change in overall hydro-
genation geometry leads to the preferential and selective
In the light of the piperidyl-substituted tetrahydroindan-
1,3-dione compound 1 possessing marked analgesic activity,
we prepared additional variants of this compound. The
new quinoline variant 17 was similarly synthesized (see
above) by a Michael-type addition of 2-phenyl-4,5,6,7-
tetrahydroindan-1,3-dione 7 to 2-vinylquinoline (63%)
(Scheme 3) to yield the adduct 17. Selective catalytic
hydrogenation of the pyridine ring of the quinoline nucleus,
over other present functionalities, could not be realized
under identical hydrogenation conditions to those used
for the unsaturated-indan 15. The sole product isolated
had surprisingly undergone reduction to the 5,6,7,8-
tetrahydroquinoline, with concomitant reduction of one of the
keto-groups to yield 18 (Schemes 4 and 5).This is an unusual
result since it is known that in the presence of acid, both
pyridine, and the pyridine portion of quinoline rings, can be

64
C. J. Roxburgh and L. Banting
O
O
O
Ph
Ph
N
N
O
15
17
H2, PtO2, HCl
H2, PtO2, HCl
(atmospheric pressure)
(atmospheric pressure)
O
OH
Ph
Ph
H
N
N
O
O
16
18
Scheme 3. Selective and non-selective catalytic hydrogenations of the pyridine nuclei of fully
unsaturated indan- and tetrahydroindan-1,3-diones.
O
the hydrazine across the activated ene-1,4-dione bond. In
Ph
O
addition, there are further tandem possibilities for the gener-
ated primary amine to subsequently undergo intermolecular
Michael-type additions across the activated ene-1,4-dione
functionality. However, it isinteresting to notethat thesynthe-
sis of the structurally related 3-(N-phthalimidopropyl) deriva-
tive 21 (see the Experimental), followed by hydrazinolyis,
yielded the new ring structure 23, via direct ring cycliza-
tion and condensation, in a surprisingly good 40% yield.
In this instance it is considered that initial formation of the
six-membered carbinolamine-containing intermediate 22 is
energetically far more favourable (hence the 40% yield) than
the possible five-membered carbinolamine-containing inter-
mediate 20, the latter system thus favouring polymerization
reactions over ring cyclization.
N
H2
H2
PtO2 catalyst
Fig. 4. Compound 17 is unable to adhere to the surface of the catalyst
effectively due to the slight increase in steric bulk associated with the
partially saturated nucleus (half chair conformation of the cyclohexene
ring) and, more importantly, due to its resultant overall conformation.
hydrogenation of the benzene ring. (Note: for compound 24
(see below), the introduction of a 5-ethyl substituent onto
the pyridine ring also alters the expected course of selective
catalytic hydrogentation.)
In the light of the novel ring cyclization–expansion reac-
tion of 2, we were further interested to investigate the
applicability of this reaction to systems containing both
tetrahydroindan-1,3-dione and 2-(ω-N-phthalimidoalkyl)-
1,3-dione functionalities, i.e., 19 and 21 (Scheme 4). These
compounds, on hydrazinolysis, produce primary aminoalkyl
groups, as opposed to the secondary amine groups discussed
thus far, and are, therefore, expected to behave and react
differently.
Synthesis of the 2-(N-phthalimidoethyl) compound
19 was achieved via reaction of the sodio-derivative
of 2-phenyl-4,5,6,7-tetrahydroindan-1,3-dione with N-(2-
bromoethyl)phthalimide (38%) (see Accessory Materials).
Hydrazinolysis of the 2-(N-phthalimidoethyl) compound
19 was performed in order to release the primary amine
(de-protection), which was expected to subsequently undergo
intra-molecular ring cyclization–expansion to yield an
eight-membered ring, however the reaction mixture consisted
of a polymer from which no characterizable products could
be obtained. Retrospectively, this is perhaps not surprising
since, in addition to generation of the required free primary
amine, there exists the possibility of nucleophilic attack of
We previously mentioned that the preparation of the
tetrahydro-quinoline substituted system 15, followed by
selective catalytic hydrogenation of the pyridine nucleus, to
undergo the ring cyclization–expansion reaction (in compari-
sonwith 3)failed, principallyfortworeasons. Incontinuingto
investigatethelimitationofthestructuralparametersrequired
to affect the ring-expansion transformation we decided to
make further attempts at modifying the 2-pyridyl grouping
so as to minimize these issues. It was envisaged that this
could be achieved by the placement of a small alkyl group
on the pyridine ring, in this case a 5-ethyl substituent. This
substitution would both reduce the steric effects, present in
the potential carbinolamine intermediate, and, by ‘remote’
placement (5-substitution) of a small alkyl group, minimize
interference of nucleophilic attack of the nitrogen lone-pair
at the keto-groups.
The above predictions were investigated via synthe-
sis of the 5-ethyl-pyridyl analogue 24 (Scheme 5), via a
Michael-type addition between 2-vinyl-5-ethyl pyridine and
2-phenyl-indan-1,3-dione, however, no selective catalytic
hydrogenation of the 5-ethyl-substituted pyridine ring could
be obtained over other present functionality, i.e. the

Synthesis and Reactions of 2,2-Disubstituted Indan-1,3-diones
65
O
O
Ph
NH2NH2, ꢁ
Polymer
N
O
H
N
O
NH2NH2
X
O
19
Ph
O
20
H
N
O
O
O
N
O
Ph
Hꢃ
NH2NH2, ꢁ
N
O
ꢀH2O
Ph
Ph
O
O
21
22
23
Scheme 4.
O
R
Ph
Ph
Ph
NaBH4
MeOH
H2, 5% Pd
(R ꢄ OH)
N
N
N
O
R
28
25 R ꢄ OH
26 R ꢄ OAc
27 R ꢄ p-Nitrobenzoate
24
Scheme 5.
ketone groups. We believe the unexpected non-selective
hydrogenation of the pyridine ring arises from the increase
in steric bulk associated with alkyl-substitution, which dis-
favours adhesion to the catalyst surface. Based on the
above results, and the fact that other appropriately substi-
tuted 2-vinyl pyridine analogues were not readily accessible,
no attempts to synthesize further derivatives were made.
(Some attempts were made at synthesizing α- and β-methyl-
substituted 2-vinyl pyridine analogues; however, the yields
were low and thus did not lend themselves to a large-
scale synthesis required in order to further investigate the
ring-expansion transformation.)
With the aim to produce some further new indan-1,3-dione
structures, compound 24 was reduced using sodium boro-
hydride to yield a mixture of the cis- and trans-alcohols 25
which could not be separated using chromatography. Deriva-
tization of the diol as its di-acetate 26 or di-p-nitrobenzoate
27 allowed partial isomer separation. Hydrogenolysis of the
diol over 5% Pd/C cleanly gave the new indan system 28 in
good yield.
addition to 2-vinylpyridine, to yield the new indan-1,3-dione
61 (see Accessory Materials). Selective catalytic hydrogena-
tion of the pyridine ring, over other present functionalities,
could not be achieved (the resulting mixture produced a poly-
meric material and column chromatography yielded no char-
acterizable products). The authors attribute this instability
to the 2-methyl-1,3-dione functionality, and its containment
within a strained five-membered ring.
