Selective electrochemical reduction of cinnamyl ethers in the presence of other allylic C-O bonds

Article abstract of DOI:10.1016/S0040-4039(00)02250-4

Several conduritol derivatives protected as allyl and cinnamyl ethers were subjected to electrochemical reduction at a mercury cathode, resulting in selective removal of the cinnamyl group.

Full text of DOI:10.1016/S0040-4039(00)02250-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 1241–1243
Selective electrochemical reduction of cinnamyl ethers in the
presence of other allylic CꢀO bonds
Aida Solis-Oba,† Tomas Hudlicky,* Lukasz Koroniak and Dean Frey
Department of Chemistry, University of Florida, Gainesville, FL 32611-7200, USA
Received 14 November 2000; revised 30 November 2000; accepted 5 December 2000
Abstract—Several conduritol derivatives protected as allyl and cinnamyl ethers were subjected to electrochemical reduction at a
mercury cathode, resulting in selective removal of the cinnamyl group. © 2001 Published by Elsevier Science Ltd.
The use of allyl and cinnamyl ethers as protecting
groups for alcohols is well known1 as are the methods
for their removal.2,3 These can involve oxidative
removal, as in the case of deallylations with Ce(IV),4
for example, a reaction proceeding via cationic or radi-
cal cationic intermediates, thus also allowing a further
substitution of such intermediates. In such ‘deprotec-
tions’, the oxygen of the original alcohol is replaced
with one from the trapping nucleophile, with the poten-
tial for scrambling of stereochemical configuration.
Reductive removal is most frequently accomplished by
such means as dissolving metal reduction, where the
original alcohol oxygen and its configuration remain
intact. For cinnamyl groups, their electrochemical
reduction in the presence of Ni(II) is preferred to
radical cyclization from neighboring aryl halides.5 Such
has been our experience in attempted electrochemical
cyclizations of vinyl bromides onto cinnamyl ethers,6
where cleavage of the cinnamyl group and reduction of
a vinyl6 or aryl5 bromide occur essentially simulta-
neously. Recently palladium-catalyzed electrochemical
reduction of both allyl and cinnamyl groups in the
presence of aryl halides has been reported.7
We found no mention in the literature of any attempts
to control selectively a cleavage of only one of several
allylic systems or attempts to control allylic versus
cinnamyl cleavage. A comparison of allylic and benzylic
reduction has been published.2i We chose several exam-
ples for study, the most complex being the conduritol
derivative 1 containing an endocyclic (position (a)), as
well as exocyclic (b) cinnamyl units in addition to
endocyclic (c) and exocyclic (d) allyl groups. In most of
the cases shown in Table 1, the dissolving metal reduc-
tion of the two ‘exocyclic’ groups occurs without selec-
tivity, whereas the electrochemical reduction at −2.9 V
(Hg or C cathode) cleaves only the exo cinnamyl ether
to afford 3.
Other examples of selective removal are shown in Table
1. In most cases the treatment of the cinnamyl–allyl
ethers with excess Na in NH3 resulted in the removal of
both groups. In attempts with a ‘stoichiometric’
amount of the metal (2 equiv.), cinnamyl cleavage
occurred at a faster rate, but the reaction mixtures
contained starting material, allyl ether, and diol in all
cases. However, the electrochemical reduction of the
* Corresponding author.
†
On leave from: Department of Sistemas Biologicos, Universidad Autonoma Metropolitana, Unidad Xochimilco, Calz. Del Hueso No. 1110,
Cp 04960, Mexico, D. F.
0040-4039/01/$ - see front matter © 2001 Published by Elsevier Science Ltd.
PII: S0040-4039(00)02250-4

1242
A. Solis-Oba et al. / Tetrahedron Letters 42 (2001) 1241–1243
cinnamyl group was in all cases selective. In the cases of
the phenyl-substituted conduritol 1b and a simpler
model 10, the Na–NH3 reaction yielded only recovered
starting material (in the case of 1b) or over reduction of
both cinnamyl groups (in the case of 10). Electrochem-
ical reduction, on the other hand, gave clearly the
deprotected alcohols. We attribute these results to solu-
bility problems and a different reduction regime. With
co-solvents (THF) the reduction was sluggish and non-
selective, and in the case of 10 it provided alcohol 12
(mixture of diastereomers) exclusively.
The electrochemical reductions were performed with a
Hg pool cathode or a vitrified carbon cathode at the
indicated potentials. The linear sweep voltammograms
(LSV) of the starting materials showed reduction waves
around −2.9 V for the cinnamyl CꢀO bond cleavage
and the reduction to 1-phenyl-1-propene. Additional
reduction wave, assumed to be that corresponding to
the reduction of an allyl group, was separated by ca.
300 mV.
Cyclic and linear sweep measurements were determined
with a platinum glassy carbon electrode and a platinum
auxiliary electrode. Measurements were referenced to a
Ag/Ag+ couple, constructed by placing a silver wire in a
0.1 M solution of silver nitrate in CH3CN, producing a
reference electrode 0.5785 V versus NHE (normal hy-
drogen electrode). The voltammetry of all compounds
was hampered by adsorption of material to the working
electrode, shown by the cathodic shift and the decrease
in current exhibited in the reduction waves of successive
sweeps. Reproducible voltammograms were obtained
by polishing the electrode surface between successive
sweeps, and at the same time stirring and degassing the
solution. Several general characteristics were observed
in the voltammograms, such as typically broad reduc-
In spite of the fact that these experiments were not
optimized, it is clear that the electrochemical removal
offers advantages over other methods. First, it retains
the stereochemical configuration of the alcohol, which
is not possible by employing methods performed under
oxidizing conditions. Second, it is selective not only
with respect to allyl groups but also with respect to
other cinnamyl units where the reductively generated
anion would be more basic (i.e. at secondary cyclic
carbon as in a in 1). Third, alkylative methods of
removal, such as the recently reported TMSCl/NaI
procedure,8 offer no selectivity in differentiating
between allylic and cinnamyl carbons.
Table 1. Electrochemical versus dissolving metal reductions of cinnamyl–allyl systems

