A novel [2,3] intramolecular rearrangement of N-benzyl-O-allylhydroxylamines

Article abstract of DOI:10.1039/a806200e

A novel [2,3]-sigmatropic rearrangement whereby N-benzyl-O-allylhydroxylamines undergo transformation to the corresponding N-allylhydroxylamines, which can subsequently be reduced to the corresponding allylamines, is described and evidence for the intramolecular nature of this process presented.

Full text of DOI:10.1039/a806200e

View Article Online / Journal Homepage / Table of Contents for this issue
A novel [2,3] intramolecular rearrangement of N-benzyl-O-allylhydroxylamines
Stephen G. Davies,*† Simon Jones, Miguel A. Sanz, Fátima C. Teixeira and John F. Fox
The Dyson Perrins Laboratory, University of Oxford, South Parks Road, Oxford, UK OX1 3QY
A novel [2,3]-sigmatropic rearrangement whereby N-benzyl-
O-allylhydroxylamines undergo transformation to the corre-
sponding N-allylhydroxylamines, which can subsequently be
reduced to the corresponding allylamines, is described and
evidence for the intramolecular nature of this process
presented.
afforded the [2,3] rearrangement product 4a in essentially
quantitative yield by examination of the crude 1H NMR
spectrum. Purification of this compound proved to be difficult
due to decomposition, although it was finally achieved by
chromatography on previously deactivated silica gel (1% Et3N).
The hydroxylamine thus obtained as an oil was isolated in 61%
yield and the structure confirmed by 1H and 13C NMR
spectroscopy and HRMS. With conditions for the rearrange-
ment identified,‡ it was repeated for the other substrates
(Scheme 2). All gave the [2,3] rearrangement product, although
in the case of the rearrangement to a trisubstituted centre (3c ?
Intramolecular sigmatropic rearrangements have found wide-
spread use in synthetic organic chemistry primarily due to the
high selectivities observed in these transformations. In partic-
ular, effective use has been made of a variety of [2,3] processes,
such as the Meisenheimer1 and Stevens rearrangements2 to
control stereochemistry as they tend to proceed at significantly
lower temperatures than [3,3] processes3 and thus lead to better
observed selectivities. For example, Anderson et al. have
recently developed an aza-[2,3]-Wittig rearrangement which
proceeds with excellent stereocontrol.4
1
4c), the reaction was found to be only 10% complete by H
NMR spectroscopy after 48 h. However, by heating the reaction
mixture to reflux for 2 h after the initial addition of n-BuLi at
278 °C, the reaction proceeded to completion and the [2,3]
product was obtained in 60% isolated yield. The hydroxyl-
amines formed in all of these reactions were all very prone to
decomposition during purification. Thus, isolated yields of the
allylhydroxylamines were always considerably lower than the
quantitative yields for the crude reaction observed by 1H NMR
spectroscopy. Attempted thermal rearrangement of 3a by
heating under reflux in xylene for 7 h led to only a trace amount
of the desired rearrangement product in the 1H NMR spectrum
( < 5%).
During continuation of our lithium amide studies5 we
discovered that N-benzyl-O-allylhydroxylamine gives N-allyl-
N-benzylhydroxylamine when treated with n-BuLi, a process
which may be attributed to a novel [2,3] sigmatropic rearrange-
ment analogous to the [2,3] Wittig rearrangement.6 We now
report our initial investigations into the nature of this rearrange-
ment and demonstrate its synthetic use.
The substrates for all of the rearrangements were easily
prepared in two steps and high yield starting from syn-
benzaldehyde oxime 1 (Scheme 1). O-Allylation was achieved
by formation of the potassium salt of the oxime and subsequent
quenching by treatment with the appropriate allyl bromide.7
The O-allyl oximes 2a–d were reduced with pyridine–borane
complex in EtOH–10% HCl to give the desired substrates 3a–d
in excellent overall yield.8 In most cases purification of the
intermediates and substrates was achieved by distillation.
The rearrangement was carried out by treatment of the simple
N-benzyl-O-allylhydroxylamine 3a in dry THF with 1 equiv. of
n-BuLi at 278 °C for 1 h, followed by warming to room
temperature for 30 min before quenching with water. This
The rearrangement of the crotylhydroxylamine 3b rules out
the possibility of a 1,2 anionic shift which would have given rise
to the N-benzyl-N-crotylhydroxylamine instead of the observed
product. However the possibility still existed that the reaction
was intermolecular and not intramolecular. In order to demon-
strate that the process was indeed intramolecular the rearrange-
ment was carried out with two different substrates 3b and 5 with
2 equiv. of n-BuLi (Scheme 3). If an intermolecular process was
being observed then the mixed products 4a and 7 from this
rearrangement would be observed. All four of the possible
products that could arise from this reaction were prepared in an
analogous manner to the hydroxylamines prepared earlier.
1
When the crude H NMR spectrum of the mixed reaction was
analysed, only peaks due to the respective intramolecular
rearrangement products 4b and 6 were present, and not the
crossover products 4a and 7 confirming that this reaction was
indeed an intramolecular process.
Based on these results the reaction is comparable to a
[2,3]-Wittig rearrangement, and thus reasonably proceeds via a
transition state which is similar to that suggested for this process
(Scheme 4).6 Deprotonation of the N–H proton affords the
OH
O
R1
t-BuOK, THF
Br
R1
N
N
R2
Ph
Ph
Compound
2a R1 = R2 = H
2b R1 = H, Me; R2 = Me, H
2c R1 = R2 = Me
2d R1 = Ph; R2 = H
R2
Yield
91%
86%
89%
95%
1
R1 R2
O
R1
Pyr-BH3
i) n-BuLi–THF
ii) H2O
OH
HN
10% HCl–EtOH
N
R2
O
R1
Ph
Ph
HN
R2
Yielda
61%
59%
60%
40%
Compound
Compound
4a R1 = R2 = H
4b R1 = H, Me; R2 = Me, H
4c R1 = R2 = M e
4d R1 = Ph; R2 = H
aRefers to isolated yields
Ph
Compound
3a R1 = R2 = H
3a R1 = R2 = H
Yield
94%
85%
80%
84%
3b R1 = H, Me; R2 = Me, H
3c R1 = R2 = Me
3d R1 = Ph; R2 = H
3b R1 = H, Me; R2 = Me,H
3c R1 = R2 = Me
3d R1 = Ph; R2 = H
Scheme 1
Scheme 2
Chem. Commun., 1998
2235

