1,7-Diazaspiro[5.5]undecane - A neglected heterocycle

Article abstract of DOI:10.1021/ol402301g

A convenient and simple three step synthesis of 1,7-diazaspiro[5.5]undecane via Claisen condensation and acid catalyzed decarboxylation and spirocyclization of N-Boc-δ-valerolactam is described. Reactions of this spiroaminal with electrophiles including alkyl halides, alkane dihalides, acid chlorides, and sulfonyl chlorides gave either spirocyclic adducts or tetrahydropyridine derivatives. Additionally, the parent heterocycle is a novel bidentate ligand and formed complexes with ruthenium(II) and copper(II).

Full text of DOI:10.1021/ol402301g

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
1,7-Diazaspiro[5.5]undecane ꢀ
A Neglected Heterocycle
Jens Cordes, Philip R. D. Murray, Andrew J. P. White, and Anthony G. M. Barrett*
Department of Chemistry, Imperial College London, London SW7 2AZ, U.K.
agm.barrett@imperial.ac.uk
Received August 12, 2013
ABSTRACT
A convenient and simple three step synthesis of 1,7-diazaspiro[5.5]undecane via Claisen condensation and acid catalyzed decarboxylation and
spirocyclization of N-Boc-δ-valerolactam is described. Reactions of this spiroaminal with electrophiles including alkyl halides, alkane dihalides, acid
chlorides, and sulfonyl chlorides gave either spirocyclic adducts or tetrahydropyridine derivatives. Additionally, the parent heterocycle is a novel
bidentate ligand and formed complexes with ruthenium(II) and copper(II).
In contrast to spiroketals (O,O-ketals) and spirohe-
miaminals (N,O-ketals), little is known about the chemis-
try of the corresponding spiroaminals (N,N-ketals).1,2 Few
papers have been published on synthesis and reactions of
1,7-diazaspiro[5.5]undecane (1) and its derivatives.3ꢀ6 One
simple synthesis is known of the parent compound 1 from
piperidin-2-one (2) in 36% yield, via heating with lithium
hydroxide at 320 °C (Scheme 1).3 Two N-substituted
analogues 5 and 9 were synthesized from the correspond-
ing N-substituted lactams 4 and 8 in 39% and 47% yield
over two steps.4 Only three reactions are known for
spiroaminal 1 and its derivatives. The parent heterocycle
1 can be oxidized to form diaziridine 3.3 The dimethyl-
analogue 5 was hydrogenated with ring cleavage to form
diamine 6 or, under more forcing conditions, to generate
diamine 7 via two reductive ring-opening reactions.5
Scheme 1. Known Syntheses and Conversions of 1 and Simple
Analogues3ꢀ5
(1) (a) Kibayashi, C.; Yamazaki, N. Sci. Synth. 2006, 30, 639. (b)
Pawlenko, S.; Lang-Fugmann, S. In Houben-Weyl Methods of Molecu-
lar Transformations; Hagemann, H., Klamann, D., Eds.; Georg Thieme
Verlag: Stuttgart, 1992; E 14a/3, pp 545ꢀ547.
(2) For recent examples of aminals in natural product syntheses, see:
(a) Padwa, A.; Flick, A. C.; Lee, H. I. Org. Lett. 2005, 7, 2925. (b)
Bobeck, D. R.; Lee, H. I.; Flick, A. C.; Padwa, A. J. Org. Chem. 2009, 74,
7389. (c) Miura, Y.; Hayashi, N.; Yokoshima, S.; Fukuyama, T. J. Am.
Chem. Soc. 2012, 134, 11995.
(3) Denisenko, S. N.; Pasch, E.; Kaupp, G. Angew. Chem. 1989, 101,
1397. Angew. Chem., Int. Ed. Engl. 1989, 28, 1381.
In light of the limited information available on this
fascinating spirocyclic compound,3ꢀ6 we sought to develop
an efficient synthesis of heterocycle 1 and to investigate its
reactivity toward electrophiles and as a novel bidentate
ligand.
Following the general strategy of the known syntheses,3ꢀ5
we developed a method that allows the synthesis of
€
(4) Buchel, K. H.; Bocz, A. K.; Korte, F. Chem. Ber. 1966, 99, 724.
€
(5) Korte, F.; Bocz, A. K; Buchel, K. H. Chem. Ber. 1966, 99, 737.
(6) For syntheses of other derivatives of 1, see: (a) Zondler, H.;
ꢀ
Pfleiderer, W. Helv. Chim. Acta 1975, 58, 2247. (b) Malon, P.; Barness,
ꢀꢀ
C. L.; Budesinsky, M.; Dukor, R. K.; van der Helm, D.; Keiderling,
T. A.; Koblicova, Z.; Pavlikova, F.; Tichy, M.; Blaha, K. Collect. Czech.
Chem. Commun. 1988, 53, 2447. (c) Smolikova, J.; Koblicova, Z.; Blaha,
K. Collect. Czech. Chem. Commun. 1973, 38, 532.
r
10.1021/ol402301g
XXXX American Chemical Society

