Chemo- and regioselective reductive opening of azetidinium ions

Article abstract of DOI:10.1016/j.tetlet.2006.11.175

Enantiomerically pure azetidinium trifluoromethanesulfonates were opened by various hydride reagents. LiAlH₄ and NaBH₃CN reacted with a complete regioselectivity and the latter reagent also reacted in a chemoselective way, leaving un

Full text of DOI:10.1016/j.tetlet.2006.11.175

Tetrahedron Letters 48 (2007) 1027–1031
Chemo- and regioselective reductive opening of azetidinium ions
*
Fran c¸ ois Couty, Olivier David and Fran c¸ ois Durrat
Institut LAVOISIER, UMR CNRS 8180, Universit e´ de Versailles, 45, avenue des Etats-Unis, 78035 Versailles C e´ dex, France
Received 26 October 2006; revised 23 November 2006; accepted 29 November 2006
Available online 22 December 2006
Abstract—Enantiomerically pure azetidinium trifluoromethanesulfonates were opened by various hydride reagents. LiAlH
4
and
NaBH CN reacted with a complete regioselectivity and the latter reagent also reacted in a chemoselective way, leaving unaffected
3
an ester or a cyano moiety present in the substrate. This reaction provides 1,2-diamines, 1,2- and 1,4-amino alcohols or a-amino
esters by combining proper choice of substrate and hydride reagent.
Ó 2006 Elsevier Ltd. All rights reserved.
1
Azetidinium ions 1 are electrophilic synthons able to
the reductive opening of azetidinium ions, (i.e.,
2
À
À
react with a range of heteronucleophiles such as amines,
Nu = H in Scheme 1), a reaction that proceeds in a
highly chemo- and regioselective manner giving rise to
enantiopure functionalized amines.
3
3
4
azide anions, acetate anions, thiols, sodium thiosul-
5
6
fate and phosphorous-containing nucleophiles, leading
to the corresponding opened product in a good yield.
Recently, we have studied in details the regioselectivity
of this nucleophilic opening starting with enantiopure
In order to determine the scope of this reduction we first
selected a series of azetidinium ions (trifluoromethane-
sulfonate salts) and these substrates can be divided into
three categories. Compounds 4–6, derived from (R)-phen-
yl glycinol are bearing a representative set of moieties
at C-2 and are unsubstituted at C-4. On the contrary,
compounds 7–10 prepared from (À)-ephedrine are
substituted at C-4. Finally, compounds 11 and 12 (pre-
pared respectively from (S)-phenylethylamine or benzyl-
amine) are unsubstituted at C-3,4. All these substrates
were prepared from the corresponding azetidine by
functionalized azetidines such as 2 (FG = CN, CO Et)
or 3, and we have demonstrated that heteronucleophiles
azide and acetate anions or amines) preferably attack at
the unsubstituted position in 2 and at the carbon bearing
2
(
3
the functional group in type 3 substrates (Scheme 1).
Considering the new methodologies available for the
synthesis of functionalized azetidines in enantiomeri-
cally pure form, and the efficiency in terms of yield
7
and regioselectivity associated with the above reaction,
it is surprising to note that only scarce examples of such
8
HO
nucleophilic opening involving carbon nucleophiles or
9
Ph
CO
2
Et
Ph
CN
Me
Ph
hydrides have been reported so far. In this letter, in
Ph
continuation with our interest in the reactivity of azetid-
N
N
N
1
0
Me
Me
inum ions, we wish to report some examples involving
Bn
4
Bn
5
Bn
6
Ph
CO Et
Ph
CN
Ph
CN
Ph
2
OH
Me
Nu-
Nu-
N
N
N
N
FG
Me
FG
Me
Me
Me
Me
Me
Me
Me
Me
N+
N+
Me
2
N+
3
Me
Me
1
8
10
7
9
CN
CN
Scheme 1. Azetidinium ions are regioselectively opened by
heteronucleophiles.
N
Me
N
Me
Me
1
1
12
Ph
Ph
Keywords: Azetidines; Azetidinium ions; Strain; Reduction.
*
Corresponding author. Tel.: +33 139254455; fax: +33 139254452;
e-mail: couty@chimie.uvsq.fr
Scheme 2. Selected azetidinium trifluoromethanesulfonates for reduc-
tive opening.
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.11.175