View Article Online
O
O
Me
which was prepared in a two step process by rearrangement of
the parent hydroxylamine 3a and subsequent reduction of the
crude material to afford the desired allylamine 9a in excellent
overall yield (91%).§ In a similar manner the crotyl- and
cinnamyl-hydroxylamines 3b and 3d were rearranged then
immediately reduced to give the allylic amines 9b and 9d in 93
and 92% overall yield respectively.
HN
HN
Ph
Ph
3b (87:13 E:Z)
5
i) 2eq. n-BuLi–THF
ii) H2O
In conclusion we report a novel [2,3] rearrangement of
N-benzyl-O-allylhydroxylamines to give the corresponding
N-allyl-N-benzylhydroxylamines, which may be reduced to the
N-benzylallylamines with Zn/HCl. This provides a novel
approach to the synthesis of allyl substituted hydroxylamines
and amines.
We would like to thank the Ministerio de Educación y
Ciência (M. A. S.) and J.N.I.C.T. (Programa Ciência and Praxis
XXI) (F. C. T.) for provision of funding and Oxford Asymmetry
for technical and financial support.
Intermolecular
Intramolecular
Me
Me
OH
OH
OH
OH
N
N
N
N
Ph
Ph
Ph
Ph
7
4a
4b
6
Scheme 3
Notes and References
–
R1 R2
R1
R2
O
R1
† E-mail: steve.davies@dpl.ox.ac.uk
O–
n-BuLi
HN
N
‡ Typical experimental procedure: n-BuLi (1.66 M, 0.81 mL, 1.35 mmol)
was added to a solution of 3a (200 mg, 1.23 mmol) in anhydrous THF (15
mL) at 278 °C under N2. After stirring for 1 h the reaction was allowed to
warm to room temp. and stirred for a further 30 min. The reaction was
quenched with distilled water (25 mL) and extracted with Et2O (3 3 25
mL). The combined organic extracts were dried (MgSO4) and concentrated
in vacuo. The residue was purified by column chromatography on
deactivated silica gel (1% Et3N–petroleum ether) eluting with petroleum
ether–Et2O (6:1) to give the desired hydroxylamine 4a as a clear yellow oil
(122 mg, 61%).
§ Typical proceedure for the reduction of the hydroxylamines: 4a (0.100 g,
0.61 mmol) was dissolved in 2 M HCl (5 mL) and zinc dust (0.200 g, 3.1
mmol) added cautiously. The reaction was heated at 80 °C for 1 h, cooled
and neutralised with 2 M NaOH. The white suspension was extracted with
Et2O (3 3 15 mL) and dried (MgSO4). Evaporation afforded 9a (0.081 g,
90%).
O
R2
N
Ph
Ph
Ph
4
3
8
Scheme 4
lithium amide which rearranges through an envelope transition
state 8. The driving force for this reaction lies in the relative
stability of the lithium oxy-anion 4 formed compared to the
lithium amide precursor 8, which is illustrated in the pKa values
of the corresponding alcohol and amine (pKa EtOH = 15.9, pKa
EtNH2 = 35). A similar [1,2] anionic process has been
previously reported in which the anion derived from N,O-
bis(trimethylsilyl)hydroxylamine rearranges to N,N-bis(tri-
methylsilyl)hydroxylamine.9
As already eluded to, the hydroxylamines produced were
relatively unstable and were thus difficult to isolate or store for
prolonged periods of time. However, the allylamines from
which the hydroxylamines are derived are stable and are also
useful synthons.10 Thus, reduction of the hydroxylamines 4a–d
with Zn/2 M HCl at 80 °C for 1 h followed by neutralisation
(NaOH) and extraction with Et2O gave the respective amines
9a–d in excellent yields (Scheme 5).
Alternatively, the crude product 4a obtained from rearrangement of 3a
was directly reduced by addition of 2 M HCl (3 mL) and zinc dust (0.200
g, 3.065 mmol) and heating at 80 °C. After 1 h the reaction was cooled,
neutralised with 2 M NaOH, extracted with Et2O (3 3 20 mL) and dried
(MgSO4). Evaporation afforded 9a (0.082 g, 91% overall yield).
1 S. G. Davies and G. D. Smyth, Tetrahedron: Asymmetry, 1996, 7,
1001.
2 P. A. Grieco, D. Boxler and K. Hirori, J. Org. Chem., 1973, 38,
2572.
3 R. W. Hoffmann, Angew. Chem., Int. Ed. Engl., 1979, 18, 563.
4 J. C. Anderson, D. C. Siddons, S. C. Smith and M. E. Swarbrick,
J. Chem. Soc., Chem. Commun., 1995, 1835.
5 S. G. Davies and O. Ichihara, Tetrahedron: Asymmetry, 1991, 2, 183.
6 K. Mikami and T. Nakai, Synthesis, 1991, 594.
7 E. Buehler, J. Org. Chem., 1967, 32, 261.
The versatility of this method for the synthesis of allylamines
was demonstrated in the synthesis of N-allyl-N-benzylamine 9a
R1 R2
N
R1 R2
NH
OH
Zn–2M HCl
80°C
8 M. Kawase and Y. Kikugawa, J. Chem. Soc., Perkin Trans. 1, 1979,
643.
9 R. West, P. Boudjouk and T. A. Matuszko, J. Am. Chem. Soc., 1969, 91,
5184.
10 For example, K. Burgess and M. J. Ohlmeyer, J. Org. Chem., 1991, 56,
1027; G. R. Cook and J. R. Stille, J. Org. Chem., 1991, 56, 5578.
Ph
Ph
Yield
90%
94%
92%
91%
Compound
4a R1 = R2 = H
Compound
9a R1 = R2 = H
4b R1 = H, Me; R2 = Me, H
4c R1 = R2 = Me
9b R1 = H, Me; R2 = Me, H
9c R1 = R2 = Me
4d R1 = Ph; R2 = H
9d R1 = Ph; R2 = H
Scheme 5
Received in Liverpool, UK, 5th August 1998; 8/06200E
2236
Chem. Commun., 1998