spiroaminal 1 under significantly less harsh reaction condi-
tions. Commercially available Boc-lactam 10 was converted
into β-ketolactam 11 via Claisen condensation,7 by the slow
addition of lithium hexamethyldisilazide (LiN(SiMe3)2)
to a solution of lactam 10 in THF at 0 °C (Scheme 2).
The resulting product exists as both β-ketolactam 11 and
hemiaminal 12 in equilibrium and, as a crude product, is
prone to decomposition via retro-Claisen reaction. Fortu-
nately, deprotection and dehydration of the mixture directly
using trifluoroacetic acid (TFA) gave enamino-lactam 13,
in sufficient purity for the next step. Lactam hydrolysis,
decarboxylation, and aminal formation by heating in
hydrochloric acid gave, after column chromatography,
spiroaminal 1 in 84% yield over the three steps.
The monoalkylated products 14a and 14b shed some
light on the reactivity of 1,7-diazaspiro[5.5]undecane.
The NMR spectra of spiroaminal 1 clearly show that this
substance has predominantly the spirocyclic constitution.9
The NMR spectra of compounds 14a and 14b are con-
sistent with these substances existing predominantly in
the ring-opened imine/amine form. It is a reasonable
assumption, though, that for both compounds a small
concentration of the other tautomer (i.e., imine/amine 16 is
in equilibrium with spiroaminal 1, and spiroaminal 17 is in
equilibrium with imine/amine 14, respectively, Scheme 3)
exists.
Scheme 3. Reasonable Mechanism for the Formation of 14 and
15 by Alkylation of 1, and the Failure To Form Spirane 18
Scheme 2. Synthesis of 1,7-Diazaspiro[5.5]undecane (1)
With spiroaminal 1 in hand, we turned our attention to
the synthesis of derivatives by reaction with electrophiles.
Among the simplest possible reactions are alkylations,
acylations, and sulfonylations.8 Reaction of spirane 1 with
allyl bromide or benzyl bromide, and potassium carbonate
in acetonitrile did not yield spirocyclic products but gave
the tetrahydropyridines 15a (62%) and 15b (72%) respec-
tively (Table 1, entries 1 and 3). Reaction with only 1.1
equiv of electrophile gave the same compounds, but this
time monosubstituted analogues 14 were also isolated.
The consequence of this is obvious: spiroaminal 1 is a
sterically rather hindered nucleophile (both amines are 2,2-
disubstituted piperidines) and should react slowly, whereas
the primary amine 16 is both a good nucleophile and
sterically much less hindered and should therefore react
rapidly. The first alkylation presumably takes place on the
primary amine of isomer 16, even though its concentration
is very low. The equilibrium between the resultant amines
14 and 17 is now clearly on the side of the imine/amine 14.
Following the same reasoning asbefore, but withincreased
steric bulk on 17 and increased nucleophilicity on 14, the
product of the second alkylation is consequently 15, rather
than 18.
We next turned our attention to alkylation reactions
with dielectrophiles. Depending on the dielectrophile
structure, the product was either tricyclic diamine 21 or
bicyclic imine-amine 22, and mixtures of these were not
observed (Table 2). This interesting result again can be
rationalized by considering spiroaminal tautomerism.
Reaction of diaminal 1 with 1,2-dibromoethane, 3-bro-
mo-2-bromomethyl-1-propene, and 1,3-diiodopropane
gave rise to the tricyclic adducts 21a, 21b, and 21c respec-
tively. In contrast, reaction of spirane 1 with 1,4-diiodo-
butane, 1,5-diiodopentane and 1,6-diiodohexane gave the
Table 1. Alkylation Reactions of 1 with Allyl Bromide and
Benzyl Bromide
entry
R
equiv RꢀBr
products (%)a
1
2
3
4
allyl
allyl
Bn
2.2
1.1
2.2
1.1
14a (trace) þ 15a (62)
14a (12) þ 15a (32)
14b (trace) þ 15b (72)
14b (30) þ 15b (35)
Bn
a Isolated yield.
(8) Alkylations: (a) Lawrence, S. A. Sci. Synth. 2008, 40, 526.
Acylations: (b) Ziegler, T. Sci. Synth. 2005, 21, 43. Sulfonylations: (c)
ꢁ
_
Drabowicz, J.; Kiezbasinski, P.; Łyzwa, P.; Zajac, A.; Mikozajczyk, M.
)
Sci. Synth. 2007, 39, 77.
(7) Yasuda, N.; Hsiao, Y.; Jensen, M. S.; Rivera, N. R.; Yang, C.;
Wells, K. M.; Yau, J.; Palucki, M.; Tan, L.; Dormer, P. G.; Volante,
R. P.; Hughes, D. L.; Reider, P. J. J. Org. Chem. 2004, 69, 1959.
(9) Interestingly, this distinguishes spiroaminal 1 from its analogues
with 5- and 7-membered rings, which exist predominantly in the form of
the ring-opened tautomer (see ref 3).
B
Org. Lett., Vol. XX, No. XX, XXXX