1028
F. Couty et al. / Tetrahedron Letters 48 (2007) 1027–1031
Table 1. Reaction of azetidinium ions with hydride donors
a
Entry
1
Substrate
Conditions
Product(s)
Yield (%)
Me Ph
4
4
LiAlH , 2 equiv, THF, 0 °C, 1 h
N
40
Bn
OH
13
Ph
Me
Me
N
N
2
3
7
LiAlH
4
, 2 equiv, THF, 0 °C, 1 h
34
OH
1
4
Me
Ph
Me
Me
10
LiAlH
4
, 2 equiv, THF, 0 °C, 1 h
42
59
OH
1
4
Me
Me Ph
N
4
5
LiAlH
4
, 2 equiv, THF, 0 °C, 1 h, then Boc
2
O,
Bn
N
AcOEt, reflux
15
Boc
Ph OH
5
6
7
6
9
LiAlH
4
, 2 equiv, THF, 0 °C, 1 h
Me
Ph
49
1
6
N
N
Me
Bn
b
LiAlH
4
, 2 equiv, THF, 0 °C, 1 h, then Boc
2
O,
—
AcOEt, reflux
Me
NHTr
11
LiAlH
4
, 2 equiv, THF, 0 °C, 1 h, then TrCl, Et
3
N
Me 17
82
Me
N
Ph
Ph
Me
Me
CN
8
9
9
9
NaBH
4
, 2 equiv, THF, rt, 6 h
39
73
1
8
Me
Ph
Me
Me
N
N
CN
CN
NaBH
3
CN, 5 equiv, THF, reflux, 48 h
1
8
Me
Ph
Me
Me
1
1
0
1
9
8
NaBH(OAc)
3
, 5 equiv, THF, reflux, 24 h
19
Quant.
Me OAc
(
8:2 mixture of
stereoisomers)
Ph
Me
Me
N
CN
NaBH
3
CN, 5 equiv, THF, reflux, 48 h
67
1
8
Me
CN
Me
Me
BH3
Ph
N
20
N
21
Me
Me
c
12
12
NaBH
3
CN, 5 equiv, THF, reflux, 48 h
Ph
63 (Ratio 20:21:22 = 1:4:3)
Me
N
2
2
Me
Ph

F. Couty et al. / Tetrahedron Letters 48 (2007) 1027–1031
1029
Table 1 (continued)
Entry Substrate
a
Conditions
Product(s)
Yield (%)
CN
Me
Me
BH3
Ph
d
13
12
3
NaBH CN, 5 equiv, THF, reflux, 2 h
75 (Ratio 20:21 = 2:1)
N
20
N
21
Me
Me
Ph
Ph
CO2Et
Ph
14
4
3
NaBH CN, 5 equiv, THF, reflux, 48 h
Me
67
N
Me
23
a
b
c
Yield of isolated product.
A complex mixture was obtained.
Combined yields. Compounds 20/21 on one side and 21/22 on the other side were obtained as mixtures.
d
These compounds were obtained as mixtures.
methylation with methyltrifluoromethanesulfonate11
or cyano moiety) could be achieved. The issues of these
experiments are collected in Table 1.
(
Scheme 2).
Two sources of hydride ions were then selected in order
to test the reductive opening with these substrates.
These results first bring one general comment: except
entries 6, 12, 13 a single compound is obtained resulting
from the regioselective opening of the four-membered
ring by a hydride ion. This demonstrates that both
reagents are able to perform the reductive opening of
the azetidinium ring. The yields of reduced compounds
LiAlH was first chosen since it has been previously
4
9
12
described to open azetidinium or aziridinium ions.
Then, NaBH4 and related reagents NaBH(OAc) and
1
3
3
NaBH CN were next tested in order to determine
3
whether some chemoselectivity (i.e., reductive opening
of the azetidinium ion without reduction of the ester
are quite low (especially when LiAlH is used, mostly
4
because of the highly polar nature of the produced com-
pounds) to fair. In all cases, no significant amount of
regioisomer other than the one depicted in Table 1 could
be detected in the crude reaction mixture. Whereas
-
[
]Al O
Ph
Ph
Me Ph
N
LiAlH leads to a concomitant reduction of the adjacent
4
Ph
LiAlH4
moiety (ester or nitrile) at C-2, the use of NaBH (or
N+
16
14
4
6
7
Bn
Me
more efficiently NaBH CN: compare entries 8 and 9)
O
3
Bn
permits to preserve these moieties in the substrate. How-
2
4
26
ever, NaBH(OAc) induces an opening of the azetidi-
3
Ph
nium by an acetate anion, as previously observed
LiAlH4
Ph
-
O Al[ ]
3
Me N
2
(entry 10).
N+
O
Me
Me
Me
27
Me
The regioselectivity of the opening depends both on the
substitution pattern of the azetidinium ring and on the
nature of the hydride reagent. Hence, when boron-
derived reagents NaBH and NaBH CN are used, the
2
5
Scheme 3. Reduction of azetidinium ions might proceed through an
4
3
intermediate epoxide.
regioselectivity follows the one previously observed with
Ph
OAl[ ]
Ph
Ph
H
OH
Ph
LiAlH4
BnMeN
NMeBn
2
6
+
N+
Me Bn
O Al[ ]
Ph Ph
2
9
2
8
Ph
b
a
NMe2
Ph
H
OAl[ ]
Ph
OH
3
1
LiAlH4
Me N
2
H
2
7
b
+
N+
Me Me
O Al[ ]
Ph
a
Me
NMe2
30
Me
OH
32
Scheme 4. Reduction of c-epoxy amines 26 and 27 produces alkenes.