Further structural modifications, in order to ascertain
the necessary functional groups and structural paramenters
for the ring cyclization–expansion to occur, were made. In
particular, we were interested in determining whether a short-
ening of the 2-ethyl linking group to a 2-methyl group,
for the fully unsaturated indan-1,3-dione, would still allow
the ring cyclization–expansion reaction to progress. This
was achieved via the initial synthesis of 2-picolyl chloride
hydrochloride, mentioned previously, and alkylation of this
molecule with the sodio-derivative of 2-phenyl-indan-1,3-
dione in the presence of two equivalents of base to yield
62 (see Accessory Materials). Selective catalytic hydrogena-
tion of the pyridine ring yielded the piperidyl compound
63, however, no products corresponding to a possible ring
cyclization–expansion reaction, as for the 2-ethyl-piperidyl
compound 3, were identified. We believe this is due to the
piperidyl group not being able to adopt the conformation
required to allow nucleophilic attack of the nitrogen lone pair
at the carbonyl groups, and simultaneously, it not being able
to form the required, strained, and energetically unfavourable
The structural variations made thus far have primarily
centred on modifications to the indan- and 2-pyridylalkyl-
groupings. We were further interested to test the generality
and applicability of the novel ring expansion reaction to sys-
tems possessing a 2-alkyl substituent on the indan structure,
as opposed to the 2-phenyl of the parent molecule.
This was achieved by the initial synthesis of
2-methylindan-1,3-dione,[22] and subsequent Michael-type

66
C. J. Roxburgh and L. Banting
Ph
HO
O
O
Ph
NH2NH2
N
N
ꢁ
O
O
O
29
30
LiAlH4,
THF, ꢁ
Ph
4
3
N
Ph
Hꢀ
Ph
R3AlO
6
7
1R
2
10bR
10a
Nꢃ
N
10
8
9
33
32
Scheme 6.
31
five-membered-ring carbinolamine-intermediate necessary
for ring cyclization and subsequent expansion.
−12, J4,3 12) and δ 3.2 (Jgem −12), which supports a chair
conformation for the piperidine ring. The stereochemistry
is thus depicted as shown in 33 and is assigned as rel-(1R,
10bR)-1,2,3,4,6-hexahydro-1-phenylpyrido[2,1-a]isoindole.
Synthesis of the New Bridged Ring System rel-(1R,10bR)-
1,2,3,4,6-Hexahydro-1-phenylpyrido[2,1-a]isoindole 33
by Selective Hydride Reduction of 1,2,3,10b-Hydroxy-
1-phenylpyrido[2,1-a]isoindol-6(4H)-one 30
Synthesis of 2-Phenyl-2-[2-(2-pyridyl)ethyl]-
Substituted 1,3-Diones Possessing Six- and
Seven-Membered Ring-Containing 1,3-Dione Units
Pyridine, quinoline, and their selectively reduced secondary
amine containing structures have been extensively explored,
as described throughout this paper. Also of interest, par-
ticularly in ascertaining their ability to undergo the ring
cyclization–expansion reaction sequence, are primary amine
containing structures.
The synthesis of 2-phenyl-2-(3-N-phthalimidopropyl)-
indan-1,3-dione 29 has been reported, and its hydrazinolysis
yielded 1,2,3,4,6-tetrahydro-10b-hydroxy-1-phenylpyrido-
[2,1-a]isoindol-6(4H)-one 30 as the major product
(Scheme 6).[23]
In the light of the novel route to medium-ring nitrogen-
containing heterocycles, afforded by catalytic hydrogenation
of the pyridine ring of 2, we decided to explore the appli-
cability of this reaction to analogous systems possessing the
1,3-dione functionality within both six- and seven-membered
rings, in the hope of obtaining, by a similar transformation,
ten- and eleven-membered rings.
Modification of the size of the ring containing the 1,3-
dione functionality 35, 36, 38, and 39 (Scheme 7) was
achieved by the initial syntheses of the six-membered
34[24–26] and seven-membered 37[27] 1,3-dione-containing
units. Michael-type addition of 2-vinylpyridine to these
nuclei under aprotic conditions, in a seven day benzene
reflux, using Triton B as the catalyst gave the new pyri-
dine adducts 35 and 38, respectively. Selective catalytic
hydrogenation of the pyridine rings in ethanol solvent over
platinum oxide catalyst, containing hydrochloric acid, gave
the piperidyl variants 36 and 39, respectively, after basifi-
cation. Both systems were isolated and stable as their free
bases. It is interesting to note that unlike the five-membered
1,3-dione-containing ring system, both the six- and seven-
membered 1,3-dione-containing rings did not undergo the
potential ring cyclization–expansion transformation to ten-
and eleven-membered nitrogen heterocycles. The authors
believe that the failure of the above two systems to undergo
the ring cyclization–expansion transformation are almost
entirely related to the relief of ring strain on ring-expansion.
(The fully unsaturated indan-1,3-dione, possessing a five-
membered 1,3-dione-containing ring is considerably more
strained in both the six- and seven-membered 1,3-dione-
containing rings.)These two results proved useful during this
In an attempt to produce some further new and unexplored
nitrogen-containing heterocycles, we attempted a lithium alu-
minium hydride reduction of this compound, which produced
a surprising result (Scheme 6). It is postulated that reduction
of the tertiary amide occurs with concomitant coordination of
the alcohol group to the aluminium to give the intermediate
31. Displacement of the aluminium-coordinated alcohol by
anchimeric assistance from the tertiary amine probably yields
the iminium ion 32, which is further reduced by the hydride
to yield the new 1,2,3,4,6-hexahydro-1-phenylpyrido[2,1-a]-
isoindole 33. The IR spectrum of 33 showed no Bohlman
bands (found in systems containing trans-ring fusions) thus
indicating a cis-ring fusion. The C(10b) bridgehead proton
1
in the H NMR spectrum at δ 3.8 appeared as a slightly
broadened singlet, which indicated a torsional angle of
approx. 60◦ between the C(1)-H and C(10b)-H bonds. The
phenyl substituent was thus assigned the axial orientation.
Further evidence for the axial phenyl group is the magnitude
of couplings between the C(1) and C(2) protons, of 5 Hz,
consistent with bisection of the C(2) methylene by the C(1)-
H bond. The C4Hax and C4Heq protons appear at δ 2.57 (Jgem

Synthesis and Reactions of 2,2-Disubstituted Indan-1,3-diones
67
O
O
O
2-Vinylpyridine,
benzene, Triton B, ꢁ
Ph
Ph
N
O
34
35
H2, PtO2,
EtOH, Hꢃ
O
Ph
H
N
O
36
O
O
2-Vinylpyridine,
Ph
benzene, Triton B, ꢁ
Ph
N
O
O
37
38
H2, PtO2,
EtOH, Hꢃ
O
Ph
H
N
O
39
Scheme 7.
synthetic program in dictating the structure and functionality
that is needed in the design of further possible chemical enti-
ties that may potentially undergo the novel ring-expansion
reaction.
In particular, we were interested in applying this reaction
to the three new 2-methyl-2-(ω-(N-phthalimidoalkyl))indan-
1,3-dione structures, and specifically those containing
2-(N-phthalimidoethyl) 41, 3-(N-phthalimidopropyl) 45, and
4-(N-phthalimidobutyl) 53 groups, which possess the poten-
tial to generate eight-, nine-, and ten-membered nitrogen-
containing heterocycles by the aforementioned reaction
sequence.
Ring Cyclization–Expansion–Transannular Contraction,
and Cyclization Reactions of Hydrazinolyzed
2-Methyl-2-(ω-N-phthalimidoalkyl)indan-1,3-diones
Initial syntheses of the new 2-methyl-2-(ω-(N-
phthalimidoalkyl))indan-1,3-diones were achieved using
the method previously and successfully adopted for the
syntheses of 2-phenyl-2-(ω-(N-phthalimidoalkyl))indan-1,3-
diones,[28] however, the yields were much lower and were
accompanied by significant quantities of side products.
