A conceptually new and straightforward method for the one-pot transformation of alcohols into amines

Article abstract of DOI:10.1055/s-2002-20487

A conceptually new amination method of alcohols has been developed using α-hydroxy esters as substrates and N-phenyl bis-trifluoromethanesulfonimide (PhNTf₂) as a test reagent. Playing two roles at once, PhNTf₂ activates the hydroxyl group as a highly reactive triflate intermediate and introduces the amino functionality through an in situ nucleophilic substitution by the anionic residue PhTfN^-. Two complementary procedures (methods A and B herein) have been developed, the latter permitting reaction of substrates with unstable alcoxides.

Full text of DOI:10.1055/s-2002-20487

528
LETTER
A Conceptually New and Straightforward Method for the One-pot Transfor-
mation of Alcohols into Amines
O
ne-pot Transfor
m
.
ation of
Alc
G
ohols into
A
minesasparik, V. Dalla,* B. Decroix
Unité de Recherche en Chimie Organique et Macromoléculaire, Faculté des Sciences et Techniques de l’Université du Havre, 25 rue Phil-
ippe Lebon, BP 540, 76058 Le Havre cedex, France
Fax +33(2)32744391; E-mail: vincent.dalla@univ-lehavre.fr
Received 18 December 2001
for trapping enolates to form vinyl triflates,7,8 in conjunc-
Abstract: A conceptually new amination method of alcohols has
tion with the good nucleophilic properties of triflamides
been developed using -hydroxy esters as substrates and N-phenyl
bis-trifluoromethanesulfonimide (PhNTf2) as a test reagent. Playing
two roles at once, PhNTf2 activates the hydroxyl group as a highly
RNHTf9 and their anions,10 we reasoned that triflimides of
general formula RNTf2 could met the criteria of our pur-
reactive triflate intermediate and introduces the amino functionality pose. The first step of our planned reaction is proposed as
through an in situ nucleophilic substitution by the anionic residue
follows: there is an initial addition of the alcoholate of the
PhTfN–. Two complementary procedures (methods A and B herein)
starting -hydroxyester to a triflimide RNTf2, which may
have been developed, the latter permitting reaction of substrates
provide a transient triflate along with a nucleophilic resi-
with unstable alcoxides.
due –NRTf. Subsequently, the latter may easily substitute
the triflate group to give the expected amino ester
(Scheme 1).
Key words: amination, amino esters, nucleophilic substitution, tri-
flate, one-pot
R1
R2
R1
R2
OTf
R1
R2
RTfN-
-Amino acids and their derivatives play a pivotal role in
organic synthesis. The ease of preparation of -hydroxy
esters makes them ideal substrates for the preparation of
-amino esters.1 Originally, such a functional group inter-
conversion was achieved by means of the Mitsunobu re-
action.2 More recently, it has been demonstrated that a
two-step procedure based on the activation of a hydroxyl
followed by displacement by a nucleophilic nitrogen
source was more widely applicable. In this case, the tri-
flate group has been recognized as the most convenient
leaving group by comparison with usual bromide, mesy-
late, sulfonate, both in terms of efficiency and chirality.3
Base
RNTf2
OH
NRTf
Scheme 1
Herein we wish to communicate our preliminary results
establishing that the new concept exposed above can be
accomplished, with the use of PhNTf2 as a test reagent and
-hydroxy esters as substrates. To the best of our knowl-
edge, the utilization of the crystalline, stable and commer-
cially available PhNTf2 for such a dual role of activating
and aminating an hydroxyl group has not been previously
reported.
Both methods, however, are not suitable for all circum-
stances.4,5 Therefore, the development of new methodolo-
gies allowing for the direct transformation of an alcohol
into an amine remains a subject of current interest.6
The first part of this work was aimed at optimizing the
conditions for the amination reaction.
Except for the commercially available methyl mandelate
1a and ethyl lactate 3, the -hydroxy esters used in this
study were readily prepared by conventional means
(Scheme 2).11,12
As part of an ongoing programme aimed at developing
new methodologies in the chemistry of the trifluorometh-
anesulfonyl (triflyl) group, we envisaged that a process
combining the advantages of both the Mitsunobu reaction
(one step) and the triflate method (exceptional leaving
group properties of OTf) would be of great benefit.
