Synthesis of enantiopure amino alcohols by ring-opening of epoxyalcohols and epoxyethers with ammonia

Article abstract of DOI:10.1016/j.tetlet.2003.09.124

Arylglycidols and their corresponding ethers undergo regioselective ring-opening with aqueous ammonia in the presence of organic co-solvents (isopropanol or 1,4-dioxane) to afford 1,2-amino alcohols in high yield.

Full text of DOI:10.1016/j.tetlet.2003.09.124

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 8369–8372
Synthesis of enantiopure amino alcohols by ring-opening of
epoxyalcohols and epoxyethers with ammonia
Mireia Pasto´, Bele´n Rodr´ıguez, Antoni Riera and Miquel A. Perica`s*
Unitat de Recerca en S´ıntesi Asime`trica (URSA-PCB) Parc Cient´ıfic de Barcelona and Departament de Qu´ımica Orga`nica,
Universitat de Barcelona, c/Josep Samitier, 1-5. E-08028 Barcelona, Spain
Received 25 July 2003; revised 16 September 2003; accepted 16 September 2003
Abstract—Arylglycidols and their corresponding ethers undergo regioselective ring-opening with aqueous ammonia in the presence
of organic co-solvents (isopropanol or 1,4-dioxane) to afford 1,2-amino alcohols in high yield.
© 2003 Elsevier Ltd. All rights reserved.
Chiral enantiopure amino alcohols are attractive com-
pounds, either as ligands for asymmetric catalysis1 or as
building blocks for the preparation of biologically
active molecules.2 Among them, those bearing a free
amino group present the opportunity of employing
different protecting group strategies (e.g. N-Boc or
N-F-moc), in the synthesis of biomolecules or, when
dedicated to catalysis to being the starting materials for
the preparation of important types of catalytic ligands
such as oxazaborolidines,3 bis(oxazolines),4 or
phosphinooxazolines.5
We have subsequently shown that 3-amino-1,2-diols (3)
are highly versatile intermediates for the enantioselec-
tive synthesis of amino acids8 and other interesting
biomolecules,9 and that properly substituted amino
alcohols (4) represent an interesting class of highly
modular chiral ligands.10
Whenever the preparation of amino alcohols containing
unprotected amino groups has been required, two dif-
ferent two-stage procedures have been used according
to common practice. In one of these procedures, benz-
hydrylamine has been employed as the source of the
amino group, and the unmasking of the amino function
has subsequently required hydrogenolysis of the doubly
benzylic carbonꢀnitrogen bond. In the other approach
−
N3 , either in the form of diisopropoxytitanium diazide
A very common method for preparing N-unprotected
1,2-amino alcohols has been the addition of
organometallic species onto the carboxyl group of natu-
ral amino acids.6 This method, however, provides only
access to a limited set of amino alcohol structures. To
overcome this limitation, we have developed methods
for the stereospecific synthesis of enantiomerically pure
b-amino alcohols (3, 4) through the regioselective and
stereospecific ring opening of epoxyalcohols (1) or
epoxyethers (2) arising from the Sharpless7 epoxidation.
or as sodium azide in the presence of lithium perchlo-
rate, has been the source of the amino group, again
with the need for reduction of the intermediate organyl
azide.
Whereas the use of benzhydrylamine as a source for
amino groups is clearly uneconomic in atomic terms,
the use of azide for the same purpose introduces some
degree of risk in the resulting sequences. Bearing in
mind the importance of enantiopure amino alcohols,
the search for greener alternatives for this considered
synthetic operation, involving a more rational use of
resources or avoiding the use of potentially dangerous
Keywords: amino alcohols; ammonia; ring-opening; epoxides.
* Corresponding author. Tel.: +34-93-4037094; fax: +34-93-4037095;
e-mail: mapericas@pcb.ub.es
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.09.124