1030
F. Couty et al. / Tetrahedron Letters 48 (2007) 1027–1031
heteronucleophiles, which is summarized in Scheme 1:
azetidinium ions 9 and 8 substituted at C-4 are regiose-
lectively opened at C-2 (entries 8, 9 and 11) and azetid-
iniums 12 or 4, devoid of substituents at C-4 are also
opened at this position (entries 12–14). It should be
noticed that in the case of substrate 12, this reduction
is not totally chemoselective, since the produced amino-
ate pyrrolidinium ions 28 or 30 that undergo b-elimina-
tion reactions. These results demonstrate that epoxides
26 and 27 are probably not produced in the course of
the reduction of 6 and 7 by LiAlH and that a direct
4
opening of the aziridinium ion by the hydride ion is
operating (Scheme 4).
1
4
nitrile 20 undergoes further reductive decyanation by
In conclusion, we have shown that azetidinium ions can
be reductively opened in a highly chemo- and regioselec-
tive way by hydride ions. By an appropriate choice of
the substrate and hydride reagent, this reaction affords
functionalized enantiomerically pure amines of high
synthetic relevance, such as a-amino ester 23, 1,2-di-
amine 17, 1,2- and 1,4-amino alcohols 16 and 13. Syn-
thetic applications of this methodology are in progress
in our group.
NaBH CN to give the corresponding amine 22, accom-
3
panied by its BH complex 22. This overreduction can
3
however be minimized by running the reduction for
2
h instead of 48 h (entry 13).
When LiAlH is used as the reducing agent, the regiose-
4
lectivity is less simple to explain due to the concomitant
reduction of the functional group linked to the azetidi-
nium ring. Compounds 4, 7 and 10 (entries 1–3) are all
resulting from a regioselective opening at C-2 and a con-
comitant reduction of the ester moiety (for 4 and 7). On
the other hand, compounds 6 and 11 are exclusively
opened at C-4 (entries 5 and 7) and in the latter case,
the cyano moiety is reduced into a primary amine, to
give (after tritylation), diamine 17 in a good yield.
Although these regioselectivities are most often in agree-
ment with the one summarized in Scheme 1, it is unex-
pected with substrate 4 that would logically be opened
at C-4 considering the precedent outcomes. As a matter
of fact the different issues of regioselectivity for sub-
strates 4 (opening at C-2, entry 1) and 6 (opening at
C-4, entry 5) are rather intriguing and suggest a strong
directing effect of the produced aluminium alkoxide
for the nucleophilic opening by the hydride. This direct-
ing effect is however radically different when the alkox-
ide is produced by the reduction of an ester
Acknowledgment
CNRS is acknowledged for the generous support.
Supplementary data
Experimental procedures and characterization for com-
pounds 13–16, 18 and 23. Supplementary data associ-
ated with this article can be found, in the online
version, at doi:10.1016/j.tetlet.2006.11.175.
References and notes
1
. For a review, see: Couty, F.; Durrat, F.; Evano, G. In
Targets in Heterocyclic Systems—Chemistry and Proper-
ties; Attanasi, O. A., Spinelli, D., Eds.; Italian Society of
Chemistry: Rome, 2005; Vol. 9, pp 186–210.
(
6
compound 4) or from a secondary alcohol (compound
). The intervention of the reduced moiety at C-2 is fur-
ther illustrated by entry 4. In this case, the aluminium
amide resulting from the reduction of the nitrile by
