´
N. Babic, F. Peyrot / Tetrahedron Letters xxx (xxxx) xxx
2
Fig. 1. Key intermediates in the synthesis of tetraethyl nitroxides.
pathways for the synthesis of tetraethyl nitroxides already exist,
but they are not without drawbacks. The methods developed by
the Utsumi group and based on intermediates 7 and 8 (Fig. 1) suffer
from low and unreliable yields while the introduction of the nitro-
gen atom occurs in the first steps of the synthesis, making the tar-
get compound very expensive [14,19,27]. Alternative strategies
through symmetric dienone 9, initially proposed by the Studer
group [28], allow introduction of the nitrogen at a late stage of
the synthesis (Fig. 1). However, these authors later dismissed their
own pathway on the grounds of low yields, before going back to
synthetic routes involving intermediate 7 [29]. More recently, the
Arimoto group successfully prepared several grams of 15N-labelled
nitroxide 2 via dienone intermediate 9 [22], but their procedure
requires specific equipment to perform a Meyer-Schuster rear-
rangement under elevated CO2 pressure. This key step is highly
sensitive to the reaction conditions [30] and difficult to reproduce.
Herein, we propose a new synthetic route for obtaining piperi-
dine-based nitroxides through dienone derivative 9, relying only
on common reagents and laboratory equipment for organic syn-
thesis. A set of tetraethyl-substituted piperidine nitroxides with
potentially different abilities to cross cell membranes was pre-
pared in view of biological applications and their resistance against
reduction by ascorbate was assessed.
Scheme 2. Synthesis of 2,2,6,6-tetraethylpiperidin-4-one. Reagents and conditions:
(a) LDA (2 eq.), THF, À 50 °C, 1 h; (b) 3-pentanone (1 eq.), THF, À 50 °C, 3 h, 68% over
two steps; (c) NaBH4 (2.5 eq.), MeOH/THF (1:3), 0 °C; (d) benzaldehyde dimethyl
acetal (10 eq.), PTSA (cat.), CH2Cl2, 22 °C, overnight, 43% over two steps; (e) EtMgBr
(8 eq.), THF, 83%; (f) 3 M HCl, EtOH, 22 °C, 16 h; (g) Jones reagent (2 eq.), acetone,
48% over two steps; (h) conc. H2SO4, CHCl3, 1.5 h; (i) NH3, EtOH, 50 °C, 3 days, 29%
over two steps. cat. = catalytic, conc. = concentrated, LDA = lithium diisopropy-
lamide, PTSA, p-toluenesulfonic acid, THF = tetrahydrofuran.
Results and discussion
Fig. 2. Reduction of compound 11 and eventually 12 by NaBH4 could be facilitated
by the formation of borate complexes 11a and 12a.
Our initial plan was to create the carbon backbone of dienone 9
and to introduce four ethyl groups in one step by submitting diox-
olane-protected diethyl 3-oxoglutarate to a Grignard reaction with
ethylmagnesium bromide. Despite several attempts, this strategy
was unsuccessful, presumably due to steric reasons, even though
similar examples exist in the literature [31]. We therefore had to
proceed via a stepwise route.
although the ester group was not reduced, the reaction was not
complete even after several days. It seems that the reduction of
the keto group is also accelerated by formation of complex 11a
(Fig. 2) and is thus very slow in its absence. Finally, a compromise
was found by performing the reaction in a mixture of tetrahydro-
furan and methanol at 0 °C with 0.5 M NaBH4. The conversion
was complete after a few hours with little side product formation.
The crude product was thus used in the next step without
purification.
Direct Grignard reaction on diol 12 was prevented, presumably
due to a negative charge close to the reactive ester functional
group. Our next move was thus to prepare the isopropylidene ketal
of diol 12. The typical procedure involves dissolving the substrate,
in this case diol 12, in 2,2-dimethoxypropane and stirring with a
catalytic amount of p-toluenesulfonic acid. Unfortunately, after
several trials, only a low yield (<20%) of ketal 17 was isolated
(Fig. 3). Various attempts at changing the solvent or the reagent
(to either 2-methoxypropene or acetone in the presence of dehy-
drating agents such as aluminum chloride, trimethylsilyl chloride,
or molecular sieves) failed to increase the yield. Therefore, benzyli-
dene acetal protection was tested instead. Diol 12 was reacted with
ten equivalents of benzaldehyde dimethyl acetal in dry dichloro-
methane and a catalytic amount of p-toluenesulfonic acid. After
stirring at 22 °C for twenty-four hours, acetal 13 was isolated by
column chromatography in an approximately 70% yield (calculated
from the 1H NMR spectra).
Our synthetic route towards 2,2,6,6-tetraethylpiperidin-4-one 1
is presented in Scheme 2. First, the 1,3-dianion of ethyl acetoac-
etate was generated by deprotonation with two equivalents of
freshly prepared lithium diisopropylamide at À50 °C in tetrahydro-
furan. When one equivalent of 3-pentanone was added, regioselec-
tive addition of the terminal carbon of the dianion to the ketone
provided d-hydroxy-b-keto ester 11 in 68% yield, comparable to
similar examples in the literature [32]. Compound 12 was then
obtained by reducing 11 with sodium borohydride (NaBH4). NaBH4
is used commonly for converting ketones and aldehydes to alco-
hols, whereas esters are usually stable under the typical reaction
conditions. However, when neighboring functional groups such
as oxo, hydroxyl, or carboxylic acids are present, sodium borohy-
dride can reduce esters to alcohols [33]. Indeed, when we
attempted the reduction of the keto group of compound 11 using
typical conditions with several equivalents of sodium borohydride
in methanol at room temperature, a significant amount of ester
was also reduced to give the corresponding triol. It is likely that
NaBH4 partially reacts with methanol and the hydroxyl groups in
compound 11 to form a trialkylborate [32] that acts as a Lewis acid,
activates the ester group, and facilitates its reduction, as illustrated
by structures 11a and 12a in Fig. 2.
It has been reported [32] that performing the reduction in
tetrahydrofuran at a low concentration of NaBH4 (0.2 M) prevents
reduction of the ester functional group. In our hands, however,
Steric effects can explain the unsuccessful acetonide protection
shown in Fig. 3. In the case of six-membered cyclic ketal 17, 1,3-
Please cite this article as: N. Babic´ and F. Peyrot, New synthetic route to 2,2,6,6-tetraethylpiperidin-4-one: A key-intermediate towards tetraethyl nitrox-
Babi?, Nikola