We chose the crystalline methyl mandelate 1a as our mod-
el substrate for reasons of ease of handling and availabili-
ty. Bearing in mind that the elimination of TfH from
triflamides (a class of compounds to which our desired
amino esters belong) is a possible reaction,10 NaH was
preferred as the base over the usual metallic bis-trimethyl-
silylamidures (MHMDS, M = K, Na, Li) or lithium diiso-
propylamidure (LDA) or lithium isopropylcyclo-
hexylamidure (LICA) because NaH does not generate a
weak basic organic residue such as the others do. Follow-
ing this logic, only a slight excess of NaH (1.2 equiv) was
used. With the expectation that it would correctly solvate
the anionic intermediates, THF was selected as the sol-
vent. All the exploratory reactions were achieved on a 100
We have devised a conceptually new amination reaction
whereby the same reagent would serve a dual function of
triflating and aminating an alcohol in a single operation.
By capitalizing on the routine use of bis-trifluoromethane-
sulfonimides (triflimides) PhNTf2, N-pyridine bis-trifluo-
romethanesulfonimide (PyNTf2) and m-chloro-N-
pyridine bis-trifluoromethanesulfonimide (m-Cl-PyNTf2)
Synlett 2002, No. 3, 04 03 2002. Article Identifier:
1437-2096,E;2002,0,03,0528,0530,ftx,en;G34201ST.pdf.
© Georg Thieme Verlag Stuttgart · New York
ISSN 0936-5214

LETTER
One-pot Transformation of Alcohols into Amines
529
( )n
CHO
CO2Me
OHC
lit. 12
lit. 11
R
R
R
OH
n = 1 R = 2-Br
1b
1c
1d
1e
n = 2 R = H
2a
2b
R = 2-I
R = 4-OMe
R = 4-CO2Me
R = 3-NO2
R = 3,4-di-OMe 2c
Scheme 2
mg scale. After exploring numerous sets of conditions, it along with the usual side-products. Apparently, electronic
was found that the isolated yield of amino ester 4a could donation of a para methoxy group was not sufficient to
not exceed 40% (Scheme 3).13 All preliminary attempts prevent the side-reactions. In the two former cases, the
featured a rapid consumption of starting material (less starting material was totally consumed and the imines 7b–
than 5 min, TLC monitoring). GC/MS and 1H NMR anal- c were only detected in trace amount. Scaling-up these
yses on the crude mixtures revealed a partial recovery of successful reactions to 1 g did provide 4b and 4c in yields
both 1a (5–50%) and triflimide PhNTf2, as well as mar- superior to those obtained from the test reactions. As ex-
ginal production of both triflamide PhTfNH and imine 7a pected, the diester 1e led to 4e also in poor yield, the imine
plus other unidentified by-products. As judged from the 7e plus other side-products being dominant (entry 4).
1H NMR spectra, 7a was present as a sole stereomer
Apparently, only those mandelates having the ortho posi-
whose stereochemistry has not been determined at this
tion congested proved synthetically useful in this reaction.
time. The temperature proved, at least to a certain extent,
to be an important factor, which permitted us to minimise
the formation of the by-products without affecting the re-
action rate (Scheme 3). Under optimal conditions (meth-
od A),13 80% of conversion was achieved and the imine
was isolated in 15% yield.
Table 1 Reaction of substituted Mandelates 1b–e using Method A
Entry
1b
-Hydroxy Ester
1b
Product (Yield %)a
4b (65)
2b
1c
1d
1e
1b
1c
4c (60)
From these results, it clearly appears that the amino ester
4a formed was continuously undergoing a basic-driven
elimination of TfH, thus leading to the imine 7a. The re-
covery of amounts of both 1a and PhTfNH in the medium
suggests that the related anionic forms (alcoholate and Ph-
TfN–) were mainly responsible for this undesired process.
As a result, we felt that introducing some differences in
the electronic and steric environment surrounding the
enolizable position in the benzylic series (type 1 sub-
strates) may have a deep impact on the reaction profile.
For instance, it was anticipated that substrates bearing an
electron-donating substituent on the aromatic nucleus
would decrease the rate of the deprotonation, and subse-
quently provide the amino esters in higher yields. Consid-
ering the steric demand of the bases mediating the
elimination, it was also anticipated that more crowded
substrates such as ortho substituted alcohols 1b and 1c
would also prevent the imine formation.