Table 2. Alkylation Reactions of 1 with Dihalides
Scheme 4. Plausible Mechanism for the Formation of
Compounds 21 and 22 by Alkylation of 1 with Dihalides
bicyclic imino-amines 22a, 22b, and 22c. Clearly, the first
alkylation step (Scheme 4) does not differ from the allyla-
tion or benzylation discussed above. But the resultant
intermediate 19/20 can now undergo a second intramole-
cular alkylation step. By reaction with the secondary
amine in 20, it forms imino-amine 22. This is the case
for dihalides that can form 5-membered or larger rings.
It is not observed, however, for dihalides that would result
in 4-membered or smaller rings. These result in tricyclic
spiroaminals, most likely via alkylation of the second
amine via the spiroaminal tautomer 19. The exclusive
formation of either 21 or 22 can therefore be explained
by ring strain (which prevents the formation of 22 with
4-membered or smaller cyclic amines) and fast ring closing
of 20 to 22 (which prevents formation of tricyclic spiroam-
inals 21 for R,ω-diiodoalkanes with four or more carbon
atoms, which would have resulted in 7- to 9-membered
heterocycles).
a Reaction conditions: 70 °C, 3 h.
Reaction of diaminal 1 with acetyl chloride and 4-tolue-
nesulfonyl chloride again resulted in spirocycle scission
(Scheme 5). With the first amine converted into a rather
non-nucleophilic amide, the next best nucleophile is now
the imine, which leads (via an iminium ion intermediate)
to N-acylated or N-toluene-4-sulfonated enamines 24
respectively.
Scheme 5. Reactions of 1 with Acetyl Chloride and
4-Toluenesulfonyl Chloride
The formation of an acylated or N-toluene-4-sulfonated
spiroaminal as an intermediate is less likely to occur, and
its reaction with acetyl or toluene-4-sulfonyl chloride even
more so. The NMR spectra of compounds 23 and 24 show
clearly the ring opened tautomer structures. In addition to
that, an X-ray crystal structure determination of sulfona-
mide 23b confirms this structure to be present for the solid
state, too.
Consistent with the reactions of spirane 1 with the
dihalides in entries 1ꢀ3 in Table 2, reaction of 1 with
bromide 25 gave the corresponding adduct 26, retaining
the spiroaminal unit as a single diastereoisomer (Scheme
6). Clearly, the product 26 was formed via N-alkylation
followed by Michael addition. In contrast, reaction of
the isomeric bromide 27 with spirane 1 gave the doubly
alkylated amino-imine 15c. Presumably the difference
between the two reaction pathways is that the examined
system 19 apparently undergoes 5-exo-trig cyclizations
easily, while 6-endo-trig cyclizations seem to be a disfa-
vored process. Reaction of spirane 1 with chlorosulfonyl
isocyanate (28) resulted in the formation of a thiatriazine
ring system. The structure of tricyclic sulfonyl-urea 29
was confirmed by an X-ray crystallographic structure
determination.
Preliminary experiments also demonstrate the utility of
spiroaminal 1 as a bidentate ligand in transition metal
complexes. The molecule can bind to a metal with either
retention or scission of the spirane ring system, as X-ray
Org. Lett., Vol. XX, No. XX, XXXX
C