This is exemplified for the synthesis of 2-methyl-2-(3N-
phthalimidopropyl)indan-1,3-dione 45, which produced, in
addition to the required C-alkylated product 45 (14%),
the side product (40%) 40 (Fig. 5) which was isolated
and identified as the O-alkylated compound. This is a
contrasting result to that obtained when the analogous
alkylation reactions, i.e., the generated carbanion of
2-phenyl-indan-1,3-dionewasreactedwiththecorresponding
In continuing to investigate and explore the generality of
the novel ring cyclization–expansion reaction found to occur
for structure 3 (Scheme 1), we decided to replace both the
2-aryl substituent with a 2-alkyl group, and, simulta-
neously, the 2-ethyl-piperidyl group with various ω-(N-
phthalimidoalkyl) groups (Scheme 8). This would further
enable us to identify the applicability and scope of this
reaction to indan-1,3-diones possessing both 2-alkyl and
ω-(N-phthalimidoalkyl) groups to potentially lead to the
generation of several new medium-ring alkyl-substituted
nitrogen-containing heterocycles. In addition, the structures
and reactions of these particular compounds had not previ-
ously been explored.

68
C. J. Roxburgh and L. Banting
O
N
O
O
O
O
O
O
O
H
N
Me
Me
NH2NH2
ꢁ
N
O
NH2
O
O
Me
HO
Me
41
42
43
44
O
O
O
O
Me
H
N
Me
NH2NH2
N
N
NH2
ꢁ
HO
O
O
O
Me
Me
45
46
47
48
N
O
N
O
H2N
NH2NH2
Me
ꢀH2O
Me
Me
N
NH2
O
49
50
51
NH2NH2
N
Me
N
N
H2N
52
O
O
O
O
HN
HO
O
Me
Me
NH2NH2
ꢀH2O
N
ꢁ
NH2
Me
Me
O
O
O
53
54
Scheme 8.
55
56
2-(N-phthalimidoalkyl)bromo compounds (alkyl = ethyl,
propyl, butyl), were performed. In this latter instance we
believe the generated carbanions possess their charge local-
ized and stabilized on C-2 by the 2-phenyl group, which
results predominantly in C-alkylation. However, in the case of
the 2-methyl substituted indan-1,3-diones, a destabilization
of the generated carbanion occurs (via an inductive effect
from the 2-methyl group), which results in an equilibrium
situation favouring predominant formation of the enolate
structure (Scheme 9) and thus preferential O-alkylation.
Synthesis of 44 was achieved via hydrazinolysis of
41 to yield the new heterocyclic system 44 (25%). The
intermediate primary amino structures 42/43 were not
isolated, forming, in situ, the ring structure 44. This is
postulated to occur by initial nucleophilic attack of the
amine on one of the keto-groups, to yield an intermediate
cyclic-carbinolamine (compare with 4, Scheme 1). Further
rearrangement, by ring-expansion, of this carbinolamine
produces the eight-membered ring intermediate 43, which
subsequently, by transannular-cyclization, yields the new
tricyclic system 44. (This eight-membered ring-containing
intermediate 43, unlike 5, possesses a secondary amide
functionality, which allows further transannular reaction to
produce the 5,5-ring-contracted amide structure 49 to occur.)
Hydrazinolysis of the 3-(N-phthalimidopropyl) derivative
45 yielded 48 via
a similar cyclization–expansion-
transannular reaction, however, in addition, the two hydra-
zone products 51 and 52 were produced. In this instance it
is possible that the six-membered carbinolamine-containing
ring intermediate (compare with 4) is energetically more
stable than that of the five-membered ring-intermediate pro-
duced upon the hydrazinolysis of 41. This intermediate,
therefore, probably does not have to so readily (or possibly
more slowly) undergo ring expansion to relieve this strain,
i.e., formation of a 5,6-ring-fused system, as opposed to
a 5,5-ring-fused system: This facilitates dehydration of the
carbinolamine intermediate to the imine 50. Subsequent reac-
tion of the keto group with hydrazine yields the two isomeric
hydrazones 51 and 52.
The transition of reaction products throughout this
sequence of compounds is further exemplified by hydrazi-
nolysis of the 4-(N-phthalimidobutyl) compound 53, which

Synthesis and Reactions of 2,2-Disubstituted Indan-1,3-diones
69
Oꢀ
O
O
ꢀ
Me
O
Me
Me
(CH2)3
O
O
N
O
Scheme 9.
O
40
Fig. 5.
O
O
O
O
N
H
(i) NaBH4, MeOH
(ii) dil HCl
yields the azepine-containing structure 56 as the sole
product. In this instance it is interpreted that the seven-
membered carbinolamine intermediate 55 does not energet-
ically favour/require ring expansion, but simply undergoes
ring cyclization followed by condensation.
HCl
N
H
H
H
Ph
Ph
H
H
6
57
Overall, in transcending this series of homologous
ω-(N-phthalimidoalkyl)-substituted indan-1,3-diones there
is a very subtle, but complex, interplay between steric, ther-
modynamic, and electronic factors. This combination of
factors dictates the outcome of these reactions by influencing
both the nature of products and their relative amounts.
Scheme 10.
may wish to further explore these compounds as potential
drug candidates. Specifically, we reported the synthesis of 1
which possessed marked analgesic activity (100% inhibition
referenced to codeine) and a further five compounds exhibit-
ing moderate activity to several pharmacological tests.
In addition, we have thoroughly explored and inves-
tigated the applicability and generality of the novel
ring cyclization–expansion reaction of 2-phenyl-2-[2-(2-
piperidyl)ethyl]indan-1,3-dione 3, described in Scheme 1,
to systems containing a variety of functionalities. We con-
clude that in order for the ring cyclization–expansion reaction
sequence to occur, a fully unsaturated indan-1,3-dione group
needs to be present. A 2-phenyl group is not a requirement
as ring-expansion occurs for the 2-methyl substituted com-
pounds 42 and 46. Furthermore, a 2-(piperidyl)ethyl group is
also not a necessity as ring expansion was found to occur with
both 2-N- and 3-N-phthalimidopropyl groups, however, in
these latter two cases, the intermediates possessing secondary
amide functionality undergo subsequent ring-contraction
reactions to form the new nitrogen-containing heterocycles
44, 48, 51, and 52. A limitation of this ring-expansion
reaction, related to the length of the ω-N-phthalimidoalkyl
group, is noted, i.e., the 4-N-phthalimidobutyl-substituted
compound 53 undergoes simple ring cyclization to form the
new azepine-containing heterocycle 56 (azepine-containing
structures are also of medicinal interest since they also pos-
sess many of the biological properties mentioned above for
medium-sized rings).[14–16]
In spite of the fact that structures containing the 1,2,3,4-
tetrahydroindan-1,3-dione nuclei failed to undergo the ring-
expansion reaction sequence, they were not totally with-
out reactivity since the 2-phenyl substituted compound 21
underwent direct cyclization to form the new heterocyclic
system 23.
Overall, we have defined many of the functional
group parameters necessary for the novel ring cyclization–
expansion reaction sequence to occur. This will be of use to
chemists who may wish to consider and utilize this useful
ring-expansion reaction as part of a wider synthesis, or to
synthesize other medium-sized rings.
Hydride-Induced Transannular–Contraction Reaction
of the Nine-Membered Ring-Containing Nitrogen
Heterocycle 6 (Lactam-to-Lactone Interchange)
Lastly, in continuing to investigate the chemistry of some
of the final ring structures we turned our attention to
the nine-membered ring-containing structure 6,[11,29] whose
modified structures, and resultant stereochemistries have
been studied.[19,30] Further attempts to modify the struc-
ture of this medium-sized ring, and explore its reactivity and
behaviour, led us to investigate a sodium borohydride reduc-
tion (Scheme 10) with the expectation of realizing selective
reduction of the keto-group.This was initially achieved, how-
ever, the resulting alcohol was surprisingly found to undergo
a lactam-to-lactone transannular atom interchange to yield
the novel ring-contracted lactone-containing system 57.
Pharmacological Test Results of Analgesic
Screening of Selected Compounds
The hydrochloride of 2-phenyl-2-[2-(2-piperidyl)ethyl]-
4,5,6,7-tetrahydroindan-1,3-dione 1 possesses marked anal-
gesic activity (100% inhibition referenced to codeine).[10]
The previously unreported pharmacological test results of
ten selected compounds, for analgesic activity, are detailed in
Table 2, of which six compounds possess between 58% and
8% inhibition, with a further two exhibiting moderate and
‘some’ activity. (Note: we have/had no control over the deci-
sion of which compounds were pharmacologically tested, or
over the limited results with which we were furnished.)
Conclusions
We have described the synthesis of thirty-eight new nitrogen-
containing heterocyles, which are related in structure to many
current classes of known drugs. These structures possess the
potential to be biologically active and may, therefore, be of
benefit to medicinal chemists working in these areas, and who

70
C. J. Roxburgh and L. Banting
Table 2. Pharmacological test results of analgesic screening of selected compounds
All compounds administered orally.
Compound
Ref.
Test (Mice)
Dose [mg kg−1
]
ResultA
RemarksB
O
Acetylcholine induced writhing
Hot plate
Inflamed paw pressure
Normal paw pressure
Antipyretic activity
100
100
75
75
75
++
−
100% inhibition
Inactive
Marked activity
Negligible activity
Inactive
Ph
H
N
HCl
++
O
O
−
Ph
Acetylcholine induced writhing
Hot plate
100
100
42% inhibition
0% inhibition
−
N
O
N
[2]
Acetylcholine induced writhing
Hot plate
100
100
−
−
H
O
H
Ph
O
Ph
[23]
[23]
Acetylcholine induced writhing
Hot plate
100
100
25% inhibition
8% inhibition
−
N
O
O
N
Acetylcholine induced writhing
Hot plate
100
100
−
−
Ph
N
[23]
Acetylcholine induced writhing
Hot plate
100
100
++
−
58% inhibition
Moderate activity
Ph
O
N
N
[23]
[28]
Acetylcholine induced writhing
Hot plate
100
100
−
−
17% inhibition
0% inhibition
HO
O
Ph
Ph
Acetylcholine induced writhing
Hot plate
100
100
−
−
O
N
[28]
[28]
Acetylcholine induced writhing
Hot plate
100
100
−
−
0% inhibition
HO
O
Ph
N
Acetylcholine induced writhing
Hot plate
100
100
25% inhibition
Some activity
+
Ph
A
++ Marked activity, + moderate activity, negligible activity, − inactive.
B Reference drug codeine.

Synthesis and Reactions of 2,2-Disubstituted Indan-1,3-diones
71
Lastly, several interesting new compounds and selec-
tive functional group interconversions have been uncovered
during the course of this programme, these include: the
reductive formation of the new bridged nitrogen-containing
heterocyclic system 33, the transannular-contraction reac-
tion of 6 to form the new lactone-containing system 57, and
finally, the selective catalytic hydrogenation, under identical
physical conditions, of the pyridine and benzene rings of the
quinoline-substituted systems 15 and 17.
recrystallizations from ether/ethanol gave hexahydro-2-phenyl-2-[2-(2-
pyridyl)ethyl]-indan-1,3-dione 12 as white plates (1.88 g, 60%), mp
116–117◦C. (Found: C 79.3, H 6.9, N 4.2%. C22H23NO2Cl requires
C 79.3, H 7.0, N 4.2%.) νmax (Nujol)/cm−1 1755, 1730, 1600, 750.
δH (CDCl3) 7.3 (9H, ArH), 3.0 (2H, m, (COCH)2), 2.3–2.8 (4H, m,
(COCHCH2)2), 1.7 (4H, m, (COCHCH2CH2)2).
Hexahydro-2-phenyl-2-[2-(2-piperidyl)ethyl]-indan-
1,3-dione Hydrochloride 13
Hexahydro-2-phenyl-2-[2-(2-pyridyl)ethyl]4,5,6,7-tetrahydroindan-1,3-
dione 12 (0.5 g, 1.5 mmol) was dissolved in methanol and concentrated
hydrochloric acid (0.2 g). PtO2 catalyst (0.15 g) was added and the
mixture was hydrogenated at atmospheric pressure and room tem-
perature until the calculated amount of hydrogen had been taken up.
The catalyst was filtered off and the methanol was reduced in vol-
ume. On allowing the solution to stand, a white crystalline solid was
deposited. Recrystallization of this from methanol gave hexahydro-2-
phenyl-2-[2-(2-piperidyl)ethyl]-indan-1,3-dione hydrochloride 13 as a
white powder (0.32 g, 56%), mp 177–179◦C. (Found: C 67.8, H 7.8,
N 3.6%. C22H30NO2Cl·0.75H2O requires C 67.8, H 7.7, N 3.6%.) νmax
(Nujol)/cm−1 3500, 1740, 1710, 1600, 750. δH (CDCl3) 9.5 (2H, s,
NH2), 7.3 (5H, m, ArH), 3.5 (1H, m, (Ph)CCH2CH2CHN).
Experimental
Elemental analyses were carried out at Glaxo Group Research Limited.
1H NMR spectra were determined at 60 MHz with a Varian T60 spec-
trometer and for 270 MHz with a Bruker Spectrospin WH-270 FT NMR
spectrometer. 13C NMR spectra were determined with a JEOL GSX 270
FT NMR spectrometer. Mass spectra were recorded on a JEOL JMS-
DX303 GC/mass spectrometer. Column chromatography was carried
out over silica. Melting points were determined on a Kofler hot-stage
apparatus and are uncorrected.
2-Phenyl-2-[2-(2-piperidyl)ethyl]-4,5,6,7-tetrahydroindan-
1,3-dione 10
2-Phenyl-2-[2-(2-piperidyl)methyl]-4,5,6,7-tetrahydroindan-
A solution of 2-phenyl-2-[2-(2-pyridyl)ethyl]4,5,6,7-tetrahydroindan-
1,3-dione 9 (4 g, 12 mmol) in methanol (100 mL) and concentrated
hydrochloric acid (1.5 mL) were hydrogenated at atmospheric pres-
sure and room temperature in the presence of PtO2 (0.3 g) until the
calculated amount of hydrogen had been taken up. The catalyst was fil-
tered off and the solvent reduced in volume. Recrystallization of the
yellow solid, which separated on standing, from ethanol gave 2-phenyl-
2-[2-(2-piperidyl)ethyl]-4,5,6,7-tetrahydroindan-1,3-dione hydrochlo-
ride 10asapaleyellowcrystallinesolid(4.1 g, 91%), mp240◦C. (Found:
1,3-dione Hydrochloride 59
2-Phenyl-2-[2-(2-pyridyl)methyl]4,5,6,7-tetrahydroindan-1,3-dione 58
(4 g, 13 mmol) was dissolved in methanol and concentrated hydrochlo-
ric acid (0.5 mL). PtO2 Catalyst (0.15 g) was added and the mix-
ture was hydrogenated on a Parr Hydrogenator until the calculated
pressure change had occurred. The catalyst was filtered off and
the solvent was removed under reduced pressure to give an off-
yellow oil. On standing at low temperature for one week in absolute
ethanol, a yellow crystalline solid was formed. Several recrystallizations
from ethyl acetate/methanol gave 2-phenyl-2-[2-(2-piperidyl)methyl]-
4,5,6,7-tetrahydroindan-1,3-dione hydrochloride 59 (0.7 g, 16%), mp
178–180◦C. (Found: C 68.4, H 7.2, N 3.7%. C21H26NO2Cl·0.5H2O
requires C 68.4, H 7.1, N 3.7%.) νmax (Nujol)/cm−1 3400, 1760, 1710,
1625, 1600, 750. δH (CDCl3) 9.6 (1H, s, N+Ha), 8.1 (1H, s, N+Hb),
7.3 (5H, s, ArH), 3.5 (1H, d, N+CHeq), 2.9 (1H, m, N+CH(CH2)2),
2.8 (1H, m, N+CHaxCH2).
