Mechanistic and stereochemical aspects of the Lewis acid mediated cleavage of α-aminoacetals

Article abstract of DOI:10.1016/S0040-4039(01)00292-1

The TMSOTf mediated nucleophilic cleavage of α-aminoacetals can be used to prepare a variety of substituted amines, with variable levels of stereocontrol depending on the substitution patterns. The reaction most likely proceeds via either an α-alkoxy aziridinium ion or an α-oxocarbenium ion depending on the type of nucleophile.

Full text of DOI:10.1016/S0040-4039(01)00292-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 2865–2868
Mechanistic and stereochemical aspects of the Lewis acid
mediated cleavage of a-aminoacetals
Mark A. Graham,a Alan H. Wadsworth,b Mark Thornton-Pett,a,† Benedetta Carrozzini,c
Giovanni L. Cascaranoc and Christopher M. Raynera,*
aSchool of Chemistry, University of Leeds, Leeds LS2 9JT, UK
bGlaxoWellcome Medicines Research Centre, Gunnels Wood Road, Stevenage, Herts. SG1 2NY, UK
cIRMEC, Dipartimento Geomineralogico, Campus Universitario, Via E. Orabona, 4 I-70125 Bari, Italy
Received 8 February 2001; revised 9 February 2001; accepted 20 February 2001
Abstract—The TMSOTf mediated nucleophilic cleavage of a-aminoacetals can be used to prepare a variety of substituted amines,
with variable levels of stereocontrol depending on the substitution patterns. The reaction most likely proceeds via either an
a-alkoxy aziridinium ion or an a-oxocarbenium ion depending on the type of nucleophile. © 2001 Elsevier Science Ltd. All rights
reserved.
The Lewis acid mediated nucleophilic cleavage of
acetals is a reaction of considerable mechanistic and
synthetic interest. In particular, the nature of the
reactive species involved is determined by an intricate
balance of electronic and steric effects, and reaction
conditions.1 Our interest in acetal chemistry stems
from its potential for stereocontrolled synthesis, as
demonstrated by our synthesis of the unusual but
very important nucleoside analogue Lamivudine,
which is currently used for the treatment of HIV and
hepatitis B.2 We also have an interest in the chem-
istry of aziridinium ions, and we, and other groups,
have shown them to be versatile, yet underused inter-
mediates in synthetic organic chemistry.3 In an exten-
sion of this work, and analogous to our recent
a-thiothiiranium ion chemistry,4 is the potential for
the formation of the little studied a-alkoxy aziri-
dinium ions5 such as 3, which would be expected to
be in equilibrium with open chain a-oxocarbenium
ions 2 and various intermediate forms, generated by
Lewis acid mediated cleavage of a-aminoacetals6
1
(Scheme 1).
We believed this work to be of mechanistic interest,
particularly if reaction pathways involving either 2 or
3 could be distinguished, but it was also important
synthetically due to the potential for the formation of
a variety of highly functionalised amines, which often
show interesting biological activity. We thus
embarked on a research programme to investigate the
reactivities of intermediates such as 2 and 3, and to
exploit them in the area of stereocontrolled synthesis.
Synthesis of the appropriate acetal precursors was
readily accomplished by reacting the required sec-
ondary amine (diallylamine, dibenzylamine) with bro-
moacetaldehyde diethyl acetal in DMF at 90°C in the
presence of catalytic potassium iodide (71–74% yield).
Substituents on nitrogen were chosen such that they
would allow subsequent deprotection to the corre-
R2
R2
N
R2 OR1
N
R2 Nu
Nu
Lewis acid
R2
N+
N
R2
OR1
OR1
OR1
R2
R2
OR1
+
1
4
2
3
Scheme 1.
* Corresponding author. E-mail: c.m.rayner@chem.leeds.ac.uk
† Author to whom communications regarding X-ray crystallography should be addressed.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00292-1