via formation of a four-membered chelate structure
(RuꢀNꢀCꢀNꢀring). The spiroaminal adopts an unusual
double boat conformation in complex 30. In contrast,
reaction of spiroaminal 1 with copper(II) benzoate gave
a binuclear complex [(BzO)2Cu(16)]2 (31) with two imine/
amine ligands 16 bridging two copper cores, each copper
atom being coordinated by the amine of the one ligand and
the imine of the other.
Scheme 6. Reactions of 1 with Unsymmetric Dielectrophiles
In conclusion, we describe an improved synthesis of 1,7-
diazaspiro[5.5]undecane (1) and the formation of deriva-
tives by reaction of 1 with mono- and dihalo-alkanes,
acetyl chloride, and 4-toluenesulfonyl chloride. Most of
these reactions resulted in conversion of the spiroaminal
into 3,4,5,6-tetrahydropyridines. Reaction with 1,2- and
1,3-dihalides, (E)-methyl 4-bromobut-2-enoate (25), and
chlorosulfonyl isocyanate (28), however, resulted in N,N0-
substitution and formation of tricyclic adducts 21aꢀc, 26,
and 29, retaining the spirane core. Two transition metal
coordination complexes of 1 were formed, a ruthenium
complex (30) of the spiroaminal itself, and a binuclear
copper complex (31) of the ring-opened tautomer 16.
The synthesis of core-substituted analogues of 1 and
analyses of the conformational and tautomeric preferences
of spiroaminals like 1 are the subject of current investiga-
tions and will be reported in due course.
Acknowledgment. We thank the European Research
Council(ERC) forgenerous grant support, GlaxoSmithK-
line for the Glaxo endowment, and P. R. Haycock and
R. N. Sheppard (both Imperial College London) for high-
resolution NMR spectroscopy.
Figure 1. Ruthenium complex 30 and copper complex 31.
Supporting Information Available. Experimental pro-
cedures and analytical data for all new compounds and
X-ray crystallographic data (CIF) for 23b, 29, 30, and 31.
This material is available free of charge via the Internet at
http://pubs.acs.org.
crystal structure determinations of two complexes have
shown. Reaction of spiroaminal 1 with (Ph3P)3RuCl2 gave
the corresponding monomeric complex (Ph3P)2RuCl2(1)
(30) which retained the intact spiroaminal unit (Figure 1),
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX