An unexpected chelation-controlled Yb(OTf)3-catalyzed aminolysis and azidolysis of cyclitol epoxides

Article abstract of DOI:10.1021/jo0261146

A chelation-controlled aminolysis and azidolysis of cyclitol epoxides with Yb(OTf)₃ has been disclosed. The presence of a free OH group able to direct the coordination with the lanthanide seems essential for an efficient regiocontrol of the pro

Full text of DOI:10.1021/jo0261146

An Un exp ected Ch ela tion -Con tr olled
Yb(OTf) -Ca ta lyzed Am in olysis a n d
SCHEME 1
3
Azid olysis of Cyclitol Ep oxid es
Pedro Serrano,†,‡ Amadeu Llebaria, and
‡
Antonio Delgado*,
†,‡
Universitat de Barcelona, Facultat de Farm a` cia, Unitat de
Qu ı´ mica Farmac e` utica (Unitat Associada al CSIC), Avgda.
J oan XXIII, s/ n, 08028 Barcelona, Spain, and RUBAM,
Departament de Qu ı´ mica Org a` nica Biol o` gica, Institut
d’Investigacions Qu ı´ miques i Ambientals de Barcelona
bias to give the corresponding C2 adducts 3 arising from
trans-diaxial opening of an “all equatorial” conformation
such as 5B (Scheme 2).
8
(IIQAB-CSIC), J ordi Girona 18-26, 08034 Barcelona, Spain
This assumption was reinforced by the work of Crotti
6
et al. on cis-4-benzyloxy-1,2-epoxycyclohexane, which
adelgado@far.ub.es
3
emphasizes the nonchelating properties of Yb(OTf) to
Received J une 28, 2002
account for the observed regioselective aminolysis to give
the corresponding C2 adducts.
Initial aminolysis experiments from epoxide 5 and Et
2
-
Abstr a ct: A chelation-controlled aminolysis and azidolysis
of cyclitol epoxides with Yb(OTf)3 has been disclosed. The
presence of a free OH group able to direct the coordination
with the lanthanide seems essential for an efficient regio-
control of the process.
NH in toluene at room temperature in the presence of
6
3
50mol % Yb(OTf) were unsuccessful, since only starting
material was recovered.9 Harsher conditions were re-
quired for the complete aminolysis of the starting epoxide
1
0
5
1
5
.
However, contrary to our expectations, amino alcohol
h (Table 1, entry 9), arising from C1 opening of epoxide
Over the last few years, aminocyclitols have gained
importance as pharmacological tools for the study of
1
1,12
, was obtained instead as a single regioisomer.
The
1
products arising from a C1 opening could be easily
distinguished from its C2 opening regioisomers, since
cellular processes linked to the inositol cycle, as well as
2
potential glycosidase inhibitors. As part of our ongoing
1
13
both H and C-RMN were in the former case simplified
due to the symmetry of the molecule. Spectral data
obtained for the compounds obtained are consistent with
a trans-diaxial epoxide opening following the F u¨ rst-
Plattner rule. The same regio- and stereochemistry was
found on aminolysis of 5 with a variety of primary and
secondary amines (Table 1, entries 3-12) with the
exception of the bulky amines i and k , which failed to
react under these conditions. This unexpected outcome
can be interpreted as a result of a stereocontrolled trans-
diaxial opening of 5 through an “all-axial” conformation
research directed toward the design and synthesis of new
amino and diaminocyclitols as modulators of sphingolipid
metabolism, we became interested in the synthesis of
3
libraries of aminocyclitols 3 arising from regio- and
4
stereoselective C2 opening of epoxide 5 (Scheme 1).
Toward this end, we considered the combination of a
nucleophilic amine and a metal triflate Lewis acid
5
catalyst as the most suitable system for our purposes.
6
Following the seminal papers of Crotti et al. and
7
Yamamoto et al., lanthanide triflates seem the catalysts
of choice for the efficient aminolysis of epoxides in
nonprotic solvents. On the basis of these precedents, we
reasoned that aminolysis of epoxide 5 in the presence of
1
3
5
A stabilized by lanthanide chelation (Scheme 2). This
(8) For stereoelectronic effects in epoxide opening with nucleophiles,
3
Yb(OTf) would be mainly dictated by stereoelectronic
see: Rickborn, B.; Murphy, D. K. J . Org. Chem. 1969, 34, 3209. For
related fully benzylated conduritol epoxide opening through a confor-
mation similar to 5A with all substituents in the axial position see:
Montchamp, J . L.; Migaud, M. E.; Frost, J . W. J . Org. Chem. 1993,
58, 7679-7684. Takahshi, H.; Iimori, T.; Ikegami, S. Tetrahedron Lett.
1998, 39, 6939-6942.
*
Address correspondence to this author at IIQAB-CSIC. Fax: int
+
34-932045904.
†
Unitat de Qu ´ı mica Farmac e´ utica.
Institut d’Investigacions Qu ´ı miques i Ambientals de Barcelona.
‡
(1) (a) Brunn, G.; Fauq, A. H.; Chow, S.; Kozikowski, A. P.; Gallegos,
(9) A control aminolysis with cis-4-benzyloxy-1,2-epoxycyclohexane
A.; Powis, G. Cancer Chemother. Pharmacol. 1994, 35, 71-79. (b)
Powis, G.; Aksoy, I. A.; Melder, D. C.; Aksoy, S.; Eichinger, H.; Fauq,
A. H.; Kozikowski, A. P. Cancer Chemother. Pharmacol. 1991, 29, 95-
2
and Et NH, as described in ref 6, afforded the expected C2 aminolysis
product. The observed lack of reactivity from 5 can be thus attributed
to the nature of the substrate.
1
04.
(10) In general, aminolysis of cyclitol epoxides requires harsher
reaction conditions than that of simpler epoxycycloalkanes. See, for
example: (a) Nakata, M.; Tatsuta, K. J . Antibiotics 1993, 46, 1919-
1922. (b) Letellier, P.; Ralainirina, R.; Beaup e` re, D.; Uzan, R. Tetra-
hedron Lett. 1994, 35, 4555-4558. (c) Paulsen, H.; Burkhard, M.
Liebigs Ann. Chem. 1990, 169-180.
(11) The regio- and the stereochemistry of all compounds was
carefully checked by mono- and bidimensional NMR methods. After
peak assignment and assuming a chair conformation for the final
products, the coupling constants of cyclohexane hydrogen atoms were
of diagnostic value (see Supporting Information).
(12) Additional evidence of the structure of this amino alcohol has
also been obtained by comparison with the product of the chelation-
4
controlled LiClO promoted C1 aminolysis of epoxycyclitol 5. This
stereochemically well-defined process (see ref 13) gives a single amino
alcohol identical in all respects to the major product obtained with
(2) (a) Lillelund, V. H.; J ensen, H. H.; Liang, X. F.; Bols, M. Chem.
Rev. 2002, 102, 515-553; (b) Berecibar, A.; Grandjean, C.; Siriwardena,
A. Chem. Rev. 1999, 99, 779-844. (c) Potter, B. V. L.; Lampe, D. Angew.
Chem., Int. Ed. Engl. 1995, 34, 1933-1972.
(3) For a preliminary communication, see: Delgado, A.; Serrano, P.
In XII Congreso Nacional de la Sociedad Espa n˜ ola de Qu ı´ mica
Terap e´ utica; Book of Abstracts: Sevilla, Spain, 2001; p 61.
(
4) J aramillo, C.; Chiara, J .; Mart ´ı n-Lomas, M. J . Org. Chem. 1994,
9, 3135-3141.
5) For an overview of metal triflates as catalysts for aminolysis of
epoxides, see: Sekar, G.; Singh, V. K. J . Org. Chem. 1999, 64, 287-
89 and references therein.
6) Chini, M.; Crotti, P.; Favero, L.; Macchia, F.; Pineschi, M.
Tetrahedron Lett. 1994, 35, 433-436.
7) Meguro, M.; Asao, N.; Yamamoto, Y. J . Chem. Soc., Perkin Trans.
5
(
2
(
(
1
1994, 2597-2601.
3
Yb(OTf) .