8370
M. Pasto´ et al. / Tetrahedron Letters 44 (2003) 8369–8372
these reactions (Scheme 1) and, for solubility reasons, a
co-solvent was also used.
reagents is fully warranted. In line with these ideas, we
wish to report here on the development of a safe and
straightforward procedure for the regioselective and
stereospecific synthesis of amino alcohols from arylgly-
cidols and ethers, involving the use of inexpensive
aqueous ammonia as the sole reagent and isopropyl
alcohol or 1,4-dioxane as reaction co-solvents.
The results and reaction conditions employed for these
reactions15 are summarized in Table 1.
Interestingly, when the benzhydryl ether of phenylgly-
cidol (5b) was heated for 7 h at 85°C in ammonium
hydroxide/methanol in the presence of a 0.15% equiv.
molar amount of ytterbium triflate (pressure tube)
regioselective ring-opening took place, and amino alco-
hol 6b was obtained in 49% yield. Under these condi-
tions, methanol competed with ammonia as the
nucleophile, and 3-benzhydryloxy-1-methoxy-1-phenyl-
propan-2-ol arising from C-3 attack was also obtained
(10% yield). To avoid this problem, the use of a less
nucleophilic co-solvent was considered. Changing to
isopropanol (entry 2) led to an increase in the yield of
6b, but with longer reaction time. At this stage, in order
to analyze the relative importance of the Lewis acid on
the reaction course, we decided to perform the reaction
without adding the ytterbium salt (entry 3). To our
surprise, the reaction rate showed no decrease, 6b being
obtained in an optimal 71% yield with complete
regiocontrol.
Although the possibility of preparing amino alcohols
by ring-opening of epoxides with ammonia is discussed
in textbooks, the use of this reaction for synthetic
purposes is scarce.11 Thus, even terminal epoxides are
reported to be attacked slowly by ammonia in the
absence of special activation, while most examples with
disubstituted epoxides refer to situations where an adja-
cent carbonyl group activates the ring-opening pro-
cess.12 In any case, no example of ring-opening with
ammonia of a non-terminal epoxyalcohol or epoxyether
has been reported in the literature.13
For the initial attempts of direct ring-opening of these
substrates, we reasoned that the use of the water-com-
patible lanthanide triflates introduced by Kobayashi14
in conjunction with aqueous ammonia could greatly
facilitate the reaction by Lewis acid activation of the
epoxide. Epoxyethers 5a–d were used as substrates for
The same reaction conditions were subsequently
applied to 5a, 5c and 5d. In the case of 5c, 1,4-dioxane
was used as a co-solvent instead of isopropyl alcohol to
improve substrate solubility. As can be observed in
Table 1 (entries 3–6) the reactions took place very
cleanly, leading in high yield to amino alcohols 6a–d.
The regioselectivities in these ring-opening reactions
were in all cases very good, with no C-2 ring-opened
products being detected (analysis by 13C NMR). More-
over, since no syn diastereomers of the amino alcohols
6 could be detected by 13C NMR, and the enantiomeric
purity of amino alcohols 6a–d was coincident with that
of the starting phenylglycidyl ethers 5a–d (>99%), the
ring-opening reactions are stereospecific. When R4 was
a rather bulky protecting group, the process turned out
to be slower and some starting material was recovered
(entries 5 and 6). Conversely a very high yield was
recorded for the substrates bearing a non-bulky R4
group. All reaction products were fully characterised by
1H and 13C NMR data and by comparison with the
known compounds16 obtained via the indirect, two-
stage procedure.
Scheme 1. Ring-opening of phenylglycidyl ethers 5 with aq.
ammonia.
Table 1. Ring-opening of phenylglycidyl ethers 5 with
aqueous ammonia
Entry Product R4
Reaction conditions
Yield (%)
1
2
6b
6b
CHPh2 MeOH, Yb(OTf)3, 85°C,
49
61
7 h
CHPh2 iPrOH, Yb(OTf)3, 85°C,
20 h
3
4
5
6
6b
6a
6c
6d
CHPh2 iPrOH, 80°C, 23 h
71
Me
iPrOH, 85°C, 15 h
1,4-dioxane, 85°C, 60 h
iPrOH, 80°C, 17 h
95
CPh3
82a
60b
SiiPr3
Also interesting is the direct ring-opening of epoxyalco-
hols with ammonia, since the resulting aminodiols
could be transformed in a very straightforward and
practical manner, following our previously described
methodology,8a,c into a variety of biologically impor-
tant amino acids and dipeptide isosteres. Accordingly,
epoxyalcohols 7a–c were evaluated in this epoxide-
opening process (Scheme 2). The results obtained in
these experiments are summarized in Table 2.
a Conversion was 83%.
b Conversion was 74%.
Again in this case, the reactions carried out on aryl-
glycidol substrates took place stereospecifically and
with complete regioselectivity leading to the target
Scheme 2. Ring-opening of epoxyalcohols 7 with aqueous
ammonia.

M. Pasto´ et al. / Tetrahedron Letters 44 (2003) 8369–8372
8371
protected arylglycines8 renders these products available
in three simple steps and high overall yield (Scheme 3).
Table 2. Ring-opening of epoxyalcohols 7 with aqueous
ammonia
Entry
Product
R
Reaction
Yield (%)
conditions
Acknowledgements
1
2
3
8a
8b
8c
Phenyl
Mesityl
iPrOH, 70°C, 86
18 h
iPrOH, 80°C, 71a
20 h
Financial support from DGI-MCYT (BQU2002-02459)
and DURSI (2001SGR50). B.R. thanks the University
of Barcelona for a fellowship.
Biphenyl-4-yl 1,4-Dioxane, 95
80°C, 23 h
a Conversion was >95%.
References
1. Noyori, R. Asymmetric Catalysis in Organic Synthesis;
John Wiley & Sons: New York, 1994.
2. (a) Coppola, G. M.; Schuster, H. F. Asymmetric Synthe-
sis. Construction of Chiral Molecules Using Amino Acids;
Wiley: New York, 1987; (b) Bergmeier, S. C. Tetrahedron
2000, 56, 2561–2576.
3. Corey, E. J.; Bakshi, R. K.; Shibata, S. J. Am. Chem.
Soc. 1987, 109, 5551–5653.
4. Desimoni, G.; Faita, G.; Mella, M. Tetrahedron 1996, 52,
13649–13654.
5. Helmchen, G.; Pfaltz, A. Acc. Chem. Res. 2000, 33,
336–345.
Scheme 3. Three-step synthesis of enantiopure arylglycines.
6. Dauelsberg, C.; Behnen, W.; Martens, J. DE. Patent
4341605, 1995 (CAN 123: 285757).
7. Gao, Y.; Klunder, J. M.; Hanson, R. M.; Masamune, H.;
Ko, S. Y.; Sharpless, K. B. J. Am. Chem. Soc. 1987, 109,
5765–5780.
aminodiols17 in good yield. As already noted for
phenylglycidyl ethers (Table 1) the reaction is some-
what sensitive to steric hindrance near the epoxide ring,
and mesitylglycidol 7b underwent ring-opening at a
slightly slower rate despite of the strong electron donat-
ing nature of the mesityl substituent.
8. Selected references: (a) Medina, E.; Moyano, A.; Perica`s,
M. A.; Riera, A. Helv. Chim. Acta 2000, 83, 972–988; (b)
Alco´n, M.; Moyano, A.; Perica`s, M. A.; Riera, A. Tetra-
hedron: Asymmetry 1999, 10, 4639–4651; (c) Medina, E.;
Vidal-Ferran, A.; Moyano, A.; Perica`s, M. A.; Riera, A.
Tetrahedron: Asymmetry 1997, 8, 1581–1586; (d) Poch,
M.; Alco´n, M.; Moyano, A.; Perica`s, M. A.; Riera, A.
Tetrahedron Lett. 1993, 34, 7781–7784.
9. Selected references: (a) Mart´ın, R.; Alco´n, M.; Perica`s,
M. A.; Riera, A. J. Org. Chem. 2002, 67, 6896–6901; (b)
Ginesta, X.; Perica`s, M. A.; Riera, A. Tetrahedron Lett.
2002, 43, 779–782; (c) Pasto´, M.; Moyano, A.; Perica`s,
M. A.; Riera, A. Tetrahedron Lett. 1998, 39, 1233–1236;
(d) Pasto´, M.; Castejo´n, P.; Moyano, A.; Perica`s, M. A.;
Riera, A. J. Org. Chem. 1996, 61, 6033–6037.
10. Selected references: (a) Pasto´, M.; Riera, A.; Perica`s, M.
A. Eur. J. Org. Chem. 2002, 2337–2341; (b) Puigjaner, C.;
Vidal-Ferran, A.; Moyano, A.; Perica`s, M. A.; Riera, A.
J. Org. Chem. 1999, 64, 7902–7911; (c) Vidal-Ferran, A.;
Moyano, A.; Perica`s, M. A.; Riera, A. J. Org. Chem.
1997, 62, 4970–4982.
11. (a) Favretto, L.; Nugent, W. A.; Licini, G. Tetrahedron
Lett. 2002, 43, 2581–2584; (b) Lindstrom, U. M.; Olofs-
son, B.; Somfai, P. Tetrahedron Lett. 1999, 40, 9273–
9276.
12. (a) Kurokawa, N.; Ohfune, Y. Tetrahedron 1993, 49,
6195–6222; (b) Valpuesta, M.; Durante, P.; Lopez-Her-
rera, F. J. Tetrahedron Lett. 1995, 36, 4681–4684; (c)
Wuts, P. G. M.; Gu, R. L.; Northuis, J. M. Tetrahedron:
Asymmetry 2000, 11, 2117–2123.
On the other hand, and not unexpectedly when the
reaction was attempted on an aliphatic substrate such
as (2S,3S)-2,3-epoxyhexanol the process was not as
effective and was poorly regioselective. Thus, after
heating the reaction mixture at 85°C for 60 h under the
standard reaction conditions a mixture of regioisomeric
aminodiols was obtained in low yield (29%). The ben-
zylic nature of the carbon atom being attacked thus
appears to be a requisite for the reaction to proceed.
In summary, the ring-opening of arylglycidyl ethers and
arylglycidols with aqueous ammonia is a feasible reac-
tion; it takes place stereospecifically and with complete
regiocontrol, and provides a convenient alternative to
the classical two-stage sequences involving either poorly
atom-economic conditions or the use of potentially
hazardous reagents. Bearing in mind both the potential
uses in catalysis of the amino alcohols derived from
arylglycidyl ethers, and the interest in arylglycines with
aromatic groups covering a range of electronic or steric
characteristics, it is to be expected that the newly
developed procedure will find, for its simplicity and
sustainable characteristics, ample application among
synthetic chemists.
As a simple illustration of its potential, the integration
of the aqueous ammonia-mediated regioselective ring-
opening of enantiopure arylglycidols into well estab-
lished synthetic schemes for the preparation of
13. Methods for the ring-opening of terminal epoxides have
been reported in the patent literature. See: Eicher, J.;