C 70.5, H 7.5, N 3.7%. C22H28NO2Cl requires C 70.6, H 7.5, N 3.7%.)
νmax(CHCl3)/cm−1 1740, 1700, 1660, 1600. δH (CDCl3) 8.5–8.7 (5H,
m, aromatic), 1.1–2.2 (18H, m, aliphatic).
N-Acetyl-2-phenyl-2-[2-(2-piperidyl)ethyl]-4,5,6,7-
tetrahydroindan-1,3-dione 11a/11b
2-Phenyl-2-[2-(2-piperidyl)ethyl]-4,5,6,7-tetrahydroindan-1,3-dione10
(0.5 g, 1.5 mmol) was dissolved in a 1 : 1 mixture of dried triethylamine
and pyridine (20 mL) and the mixture was cooled to 0◦C. Acetylchlo-
ride (0.6 g, 15 mmol) was slowly added and the mixture was allowed
to stand for 24 h at room temperature. After this period, the mix-
ture was extracted with ether and the combined extracts were dried
(Na2SO4). The ether was removed under reduced pressure to give an
oil. Chromatography of the oil over silica using ether/petroleum ether
(20/80) as the elutant gave N-acetyl-2-phenyl-2-[2-(2-piperidyl)ethyl]-
4,5,6,7-tetrahydroindan-1,3-dione 11a/11b as a yellow oil (0.31 g,
55%). (Found: C 76.0, H 7.7, N 3.6%. C24H29NO3 requires C 75.9,
H 7.7, N 3.7%). νmax (CHCl3)/cm−1 1730, 1685, 1630, 1600. δH
(CDCl3) 4.7 (1H, m, PhCCH), 4.5 (1H, dd, PhCCH2CH), 3.7 (1H,
m, NCH(CH2)2), 3.5 (1H, br d, NCHeq), 3.0 (1H, br t, NCHax), 2.1
(3H, s, CH3), 2.06 (3H, s, CH3). m/z 379 (M+), 336 (M+ − CH3CO),
154 (M+ − CH2CH2CH(CH2)4NCOCH3), 91 (PhCH2), 77 (Ph), 43
(CH3O).
2-Phenyl-2-(2-[2-(1,2,3,4-tetrahydroquinolin-yl)]ethyl)indan-
1,3-dione 16
2-Phenyl-2-(2-quinolylethyl)indan-1,3-dione 15 (11.7 g, 31 mmol) was
dissolved in methanol and concentrated hydrochloric acid (3.7 mL).
PtO2 catalyst (0.7 g) was added and the mixture hydrogenated at atmo-
spheric pressure and room temperature until the calculated amount of
hydrogen had been taken up.The catalyst was filtered off and the solvent
was removed under reduced pressure to give a thick dark oil.The residue
was treated with aqueous sodium bicarbonate solution and extracted
with chloroform. Passage of the organics over silica gave 2-phenyl-
2-(2-[2-(1,2,3,4-tetrahydroquinolin-yl)]ethyl)indan-1,3-dione 16 as a
slightly yellow oil (7.6 g, 64%), Rf (chloroform) 0.54. (Found: C 82.2,
H 6.2, N 3.5%. C26H23NO2 requires C 81.9, H 6.0, N 3.7%.) νmax
(liquid film)/cm−1 3400, 1735, 1700, 1600, 750. δH (CDCl3) 6.8–7.8
(15H, ArH), 3.72 (1H, s, NH), 3.2 (1H, m, NCHeq), 2.74 (2H, m,
ArCH2), 2.4 (2H, m, C(Ph)CH2), 1.95 (1H, m, ArCH2CHeq), 1.66 (1H,
m, ArCH2CHax), 1.42 (2H, m, C(Ph)CH2CH2). m/z 381 (M+).
Hexahydro-2-phenyl-2-[2-(2-pyridyl)ethyl]-indan-1,3-dione 12
Zinc dust (5 g) was added to
a solution of 2-phenyl-2-[2-(2-
pyridyl)ethyl]4,5,6,7-tetrahydroindan-1,3-dione 9 (2.6 g, 7.9 mmol) in
glacial acetic acid (400 mL) and the mixture heated in a boiling
water bath for 25 min. The zinc was filtered off and the solvent
was removed under reduced pressure. Saturated sodium bicarbonate
solution was added and the mixture was extracted with chloroform.
The combined extracts were dried (Na2SO4) and the solvent was
removed under reduced pressure to give a thick oil. Crystallization
occurred from ether/ethanol, 50/50, to give a white solid. Several
2-Phenyl-2-[2-(2-quinolyl)ethyl]-4,5,6,7-tetrahydroindan-
1,3-dione 17
2-Phenyl-4,5,6,7-tetrahydroindan-1,3-dione(10 g, 44 mmol)washeated
under reflux for 40 h with 2-vinylquinoline (8.5 g, 55 mmol) in abso-
lute ethanol (250 mL). The solution was concentrated to give an
orange/yellow oil that was left to stand overnight whereupon a
bright yellow crystalline solid formed. Recrystallization from absolute

72
C. J. Roxburgh and L. Banting
ethanol gave 2-phenyl-2-[2-(2-quinolyl)ethyl]-4,5,6,7-tetrahydroindan-
1,3-dione 17 as a yellow crystalline sold (10.6 g, 63%), mp 111–112◦C.
(Found: C 76, H 6.6, N 3.9%. C22H23NO3 requires C 75.8, H 6.59,
N 4.0%.) νmax (Nujol)/cm−1 1730, 1680, 1630, 1600. δH (CDCl3) 1.8
(4H, m, (COCCH2CH2)2), 2.4 (4H, m, (COCH2)2), 2.6–3.0 (4H, m,
CH2CH2). m/z 379 (M+).
1,2,3,10b-Tetrahydro-10b-hydroxy-1-phenylpyrido[2,1-a]isoindol-
6-(4H)-one 30
2-Phenyl-2-(3-phthalimidopropyl)indan-1,3-dione 29 (20 g) was stirred
and heated under reflux with hydrazine hydrate (6 mL) in ethanol
(500 mL) for 1 h.The phthaldydrazideformed on cooling was filtered off
andthefiltrateevaporatedtodryness. Chromatographicseparationofthe
residue(10 g)oversilicausingdiethyletherastheelutantgave 1,2,3,10b-
tetrahydro-10b-hydroxy-1-phenylpyrido[2,1-a]isoindol-6-(4H)-one 30
as a white crystalline solid (3 g, 40%), mp 224–225◦C. (Found: C
77.5, H 6.2, N 5.2%. C18H17NO2 requires C 77.4, H 6.1, N 5.0%).
νmax (CHBr3)/cm−1 3560, 3320, 1680, 1600. λmax (EtOH)/nm 261.