1
0.1021/jo0261146 CCC: $22.00 © 2002 American Chemical Society
Published on Web 09/11/2002
J . Org. Chem. 2002, 67, 7165-7167
7165

SCHEME 2
TABLE 1
RRa
yield (%)
b
entry substrate nucleophile major isomer
RRa
yield (%)
b
entry substrate nucleophile major isomer
1
2
3
4
5
6
7
8
5
5
5
5
5
5
5
5
a
a
b
c
d
e
f
-
-
-
89
87
85
88
80
81
85
9
10
11
12
13
14
15
5
5
5
5
6
6
6
h
i
j
k
a
b
h
1h
-
>95:5
-
79
-
80
-
-
e
c
1a
1b
1c
1d
1e
1f
1g
>95:5
>95:5
80:20
90:10
90:10
>95:5
80:20
1j
90:10
-
-
-
2b/4b
-
c
-
55:45
-
d
-
g
a
RR: regioisomeric ratio determined by HNMR or HPLC. Isolated (major isomer). c Et3N (15 mol‚equiv) was added. d Together with
1
b
e
variable amounts of starting material. Not determined
chelation depends largely on the presence of the free C6-
OH group and its activation by an external amine,14 as
evidenced by comparing the azidolysis of 5 in the pres-
action.17 Additional evidence in favor of the dramatic
effect of the free OH group on Yb chelation and epoxide
activation toward nucleophilic attack is found on com-
parison with the reactivity of perbenzylated epoxide 6.
In this case, scarce (entry 14) or no reactivity (entries 13
and 15) was observed under our conditions.18 Moreover,
the lower diastereoselectivity observed with BuNH2
ence or in the absence of Et
(
3
N, a nonnucleophilic base
see Table 1, entries 1 and 2).15 In addition, the enhanced
acidity of the free OH group on coordination with the
lanthanide16 does not rule out an ionic stabilizing inter-
(compare entries 3 and 14) indicates a competition
(13) For the use of chelation-controlled aminolysis and azidolysis
of 1,2-epoxycyclohexanes bearing remote polar groups, see: Calvani,
F.; Crotti, P.; Gardelli, C.; Pineschi, M. Tetrahedron 1994, 50, 12999-
(15) The nucleophilic amines (10 equiv) used in the aminolysis of 5
(Table 1) can also play the role of the external base required to promote
OH-Yb coordination.
1
3022 and references therein.
(
14) (a) For a related amine-promoted coordination of a free OH with
Yb, see: Kobayashi, S.; Ishitani, H. J . Am. Chem. Soc. 1994, 116,
083-4084. (b) For the use of lanthanides in organic synthesis, see:
i) Lanthanides: Chemistry and use in organic synthesis; Kobayashi,
(16) (a) Neverov, A. A.; Montoya-Pelaez, P. J .; Brown, R. S. J . Am.
Chem. Soc. 2001, 123, 201-217. (b) Geyer, C. R.; Sen, D. J . Mol. Biol.
1998, 275, 483-489. (c) Breslow, R.; Zhang, B. J . Am. Chem. Soc. 1994,
116, 7893-7894.
4
(
S., Ed.; Springer-Verlag: Berlin, Heidelberg, New York, 1999; (ii)
Lanthanides in Organic Synthesis; Imamoto, T., Ed.; Academic Press
Limited: London, UK, 1994.
(17) For a related example, see: Morrow, J . R.; Aures, K.; Epstein,
D. J . Chem. Soc., Chem. Commun. 1995, 2431-2432.
7
166 J . Org. Chem., Vol. 67, No. 20, 2002

SCHEME 3a
an effective chelation-controlled Yb catalysis. All these
results are in full agreement with our postulated chelat-
ing model shown in Scheme 2.
In summary, a chelation-controlled aminolysis and
azidolysis of cyclitol epoxides with Yb(OTf) has been
3
disclosed. The presence of a free OH group able to direct
the coordination with the lanthanide seems essential for
an efficient regiocontrol of the process. This chelation
ability of a lanthanide triflate was hitherto unprec-
edented and should be added to the repertoire of other
well-established procedures for the chelation-controlled
aminolysis of epoxides.13
a
Conditions: (a) BuNH2 (10 mol‚equiv), Yb(OTf)3 (0.5 mol‚equiv),
toluene, 80 °C, 18 h; (b) NaN3 (10 mol‚equiv), Et3N (15 mol‚equiv),
Yb(OTf)3 (0.5 mol‚equiv), toluene, 80 °C 18 h.
Exp er im en ta l Section 19
Sta r tin g Ma ter ia ls. Cyclitol epoxides 6 and 8 were obtained
4
20
from benzylation of 5 and 7 with NaH/BnBr in DMF. Epoxide
(
(
(
()-11b was obtained as described in the literature.21 Epoxide
()-11a was obtained from debenzylation of (()-11b with Pd-
OH)
/H
2
in MeOH or, alternatively, by deprotection (DDQ in CH
O 18:1) of c-3-[(p-methoxybenzyloxy)methyl]-r-1,2-epoxy-
2
-
F IGURE 1.