A. Solis-Oba et al. / Tetrahedron Letters 42 (2001) 1241–1243
1243
(CHE-9615112, CHE-9910412), and TDC Research,
Inc., for financial support of this work, and CONA-
CyT, Mexico, for a postdoctoral fellowship for A.S.-O.
References
1. Greene, T. W.; Wuts, P. G. M. Protective Groups in
Organic Synthesis, 2nd ed.; Greene, T. W., Ed.; Wiley:
New York, 1991.
2. (a) Yasuhara, A.; Kasano, A.; Sakamoto, T. J. Org. Chem.
1999, 64, 4211–4213; (b) Yu, B.; Xhang, J.; Lu, S.; Hui, Y.
Synlett 1998, 29–30; (c) Alonso, E.; Ramon, D. J.; Yus, M.
Tetrahedron 1997, 53, 14355–14368; (d) Thomas, R. M.;
Mohan, G. H.; Iyengar, D. S. Tetrahedron Lett. 1997, 38,
4721–4724; (e) Yadav, J. S.; Chandrasekhar, S.; Sumithra,
G.; Kache, R. Tetrahedron Lett. 1996, 37, 6603–6606; (f)
Robles Diaz, R.; Rodriguez Melgarejo, C.; Plaza Lopez-
Espinosa, M. T.; Izquierdo Cubero, I. J. Org. Chem. 1994,
59, 7928–7929; (g) Beugelmans, R.; Bourdet, S.; Bigot, A.;
Zhu, J. Tetrahedron Lett. 1994, 35, 4349–4350; (h) Ito, H.;
Taguchi, T.; Hanzawa, Y. J. Org. Chem. 1993, 58, 774–
775; (i) Kadam, S. M.; Nayak, S. K.; Banerji, A. Tetra-
hedron Lett. 1992, 33, 5129–5132; (j) Corey, E. J.; Suggs,
W. J. Org. Chem. 1973, 38, 3224; (k) Akiyama, T.; Hiro-
fuji, H.; Ozaki, S. Tetrahedron Lett. 1991, 32, 1321–1324;
(l) Kariyone, K.; Yazawa, H. Tetrahedron Lett. 1970, 11,
2885–2888.
3. (a) Mairanovsky, V. G. Angew. Chem., Int. Ed. Engl. 1976,
15, 281; (b) Veinberg, A. Y.; Mairanovskii, V. G.;
Samokhavalov, G. I. J. Gen. Chem. USSR 1968, 38, 643;
(c) Espanet, B.; Dunach, E.; Perichon, J. Tetrahedron Lett.
1992, 33, 2485–2488; (d) Olivero, S.; Dunach, E. Synlett
1994, 531–532; (e) Olivero, S.; Dunach, E. J. Chem. Soc.,
Chem. Commun. 1995, 2497; (f) for a recent summary of
electrochemical reductions of alkyl ethers, see: Olivero, S.;
Franco, D.; Clinet, J.-C.; Dunach, E. Coll. Czech. Chem.
Commun. 2000, 65, 844.
Figure 1. Linear sweep voltammogram of 7.
tion waves with half peak potentials greater than 100
mV and with a cathodic shift that increased with sweep
rate.
The linear sweep voltammogram of conduritol deriva-
tive 7 is shown in Fig. 1. The voltammogram is charac-
terized by two broad reduction waves at E1=−2.9 V
and E2=−3.2 V (half-peak potentials). These reduction
waves were found to be irreversible through cyclic
voltammetric analysis. Attempts to extend the prepara-
tive reductions to carbon electrodes have thus far not
succeeded.
In conclusion, we have shown selectivity in removal of
cinnamyl groups over allyl groups under extremely mild
conditions. Because the cinnamyl unit is inexpensive
and because the deprotection occurs in an environmen-
tally benign manner, this method is superior to the
classical procedures reported in the literature.
Worthwhile future endeavors include the study of mul-
tiple selectivities to determine the ease of removal of
cinnamyl versus cinnamate, cinnamyl (or cinnamate) or
benzoate, for example. Since hydrogenations of benzyl
groups are sometimes troublesome, cinnamyl group
offers the advantages of easy removal. We will report
on these endeavors in due course.
4. Iranpoor, N.; Mottaghinejad, E. Tetrahedron 1994, 50,
7299.
5. Olivero, S.; Dunach, E. Synlett 1994, 531.
6. Hudlicky, T.; Claboe, C. D.; Brammer, L. E.; Koroniak,
L.; Butora, G.; Ghiviriga, I. J. Org. Chem. 1999, 64, 4909.
7. Franco, D.; Pahyella, D.; Rocamora, M.; Goez, M.;
Clinet, J. C.; Muller, G.; Dunach, E. Tetrahedron Lett.
1999, 40, 5685.
Acknowledgements
8. Kamal, A.; Laxman, E.; Rao, V. Tetrahedron Lett. 1999,
40, 371.
The authors thank the Environmental Protection
Agency (R826113), the National Science Foundation
.
.