δH [(CD3)2SO] 7.1–7.7 (8H, m, ArH), 6.5 (1H, s, OH), 6.3 (1H,
d, ArH), 4.1 (1H, J4ax,4eq 12.5, J4eq,3ax 3.75, 4eq-H), 3.15 (1H,
3-Hydroxy-2-phenyl-2-[2-(2,5,6,7,8-tetrahydroquinolyl)ethyl]-
4,5,6,7-tetrahydroindan-1-one Hydrochloride 18
2-Phenyl-2-[2-(2-quinolyl)ethyl]-4,5,6,7-tetrahydroindan-1,3-dione 17
(6 g, 16 mmol) was dissolved in methanol and concentrated hydrochloric
acid (5 mL). PtO2 (0.5 g) was added and the mixture was hydro-
genated on a Parr Hydrogenator until the calculated pressure change
had occurred. The catalyst was filtered off and the solvent was removed
under reduced pressure. On standing in ethanol overnight a white crys-
talline solid was formed. Recrystallization from methanol/ethyl acetate
gave 3-hydroxy-2-phenyl-2-[2-(2,5,6,7,8-tetrahydroquinolyl)ethyl]-4,5,
6,7-tetrahydroindan-1-one hydrochloride 18 as a white crystalline
solid (1.7 g, 26%), mp 176–178◦C. (Found: C 72.0, H 7.2, N
3.2%. C26H28NO2Cl·0.5H2O requires C 72.1, H 6.9, N 3.2%.) νmax
(Nujol)/cm−1 3500, 3300, 1700, 1625, 1600. δH (CDCl3) 16.5 (1H, s,
OH), 7.1–7.9 (9H, ArH), 6.8 (1H, s, CHOH).
J4ax,4eq = J4ax,3ax = 12, 4ax-H), 2.5 (1H, m, 1ax-H), 2.35 (1H, J2ax,2eq
=
J2ax,1ax = J2ax,3ax = 12.5, J2ax,3eq = 3.75, 2ax-H), 1.6–2.0 (2H, m, 2eq
-
H, 3eq-H), 1.45 (1H, m, 3ax-H).
A second fraction eluted with diethylether gave 2,3,4,4a-tetrahydro-
4a-phenylindeno[1,2-b]pyridine-5-one as a yellow crystalline solid
(1.5 g), mp 95–96◦C. (Found: C 82.9, H 5.9, N 5.4%. C18H15NO
requires C 82.7, H 5.8, N 5.4%). νmax (CDCl3)/cm−1 1720, 1660, 1600.
δH (CDCl3) 7.2–8.2 (9H, m, ArH), 3.95 (2H, t, CH2N), 2.4–2.8 (2H, m,
C(Ph)CH2), 1.45–2.2 (2H, m, CH2CH2CH2). A third fraction eluted
with diethylether gave 2,3-dihydro-1-phenyl-pyrido[2,1-a]isoindol-
6(4H)-one as a pale yellow crystalline solid (0.5 g), mp 142◦C. (Found:
C 82.6, H 5.9, N 5.3%. C18H15NO requires C 82.7, H 5.8, N 5.4%).
2,3,4,4a,6,7,8,9-Octahydro-4a-phenylindeno-
[1,2-b]pyridine-5-one 23
2-Phenyl-2-(3-phthalimidopropyl)-4,5,6,7-tetrahydroindan-1,3-dione21
(8.1 g) was stirred and heated under reflux with hydrazine hydrate
(1 mL) in ethanol (200 mL) for 1.5 h. The phthalhydrazide formed
was filtered off and the filtrate was evaporated to dryness to give
a yellow oil (3 g). The oil was chromatographed over silica using
diethyl ether as the elutant to give to give 2,3,4,4a,6,7,8,9-octahydro-
4a-phenylindeno[1,2-b]pyridine-5-one 23 as a white crystalline solid
(2 g, 40%), mp 129–130◦C. (Found: C 81.8, H 7.2, N 5.2%. C18H19NO
requires C 81.5, H 7.2, N 5.2%.) νmax (CHBr3)/cm−1 1700, 1650.
λmax (EtOH)/nm 261. δH (CDCl3) 7.1–7.4 (5H, m, ArH), 3.75 (2H,
m, CH2CH2N(CH2)2), 2.0–2.8 (5H, m, C-6H, C-9H, C-4Heq), 1.4–1.6
(7H, m, C-3H, C-4Hax, C-7H, C-8H).
rel-(1R,10bR)-1,2,3,4,6-Hexahydro-1-phenylpyrido-
[2,1-a]isoindole 33
1,2,3,10b-Tetrahydro-10b-hydroxy-1-phenylpyrido[2,1-a]isoindol-6-
(4H)-one 30 (2 g, 7.1 mmol) in tetrahydrofuran (THF; 50 mL) was
added slowly to a stirred slurry of lithium aluminium hydride (0.5 g)
in THF (25 mL). The mixture was heated and boiled under reflux for
30 h during which time the solution darkened. Water was carefully
added and the solid precipitate filtered off and washed with chloro-
form. The solution was extracted with chloroform and the combined
extracts were dried (Na2SO4). Removal of the solvent gave a thick dark
oil. Chromatography over silica using diethylether as the elutant gave
rel-(1R,10bR)-1,2,3,4,6-hexahydro-1-phenylpyrido[2,1-a]isoindole 33
as colourless needles (71 mg), mp 121–123◦C. (Found: C 86.7, H 7.6,
N 5.6%. C18H19N requires C 86.7, H 7.7, N 5.6%). νmax (Nujol)/cm−1
1600, 750. δH (CDCl3) 6.73–8.6 (9H, m, ArH), 4.2 & 3.55 (2H, AB,
NCH2Ar), 3.83 (1H, d, NCHCHPh), 3.55 (1H, m, CHPh), 3.19 (1H, m,
NCHeq), 2.57 (1H, m, NCHax). m/z 249 (M+).
2-Phenyl-2-(2-(5-ethyl-2-pyridyl)ethyl)indan 28
Cis- and trans-2-phenyl-2-(2-(5-ethyl-2-pyridyl)ethyl)indan-1,3-diol
25 (4 g, 11 mmol) were dissolved in methanol (100 mL) and concen-
trated hydrochloric acid (4 mL). Pd/C (5%, 0.4 g) was added and the
mixture was hydrogenated at room temperature and atmospheric pres-
sure until the required amount of hydrogen had been taken up (7 days).
The catalyst was filtered off and the solvent was removed under reduced
pressure. Saturated sodium bicarbonate was then added and the mix-
ture was extracted with chloroform. The combined extracts were dried
(Na2SO4) and the solvent was removed under reduced pressure to give
an oil which solidified on standing. Recrystallization from ethanol
gave 2-phenyl-2-(2-(5-ethyl-2-pyridyl)ethyl)indan 28 as a white crys-
talline solid (2.3 g, 62%), mp 88–89◦C. (Found: C 87.7, H 7.5, N 4.2%.
C24H25N requires C 88.1, H 7.6, N 4.3%.) νmax (Nujol)/cm−1 1600,
750. δH (CDCl3) 6.9–8.3 (12H, ArH), 2.7 (4H, s, CH2CH2), 2.6 (2H, q,
CH2CH3), 1.2 (3H, t, CH2CH3). m/z 327 (M+).
2-Phenyl-2-[2-(2-piperidyl)ethyl]phenalene-indan-1,3-dione 36
A
solution of 2-phenyl-2-[2-(2-pyridyl)ethyl]phenalene-indan-1,3-
dione 35 (2.1 g, 6.6 mmol) in absolute ethanol/water (50/50, 100 mL)
and concentrated hydrochloric acid (1 mL) was hydrogenated in the
presence of PtO2 (Adam’s catalyst) (0.25 g) at room temperature and
atmospheric pressure until the required amount of hydrogen had been
taken up. The catalyst was filtered off and the solvent was removed
under reduced pressure to give an oil. Chromatography of this oil
over alumina, using diethyl ether as the elutant gave 2-phenyl-2-[2-(2-
piperidyl)ethyl]phenalene-indan-1,3-dione 36 as a yellow oil (0.3 g).