Cl
2
2
cyclohexane, obtained by a modification of the method described
for (()-11b.
between the chelating and the nonchelating postulated
conformations.
The ability of Yb to effectively assist the opening of a
cyclitol epoxide through a chelating-controlled process
was also investigated on the homologue 7 bearing a
hydroxymethyl group at C6 (Scheme 3). Aminolysis of 7
Typ ica l Exp er im en ta l P r oced u r e for Ep oxid e Op en in g.
A solution of the starting epoxide 5 or 6 (0.5 mmol) in toluene
(
10 mL) is added dropwise under argon over Yb(OTf)
mmol, 155 mg) at room temperature. A solution of 5 mmol of
NaN and 7.5 mmol (1.0 mL) of Et N or the corresponding amine
3
(0.25
3
3
in toluene (1 mL) is next added and the reaction mixture is
stirred at 80 °C under argon. After 18 h, the reaction mixture is
with BuNH
from C1 opening of the starting epoxide.
azidolysis of 7 to give azido alcohol 10 took place only in
the presence of Et N as an external base. As above, the
2
afforded a single amino alcohol 9 arising
1
1,12
As expected,
cooled to room temperature, quenched with H O (10 mL), and
2
extracted with CH
extracts are exhaustively washed with water and the residue is
filtrated through a pad of silica gel (hexane/EtOAc/Et N 48/48/
) to remove traces of metal. The crude thus obtained is used to
calculate the diastereomeric ratio (see Table 1). Flash chroma-
2
Cl
2
(3 × 20 mL). The combined organic
3
3
presence of a free OH group to promote coordination with
4
Yb was also crucial, since the perbenzylated analogue 8
failed to react under our standard conditions.1
0,18
In
tography on hexane/EtOAc/Et
in the yield given in Table 1.
3
N (78/19/3) affords pure alcohols
addition, reaction of the simplified model (()-11b with
BuNH cleanly afforded the expected “C2-like” opening
adduct (()-12b (Figure 1). The enhanced reactivity of
Yb(OTf)
2
Ack n ow led gm en t. Financial support from Comis-
sionat per a Universitats i Recerca, Generalitat de
Catalunya (Projects 2001SGR00085 and 2001SGR00342)
is gratefully acknowledged. P.S. acknowledges the Min-
isterio de Educaci o´ n y Cultura for a predoctoral fellow-
ship. The authors also thank Dr. J osefina Casas for
assistance on HPLC-MS analyses and Mrs. Meritxell
Egido for experimental contributions.
3
to promote epoxide opening in (()-11b in
comparison to 8 can be explained by the establishment
of fewer “unproductive” coordination complexes in this
less densely functionalized substrate, in agreement with
our above hypothesis.18 Hydroxymethyl analogue (()-11a
led to a mixture of “C1-like” and “C2-like” regioisomers.
Direct NMR analysis of this mixture was hampered
presumably by traces of Yb which could not be removed
even after exhaustive washings. Formation of the above
mixture can be tentatively interpreted as an indirect
proof of the need of at least three coordination sites for
Su p p or tin g In for m a tion Ava ila ble: General experimen-
13
tal methods, spectral data, and C NMR spectra for com-
pounds 1a , 1b, 1c, 1d , 1e, 1f, 1g, 1h , 1j, 2b, 6, 8, 10, 11a a n d
12b. This material is available free of charge via the Internet
at http://pubs.acs.org.
(
18) A plausible explanation to account for the observed lack of
J O0261146
reactivity should lie on the inability of the lanthanide salt to coordinate
with the epoxide in these densely oxygenated substrates. The lack of
a close free OH group as “coordinator assistant” and the well-known
oxophilicity of Yb (ref 14b) may result in alternative, “unproductive”
coordination complexes with any of the remaining benzyloxy groups.
The same results were obtained with predried Yb(OTf)
mmHg, 48 h).
(19) For general experimental data see the Supporting Information..
(20) Letellier, P.; Ralainairina, R.; Beaup e` re, D.; Uzan, R. Synthesis
1994, 925-930. See also ref 10b.
(21) Chini, M.; Crotti, P.; Flippin, L. A.; Gardelli, C.; Macchia, F. J .
Org. Chem. 1992, 57, 1713-18.
3
(160 °C, 0.05
J . Org. Chem, Vol. 67, No. 20, 2002 7167