3b
4d (29)
4b
4e (30)
5c
4b (78)
6c
4c (64)
a Isolated yields are given.
b Reaction performed on a 100 mg scale.
c Reaction performed on a 1 g scale.
Satisfied that competitive elimination was not a signifi-
cant problem, we shifted our attention to type 2 substrates
as well as to the unsubstituted ethyl lactate 3 with the ex-
pectation that imine formation should be electronically
prevented. While 3 and methyl phenyl propionate 2a did
lead predominantly to the desired products 6 and 5a in
good yields (Table 2, entries 1, 2), the substituted ana-
logues 2b–c both failed to react as expected (entries 3, 4).
During the deprotonation step of the nitro substrate 2b, the
color rapidly turned brown, indicating a decomposition of
the alkoxide. This was supported by analyses on the crude
(GC/MS and 1H NMR) from which only PhNTf2 could be
detected. The same pattern featured on spectra of products
Only the second scenario was proved as shown from the
results summarized in Table 1. Under similar conditions
as described above, the amino esters 4b–c were indeed
produced in decent yields from the suitably substituted
substrates 1b–c (entries 1–3), whereas the electron-rich
alcohol 1d gave the expected amine 4d in only 29% yield
Ph
CO2Me Ph
CO2Me
Ph
CO2Me
1)NaH 1. 2 eq., THF, 0 °C, 15 min
2) PhNTf2 1.1 eq., temperature, 15 min
3) work-up
NPhTf
4a
N
OH
Ph
1a
7a
temperature
0 °C
-20 °C
-60 °C
35%
40%
42%
21%
15%
13%
Scheme 3
Synlett 2002, No. 3, 528–530 ISSN 0936-5214 © Thieme Stuttgart · New York

530
V. Gasparik et al.
LETTER
formed from the reaction of the electron-rich alcohol 2c
although in this case no significant visual change in the
medium was observed.
References
(1) Coppola, G. M.; Schuster, H. F. -Hydroxy Acids in
Enantioselective Synthesis; Wiley-VCH: Weinheim, 1997.
(2) (a) Mitsunobu, O. Synthesis 1981, 1. (b) Wada, M.; Sano,
T.; Mitsunobu, O. Bull. Chem. Soc. Jpn 1973, 46, 2833.
(c) Arnold, L. D.; Assil, H. I.; Vederas, J. C. J. Am. Chem.
Soc. 1989, 111, 3973.
Table 2 Reaction of Ethyl Lactate 3 and -Hydroxy Esters 2a–c
under Method A
Entry
-Hydroxy Ester
Product (Yield %)a
6 (75)
(3) (a) Effenberger, F.; Burkard, U.; Willfahrt, J. Angew. Chem.,
Int. Ed. Engl. 1983, 22, 65. (b) Feenstra, R. W.;
1
2
3
4
3
Stokkingreef, E. H. M.; Nivard, R. J. F.; Ottenheijm, H. C. J.
Tetrahedron Lett. 1987, 28, 1215. (c) Feenstra, R. W.;
Stokkingreef, E. H. M.; Nivard, R. J. F.; Ottenheijm, H. C. J.
Tetrahedron 1988, 44, 5583. (d) Degerbeck, F.; Fransson,
B.; Grehn, L.; Ragnarsson, U. J. Chem. Soc., Perkin Trans.
1 1992, 245.
2a
2b
2c
5a (71)
5b (0)
5c (0)
a Isolated yields are given.
(4) For a recent article wherein the drawbacks associated with
the Mitsunobu reaction were discussed, see: Kiankarimi, M.;
Lowe, R.; McCarthy, J. R.; Whitten, J. P. Tetrahedron Lett.
1999, 40, 4497.
An alternative method (method B) was then devised
whereby the alkoxide formed would be immediately
trapped by the triflimide. This could be made possible by
premixing the substrate and triflimide and then adding the
resulting THF solution to a suspension of NaH. When
method B was applied to 2b,c,14 5b,c could now be se-
cured in fair yields (Table 3, entries 1, 2). This procedure
could also be successfully scaled-up to 1 g (entries 3, 4).
(5) The stability of triflates of alcohols is marginal and closely
depends on the nature of the starting material. For some
related descriptions, see: (a) Vedejs, E.; Engler, D. A.;
Mullins, M. J. J. Org. Chem. 1977, 42, 3109. (b) Beard, C.