2866
M. A. Graham et al. / Tetrahedron Letters 42 (2001) 2865–2868
Table 1. TMSOTf mediated nucleophilic cleavage of a-
Treatment of the substrates with trimethylsilyl trifl-
uoromethanesulfonate (TMSOTf) at −78°C followed by
addition of the nucleophile and slowly warming to
room temperature gave good yields of the acetal substi-
tution products 6 (Table 1).
aminoacetals
i) TMSOTf, CH2Cl2
-78 °C
ii) Nucleophile,
-78 °C-rt
R
N
Nu
OEt
R
N
OEt
OEt
R
R
5
6
Reaction with organometallic species such as dialkyl
zinc and Grignard reagents proceeded efficiently, for
both the N-allyl and N-benzyl substrates. Similarly,
introduction of the uracil group and 2-pyridone was
also efficient. Enamines reacted smoothly to introduce
carbonyl functionality, but with silyl enol ethers, the
Lewis acidic reaction conditions led to further loss of
ethanol to give the a,b-unsaturated ketone as the major
product.
Entry
Nucleophile
Et2Zn
Product
Yield (%)
R
1
2
allyl
Bn
72
86
R
Et2Zn
N
R
OEt
Bn
EtMgBr
3
71
O
N
OTMS
N
NH
4
5
allyl
Bn
75
77
R
N
O
N
OTMS
These encouraging preliminary results led us to con-
sider investigating more complex systems, which had
greater potential for stereocontrolled synthesis, but
would also provide more revealing mechanistic data to
help determine the dominant reactive species. The sub-
strates were synthesised from the corresponding opti-
cally active amino alcohols by N-dibenzylation (BnBr,
KI, DMF, 80°C), Swern oxidation, and acetal forma-
tion (EtOH, (EtO)3CH, H2SO4). They were then sub-
jected to the same reaction conditions as before. The
results of these reactions are shown in Table 2.
R
R
OEt
6
7
allyl
Bn
64
84
R
N
N
O
N
OTMS
OEt
8
9
86a
N
O
Bn
Bn
R
N
O
13a,b
OTMS
R
OEt
a78:22 mixture of isomers; bMajor product (57%) is enone from further loss of EtOH.
Table 2 shows a number of interesting trends. Firstly,
the use of diethyl zinc as nucleophile gives high levels of
diastereoselectivity, irrespective of size of the sub-
stituent R. Secondly, when R is large (Bn, Pr) high
levels of selectivity are obtained, irrespective of the
nucleophile. Thirdly, low selectivity is observed with the
sponding primary amine if required.7 We then investi-
gated the Lewis acid mediated nucleophilic cleavage of
these simple acetals.
i
Table 2. TMSOTf mediated nucleophilic cleavage of chiral a-aminoacetals
i) TMSOTf, CH2Cl2
-78 °C
ii) Nucleophile,
-78 °C-rt
Bn OEt
N
Bn Nu
N
Bn
OEt
Bn
OEt
R
R
7
8
Entry
1
Nucleophile
Et2Zn
Major product
diast. ratio Yield (%)
R
95:5a
93:7
94:6
64
61
77
Bn
iPr
Bn
N
Et2Zn
Et2Zn
2
3
Bn
OEt
Me
R
O
N
Bn
4
5
6
95:5b
89:11
55:45
70
78
71
OTMS
N
NH
iPr
Bn
N
O
Me
N
OTMS
Bn
Bn
OEt
R
7
8
Bn
95:5
89
87
Bn
N
N
O
Me
54:46
N
OTMS
OEt
R
aSterechemical assignment confirmed by comparison with literature;8 bStereochemical assignment
confirmed by X-ray crystallography, vide infra.9

M. A. Graham et al. / Tetrahedron Letters 42 (2001) 2865–2868
2867
Figure 1. X-Ray crystal structure of 9.
TfO-
Et
Zn Et
Bn OEt
N
Bn Et
N
TMSOTf
Et2Zn
Bn
Bn
Bn
OEt
Bn
OEt
N
OEt
+
R
R
R
7
10
11
Scheme 2.
1
nitrogen heterocycles with a small R group in the
substrate (Me). Finally, the sense of stereoselection is
reversed when switching from diethyl zinc to the nitro-
gen heterocycles. These intriguing observations allow
considerable insight into the nature of the reactive
intermediates involved.
between H NMR chemical shifts in the syn and anti
diastereomeric series.10
Let us first consider the high levels of stereocontrol
observed when using diethyl zinc as the nucleophile. It
is likely that the reaction proceeds via pre-coordination
of the diethyl zinc to the tertiary amine group within
the substrate (e.g. 10), as is usually required to activate
Of crucial importance to our arguments, are the stereo-
chemical assignments for the major products of the
reaction. In the case of the diethyl zinc additions, this
was assigned by comparison with the literature. The
addition of diethyl zinc to the analogous aldehyde gives
the syn diastereomer as the major product.8 Subsequent
O-ethylation shows this product to be the minor isomer
of our reactions, and hence confirms our anti assign-
ment (entries 1–3). The stereochemistry of the uracil
derivative 9 (Table 2, entry 4) was determined by X-ray
crystallography (Fig. 1), which clearly shows the syn
stereochemistry.9 There is also a very good correlation
such
organometallic
reagents.8,11
Subsequent
intramolecular delivery of the ethyl group exaggerates
the effect of the R-substituent, leading to high levels of
anti selectivity in each case (Scheme 2). Note that
because the nitrogen atom is now coordinated to the
zinc, it is no longer able to stabilise the adjacent
cationic centre, and so it would not be expected to be
able to form an a-alkoxy aziridinium ion intermediate.
The addition of nitrogen heterocycles is more complex.
Although variable levels of selectivity observed may be
TfO-
Bn
Bn Nu
N
Bn N+
R
Nu
Bn
OEt
15
OEt
R
14
Bn
N
Bn OEt
N
TMSOTf
Bn
OEt
Bn
OEt
+
13
R
R
7
Bn
Bn Nu
N
Bn N+
R
Nu
OEt
Bn
OEt
16
12
R
Scheme 3.

2868
M. A. Graham et al. / Tetrahedron Letters 42 (2001) 2865–2868
expected and are consistent with the size of the sub-
stituents R, the predominant syn stereoselectivity is
unlikely to originate by addition to the a-oxocarbenium
ion 2. Instead, the intermediacy of an a-alkoxy aziri-
dinium ion can be invoked to explain the observations
(Scheme 3). There are two possible isomeric a-alkoxy
aziridinium intermediates we can form, the cis 12 and
trans 14 isomers, which may be in equilibrium via a
small concentration of the a-oxocarbenium ion. Exten-
sive broadening observed during variable temperature
NMR experiments on intermediates derived from 7
(R=Bn) suggest this, but other equilibrium processes
such as acetal exchange cannot be discounted at this
time.10
References
1. (a) Johnson, W. S.; Kelson, A. B.; Elliot, J. D. Tetrahedron
Lett. 1988, 29, 3757; (b) Denmark, S. G.; Almstead, N. G.
J. Am. Chem. Soc. 1991, 113, 8089. For more recent work
in this area, see: (c) Bull, S. D.; Correia, L. A. M. R. B.;
Davies, S. G. J. Chem. Soc., Perkin Trans. 1 1998, 2231.
2. Milton, J.; Brand, S.; Jones, M. F.; Rayner, C. M.
Tetrahedron Lett. 1995, 36, 6961.
3. (a) Chuang, T. H.; Sharpless, K. B. Org. Lett. 1999, 1, 1435;
(b) Rayner, C. M. Synlett 1997, 11.
4. Fletcher, S. J.; Brand, S.; Jones, M. F.; Rayner, C. M.,
manuscript in preparation.
5. There appears to be only one previous literature report of
an a-alkoxy aziridinium ion according to a chemical
abstracts search although few details are given: Zarudii, F.
S.; Lazareva, D. N.; Kurmaeva, E. S.; Chalova, O. B.;
Kiladze, T. K.; Kantor, E. A.; Rakhmankulov, D. L.
Khim.-Farm. Zh. 1985, 19, 161; Chem. Abstr. 1985, 103,
71130.
6. For examples of other chemistry based on a-amino alde-
hydes and acetals see: Jurczak, J.; Golebiowski, A. Chem.
Rev. 1989, 89, 149; Reetz, M. T.; Drewes, M. W.; Schmitz,
A. Angew. Chem., Int. Ed. Engl. 1987, 26, 1141; Kano, S.;
Yokomatsu, T.; Iwasawa, H.; Shibuya, S. Chem. Lett.
1987, 1531.
Stereocontrolled SN2 nucleophilic ring opening of the
trans isomer 14 would give the observed syn product,
whereas the cis isomer 12 would give the anti product.
i
For large substituents (R=Bn, Pr), the trans aziri-
dinium ion 14 would be expected to be significantly
more stable than the cis 12 due to steric effects, hence
the high syn selectivity observed with these substituents.
The fact that little or no anti product is formed suggests
that direct addition to the a-oxocarbenium ion 13 does
not occur, and this also provides additional evidence
for the importance of the coordinated diethyl zinc
intermediate in the previous case. With small sub-
stituents (R=Me) there would be much less difference
in energy between the cis and trans isomeric aziridinium
ions, and hence both would be present in similar con-
centrations leading to the mixture of products
observed.
7. (a) Kocienski, P. J. Protecting Groups; Thieme, 1994; (b)
Jacobi, P. A.; Martinelli, M. J.; Polanc, S. J. Am. Chem.
Soc. 1984, 106, 5594.
8. Andres, J. M.; Barrio, R.; Martinez, M. A.; Pedrosa, R.;
Perez-Encabo, A. J. Org. Chem. 1996, 61, 4210.
9. X-Ray crystallography showed two different conforma-
tions in the unit cell, only one of which is shown. We are
very grateful to Professor Carrozzini and colleagues in Bari
for providing the solution to this structure using the
program SIR2000: Burla M. C.; Camalli, M.; Carrozzini,
B.; Cascarano, G.; Giacovazzo, C.; Polidori, G.; Spagna,
R. Acta Crystallogr., Sect. A 2000, 56, 451; Burla, M. C.;
Camalli, M.; Carrozzini, B.; Cascarano, G.; Giacovazzo,
C.; Polidori, G.; Spagna, R. J. Appl. Cryst. 2001, submitted
for publication. Crystallographic data (excluding structure
factors) have been deposited with the Cambridge Crystal-
lographic Data Centre as supplementary publication num-
ber CCDC 153574.
In summary, we have shown that the TMSOTf medi-
ated nucleophilic cleavage of a-aminoacetals, can be
used to introduce efficiently a variety of groups, with
variable levels of stereocontrol depending on the substi-
tution patterns. The reaction most likely proceeds via a
novel a-alkoxy aziridinium ion, but an a-oxocarbenium
ion is a likely intermediate in the presence of nucle-
ophiles capable of coordinating to the tertiary amine
group.
Acknowledgements
10. Graham, M. A.; Wadsworth, A. H.; Rayner, C. M.,
unpublished results.
11. Noyori, R.; Kitamura, M. Angew. Chem., Int. Ed. Engl.
1991, 30, 49.
We wish to thank the EPSRC and GlaxoWellcome
(Stevenage) for an Industrial CASE award.
.
.