8372
M. Pasto´ et al. / Tetrahedron Letters 44 (2003) 8369–8372
Fazniewscy, K.; Rudolph, W. DE. Patent 3830351, 1990
(CAN 113: 151818).
14. Kobayashi, S.; Nagayama, S.; Busujima, T. J. Am. Chem.
Soc. 1998, 120, 8287–8288.
epoxy-1-triisopropylsilyloxypropane and 6d, (1R,2R)-1-
amino-1-phenyl-3-triisopropylsilyloxypropan-2-ol
are
1
new compounds and gave satisfactory IR, H, 13C NMR
and MS data.
17. Since aminodiols 8a and 8b are highly polar compounds
they were purified by chromatography through a SiO2
column with AcOEt/MeOH. Epoxide 7c was prepared as
a racemate. Due to its high insolubility aminodiol 8c,
which was obtained in a relatively pure form, was directly
transformed into its diastereomerically pure (13C NMR)
N-Boc derivative. All new compounds gave satisfactory
1H and 13C NMR data.
15. Representative
experimental
procedure:
Epoxide,
NH4OH (aprox. 40 equiv. of a 30% solution in water) in
iPrOH (volume NH4OH/volume PrOH, 1/2) were stirred
i
in a sealed tube at the temperature specified and the time
indicated in Table 1 or Table 2. The solution was concen-
trated in vacuo and purified by chromatography through
an SiO2 column with hexane /EtOAc as eluent.
16. See Ref. 10b. Compounds 5d, (2S,3S)-3-phenyl-2,3-