(Found: C 81.6, H 6.4, N 3.6%. C26H25NO2 requires C 81.4, H 6.5,
N 3.65%.) νmax (Nujol)/cm−1 3400, 1710, 1680, 1585. δH (CDCl3)
7.3–8.1 (6H, m, ArH), 3.2 (1H, m, NCH(CH2)2), 3.0 (1H, dd, NCHeq),
2.65 (1H, NCHax). m/z 383 (M+), 278 (M+ − PhCO).
2-Phenyl-2-(2-(2-piperidyl)methyl)indan-1,3-dione 63
2-Phenyl-2-(2-(2-pyridyl)methyl)indan-1,3-dione62(6 g, 19 mmol)was
dissolved in glacial acetic acid and PtO2 (0.5 g) was added. The mixture
was hydrogenated on a Parr Hydrogenator until the calculated amount
of hydrogen had been taken up. The mixture was filtered and the solvent
removed under vacuum to give a thick oil. Saturated sodium bicarbon-
ate solution was added until the mixture was basic, and the mixture
was then extracted with chloroform. The combined extracts were dried
(Na2SO4) and the solvent removed under vacuum to give 2-phenyl-2-
(2-(2-piperidyl)methyl)indan-1,3-dione 63 as a dark oil (4.1 g, 67%).
(Found: C 78.9, H 6.4, N 4.4%. C21H15NO2 requires C 79.0, H 6.6,
N 4.4%.) νmax (liquid film)/cm−1 3300, 1760, 1710, 1600, 750. δH
(CDCl3) 7.2 (9H, ArH), 1.2–3.3 (12H, aliphatic).
6-Phenyl-5,7-diketo-6-[2-(2-piperidyl)ethyl]dibenzo-
[a,c]cycloheptane 39
PtO2 catalyst (0.25 g) was added to a solution of 6-phenyl-5,7-diketo-
6-[2-(2-pyridyl)ethyl]dibenzo[a,c]cycloheptane 38 (0.75 g, 2 mL) in
absolute ethanol (75 mL) and concentrated hydrochloric acid (0.3 mL)
and the mixture was hydrogenated until the required amount of hydrogen
had been taken up. The catalyst was filtered off and the solvent removed
under reduced pressure to give an oil. The oil was added to aqueous

Synthesis and Reactions of 2,2-Disubstituted Indan-1,3-diones
73
saturated sodium bicarbonate solution and the mixture was extracted
with ethyl acetate. The combined extracts were dried (Na2SO4) and
the solvent was removed under reduced pressure. Preparative thin layer
chromatography over silica using ethyl acetate as the elutant gave
6-phenyl-5,7-diketo-6-[2-(2-piperidyl)ethyl]dibenzo[ a,c]cycloheptane
39 as an oil (30 mg). νmax (Nujol)/cm−1 1760, 1690. δH (CDCl3) 7.2–7.9
(13H, m,ArH), 4.4(1H, brs, NH), 0.8–3.9(13H, m, aliphatic). Hydrogen
gas was bubbled through the solution to give 6-phenyl-5,7-diketo-
6-[2-(2-piperidyl)ethyl]dibenzo[a,c]cycloheptane hydrochloride. δH
(CDCl3) 9.0 (2H, br d, N+H2). m/z 409, 254.
phthalhydrazide residual that formed. The filtrate was made alkaline
with dilute sodium hydroxide and extracted with ethyl acetate. Sol-
vent evaporation afforded a yellow oil (0.2 g). Preparative thin layer
chromatography using silica plates and diethyl ether afforded 2,3,4,5-
tetrahydro-5a-methylindeno[1,2-b]azepin-6(5aH)-one 56 as a yellow
oil (0.1 g). (Found: C 78.8, H 7.1, N 6.6%. C14H14NO requires C 78.9,
H 7.0, N 6.6%.) νmax (CDCl3)/cm−1 1705, 1660. δH (CDCl3) 7.4–8.1
(4H, m, ArH), 4.2 (1H, dd, 2-Heq), 3.75 (1H, t, 2-Hax), 1.4–2.2 (6H, m,
–CH2CH2CH2CH2N), 1.3 (3H, s, CH3).
rel-(3ꢀS,1S,3R)-1-Phenyl-3-(piperidyl-2-yl)propylphthalide
Hydrochloride 57
2,3-Dihydro-1-methylpyrrolo[2,1-a]isoindol-5-one 44
2-Methyl-2-(2-N-phthalimodoethyl)-indan-1,3-dione 41 (20 g) was
heated under reflux with hydrazine hydrate (3 mL) in ethanol (500 mL)
for 1 h. A white gelatinous precipitate was rapidly produced and was
decomposedbywarmingwithexcessdilutehydrochloricacid.Theinsol-
uble phthalhydrazide that formed was filtered off and washed with water.
The filtrate was concentrated to remove ethanol and the cooled solution,
after filtration from a further small amount of precipitated phthalhy-
drazide, was made alkaline with dilute sodium carbonate solution, and
extracted with ethyl acetate. Solvent evaporation followed by repeated
recrystallization of the residue from ethyl acetate gave 2,3-dihydro-1-
methylpyrrolo[2,1-a]isoindol-5-one 44 as a crystalline solid (3 g), mp
210–202◦C. (Found: C 77.9, H 6.0, N 7.6%. C12H11NO requires C
77.8, H 6.0, N 7.7%.) νmax (CDCl3)/cm−1 1660, 1600. δH (CDCl3)
7.3–8.0 (4H, m, ArH), 3.9 (2H, t, CH2N), 3.1 (2H, m, CH2CH2N), 2.1
(3H, s, CH3).
Sodium borohydride (0.4 g) was added to a solution of rel-(6S,8aR)-6-
phenyl-6,7,8,8a,9,10,11,12-octahydropyrido[1,2-b][2]benzazonin-5,14-
dione 6 (1 g, 3 mmol) in dry methanol (25 mL) and the resulting
mixture was stirred for one hour. The solvent was removed under
reduced pressure and water (5 mL) was added followed by dilute
hydrochloric acid until the solution was acidic. The remaining solu-
tion was extracted with ethyl acetate, the combined extracts were dried
(Na2SO4) and the solvent was removed under reduced pressure to
give a colourless oil. The oil was dissolved in ethyl acetate/ethanol
(1/1) and left in an ice box overnight to yield white crystals. A fur-
ther recrystallization gave rel-(3ꢀS,1S,3R)-1-phenyl-3-(piperidyl-2-yl)-
propylphthalide hydrochloride 57 as shiny white crystals (0.87 g, 86%),
mp 242–244◦C. (Found: C 71.1, H 7.1, N 3.8%. C22H26NO2Cl requires
C 71.1, H 7.0, N 3.8%.) νmax (Nujol)/cm−1 3400, 1760, 1600. δH
(CDCl3) 9.45 (1H, br d, NH), 9.1 (1H, br d, NH), 6.4–7.8 (9H, m, ArH),
5.5 (1H, d, OCH), 3.1 (1H, br d, NCHeq), 2.8 (1H, m, OCHCPhH),
2.8 (1H, m, NCH(CH2)2), 2.6 (1H, m, NCHax). m/z 335 (M+), 202
(M+ − PhCHCHCH2CH(CH2)4NH), 133 (PhCOOCH), 105 (PhCO).
1,2,3,10b-Tetrahydro-10b-hydroxy-1-methylpyrido-
[2,1-a]isoindol-6(4H)-one 48, 51, 52
2-Methyl-2-(3-N-phthalimidopropyl)-indan-1,3-dione 45 (26 g) was
heated under reflux with hydrazine hydrate (4 mL) in ethanol (600 mL)
for 2 h. The phthalhydrazide formed was filtered off and the solvent was
removed by distillation to give a sticky foam (10 g). Dilute hydrochlo-
ric acid (2.5 mL) was added to a solution of the crude product (1 g) in
ethanol (10 mL) and the mixture was heated over a water bath for 20 min.