2
3
. (a) Bernstein, S.; Heller, M. J. Org. Chem. 1971, 36, 1386–
LiAlH is nucleophilic enough to override the opening
4
1
1
389; (b) Szmuszkovicz, J.; Kane, M. P. J. Org. Chem.
981, 46, 3728–3730.
by an hydride ion, and aziridine 15 is isolated in a fair
yield, after N-protection. The success of this opening
probably depends on subtle parameters, because when
this reaction was attempted with the closely related sub-
strate 9, then a complex mixture was obtained in a low
yield (entry 6), suggesting the occurrence of various
competitive pathways. Moreover, this nucleophilic
opening does not compete with substrate 11 (entry 7).
. Couty, F.; David, O.; Durrat, F.; Evano, G.; Lakhdar, S.;
Marrot, J.; Vargas-Sanchez, M. Eur. J. Org. Chem. 2006,
3479–3490.
4. Hata, Y.; Watanabe, M. Tetrahedron 1987, 43, 3881–3888.
5
6
. Jeziorna, A.; Heli n´ ski, J.; Krawiecka, B. Synthesis 2003, 2,
88–294.
2
. (a) Heli n´ ski, J.; Skrzypczy n´ ski, Z.; Michalski, J. Tetra-
hedron Lett. 1995, 36, 9201–9204; (b) Bakalarz, A.;
Heli n´ ski, J.; Krawiecka, B.; Michalski, J.; Potrzebowski,
M. J. Tetrahedron 1999, 55, 12211–12226; (c) Bakalarz-
Jeziorna, A.; Heli n´ ski, J.; Krawiecka, B. J. Chem. Soc.,
Perkin Trans. 1 2001, 1086–1090; (d) Jeziorna, A.;
Heli n´ ski, J.; Krawiecka, B. Tetrahedron Lett. 2003, 44,
3239–3243; (e) Krawiecka, B.; Jeziorna, A. Tetrahedron
Lett. 2005, 46, 4381–4384.
The participation of aluminium alkoxide 24 or 25,
resulting from the treatment of azetidinium 6 and 7 by
LiAlH could indeed result from the formation of an
4
intermediate epoxide 26 or 27 that would undergo
further regioselective reduction to furnish 16 and 14
(
Scheme 3) rather than a direct hydride opening of the
7
. For a review, see: Couty, F.; Evano, G.; Prim, D. Mini
Reviews in Organic Chemistry 2004, 1, 133–148.
azetidinium ring.
8
. For some isolated example, see: Refs. 6a,c (opening with
phosphorous-stabilized C-nucleophiles). See also: Wills,
M. T.; Wills, I. E.; Von Dollen, L.; Butler, B. L.; Porter, J.;
Anderson, A. G. J. Org. Chem. 1980, 45, 2489–2498
This hypothesis was discarded, running the following
experiments: we subjected epoxides 26 and 273 to
LiAlH4 reduction, but no traces of 16 or 14 were
detected in the crude reaction mixture. Instead, alkene
(
opening with BuLi).
2
9 could be isolated (49%) when starting from 26, and
9
. For a single example, involving LAH as hydride source,
see: Kikuchi, T.; Uyeo, S. Chem. Pharm. Bull. 1967, 15,
549–570.
alkenes 31 and 32 from 27 (16%). The formation of these
alkenes can be explained by the formation of intermedi-

F. Couty et al. / Tetrahedron Letters 48 (2007) 1027–1031
1031
1
1
1
0. Couty, F.; Durrat, F.; Evano, G.; Marrot, J. Eur. J. Org.
Chem. 2006, 4214–4223.
1. For the preparation of these substrates, see: Ref. 3 and
references cited therein.
2. See for examples: (a) Bernhard, H. O.; Snieckus, V.
Tetrahedron 1971, 27, 2091–2100; (b) Malpass, J. R.;
White, R. J. Org. Chem. 2004, 69, 5328–5344.
13. For an example of reductive opening of tetrahydroiso-
quinolinium methiodide induced by NaBH , see: Hara,
H.; Kaneko, K.; Endoh, M.; Uchida, H.; Hoshino, O.
Tetrahedron 1995, 37, 10189–10204.
14. For an example of reductive decyanation of an amino
4
nitrile by NaBH
4
, see: Guerrier, L.; Royer, J.; Grierson,
D. S.; Husson, H-P. J. Am. Chem. Soc. 1983, 105, 7754.