Formation and heron reactivity of cyclic n,n-dialkoxyamides

Article abstract of DOI:10.1071/CH13557

Cyclic N,N-dialkoxyamides have been made, for the first time, by hypervalent iodine oxidation of β- and γ-hydroxyhydroxamic esters 17, 19, and 21. The fused γ-lactam products, N-butoxy- and N-benzyloxybenzisoxazolones (22a and 22b), are stable while alicyclic γ-lactam and d-lactam products, 24 and 25, although observable by NMR spectroscopy and ESI-MS are unstable at room temperature, undergoing HERON reactions. The γ-lactam 24 undergoes exclusive ring opening to give a butyl ester-functionalised alkoxynitrene 28. The δ-lactam 25, instead, undergoes a HERON ring contraction to give butyrolactone (27). The structures of model γ- and δ-lactams 6, 7, and 8 have been determined at the B3LYP/6-31G(d) level of theory and the γ-lactams are much more twisted than the acyclic N,Ndimethoxyacetamide (5) resulting in a computed amidicity for 6 of only 25% that of N,N-dimethylacetamide (3). The HERON reactions of N,N-dimethoxyacetamide (5) and alicyclic models 6 and 8 have been modelled computationally. The facile ring opening of 6 (EA=113 kJ mol ^-1) and ring contraction of 8 (EA=145 kJ mol^-1) are predicted well, when compared with the HERON rearrangement of 5 (EA=178 kJ mol^-1). CSIRO 2014.

Full text of DOI:10.1071/CH13557

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem. 2014, 67, 507–520
Full Paper
http://dx.doi.org/10.1071/CH13557
Formation and HERON Reactivity of Cyclic
N,N-Dialkoxyamides
A B
,
A
A
Stephen A. Glover,
Adam A. Rosser, Avat (Arman) Taherpour,
A
and Ben W. Greatrex
A
Department of Chemistry, School of Science and Technology,
University of New England, Armidale, NSW 2351, Australia.
Corresponding author. Email: sglover@une.edu.au
B
Cyclic N,N-dialkoxyamides have been made, for the first time, by hypervalent iodine oxidation of b- and
g-hydroxyhydroxamic esters 17, 19, and 21. The fused g-lactam products, N-butoxy- and N-benzyloxybenzisoxazolones
(22a and 22b), are stable while alicyclic g-lactam and d-lactam products, 24 and 25, although observable by NMR
spectroscopy and ESI-MS are unstable at room temperature, undergoing HERON reactions. The g-lactam 24 undergoes
exclusive ring opening to give a butyl ester-functionalised alkoxynitrene 28. The d-lactam 25, instead, undergoes a
HERON ring contraction to give butyrolactone (27). The structures of model g- and d-lactams 6, 7, and 8 have been
determined at the B3LYP/6-31G(d) level of theory and the g-lactams are much more twisted than the acyclic N,N-
dimethoxyacetamide (5) resulting in a computed amidicity for 6 of only 25 % that of N,N-dimethylacetamide (3). The
HERON reactions of N,N-dimethoxyacetamide (5) and alicyclic models 6 and 8 have been modelled computationally. The
facile ring opening of 6 (EA ¼ 113 kJ molꢀ1) and ring contraction of 8 (EA ¼ 145 kJ molꢀ1) are predicted well, when
compared with the HERON rearrangement of 5 (EA ¼ 178 kJ molꢀ1).
Manuscript received: 15 October 2013.
Manuscript accepted: 27 November 2013.
Published online: 30 January 2014.
Crystallographic studies[2] confirm that N,N-dialkoxyamides
possess highly pyramidalised nitrogens (average angles at nitro-
˚
gen of 110.48 and 111.58) and long N–C bonds (1.42 and 1.41A).
Introduction
Recently we reported an efficient synthesis, the structure, and
the thermal reactivity of an unusual class of amide derivatives,
N,N-dialkoxyamides 1 (Chart 1).[1,2] These are the least studied
members of the class of anomeric amides 2 (Chart 1), amides
bearing two heteroatoms at the amide nitrogen.[1–26] In general,
bisheteroatom-substituted amides are characterised by a
reduced amide resonance on account of the electron demand of
the electronegative atoms on nitrogen. The amide nitrogen lone
pair, which normally resides in a 2pz orbital on an sp2 hybridised
nitrogen and strongly interacts with the corresponding orbital on
the carbonyl carbon, gains ‘s’ character and not only do the
nitrogens tend towards pyramidality, but the lone pair is loca-
lised on nitrogen with attendant loss of amide resonance.[6,16,17]
Bisalkoxyl substitution results in the largest loss of resonance of
all the classes of anomeric amides we have studied to date.[2]
In addition a reduced amide resonance is evident from dynamic
1H NMR studies in which the amide rotational barrier is too
low to measure and infrared data reveal high carbonyl vibra-
tional frequencies (between 1710–1715 cmꢀ1).[1,2,6,17] As with
almost all anomeric amides, the nitrogen lone pair, while
residing in an sp3 hybrid orbital, is still largely aligned with the
amide carbon 2pz orbital as crystal structures demonstrate that
the twist angles are small (148 and 78).[2]
These properties of N,N-dialkoxyamides are well predicted
computationally at the B3LYP/6-31G(d) level of theory.[2,5]
Furthermore, we have recently devised a quantitative measure
of amide character for a wide range of amides and lactams.[2,27]
Relative to the iconic N,N-dimethylacetamide (3), which we and
others[28] have designated as having 100 % amidicity, and the
completely twisted amide 1-aza-2-adamantanone (4), with by
definition 0 % amidicity, N,N-dimethoxyacetamide (5) is com-
puted to have an amidicity of just 46 % and to have lost more
than half the resonance embodied in N,N-dimethylacetamide.[2]
It is interesting that, while amide character is diminished by
reduced overlap as a consequence of the change in hybridisation
at nitrogen, this does not account for all of the decrease in
amidicity. Loss of resonance energy in N,N-dimethylacetamide
as a result of complete pyramidalisation at nitrogen is estimated
at ,25 kJ molꢀ1 or about one-third that of the planar struc-
ture.[27] The additional loss of resonance in N,N-dialkoxyamides
O
OMe
OMe
O
O
N
O
R
Y
N
X
O
Me
N
Me
N
N
R
Me
Me
OR
OR
1
2
3
4
5
X ꢀ Cl, Y ꢀ OR
X ꢀ Oacyl, Y ꢀ OR
X ꢀ Y ꢀ OR
X ꢀ NR2, Y ꢀ OR
X ꢀ SR, Y ꢀ OR
X ꢀ Y ꢀ Halogen
Chart 1.
Journal compilation Ó CSIRO 2014
www.publish.csiro.au/journals/ajc

508
S. A. Glover et al.
is a result of further localisation of the nitrogen lone pair on
nitrogen as a consequence of the inductive effect of both oxygen
atoms.[2]
destabilisation of an N–O bond and the geminal oxygen atom
lone pairs are too low in energy in this case.
In this paper we describe the first synthesis of cyclic
N,N-dialkoxyamides which, in certain cases, undergo thermal
HERON reactions. These reactions are described along with the
theoretical differences between acyclic and cyclic forms of these
anomeric amides.
Thermal reactivity of N,N-dialkoxyamides involves homol-
ysis to generate alkoxyl and N-alkoxyamidyl radicals.[1] Certain
bisheteroatom-substituted amides have been found to undergo
the unusual HERON (from Heteroatom rearrangements on
nitrogen) reactiony (Fig. 1).[14,15,29] N-Alkoxy-N-aminoamides
generate esters and aminonitrenes in a concerted process involv-
ing anomerically driven migration of the alkoxyl group from
nitrogen to the carbonyl.[4,8,9,14] The high energy nitrogen lone
pair strongly destabilises the adjacent N–O bond and
these reactions occur at room temperature. N-Acyloxy-N-
alkoxyamides also undergo HERON reactions yielding anhy-
drides and alkoxynitrenes but these reactions occur at higher
temperatures and in competition with homolytic reactions.[18]
The HERON in this case is driven by the weak N–O-acyl bond.
N,N-Dialkoxyamides studied to date do not undergo HERON
reactions.[1] Migration of an alkoxyl group requires anomeric
Results and Discussion
Several cyclic forms of N,N-dialkoxyamides 6–8 (Chart 2),
together with N,N-dimethoxyacetamide (5)[2] have been studied
computationally and structures of the B3LYP/6-31G(d)-
optimised geometries for 6, 7, and 8, together with that of 5
are presented in Table 1. The Winkler–Dunitz amide twist
parameter (t) and nitrogen pyramidality index (x) are defined
in Fig. 2.[30,31]
While they possess broadly similar properties to acyclic N,N-
dimethoxyacetamide (5) in that all lactam nitrogens are strongly
pyramidal, the extent of twist in the g-lactam forms is signifi-
cantly greater than in the acyclic analogue. The twist angle (t) for
model N-methoxy g-and d-lactams 6, 7, and 8 are respectively
O
O
ꢁ
ꢂ
HERON
ꢂ
R
N
N ꢀ Y
N Y
Y
b
R
X
(a)
(b)
b
a
τ
X
X ꢀ OR, Y ꢀ NR2
X ꢀ OAcyl, Y ꢀ OR
X ꢀ OR, Y ꢀ Oꢁ
X ꢀ OR, Y ꢀ N
a
O
R
X
N
O
C
R
N
Fig. 1. The generalised HERON reaction of anomeric amides.
Y
Y
O
O
O
N
X
χ ꢀ 60ꢃ: aNY ꢀ aNX ꢀ 120ꢃ
χ ꢀ 0ꢃ: aNY ꢀ aNX ꢀ 90ꢃ
OMe
N
OMe
N OMe
O
O
O
6
7
8
Fig. 2. Interpretation of the Winkler–Dunitz twist and pyramidality indi-
ces, t and x.[30,31] (a) Orbital axes in an anomeric amide and (b) angular
projections along the N–C bond.
Chart 2.
Table 1. B3LYP/6-31G(d) optimised geometries for N,N-dimethoxyacetamide (5), g-lactams 6 and 7, and d-lactam 8
N
Oendo
N
N
N
Osyn
Oanti
Oendo
Oexo
Oexo
Oendo
Oexo
5
6
7
8
tA [deg.]
xB [deg.]
yC [deg.]
˚
N–C [A]
˚
C–O [A]
˚
N–Oexo [A]
˚
N–Oendo [A]
Compound
5D
6
8.5
ꢀ36.7
ꢀ17.9
12.6
48.1
64.6
61.2
60.7
114.3
106.8
109.3
110.3
1.415
1.465
1.461
1.427
1.214
1.200
1.210
1.209
1.388(syn)
1.428
1.412(anti)
1.419
7
1.373
1.477
8
1.379
1.46
AWinkler–Dunitz twist parameter (Fig. 2).[30,31]
BWinkler–Dunitz pyramidality index (Fig. 2).[30,31]
CAverage angle at nitrogen.
DPreviously published data.[2]
yFirst presented to the 2nd Heron Island Conference on Reactive Intermediates and Unusual Molecules, Heron Island, Australia, 1994.

Properties of Cyclic N,N-Dialkoxyamides
509
O
OMe
ΔEreact
OMe
378, 188 and 128 as compared with 88 for 5. We recently reported
that amidicities of amides can be estimated computationally by
two independent isodesmic reactions, carbonyl substitution nitro-
gen atom replacement (COSNAR) and a trans-amidation
process.[27] COSNAR, as applied to lactams 6 and 8, measures
the stabilisation derived from the isodesmic reaction in Eqn 1 and
is a direct measure of resonance stabilisation.[32–34] The trans-
amidation reaction in Eqn 2 measures the destabilisation in the
lactams relative to N,N-dimethylacetamide (3).[27] Total loss
of resonance for fully twisted lactam 1-aza-2-adamantanone
O
Me
Me
Me
N
N
N
N
ꢂ
ꢂ
(CH2)n
(CH2)n
O
O
Me
Me
Me
3
10
11
13
ð2Þ
O
OMe
OMe
ΔEsteric
O
OMe
OMe
ꢂ
(CH2)n
(CH2)n
ꢂ
O
O
Me
OMe
15
Me
OMe
(4) equates with a loss of 76.0 kJ molꢀ1 [27]
thus for lactams
,
14
12
9
6 and 8 the residual resonance stabilisation is the negative of the
difference between this value and that determined from Eqn 2.
By this method, a further destabilisation of the carbonyl owing
to the inductive effect of two b-oxygens (DEinductive, estimated
ð3Þ
Residual resonance ðkJ molꢀ1
Þ
ð4Þ
¼ ꢀ76:01 þ ðDEreact ꢀ DEsteric ꢀ DEinductive
Þ
previously at 17.88 kJ molꢀ1 [2]
and specific steric effects,
)
(DEsteric) must be offset to give the residual resonance stabilisa-
tion from Eqn 4.[27] For N,N-dimethylacetamide (3) the
COSNAR and trans-amidation resonance energies are respec-
tively ꢀ77.53 and ꢀ76.01 kJ molꢀ1 and amidicities for 6 and 8
by each method are determined as a percentage of these values.
The steric correction for each lactam is estimated from the
isodesmic equation in Eqn 3. From B3LYP/6-31G(d) data in
Table 2 the amidicity for g-lactam 6 is less than half that of
N,N-dimethoxyacetamide (5) by both the COSNAR and trans-
amidation methods while that for 8 is similar to that of N,N-
dimethoxyacetamide (5). As has been demonstrated for a range
of amides and lactams, amidicities by the COSNAR and trans-
amidation methods are similar, as found for both lactam forms.
The reduction in amidicity in the g-lactam 6 can be largely
attributed to the significant degree of twist about the N–C bond
in this case (t ¼ 378, Table 1). Fig. 3 illustrates the B3LYP/6-
31G(d) energy surface generated by deformation of
N,N-dimethoxyacetamide (5). It differs markedly from that
computed for N,N-dimethylacetamide (3) where the lowest
energy form is planar (x ¼ t ¼ 08).[27] The lowest energy form
for N,N-dimethoxyacetamide has a Winkler–Dunitz pyramidal-
ity index of x ¼ 488 and twist index of t ¼ 88 and a twist of ,408
about the N–C axis in the pyramidal structure would further
reduce resonance by an estimated 20 kJ molꢀ1. In the case of 6,
˚
the twist manifests in a very long N–C bond of 1.465 A for 6 as
˚
opposed to 1.415 A for N,N-dimethoxyacetamide (5) (Table 1).
O
O
Not surprisingly, the d-lactam amidicity is very similar to
)
OMe
OMe
OMe
OMe
that of the acyclic form N,N-dimethoxyacetamide (47 %[2]
since the amide moiety in this lactam exhibits a very similar
twist and pyramidality to the acyclic form. The N–C bond
ΔECOSNAR
N
N
ꢂ
(CH2)n
(CH2)n
(CH2)n
12
(CH2)n
O
O
O
O
9
10
11
˚
lengths in d-lactam (1.427 A) and N,N-dimethoxyacetamide
˚
(1.415 A) are similar.
ð1Þ
Table 2. B3LYP/6 31G(d) energies for lactams 6 and 8 and relevant COSNAR and trans-amidation structures,
]
reaction energies from Eqns 1–4, and amidicities
Structure
Energy [au]
Energy [kJ molꢀ1
]
Amidicities [%]
N-Methoxy-g-lactam 6 [11 (n ¼ 1)]
2-Methoxy-3-tetrahydrofuranone (9) (n ¼ 1)
N-Methoxyisoxazolidine (10) (n ¼ 1)
2-Methoxytetrahydrofuran (12) (n ¼ 1)
N,N-Dimethylacetamide (3)
N,N-Dimethylethylamine (13)
1,1-Dimethoxypropanone (14)
1-Methoxy-1-propyl methyl ether (15)
N-Methoxy-d-lactam 8 [11 (n ¼ 2)]
2-Methoxy-4,5-dihydro-(2H)pyran-3-one (9) (n ¼ 2)
N-Methoxy-3,4,5,6-tetrahydrooxazine (10) (n ¼ 2)
2-Methoxy-3,4,5,6-tetrahydropyran (12) (n ¼ 2)
Eqn 1 (n ¼ 1)
ꢀ436.939062
ꢀ420.984956
ꢀ362.928153
ꢀ346.981671
ꢀ287.830189
ꢀ213.788638
ꢀ422.190664
ꢀ348.185186
ꢀ476.254615A
ꢀ460.29646A
ꢀ402.241932A
ꢀ386.298084A
ꢀ0.007624
ꢀ20.01
80.44
25.8
Eqn 2 (n ¼ 1)
0.030642
Eqn 3 (n ¼ 1)
0.002193
5.76
Eqn 4 (n ¼ 1)
ꢀ19.22
ꢀ37.56
75.78
25.3
49.4
Eqn 1 (n ¼ 2)
ꢀ0.014307
0.028868
0.007102
Eqn 2 (n ¼ 2)
Eqn 3 (n ¼ 2)
18.65
Eqn 4 (n ¼ 2)
ꢀ36.76
ꢀ35.87
48.4
47.2
Eqn 4 N,N-dimethoxyacetamide (5)B
AFor comparison with d-lactam 8, all structures bear an equatorial methoxy group.
BPreviously published data.[2]

510
S. A. Glover et al.
O
OCH3
OCH3
N
H3C
(d)
O
OCH3
140
120
N
H3C
OCH3
(c)
100
80
60
40
20
90
80
70
60
0
50
60
40
τ [deg.]
50
30
40
30
20
20
O
10
10
0
χ
[deg.]
N
H3C
O
OCH3
OCH3
OCH3
N
OCH3
H3C
(b)
(a)
Fig. 3. B3LYP/6-31G(d) energy surface for deformation of N,N-dimethoxyacetamide (5). x and t are the Winkler–Dunitz pyramidalisation
and twist parameters in degrees. Structures correspond to (a) the planar untwisted form with an sp2 hybridised nitrogen; (b) an untwisted
conformer with an sp3 hybridised nitrogen; (c) a fully twisted form with sp2 hybridised nitrogen; and (d) a fully twisted form with an sp3
hybridised nitrogen.
O
O
(b)
OR
OR
N
H
N
O
PhI(OCOCF3)2
Oexo
Oendo
OH
Oexo
Oendo
Scheme 1.
From projections illustrated in Fig. 4a and b, it is clear that in
the g-lactam form, there is a mutual nO–s*NO anomeric overlap.
Both p-type oxygen lone pairs are nicely aligned with the
adjacent N–O s-bond and the exocyclic and endocyclic NO
6
8
(a)
˚
bonds differ in length by only 0.01 A. However, in the d-lactam,
one anomeric alignment is dominant; that between a p-type lone
pair on the exocyclic oxygen and the endocyclic N–O bond. The
p-type lone pair on the endocyclic oxygen is actually orthogonal
to the exocyclic N–O bond. Consistent with this, the exocyclic
Oendo
Oendo
˚
N–O bond is 0.08 A shorter than the endocyclic N–O bond.
Oexo
Oexo
Since N,N-dialkoxyamides are readily synthesised by phe-
nyliodine bis(trifluoroacetate) (PIFA) oxidation of hydroxamic
esters in the presence of alcohols,[1] attempted generation of
cyclic systems employed oxidative intramolecular cyclisation
of b- and g-hydroxyhydroxamic esters according to Scheme 1.
6
8
Fig. 4. Optimised geometries for model g-lactams 6 and 8 with projections
along the (a) Oendo–N and (b) Oexo–N bonds.

Properties of Cyclic N,N-Dialkoxyamides
511
O
O
O
O
O
BuO2C
H3C
CH3
O
OBu
OH
O
1. KONH2
2. BuX
N
N
N
O
OBu
N
NHOBu
OH
17
O
O
CO2Bu
OEt
16
24
25
26
27
O
O
1. KONH2
2. RX
Chart 3.
OEt
OH
18
NHOR
OH
19a R ꢀ Bu
19b R ꢀ Bn
dd, J 17.1 and 9.7) (Fig. 5a, Fig. 6a). The 5-methyl group and the
10-oxymethylene on the butyl chain appeared as a new doublet at
d1.41 and a new triplet at d4.00 respectively. g-Lactam 24
decomposed rapidly and completely upon standing at room
temperature or in CDCl3 giving a mixture of diastereomers of
26, each of which exhibited an ABX spin system centred on
d5.5–5.6 (overlapping 3-HX), d2.9ꢀ3.1 (overlapping 2-HA, dd,
JAB 17.1 and 16.6 and JAX 9.3 and 8) and d2.5ꢀ2.6 (overlapping
2-HB, dd, JBA 17.1 and 16.6 and JBX 4.7 and 6.3) (Fig. 5b,
Fig. 6b). The 10-oxymethylenes of both diastereomers over-
lapped at d4.0–4.13 and the 4-methyl protons shifted to lower
frequency and appeared as overlapping doublets at d1.35.
Analysis of the reaction mixture by electrospray ionisation–
mass spectrometry (ESI-MS) indicated two prominent ions at
m/z 347.2 and 715.5 corresponding exactly to [M þ H]þ and
[2M þ Na]þ for the hyponitrite 26. The methyne proton shift to
higher frequency in 26 is similar to that found for the methylene
of 4-chlorobenzylhyponitrite, which resonated ,0.8 ppm higher
than that in 4-chlorobenzyl alcohol.[18] A typical ester carbonyl
carbon was evident at 169.8 ppm in a 13C NMR spectrum of the
reaction mixture. Purification of a reaction mixture by centrifu-
gal chromatography afforded a clean diastereomeric mixture
(Fig. 5c), the composition of which was confirmed by high
resolution ESI-MS. A minor unidentified decomposition prod-
uct with a similar methine resonance at a slightly lower frequen-
cy than that in 26 was also present.
O
O
1. KONH2
2. BuX
1. KONH2
2. BuX
(CH2)n
NHOBu
OH
17
O
n ꢀ 1
n ꢀ 0
20
21
Scheme 2.
O
O
PIFA
NHOR
OH
OR
N
CH3CN
O
r.t.
19a R ꢀ Bu
19b R ꢀ Bn
22a R ꢀ Bu
22b R ꢀ Bn
ButOCl
R ꢀ Bu
O
Silica
Cl
N
OR
OH
23
Scheme 3.
Further support for formation of g-lactam 24 came from an
NMR study of the reaction of 17 with half an equivalent of PIFA
in CD3CN, which gave immediately a clean 50 : 50 mixture of
unreacted hydroxamic ester 17 and g-lactam 24 together with
iodobenzene (Fig. 5d). ESI-MS of this mixture showed ions at
m/z 176 and 198 attributable to [M þ H]þ and [M þ Na]þ ions
for starting material 17 and m/z 196 and 347 attributable to
[M þ Na]þ and [2M þ H]þ for g-lactam 24.
Alkoxyamination of ethyl 3-hydroxybutanoate (16), ethyl salic-
ylate (18), and b-butyro- and g-valerolactones 20 (n ¼ 0 and 1)
afforded the hydroxamic ester precursors 17, 19, and 21
(Scheme 2).
PIFA oxidation of N-butoxysalicamide (19a) in acetonitrile
afforded a 30 % yield of the g-lactam 2-butoxy-3(2H)-
benzisoxazolone (22a) together with a large quantity of uniden-
tified polymeric material which may be accounted for by
oxidation of the phenolic group. Similarly, benzyloxysalica-
mide (19b) afforded 2-benzyloxy-3(2H)-benzisoxazolone (22b)
in 20 % yield (Scheme 3). g-Lactams 22a and 22b exhibited
a characteristically high carbonyl stretch frequency at 1740
and 1743 cmꢀ1 as well as a carbonyl carbon resonance at
173 ppm very close to the chemical shift of those in acyclic
N,N-dialkoxybenzamides (174 ppm).[1] The cyclic dialkoxya-
mide 22a could also be accessed by the N-chloroamide 23.
Chlorination of N-butoxysalicamide with tert-butylhypochlorite
afforded 23, which cyclised cleanly to lactam 22a upon stirring
with, or elution on, silica gel (Scheme 3).
Treatment of b- and g-hydroxyhydroxamic esters 17 and 21
with PIFA afforded unstable intermediates which were consis-
tent with cyclised forms 24 and 25 respectively (Chart 3). NMR
analysis of the product mixture from oxidation of b-hydroxyhy-
droxamic ester 17 immediately after workup indicated the
expected formation of iodobenzene, consumption of starting
material, and generation of g-lactam 24 as evidenced by a clear
ABX spin system centred on d4.65 (5-HX as a ddq, J 9.7, 5.7, and
6.1), d2.74 (4-HA as a dd, J 17.1 and 5.7) and d2.42 (4-HB as a
Formation of 24 can be accounted for by intramolecular
cyclisation of the b-hydroxyl group onto nitrogen as initially
proposed. However, 24 appears to be thermally labile at room
temperature and an intramolecular HERON rearrangement
would afford the alkoxynitrene 28 which upon dimerisation
gives two diastereomers of 26 (Scheme 4). Alkoxynitrenes are
known to dimerise to hyponitrites although their most stable
form has a triplet ground state.[35–37] Formation of 26 must
involve a singlet to triplet (S–T) transition after the HERON
reaction.
The ring opening of g-lactam 24 is clearly a more facile
process than opening of the corresponding g-lactam ring in
benzisoxazolones 22. Ring opening in that case would require an
nO –s*NO anomeric interaction and delocalisation of the
endo
exo
endocyclic oxygen lone pair onto the aromatic ring would
oppose this effect. The ground state structure of the model 7
supports this as it indicates a strong exo-anomeric effect; the
˚
exocyclic N–O bond is a full 0.1 A shorter than the endocyclic
N–O bond (Table 1).
Treatment of g-hydroxyester 21 with PIFA afforded a
mixture of iodobenzene and two major components, one of

512
S. A. Glover et al.
5-Me
4-Me
1'-CH2
HA HB
HX
1'-CH2
3-H
(d)
4-Me
1'-CH2
HA
HB
HX
(c)
4-Me
1'-CH2
HB
HX
HA
(b)
(a)
I
Ho
5-Me
1'-CH2
Hm
HX
HB
Hp
HA
7
6
5
4
3
2
[ppm]
Fig. 5. 1H NMR spectra of mixtures from PIFA oxidationof 17. (a) Mostly g-lactam 24 (see Fig. 6a); (b) mostly hyponitrite 26 (see Fig. 6b);
(c) clean hyponitrite 26 in CDCl3; and (d) 50 : 50 mix of 17 and 24 in CD3CN (resonances for 17 in red; see Fig. 6a).
HE
Pr
2
HX
6
(a)
(b)
(c)
HD
1
HB
HB
4
N
O
HF
O
BuO2C
3
1
O
2
CH3
HA
2
O
Me
O
3
N
OBu
N
3
N
HA
H3C
4
5
HA
O
HX
O
HC
H3C
4
5
1
CO2Bu
HB
JAB ꢀ 13 Hz; JAC ꢀ JAD ꢀ 7.8 Hz; JAX ꢀ 6.3 Hz; JBX ꢀ 9.6 Hz
HX
AB ꢀ 17.1 Hz; JAX ꢀ 5.9 Hz; JBX ꢀ 9.7 Hz
(R,R), (R,S)
J
Fig. 6. Proton designations and relevant proton–proton coupling data for (a) g-lactam 24, (b) hyponitrite 26 and (c) d-lactam 25.
O
O
O
O
OBu
N
OBu
N
PIFA
HERON
NHOBu
OH
N OBu
O
O
O
H3C
2
17
24
28
26
Scheme 4.
which was g-valerolactone (27). An unstable product, from
which the lactone 27 formed upon standing, displayed reso-
protons HE and HF (Fig. 6c) at d4.0 were diastereotopic and
resonated as a tight ABX2 spin system (JAB 10.1, Jvic 6.7). This
is a further manifestation of the cyclic, chiral nature of 25 as
such diastereotopicity is not evident in analogous acyclic
N,N-dialkoxyamides.[1] Analysis of a clean multiplet at d2.1,
which was attributedtoHA, and a trans-diaxialcoupling of9.6 Hz
between HX and HB indicated that 25 was largely in the chair
conformation shown in Fig. 6c and in the B3LYP/6-31G(d)
1
nances in the H NMR spectrum characteristic of d-lactam 25
(Fig. 7, Fig. 6c). Notably, a methine proton at d4.3 exhibited
similar coupling to that of HX in the g-lactam 24, namely a ddq
(J 9.6, 6.3 and 6.6) and together with protons centred at d1.9 and
d2.1 was part of an ABX spin system. The exocyclic 6-methyl
group appeared at d1.34. Interestingly, the oxymethylene

Properties of Cyclic N,N-Dialkoxyamides
513
(c)
(b)
HE,F
HE,F
6-Me
HX
HC,D
HB
HX
HA
(a)
4
3
2
1
[ppm]
Fig. 7. 1H NMR spectra (CDCl3) of mixtures of d-lactam 25 and g-valerolactone (27). (a) Directly after PIFA oxidation of 21; (b) after
standing for several hours; and (c) reference g-valerolactone (27).
O
O
O
OBu
HERON
PIFA
NHOBu
N
O
BuOH
BuON
O
CH3CN
r.t.
OH
29
30
21
25
27
PIDA
CD3CN
ꢁAcOH
ꢁAcOH, PhI
O
OBu
Ph
N
I
OAc
OH
31
Scheme 5.
optimised geometry for model d-lactam 8 depicted in Table 1.
Signals attributable to d-lactam 25 diminished upon formation
of the g-lactone 27.
A reaction in CD3CN in the probe of the NMR machine was
analysed by ESI-MS directly after addition of PIFA. NMR
spectroscopy indicated the presence of d-lactam 25 and
g-valerolactone (27) in a 3 : 1 ratio. ESI-MS indicated a
prominent ion for 25 at m/z 188 ([M þ H]þ) together with one
for the lactone at m/z 101 ([M þ H]þ).
The facile transformation of d-lactam 25 into the g-lactone
27 suggests that 25 also undergoes a HERON reaction but in this
case the endocyclic oxygen migrates in preference to the
exocyclic oxygen (Scheme 5). No evidence for formation of
the homologue of hyponitrite 26 could be found upon complete

514
S. A. Glover et al.
decomposition of 25. Butoxynitrene products were not observed
post workup.
material coincided with formation of another intermediate,
which was a precursor to the d-lactam 25 and which was
consistent with formation of N-(acetoxyiodobenzene)-g-
hydroxyvaleramide (31) (Fig. 8a–d). Compound 31 (labelled
in Fig. 8d) exhibited a different higher frequency oxymethy-
lene signal at d4.05 as well as a new exocyclic methyl doublet
in a similar position to that of starting material 21, both of
which disappeared in concert with formation of d-lactam 25
and lactone 27. Butanol 30 (labelled in Fig. 8g), possibly from
nitrene 29, appeared to form in conjunction with the lactone 27
(labelled in Fig. 8h).
A reaction of 21 with phenyliodine diacetate (PIDA)
carried out in CD3CN and monitored by 1H NMR spectroscopy
led to slower consumption of starting material and formation
of two transient intermediates. As in the PIFA reaction, the
d-lactam 25 (labelled in Fig. 8f) was generated, as evidenced
by the characteristic methine at d4.25, oxymethylene triplet at
d3.94, and exocyclic methyl at d1.27. The d-lactam 25 was
converted into g-lactone 27 in a consecutive reaction with its
formation (Fig. 8d–h). However disappearance of starting
27
27
27
(h)
(g)
(f)
25
25
30
(e)
(d)
(c)
31
31
(b)
(a)
21
8
6
4
2
[ppm]
Fig. 8. 1H NMR spectra (CD3CN) of the reaction mixtures from PIDA oxidation of 21 over 48 h. (a) Acyclic hydroxamic ester
21; (b), (c) and (d) formation of adduct 31; (f) conversion of adduct 31 to d-lactam 25; (g) and (h) mostly g-valerolactone 27.
Table 3. B3LYP/6-31G(d) derived energies and properties of HERON reactions of N,N-dimethoxyacetamide (5), g-lactam 6, and d-lactam 8A
Activation energies (EA), enthalpies (DHz) and free energies of activation (DGz), singlet and triplet reaction energies (DES and DET), singlet and triplet reaction
enthalpies (DHS and DHT) and singlet and triplet reaction free energies (DGS and DGT) in kJ molꢀ1; transition state imaginary frequencies (TSi)
Reactant
EA
DHzB
DGzC
DEDS
DHBS,D
DGCS,D
DEET
DHBT,E
DGCT,E
TSi [cmꢀ1
]
5 Scheme 6
6 Scheme 7
6 Scheme 8
8 Scheme 9
8 Scheme 10
178.9
113.1
197.7
136.5
145.1
171.0
108.1
192.0
131.4
139.2
170.7
105.5
188.1
129.5
136.0
115.2
71.6
101.3
68.2
139.3
63.2
41.2
11.6
29.8
9.6
68.0
5.2
32 321
35 256
37 326
39 336
41 323
177.0
64.2
166.9
54.8
108.2
48.3
103.0
13.4
95.4
9.2
36.9
3.5
115.3
105.1
46.0
41.3
33.6
ꢀ25.2
AB3LYP/6-31G(d) energies of reactants, transition states, rearrangement products and thermal and entropy corrections are provided as Supplementary
Material together with geometries.
BTemperature correction at 298.15 K.
CTemperature and entropy correction at 298.15 K.
DReaction energies of singlet nitrene.
EReaction energies after nitrene singlet to triplet (S–T) relaxation.

Properties of Cyclic N,N-Dialkoxyamides
515
The ring opening of g-lactam 24 and ring contraction of
d-lactam 25 constitute the first observation of HERON
reactivity of N,N-dialkoxyamides, which hitherto have been
found to undergo thermal N–O homolysis at higher tempera-
tures.[1] The energetics for HERON reactions of models N,N-
dimethoxyacetamide (5), g-lactam 6, and d-lactam 8 computed
at the B3LYP/6-31G(d) level are given in Table 3. For the
lactams 6 and 8 the energetics of both HERON ring-opening and
ring-contraction pathways have been determined. All transition
states possessed one imaginary frequency corresponding to the
reaction coordinate.
with enthalpy and entropy correction at 298.15 K) is high when
compared with migration of a methoxyl group in bisheteroatom-
substituted amides with nitrogen involvement (N-methoxy-N-
(dimethylamino)formamide has a computed barrier of just
89.5 kJ molꢀ1, which has been experimentally verified[9,14,16]).
Migration of methoxyl in N-formyloxy-N-methoxyformamide
has a similar barrier (EA ¼ 180 kJ molꢀ1) but here the
formyloxyl group is predicted to migrate preferentially (EA ¼
163 kJ molꢀ1) as has been observed in the thermal reactions of
N-acyloxy-N-alkoxyamides.[14,18]
The activation barrier to ring opening in 6 to singlet nitrene
36 according to Scheme 7 has been computed to be only
113 kJ molꢀ1 (106 kJ molꢀ1 with enthalpy and entropy correc-
tion at 298.15 K) and although it is significantly endothermic,
the overall process is slightly endothermic after S–T relaxation.
Analysis of the ground state structure of 6 and the transition
state structure 32 for methoxyl migration in N,N-dimethoxya-
cetamide provides a rationale for this difference in activation
barriers. At the transition state for methoxyl migration in 5 there
is extensive rotation about the N–C bond and loss of most
residual stabilisation as a result of conjugation between the
nitrogen lone pair and the carbonyl carbon (Fig. 9a, b). This is
common to all HERON reactions.[9,14] Fig. 3 illustrates that such
rotation (from pyramidal, untwisted form (b) to pyramidal,
twisted form (d)) requires ,45 kJ molꢀ1. The ground-state
structure for g-lactam 6 is depicted in Table 1 and Fig. 4a, b
and the transition state for methoxyl migration in g-lactam 6 is
shown in Fig. 10a, b. It is clear that the transition state can be
achieved without any significant degree of twist about the N–C
bond. Clearly this contribution to the activation barrier in
rearrangement of N,N-dimethoxyacetamide (5) to methylacetate
The computed activation barrier for formation of methyl
acetate (33) and methoxynitrene (34) from N,N-dimethoxyace-
tamide (5) according to Scheme 6 (EA ¼ 179 or 171 kJ molꢀ1
O
O
Me
O
Me
MeO
N
OMe
ꢂ
OMe
N
OMe
Me
MeO N
O
Me
5
32
Scheme 6.
33
34
O
O
O
O
OMe
N
N
OMe
N
O
O
O
Me
6
35
36
Scheme 7.
(a)
(b)
Osyn
1.860 Å
1.21 Å
O
syn 1.28 Å
N
48.2ꢃ
N
111ꢃ
1.93 Å
1.546 Å
Oanti
Oanti
Fig. 9. B3LYP/6-31G(d) transition state geometry for HERON reaction of N,N-dimethoxyacetamide (5).
(a) Projection along the N–C bond; (b) side-on view.
(a)
(b)
(c)
1.74 Å
Oendo
N
1.28 Å
1.94 Å
N
44.6ꢃ
1.22 Å
Oexo
1.99 Å
46ꢃ
Oendo
1.54 Å
N
Oendo
1.50 Å
Oexo
111ꢃ
1.24 Å
2.10 Å
Oexo
1.19 Å
Fig. 10. B3LYP/6-31G(d) transition state geometry for HERON reactions of g-lactam 6. (a) Ring opening projection along the N-C
bond in 35; (b) ring opening side-on view; and (c) ring contraction in 37.

516
S. A. Glover et al.
(33) is largely absent in the case of 6 resulting in the lower
activation barrier. Fig. 4 also shows that anomeric destabilisa-
tion of the endocycic and exocyclic N–O bonds is similar and
this is reflected in nearly equivalent bond lengths (Table 1).
However, the alternative HERON (Scheme 8) leading to loss of
methoxynitrene (34) and formation of b-propiolactone (30) is
computed to have a much larger activation barrier of (198 or
188 kJ molꢀ1 with enthalpy and entropy correction at 298.15 K)
and is not competitive with ring opening rearrangement to
nitrene 36. Presumably formation of the four-membered ring
in transition state 37 (Scheme 8, Fig. 10c) invokes too much
strain for this process to be viable.
Reactivity of g-lactam 6 is in contrast to the rearrangement
of model d-lactam 8. Ring opening and ring contraction
HERON processes of 8, shown in Schemes 9 and 10 respec-
tively, have similar activation barriers (Table 3) and both are
measurably lower than that for the HERON reaction of N,N-
dimethoxyacetamide (5). However, two factors favour ring
contraction to 42 over ring opening to 40 in this case. First, ring
contraction to g-lactone 42 and singlet methoxynitrene (34) is
slightly less endothermic than ring opening but collapse of
methoxynitrene to the lower energy triplet state results in overall
exothermicity. Second, the reaction may be under stereoelec-
tronic control; the chair conformation for 8 in the ground state
(Table 1 and Fig. 4b) exhibits an unusually strong anomeric
destabilisation of the endocyclic N–O bond by the p-type lone
pair on the exocyclic oxygen, as evidenced by the very large
˚
difference in the N–O bond lengths (0.08 A, Table 1). A strong
nY–s*NX interaction in Fig. 1 is intrinsic to the HERON
process.[9,14] However, the reverse effect is switched off
(Fig. 4a) as the p-type lone pair on the endocyclic oxygen is
orthogonal to the exocyclic N–O bond. Transition states for
HERON reactions of d-lactam 8 are shown in Fig. 11. The greatly
reduced activation barrier for the g-lactone formation from
d-lactam 8 relative to the b-lactone formation from g-lactam 6
is clearly a reflection of the reduced strain in the five-membered
ring transition state 41. A clear reason for the reduced activation
barriers for both HERON reactions of d-lactam 8 relative to N,N-
dimethoxyacetamide (5) as well as the preference of ring con-
traction over ring opening is not immediately obvious from the
ground state or transition state structures.
O
OMe
O
N
O
Conclusion
N
OMe
MeON
ꢂ
O
Hypervalent iodine oxidation of b- and g-hydroxyhydroxamic
esters has been demonstrated to effect ring closure to cyclic
N,N-dialkoxyamides. Cyclisation of the salicamides 19 afforded
the first stable g-lactams, N-alkoxybenzisoxazolones 22. The
g-lactam 24 formed from the saturated b-hydroxybutanamide
(17) was however unstable, undergoing an intramolecular
O
O
6
37
38
34
Scheme 8.
O
HERON ring opening driven by an n
interaction. This anomeric process is disfavoured in the benzi-
–s*
anomeric
O
O
Oendo
NOexo
OMe
N
N
OMe
N
O
soxazolones on account of nO lone pair delocalisation onto the
endo
O
O
benzene ring. The d-lactam 25, while formed upon PIFA and
PIDA oxidation of 21, is also unstable undergoing the reverse
HERON reaction leading to ring contraction to a lactone.
The computed activation barriers for ring opening and ring
contraction of g-lactam 6 support the experimental findings;
only ring opening of g-lactam intermediate 24 to nitrene 28 was
observed. In the case of formation of d-lactam intermediate 25
using PIFA or PIDA, no ring opening in competition with
lactone formation was evident. While stereoelectronic factors
may be a factor, the influence of substituent effects cannot be
discounted.
O
Me
8
39
40
Scheme 9.
O
OMe
N
O
MeON
ꢂ
O
N
OMe
O
O
O
The search for stable cyclic and bicyclic forms of N,N-
dialkoxyamides is of interest since structural properties could
generate lactams with very low amidicity. We recently showed
8
41
42
34
Scheme 10.
(a)
(b)
1.28Å
Oexo
N
1.80Å
Oendo
1.21
1.29Å
49ꢃ
1.88Å
48ꢃ
N
1.54Å
Oexo
1.90Å
1.21Å
1.94Å
1.55Å
Oendo
Fig. 11. B3LYP/6-31G(d) transition states for HERON reactions of 8. (a) Ring opening in 39 and (b) ring
contraction in 41.

Properties of Cyclic N,N-Dialkoxyamides
517
that b-lactam antibiotics such as those derived from penam
(87 %) and penem (73 %) as well as the cepham (123 %) and
cephem (95 %) bicyclic structures possess amide linkages with
much higher amidicities.[27] Cyclic and bicyclic N,N-dialkox-
yamide structures have greater potential to mimic the behaviour
of b-lactam antibiotics as serine acylators.[38,39]
chromatography (15 % ethyl acetate/n-hexane) to give N-butoxy-
2-hydroxybenzamide (19a) (0.59 g, 2.8 mmol, 35 %). nmax
(CHCl3)/cmꢀ1 3417 (N–H), 3256br (O–H), 1653 (CO), 1607
(Ar). dH 8.86 (br, N–H), 7.40 (d, 1H), 7.38 (t, 1H), 7.00 (d, 1H),
6.85 (t, 1H), 4.02 (t, 2H), 1.86 (br, OH), 1.70 (quin, 2H), 1.46
(sextet, 2H), 0.95 (t, 3H). dC 168.95(CO), 161.29 (C–OH),
134.75, 125.05, 118.94, 118.80 112.22, 77.20 (C–O), 30.04
(CH2CH2O), 19.06 (CH2CH3), 13.86 (CH3). m/z (EI), 209,
137, 121, 93, 92, 88, 86, 84; (HR-EI) 209.1052 (Mþ),
C11H15NO3 requires 209.1052.
Experimental
Materials and Methods
NMR spectra were recorded in CDCl3 or CD3CN on a Bruker
Avance 300P FT NMR spectrometer with a 5 mm 1H inverse/BB
z-gradient probe, operating at 300.13 (1H) and 75.46 MHz (13C).
Mass spectra were recorded by direct injection of samples onto a
Varian 1200 L triple quadrupole mass spectrometer with an
electrospray ionisation interface using 1 % formic acid in
methanol as solvent matrix. Centrifugal chromatographic
separations were performed on a 7294T model Harrison
Research chromatatron with plates coated with 2.0 mm of silica
gel 60 F254 (Merck). Flash chromatography columns were
charged with Silica gel 60, 230–400 mesh. Hypervalent iodine
oxidants, PIFA and PIDA, were available commercially.
Potassium salts of hydroxamic acids and hydroxamic esters
were synthesised according to established protocols.[7,40–43]
Method 2. Potassium 2-hydroxybenzohydroxamate (1 g,
5.9 mmol), butyl bromide (0.81 g, 5.9 mmol), and sodium car-
bonate (0.75 g, 7.1 mmol) were stirred overnight in 50 % aq.
methanol (50 mL) and then refluxed for 2 h. Excess methanol
was removed under reduced pressure and the mixture extracted
with dichloromethane and the extract was dried over anhydrous
sodium sulfate, filtered, and concentrated to give an oil. Purifi-
cation by centrifugal chromatography (15 % ethyl acetate/
n-hexane) afforded clean N-butoxy-2-hydroxybenzamide
(19a) (0.57 g, 2.7 mmol, 48 %).
N-Benzyloxy-2-hydroxybenzamide (19b)
2-Hydroxybenzoyl chloride (1.25 g, 8 mmol) and benzyloxy-
amine hydrochloride (1 g, 8 mmol) were reacted according to
Method 1. Workup produced a brown oil, which was purified by
centrifugal chromatography (15 % ethyl acetate/n-hexane) to
give N-benzyloxy-2-hydroxybenzamide (0.34 g, 1.4 mmol,
17.53 %). vmax (CHCl3)/cmꢀ1 3413 (N–H), 3277br (OH),
1654 (CO), 1607 (Ar). dH 9.25 (br, 1H), 7.30–7.50 (m, 7H),
6.98 (d, 1H), 6.80 (t, 1H), 5.04 (s, 2H), 0.9 (br, 1H). dC 168.72
(CO), 161.25 (C–OH), 134.74, 129.84, 129.39, 129.00, 128.74,
125.60, 118.89, 118.59 112.38, 78.74 (C–O). m/z (HR-ESI)
244.0974 ([M þ H]þ), C14H14NO3 requires 244.0974; 266.0792
([M þ Na]þ), C14H13NO3Na requires 266.0793.
Computational Methods
Fully optimised ground states of models of all structures in
Table 1 and those listed in Table 2 and Table 3, as well as the
energies of distorted N,N-dimethoxyacetamide shown in Fig. 3,
were computed at the B3LYP/6-31G(d) level using MacSpartan
10[44] and Gaussian 03.[45] B3LYP/6-31G(d) has been shown to
perform well in determinations of structure and properties of a
wide range of amides, lactams, and anomeric amides,[2,5,9,10,14,27]
and resonance energies at this level have been demonstrated to
be in relatively good agreement with those obtained with much
larger basis sets and higher levels of correlation energy cor-
rections.[27] Energies of model structures in isodesmic reactions
Eqn 1–Eqn 4 were computed without zero point energies (ZPEs)
and thermal corrections since these largely cancel. For the
HERON processes depicted in Schemes 6–10, the determination
of activation energies and reaction energies with ZPE, vibra-
tional, and entropy corrections required full frequency calcula-
tions on all reactants, transition states, and singlet and
triplet nitrenes and lactone products. Transition states for
HERON processes possessed one imaginary frequency. Abso-
lute energies for these reactions as well as those for deformation
of N,N-dimethoxyacetamide (5) are provided as Supplementary
Material.
N-Butoxy-3-hydroxybutanamide (17)
Method 1. To a stirred solution of ethyl 3-hydroxybutano-
ate[46] (1.0 g, 7.56 mmol) dissolved in methanol (10 mL) was
added hydroxylamine hydrochloride (0.53 g, 7.56 mmol) fol-
lowed by solid potassium hydroxide (0.85 g, 15.1 mmol). The
mixture was heated to 658C for 1 h, allowed to cool followed by
addition of butyl iodide (2.78 g, 15.1 mmol) and the mixture was
stirred at 258C for 16 h and then warmed to reflux for 1 h.
Removal of the volatiles and purification of the residue by flash
chromatography (ethyl acetate) afforded a colourless oil which
crystallised upon standing (280 mg, 28 %). Mp 47–498C. Rf 0.3
(ethyl acetate). nmax (neat)/cmꢀ1 3154, 1661. dH 9.71 (br s, 1H,
NH), 4.14 (br s, 1H, CH(OH)), 4.00 (br s, 1H, OH), 3.86 (br t,
2H, NOCH2), 2.25 (br dd, J 14.8, 3.3, 1H, CHHCO), 2.18 (br dd,
J 14.8, 8.5, 1H, CHHCO), 1.59 (quintet, 2H, OCH2CH2), 1.36
(sextet, 2H, CH2CH3), 1.19 (d, J 6.2 Hz, 3H, CH(OH)CH3), 0.89
(t, 3H, CH2CH3). dC 169.9 (CO), 76.5 (CH2–O), 64.7 (3–C),
41.56 (2–C), 29.95 (O–CH2CH2), 22.95 (4-C), 18.97 (CH2CH3),
13.77 (CH2CH3). m/z (ESI) 176 ([M þ H]þ), 351 ([2M þ H]þ),
373 ([2M þ Na]þ); (HR-ESI) 176.1286 ([M þ H]þ), C8H18NO3
requires 176.1287; 198.1105 ([M þ Na]þ), C8H17NO3Na
requires 198.1106.
Synthesis of Hydroxamic Esters 17, 19, and 21
N-Butoxy-2-hydroxybenzamide (19a)
Method 1. 2-Hydroxybenzoyl chloride (1.25 g, 8 mmol)
and butoxyamine hydrochloride (1 g, 8 mmol) were added to
100 mL of diethyl ether in a two neck flask in an ice bath.
Triethylamine (1.62 g, 0.016 mol) was added dropwise over 1 h
so that the temperature remained below 58C. After stirring
overnight, the ether was removed under reduced pressure. The
solid was treated with saturated sodium bicarbonate and
extracted with diethylether (2 ꢁ 15 mL). The extract was
washed with 50 mL of HCl (0.l M), separated, and dried
over anhydrous MgSO4. Concentration afforded a yellowy-
brown oily compound, which was purified by centrifugal
Method 2. Hydroxylamine hydrochloride (4.83 g,
69.5 mmol) in boiling methanol (30 mL) was added to potassium
hydroxide (7.8 g, 0.139 mol) in boiling methanol (50 mL)
with stirring. The mixture was immediately cooled in an ice bath
for 5 min. b-Butyrolactone (5 g, 58.1 mmol) was added to the

518
S. A. Glover et al.
mixture with shaking followed by filtration. The filtrate was
monitored by TLC until no lactone was observed. Butyl bromide
(7.96 g, 58.1mmol) in methanol (10 mL) and sodium carbonate
(6.77 g, 63.9mmol) in water (40 mL) were added to the filtrate
and the resultant mixture was refluxed and monitored by TLC
until no further increase in non-baseline material was evident.
Excess methanol was removed under reduced pressure and the
mixture diluted with brine (50 mL). Extraction with dichloro-
methane, which was dried (anhydrous Na2SO4) and concentrated
under reduced pressure, afforded clean N-butoxy-3-hydroxybu-
tanamide (17) as a white solid (1.75 g, 9.99 mmol, 17.2 %).
Cyclisation of N-benzyloxy-2-hydroxybenzamide (19b)
N-Benzyloxy-2-hydroxybenzamide (19b) (0.24 g, l mmol)
was dissolved in 15 mL of anhydrous acetonitrile, treated with
PIFA (0.42 g, 1.5 mmol), and stirred for 5 min. NaHCO3 (10 %,
20 mL) was added with agitation for a further 3 min. The mixture
was extracted with portions of dichloromethane (2 ꢁ 20 mL),
which were washed with brine, dried over Na2SO4, and concen-
trated under reduced pressure. Purification by centrifugal
chromatography (15 % ethyl acetate/n-hexane) afforded
N-benzyloxy-3(2H)-benzisoxazolone (22b) (0.05 g, 0.2 mmol,
20 %). nmax (CHCl3)/cmꢀ1 1743 (CO). dH 7.78 (d, 1H, 4-H, 7.66
(t, 1H, 6-H), 7.51–7.54 (d, 2H, o0-H), 7.37–7.43 (m, 3H,
m0,p0-H), 7.18 (t, 1H, 5-H), 7.05 (d, 1H, 7-H) 5.31 (s, 2H). dC
173.43 (CO), 164.72 (CAr–O), 137.11 (6-C), 134.54 (C’ipso),
129.26, 128.93 (C0p), 128.58, 125.85 (4-C), 123.30 (5-C),
112.93 (C–CO), 110.64 (7-C), 79.49 (CH2). m/z (HR-ESI)
242.0815 ([MþH]þ), C14H12NO3 requires 242.0817;
264.0638 ([MþNa]þ), C14H11NO3Na requires 264.0637.
N-Butoxy-4-hydroxypentanamide (21)
Hydroxylamine hydrochloride (4.17 g, 60 mmol) in hot
methanol (30 mL) was added to potassium hydroxide (6.73 g,
120 mmol) in boiling methanol. g-Valerolactone (5 g, 50 mmol)
was added and the solution was cooled and filtered. After
standing overnight, lactone was not detected by TLC. Butyl
bromide (6.84 g, 50 mmol) was added and the reaction mixture
was stirred under reflux and monitored by TLC until no further
increase in non-baseline material was evident. Most of the
methanol was removed under reduced pressure and the mixture
diluted with brine. Extraction with dichloromethane, which was
dried (anhydrous Na2SO4) and concentrated under reduced
pressure, afforded clean N-butoxy-4-hydroxypentanamide (21)
as a yellow oil (0.729 g, 3.85 mmol, 8 %). nmax (neat)/cmꢀ1
3500ꢀ3200br (NH, OH), 1661. dH 3.85 (t, 2H, CH2–O), 3.8
(m, 1H, CH–O), 2.2–2.5 (br m, 2H, a-CH2), 1.75–1.9 (m, 1H,
b-CH), 1.65–1.75 (m, 1H, b-CH), 1.62 (quin, 2H), 1.4 (sextet,
2H), 1.2 (d, 3H), 0.95 (t, 3H). dC 171.42 (CO), 77.3(CH2–O),
67.2 (4-C), 34.07 (3-C), 29.95 (O–CH2CH2), 28.03 (2-C), 23.53
(5-C), 18.99 (CH2CH3), 13.79 (CH2CH3). m/z (ESI) 190 ([M þ
H]þ), 212 ([M þ Na]þ), 401 ([2M þ Na]þ); (HR-ESI) 190.1443
([M þ H]þ), C9H20NO3 requires 190.1443; 212.1263 ([M þ
Na]þ), C9H19NO3Na requires 212.1263.
Cyclisation of N-butoxy-3-hydroxybutanamide (17)
Treatment of N-butoxy-3-hydroxybutanamide (17) (0.027 g,
0.15 mmol) in anhydrous acetonitrile (2 mL) with an equivalent
of PIFA (0.066 g, 0.015 mmol) in anhydrous acetonitrile (2 mL)
afforded, after workup, a mixture (0.033 g) which was immedi-
ately analysed by NMR spectroscopy and was comprised mainly
of g-lactam 24 and iodobenzene (see discussion). Upon standing
overnight, 24 decomposed to a diastereomeric mixture of
hyponitrite 26. dH 5.5–5.7 (overlapping m, 2H, CH–O), 4.0–
4.15 (overlapping t, 4H, CH2–O), 3.0–3.15 (overlapping dd, 2H,
a-H1), 2.5–2.65 (overlapping dd, 2H, a-H2), 1.5–1.7 (over-
lapping quintets, 4H, O–CH2CH2), 1.4 and 1.38 (overlapping
d, 6H, CH3CH), 1.3–1.4 (overlapping sextets, 4H, CH2CH3),
0.9–0.98 (overlapping t, CH2CH3). dC 169.89 (CO), 64.80
(O–CH2), 59.90 (b-CH–O), 37.09 (a-CH2), 30.55 (CH2CH2–
O), 19.04 (CH2CH3), 16.02 (CH3CH–O), 13.64 (CH2CH3). m/z
(ESI) 347.2 ([M þ H]þ), 715.5 ([2M þ Na]þ). Purification of a
portion by centrifugal chromatography afforded clean diastero-
meric mixture 26. m/z (HR-ESI) 347.2182 ([M þ H]þ),
C16H31N2O6 requires 347.2182; 369.2002 ([M þ Na]þ),
C16H30N2O6Na requires 369.2002.
N-Butoxy-3-hydroxybutanamide (17) and PIFA in a 2 : 1
ratio were reacted in CD3CN in the probe of the NMR machine.
The reaction produced an immediate 50 : 50 mixture of
unreacted 17 and g-lactam 24 together with iodobenzene.
Compound 17: dH (CD3CN) 4.58–4.70 (ddq, J 8.9, 6.1, and
6.1, 1H, CH–O), 3.95 (t, 2H, CH2–O), 2.78 (dd, J 17.8 and 6.1, a-
H1), 2.44 (dd, J 17.8 and 8.9, a-H2), 1.59 (quin, 2H,
O–CH2CH2), 1.38 (sextet, 2H, CH2CH3), 1.37 (d, 3H, CH3CH),
0.9 (t, 3H, CH2CH3). m/z(ESI) 176 ([M þ H]þ), 198 ([M þ
Na]þ); Compound 24: dH (CD3CN) 4.0–4.15 (br m, 1H,
CH–OH), 3.83 (t, 2H, CH2–O), 2.1–2.25 (br m, 2H, a-CH2),
1.59 (quin, 2H, O–CH2CH2), 1.38 (sextet, 2H, CH2CH3), 1.15
(d, 3H, CHCH3), 0.9 (t, 3H, CH2CH3). m/z (ESI) 196
([M þ Na]þ).
PIFA Cyclisations of 17, 19, and 21
Hydroxamic esters in anhydrous acetonitrile were treated
rapidly with similar quantities of PIFA and briefly stirred. The
reaction mixture was diluted with saturated sodium carbonate.
The aqueous mixture was extracted with dichloromethane,
which was washed with brine and concentrated to afford
mixtures containing cyclised materials.
Cyclisation of N-butoxy-2-hydroxybenzamide (19a)
N-Butoxy-2-hydroxybenzamide (19a) (0.35 g, 1.7 mmol) in
50 mL of anhydrous acetonitrile was treated with PIFA (1.1 g,
2.5 mmol) and stirred for 5 min. NaHCO3 (10 %, 30 mL) was
added with agitation for a further 3 min. The mixture was
extracted with portions of dichloromethane (2 ꢁ 20 mL), which
was washed with brine, dried over Na2SO4, and concentrated
under reduced pressure. Purification by centrifugal chromatog-
raphy (15 % ethyl acetate/n-hexane) afforded N-butoxy-3(2H)-
benzisoxazolone (22a) (0.11 g, 0.5 mmol, 31.43 %). nmax
(CHCl3)/cmꢀ1 1742 (CO), 1618. dH 7.79 (d, 1H, 4-H),
7.67 (t, 1H, 6-H), 7.18 (t, 1H, 5-H), 7.06 (d, 1H, 7-H), 4.30
(t, O–CH2), 1.85 (quin, 2H, 20-CH2), 1.54 (sextet, 2H, 30-H),
0.98 (t, 3H, 40-H). dC 173.36 (CO), 164.49 (CAr–O), 137.00
(6-C), 125.82 (4-C), 123.23 (5-C), 113.06 (C–CO), 110.59
(7-C), 77.71 (10-C), 30.24 (20-C), 18.83 (30-C), 13.77 (40-C).
m/z (EI) 207, 177, 135, 121, 120, 93, 92; (HR-EI) 207.0893
(Mþ), C11H13NO3 requires 207.0895.
Cyclisation of N-butoxy-4-hydroxypentanamide (21)
Treatment of N-butoxy-4-hydroxypentanamide (21) (0.23 g,
1.2 mmol) in anhydrous acetonitrile (5 mL) with an equivalent
of PIFA (0.52 g, 1.2 mmol) in anhydrous acetonitrile (5 mL) at
room temperature afforded, after workup, a mixture (0.2 g)
which was immediately analysed by NMR spectroscopy and
was comprised mainly of d-lactam 25, g-valerolactone (27), and

Properties of Cyclic N,N-Dialkoxyamides
519
iodobenzene (see discussion). The mixture reverted to lactone
upon standing overnight. A repeat reaction in CD3CN and
analysed by NMR spectroscopy contained, along with iodoben-
zene, lactam 25, and lactone 27 in a 3 : 1 ratio. The mixture was
analysed immediately by ESI-MS. m/z (ESI) 101 (27, [M þ H]þ)
188 (25, [M þ H]þ).
[8] J. J. Campbell, S. A. Glover, J. Chem. Res. (S) 1999, 8, 474.
doi:10.1039/A903555I
[9] S. A. Glover, G. Mo, A. Rauk, Tetrahedron 1999, 55, 3413.
doi:10.1016/S0040-4020(98)01151-X
[10] S. A. Glover, A. Rauk, J. Org. Chem. 1999, 64, 2340. doi:10.1021/
JO982048P
[11] S. A. Glover, G. Mo, A. Rauk, D. Tucker, P. Turner, J. Chem. Soc.,
Perkin Trans. 2 1999, 2053. doi:10.1039/A904575I
N-Butoxy-4-hydroxypentanamide (21) (0.02 g, 0.11 mmol)
and PIDA (0.034 g, 0.11 mmol) were reacted in CD3CN (1 mL)
in the probe of the NMR machine and the complex mixture
analysed by NMR spectroscopy (see discussion).
[12] G. Mo, Properties and Reactions of Anomeric Amides 1999,
Ph.D. thesis, University of New England, Armidale.
[13] S. A. Glover, G. Mo, J. Chem. Soc., Perkin Trans. 2 2002, 1728.
doi:10.1039/B111250N
[14] S. A. Glover, A. Rauk, J. M. Buccigross, J. J. Campbell,
G. P. Hammond, G. Mo, L. E. Andrews, A.-M. E. Gillson, Can.
J. Chem. 2005, 83, 1492. doi:10.1139/V05-150
[15] S. A. Glover, in Merck Index Fourteenth Edition, ONR-43
(Ed. M. J. O’Neil) 2006 (Merck & Co., Inc.: Whitehouse Station, NJ).
[16] S. A. Glover, in Advances in Physical Organic Chemistry
(Ed. J. Richard) 2008, Vol. 42, pp. 35–123 (Elsevier: London).
[17] S. A. Glover, in The Chemistry of Hydroxylamines, Oximes and
Hydroxamic Acids, Part 2 (Eds Z. Rappoport, J. F. Liebman) 2009,
pp. 839–923 (Wiley: Chichester).
N-Butoxy-N-chloro-2-hydroxybenzamide (23)
A solution of N-butoxy-2-hydroxybenzamide (19a) and an
equimolar amount of tert-butyl hypochlorite[47] in dichloro-
methane was stirred in the dark for 3 h. The solvent and
residual tert-butyl hypochlorite were removed under reduced
pressure at room temperature to give the N-butoxy-N-chloro-2-
hydroxybenzamide (23) in quantitative yield as a yellow oil,
1
which was characterised by H NMR spectroscopy. dH 10.15
(1H, s, OH), 7.7 (d, 1H), 7.5 (t, 1H), 7.05 (d, 1H), 6.85 (t, 1H), 4.2
(t, 2H), 1.7 (quin, 2H), 1.4 (sextet, 2H), 0.95 (t, 3H). Relative to
precursor hydroxamic ester, protons a- to the alkoxyl oxygen
undergo a typical downfield shift of around 0.2 ppm in their 1H
NMR spectrum upon N-chlorination.
[18] J. P. Johns, A. van Losenoord, C. Mary, P. Garcia, D. S. Pankhurst,
A. A. Rosser, S. A. Glover, Aust. J. Chem. 2010, 63, 1717. doi:10.1071/
CH10350
[19] S. A. Glover, M. Adams, Aust. J. Chem. 2011, 64, 443. doi:10.1071/
CH10470
[20] V. G. Shtamburg, E. A. Klots, A. P. Pleshkova, V. I. Avramenko,
S. P. Ivonin, A. V. Tsygankov, R. G. Kostyanovsky, Russ. Chem. Bull.
2003, 52, 2251. doi:10.1023/B:RUCB.0000011887.40529.B0
[21] V. G. Shtamburg, A. V. Tsygankov, E. A. Klots, R. G. Kostyanovsky,
Cyclisation of N-butoxy-N-chloro-2-
hydroxybenzamide (23)
N-Butoxy-N-chloro-2-hydroxybenzamide (23) was stirred in
CHCl3 with silica gel 60 for 1 h. The mixture was filtered and
concentrated to afford clean 2-butoxy-3(2H)-benzisoxazolone
(22a).
Mendeleev
Commun.
2004,
14,
208.
doi:10.1070/
MC2004V014N05ABEH001908
[22] O. V. Shishkin, R. I. Zubatyuk, V. G. Shtamburg, A. V. Tsygankov,
E. A. Klots, A. V. Mazepa, R. G. Kostyanovsky, Mendeleev Commun.
2006, 16, 222. doi:10.1070/MC2006V016N04ABEH002195
Acknowledgements
[23] V. G. Shtamburg, O. V. Shishkin, R. I. Zubatyuk, S. V. Kravchenko,
A. V. Tsygankov, A. V. Mazepa, E. A. Klots, R. G. Kostyanovsky,
Avat (Arman) Taherpour is grateful to Razi University-Ker-
manshah-Iran (Faculty of Chemistry) for support and leave of
absence to undertake this work.
Mendeleev
Commun.
2006,
16,
323.
doi:10.1070/
MC2006V016N06ABEH002382
[24] V. G. Shtamburg, O. V. Shishkin, R. I. Zubatyuk, S. V. Kravchenko,
A. V. Tsygankov, V. V. Shtamburg, V. B. Distanov,
R. G. Kostyanovsky, Mendeleev Commun. 2007, 17, 178. doi:10.1016/
J.MENCOM.2007.05.016
[25] V. G. Shtamburg, A. V. Tsygankov, O. V. Shishkin, R. I. Zubatyuk,
V. V. Shtamburg, M. V. Gerasimenko, A. V. Mazepa,
R. G. Kostyanovsky, Mendeleev Commun. 2012, 22, 92. doi:10.1016/
J.MENCOM.2012.03.014
[26] V. G. Shtamburg, A. T. Tsygankov, O. V. Shishkin, R. I. Zubatyuk, B. V.
Uspensky, V. V. Shtamburg, A. V. Mazepa, R. G. Kostyanovsky, Men-
deleev Commun. 2012, 22, 164. doi:10.1016/J.MENCOM.2012.05.019
[27] S. A. Glover, A. A. Rosser, J. Org. Chem. 2012, 77, 5492. doi:10.1021/
JO300347K
Supplementary Material
B3LYP/6-31G(d) energies of ground state reactants, transition
states, singlet and triplet nitrene products and product lactones
relating to Schemes 6–10 for HERON reactions of 5, 6, and 8;
B3LYP/6-31G(d) energies used in Fig. 3 for deformation of N,
N-dimethoxyacetamide (5); B3LYP/6-31G(d) optimised
geometries computed for all structures in Table 1 and Table 2
and Schemes 6–10 for HERON reactions of 5, 6, and 8 and 1H
and 13C NMR spectra of all new reactants and products not
presented in the discussion are available on the Journal’s
website.
[28] Z. Mucsi, A. Tsai, M. Szori, G. A. Chass, B. Viscolcz, I. G. Csizmadia,
J. Phys. Chem. A 2007, 111, 13245. doi:10.1021/JP0759325
[29] J. M. Buccigross, S. A. Glover, J. Chem. Soc., Perkin Trans. 2 1995,
595. doi:10.1039/P29950000595
[30] F. K. Winkler, J. D. Dunitz, J. Mol. Biol. 1971, 59, 169. doi:10.1016/
0022-2836(71)90419-0
[31] J. D. Dunitz, X-Ray Analysis and Structure of Organic Molecules 1979
References
[1] K. M. Digianantonio, S. A. Glover, J. P. Johns, A. A. Rosser, Org.
Biomol. Chem. 2011, 9, 4116. doi:10.1039/C1OB00008J
[2] S. A. Glover, J. M. White, A. A. Rosser, K. M. Digianantonio, J. Org.
Chem. 2011, 76, 9757. doi:10.1021/JO201856U
[3] A. M. Bonin, S. A. Glover, G. P. Hammond, J. Chem. Soc., Perkin
Trans. 2 1994, 1173. doi:10.1039/P29940001173
[4] J. M. Buccigross, S. A. Glover, G. P. Hammond, Aust. J. Chem. 1995,
48, 353.
(Cornell University Press: London).
[32] A. Greenberg, in The Amide Linkage. Structural Significance in
Chemistry, Biochemistry and Materials Science (Eds A. Greenberg,
C. M. Breneman, J. F. Liebman) 2003, pp. 47–83 (John Wiley & Sons,
Inc.: New York, NY).
[5] A. Rauk, S. A. Glover, J. Org. Chem. 1996, 61, 2337. doi:10.1021/
JO9521817
[6] S. A. Glover, Tetrahedron 1998, 54, 7229. doi:10.1016/S0040-4020
(98)00197-5
[7] S. A. Glover, G. P. Hammond, A. M. Bonin, J. Org. Chem. 1998, 63,
[33] A. Greenberg, D. T. Moore, T. D. DuBois, J. Am. Chem. Soc. 1996,
118, 8658. doi:10.1021/JA960294H
[34] A. Greenberg, C. A. Venanzi, J. Am. Chem. Soc. 1993, 115, 6951.
doi:10.1021/JA00068A064
9684. doi:10.1021/JO980863Z

520
S. A. Glover et al.
[35] A. Srinivasan, N. Kebede, J. E. Saavedra, A. V. Nikolaitchik,
D. A. Brady, E. Yourd, K. M. Davies, L. K. Keefer, J. P. Toscano,
J. Am. Chem. Soc. 2001, 123, 5465. doi:10.1021/JA002898Y
[36] W. A. Wasylenko, N. Kebede, B. M. Showalter, N. Matsunaga,
A. P. Miceli, Y. Liu, L. R. Ryzhkov, C. M. Hadad, J. P. Toscano,
J. Am. Chem. Soc. 2006, 128, 13142. doi:10.1021/JA062712G
[37] M. S. Platz, in Reactive Intermediate Chemistry (Eds R. A. Moss,
M. S. Platz, M. Jones) 2004, pp. 501–559 (Wiley: Hoboken, NJ).
[38] M. I. Page, Acc. Chem. Res. 1984, 17, 144. doi:10.1021/
AR00100A005
[45] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, J. A. Montgomery, Jr, T. Vreven, K. N. Kudin,
J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li,
J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ocherski, P. Y. Ayala, K. Morokuma,
G. A. Voth, P. Salvador, J. J. Dannenburg, V. G. Zakrzewski,
S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. C. Ortiz,
Q. A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox,
T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong,
C. Gonzalez, J. A. Pople, Gaussian 03, Revision E.01 2005 (Gaussian,
Inc.: Wallingford, CT).
[39] M. I. Page, Adv. Phys. Org. Chem. 1987, 23, 165.
[40] C. R. Hauser, W. B. Renfrow Jr, 1943, Organic Synthesis Collective,
Vol. II, p. 67 (John Wiley & Sons, Inc.: New York, NY).
[41] J. H. Cooley, W. D. Bills, J. R. Throckmorton, J. Org. Chem. 1960, 25,
1734. doi:10.1021/JO01080A014
[42] J. J. Campbell, S. A. Glover, C. A. Rowbottom, Tetrahedron Lett.
1990, 31, 5377. doi:10.1016/S0040-4039(00)98076-6
[43] J. J. Campbell, S. A. Glover, G. P. Hammond, C. A. Rowbottom,
J. Chem. Soc., Perkin Trans.
2
1991, 2067. doi:10.1039/
[46] R. R. Vakhidov, R. N. Shakhmaev, G. Y. Ishmuratov, V. V. Zorin,
Russ. J. Electrochem. 2000, 36, 771. doi:10.1007/BF02757428
[47] M. J. Mintz, C. Walling, Org. Synth. 1963, 49, 9.
P29910002067
[44] Spartan10 for Macintosh 1.1.0 2010 (Wavefunction Inc.: Irvine, CA).