The solid that separated out was filtered off and the filtrate was basified
with dilute sodium hydroxide. Extraction with ethyl acetate followed by
recrystallization from ethanol gave 1,2,3,10b-tetrahydro-10b-hydroxy-
1-methylpyrido[2,1-a]isoindol-6(4H)-one 48 as a white crystalline solid
(0.6 g), mp 212◦C. (Found: C 71.9, H 7.0, N, 6.4%. C13H15NO2 requires
C 71.8, H 6.9, N 6.5%.) νmax (CHBr3)/cm−1 3560, 1660, 1600. δH
(CDCl3) 7.2–7.7 (4H, m, ArH), 4.22 (1H, s, OH), 3.82 (1H, dm,
C-4Heq), 2.87 (1H, dt, C-4Hax), 1.27 (3H, s, CH3), 1.2–2.1 (5H, m,
C-1H, C-2H, C-3H). The remainder of the crude hydrazinolysis product
was chromatographed over Woelm alumina (grade III). The first frac-
tion eluted with ether gave anti-2,3,4,4a-tetrahydro-4a-methlyindeno-
[1,2-b]pyridine-5-one hydrazone 51 as a yellow crystalline solid (2 g),
mp 125.5◦C. (Found: C 73.2, H 7.1, N 19.7%. C13H15N3 requires
C 73.2, H 7.0, N 19.7%.) νmax (CDCl3)/cm−1 3390, 1668, 1640,
1600. δH (CDCl3) 7.8–8.15 (2H, m, ArH), 7.4–7.7 (2H, m, aromatic),
5.67–5.8 (2H, s, =N–NH2), 3.9 (2H, t, –NCH2(CH2)2), 1.5–2.2 (4H,
m, NCH2CH2CH2), 1.27 (3H, s, CH3). A second fraction eluted
with diethyl ether gave syn-2,3,4,4a-tetrahydro-4a-methlyindeno-
[1,2-b]pyridine-5-one hydrazone 52 as a yellow crystalline solid (1 g),
mp 148–149◦C. (Found: C 73.4, H 6.9, N 19.7%. C13H15N3 requires
C 73.2, H 7.0, N 19.7%.) νmax (CDCl3)/cm−1 3410, 3320, 1660, 1640,
1600. δH (CDCl3) 7.3–7.8 (4H, m, ArH), 5.5 (2H, s, C=N–NH2), 3.4–
4.2 (2H, m, –NCH2(CH2)2), 1.5–2.6 (4H, m, NCH2CH2CH2), 1.4 (3H,
s, CH3).
Accessory Materials
Syntheses and characterization data for 9, 15, 19, 21, 24, 25,
35, 38, 41, 45, 53, 58, 60–62 available from the authors or,
until January 2011, the Australian Journal of Chemistry.
Acknowledgment
C.J.R. thanks the SERC for financial support.
References
[1] J. G. Lombardino, E. H. Wiseman, J. Org. Chem. 1968, 11, 342.
[2] U. Soderberg, C.A. Wachtmeister, J. Pharmacol. Exp.Ther. 1956,
117, 298.
[3] (a) G. Vanag, J. Pregaris, Latv. PSR Zinat. Akad. Vestis, Kim. Ser.
1964, 5, 559.
(b) L. S. Goodman, A. Gillman, The Pharmacological Basis of
Therapeutics, 3rd edn 1965, pp. 1450–1458 (MacMillan: New
York, NY).
[4] G. Pasero, N. Marini, Gazz. Med. Ital. 1958, 117, 561.
[5] N. Kubovic, M. Prazic, D.Atanackovic, Proc. Soc. Exp. Biol. Med.
1955, 90, 660.
[6] (a) J. T. Correll, US Patent 2900302 1959.
(b) D. G. Crabtree, W. H. Robinson, Pest. Control 1953, 21, 22.
(c) G. Ya. Vanag, B. Yu. Fal’kenshtein, Izvest. Akad. Nauk. Latv.
SSR Ser. Kim. 1959, 5, 99.
[7] E. Gori, Proc. Int. Conf. on Thrombosis and Embolism 1954,
p. 271 (Basel).
[8] H. Blumburg, H. B. Dayton, S. M. Gordon, Science 1958, 127,
188.
[9] L. W. Hazleton, W. H. Dolben, US Patent 2884357 1959.
[10] R. F. Newton(GlaxoGroupResearchandDevelopment), personal
communication 1985.
[11] C. J. Roxburgh, T. A. Crabb, R. F. Newton, J. Chem. Soc., Perkin
Trans. 1 1989, 2431.
2,3,4,5-Tetrahydro-5a-methylindeno[1,2-b]azepin-6(5aH)-one 56
2-Methyl-2-(4-N-phthalimidobutyl)-indan-1,3-dione 53 (0.5 g) was
heated under reflux with hydrazine hydrate (0.1 mL) in ethanol (60 mL)
for 45 min. The phthalhydrazide formed on cooling the reaction mix-
ture was filtered off and dilute hydrochloric acid (10 mL) was added to
the solution. The acidified solution was gently heated over a water bath
for 15 min, and was filtered to remove the small quantity of additional

74
C. J. Roxburgh and L. Banting
[12] L. H. Sternbach, E. Reeder, J. Org. Chem. 1961, 26, 1111.
[22] W. Mosher, R. Soeder, J. Org. Chem. 1971, 36, 1561.
doi:10.1021/JO00810A028
doi:10.1021/JO01063A034
[13] L. H. Sternbach, US Patent 2893992 1959.
[14] L. H. Sternbach, L. O. Randall, S. Gustafson, Psychopharmaco-
logical Agents (Ed. M. Gordon) 1964, Vol. 1, p. 137 (Academic:
New York, NY).
[23] T. A. Crabb, A. Patel, R. F. Newton, T. J. Cholerton, J. Chem. Soc.,
Perkin Trans. 1 1985, 191. doi:10.1039/P19850000191
[24] O. Aly, W. Awad, A. Islam, J. Org. Chem. 1957, 22, 517.
doi:10.1021/JO01356A012
[15] K. von Auwers, Chem. Ber. 1924, 57, 1723.
[16] L. H. Sternbach, E. Reeder, J. Org. Chem. 1961, 26, 4936.
doi:10.1021/JO01063A034
[25] M. Cesaris, Gazz. Chim. Ital. 1912, 42, 453.
[26] O. Aly, W. Awad, A. Islam, J. Org. Chem. 1958, 23, 1624.
doi:10.1021/JO01105A011
[17] E. Gren, G. Venags, Dokl. Akad. Nauk SSSR 1961, 139, 866.
[18] R. Budziarek, F. S. Spring, J. Chem. Soc. 1953, 956.
doi:10.1039/JR9530000956
[19] C. J. Roxburgh,Tetrahedron1994, 50, 13199. doi:10.1016/S0040-
4020(01)89329-7
[20] H. S. Mosher, J. E. Tessleri, J. Am. Chem. Soc. 1951, 73, 4925.
doi:10.1021/JA01154A135
[21] M. Kotan, O. Surnina, Zh. Obshch. Khim 1957, 27, 2974.
[27] B. Aleksiev, M. Milosev, Chem. Ber. 1976, 100, 701.
[28] T. A. Crabb, A. Patel, R. F. Newton, B. J. Price, M. J. Tucker,
J. Chem. Soc., Perkin Trans. 1 1982, 2783. doi:10.1039/
P19820002783
[29] C. J. Roxburgh, T. A. Crabb, R. F. Newton, J. Chem. Soc., Perkin
Trans. 1 1990, 2905.
[30] C. J. Roxburgh,Tetrahedron1993, 49, 10749. doi:10.1016/S0040-
4020(01)80232-5