D.; Baum, K.; Grakauskas, V. J. Org. Chem. 1973, 38,
3673. (c) See also ref.3a–d
.
(6) For a recent publication on this topic, see: Blackburn, L.;
Taylor, R. J. K. Org. Lett. 2001, 3, 1637.
(7) Ritter, K. Synthesis 1993, 735.
(8) Commins, D. L.; Dehghani, A. Tetrahedron Lett. 1992, 33,
6299.
The latter results from substrates 2a–c and 3 are extremely
encouraging and predict the effectiveness of the reaction
when simpler, less functionalized alcohols will be ad-
dressed.
(9) For the use of triflamides RTfNH as nucleophiles in
Mitsunobu reactions, see: Edwards, M. L.; Stemerick, D. M.;
McCarthy, J. R. Tetrahedron Lett. 1990, 31, 3417.
(10) (a) Hendrickson, J. B.; Bergeron, R. J.; Sternbach, D. D.
Tetrahedron 1975, 31, 2517. (b) Bergeron, R. J.; Hoffman,
P. G. J. Org. Chem. 1979, 44, 1835. (c) Bergeron, R. J.;
Sternbach, D. D.; Bair, K. W. Acc. Chem. Res. 1977, 10, 307.
(11) Dust, T.; Lear, Y. Can. J. Chem. 1997, 75, 817.
(12) Dalla, V.; Cotelle, P.; Catteau, J. P. Tetrahedron Lett. 1997,
38, 1577.
(13) Typical Procedure (Method A): A tetrahydrofuran (1 mL)
solution of methyl mandelate (100 mg, 0.6 mmol) was added
under argon at 0 °C to a suspension of sodium hydride (29
mg, 60% in oil, 0.72 mmol) in tetrahydrofuran (1 mL). After
stirring 15 min, this alkoxide solution was added dropwise at
–20 °C to a tetrahydrofuran (1 mL) solution of triflimide
(236 mg, 0.66 mmol). After 15 min, the solution was
hydrolyzed with water (5 mL) and extracted with diethyl
ether (3 5 mL). Combined organic layer was dried over
magnesium sulfate and concentrated under vacuum to give
the desired amine which was purified from side-products by
flash-chromatography on a silica gel column (eluant:
cyclohexane–ethyl acetate (9–1 to 7–3).
(14) Typical Procedure for the Amination of 3-[3-
Nitrophenyl]-2-hydroxy-methyl Propanoate (Method B)
A tetrahydrofuran (1 mL) solution of -hydroxy ester (100
mg, 0.43 mmol) and triflimide (167 mg, 0.47 mmol) was
added dropwise under argon at –20 °C to a suspension of
NaH (20 mg, 60% in oil, 0.51 mmol) in tetrahydrofuran (1
mL). After 15 min the solution was hydrolyzed with water (5
mL) and extracted with diethyl ether (3 5 mL). Combined
organic layer was dried over magnesium sulfate and
concentrated under vacuum to give the desired amine which
was purified by flash-chromatography on a silica gel column
(eluant: cyclohexane–ethyl acetate (9–1 to 7–3).
Table 3 Reaction of -Hydroxy Esters 2b,c using Method B
Entry
1b
-Hydroxy Ester
2b
Product (Yield %)a
5b (63)
2b
2c
2b
2c
5a (75)
3c
5b (85)
4c
5c (69)
a Isolated yields are given.
b Reaction performed on a 100 mg scale.
c Reaction performed on a 1 g scale.
In summary, we have accomplished the amination of -
hydroxy esters using for the first time the concept of de-
rivatization and amination by a single reagent, employing
the commercially available PhNTf2 as a test reagent. Two
complementary procedures (methods A and B) have been
developed, the latter allowing conversion of substrates
with unstable alcoholates. Further work aimed at expand-
ing the synthetic potentiel of this reaction (base, substrate,
triflimide), as well as revealing its mechanistic picture
(SN2 vs. SN1) are being pursued in our laboratory and will
be reported in due course.
Acknowledgement
The authors wish to thank Dr Adam Daich (University of Le Havre)
and Drs J. C. Plaquevent and D. Cahard (IRCOF, University of
Rouen) for their interest in this work.
Synlett 2002, No. 3, 528–530 ISSN 0936-5214 © Thieme Stuttgart · New York