The acyl nitroso Diels-Alder (ANDA) reaction of sorbate derivatives: An X-ray and 15N NMR study with an application to amino-acid synthesis

Article abstract of DOI:10.1039/b912963d

We present a study of the acyl nitroso Diels-Alder (ANDA) reaction of sorbate esters and sorbic alcohol derivatives, using alkoxycarbonyl nitroso dienophiles. An optimisation of the reaction conditions for ethyl sorbate is first presented, and the product is used in an efficient synthesis of 5-methylornithine. Structure-reactivity trends in sorbic alcohol (E,E-2,4-hexadien-1-ol) and its acylated analogues are then discussed. We present single-crystal X-ray structural proof for key adducts in both series and present in detail a novel HMBC/HSQC (¹H-¹⁵N) criterion for ready distinction of regioisomers arising from such ANDA reactions.

Full text of DOI:10.1039/b912963d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
The acyl nitroso Diels–Alder (ANDA) reaction of sorbate derivatives: an
X-ray and 15N NMR study with an application to amino-acid synthesis†
Lee Bollans,a John Bacsa,a Jonathan A. Iggo,a Gareth A. Morrisb and Andrew V. Stachulski*a
Received 1st July 2009, Accepted 30th July 2009
First published as an Advance Article on the web 2nd September 2009
DOI: 10.1039/b912963d
We present a study of the acyl nitroso Diels–Alder (ANDA) reaction of sorbate esters and sorbic
alcohol derivatives, using alkoxycarbonyl nitroso dienophiles. An optimisation of the reaction
conditions for ethyl sorbate is first presented, and the product is used in an efficient synthesis of
5-methylornithine. Structure–reactivity trends in sorbic alcohol (E,E-2,4-hexadien-1-ol) and its
acylated analogues are then discussed. We present single-crystal X-ray structural proof for key adducts
in both series and present in detail a novel HMBC/HSQC (1H–15N) criterion for ready distinction of
regioisomers arising from such ANDA reactions.
derivatives, or ‘C-nitrosocarbonyl compounds’, were introduced
Introduction
and studied by Kirby et al.7,8 More recently carbamoyl-derived
t
The Diels–Alder reaction of nitroso compounds was first reported
=
=
NO dienophiles, especially PhCH2OCON O and Bu OCON O,
have become popular.9,10 They are conveniently derived from the
corresponding hydroxylamine by in situ oxidation. Henceforward
we refer to hetero-Diels–Alder reactions of this class as the ANDA
(acyl nitroso Diels–Alder) reaction.
by Wichterle1 in 1947 and has found frequent application in
synthesis. Most often the diene substrates have been symmetrical,
particularly cycloalka-1,3-dienes:2,3 with unsymmetrical dienes of
type 1 (Scheme 1), the question of preferred regiochemistry of
the products from reaction with nitroso compounds 2 arises. A
definitive theoretical study by Leach and Houk,4 with compu-
tation of transition states and energetics, showed that the so-
called proximal isomer P (a term first used by Boger;5 Scheme 1)
was generally favoured over the distal isomer D for reactions of
1-substituted 1,3-dienes, whatever the electronic nature of the
1-substituent. The most important frontier molecular interaction
is that between HOMO (diene) and LUMO (dienophile): the
largest HOMO coefficient of a 1-substituted diene is always at
C(4).
Since it introduces 1,4-N,O-difunctionality in one step with
defined relative stereochemistry, the nitroso Diels–Alder reaction
continues to attract the synthetic chemist. Recent developments
have included the introduction of a-acetoxynitroso derivatives,
where the reaction is usefully promoted by Lewis acids in organic
media11a or by the use of organic–aqueous mixtures.11b Also vari-
ous ANDA products of sorbic alcohol and related derivatives have
been transformed into pyrroles.12 Since absolute stereochemistry
can be introduced readily by means of chiral dienophiles,13,14
the value in synthesis is heightened. In the following we report
on our investigations into the ANDA reaction of a number of
unsymmetrical diene esters and alcohols and show its value with
a new synthesis of an alkyl ornithine. We also present structural
analysis of the products using both single-crystal X-ray structure
determination and a novel 1H–15N HMBC/HSQC NMR method
of distinguishing the regioisomers.
Scheme 1 Acyl nitroso Diels–Alder reaction of unsymmetrical dienes.
For X > Y, t h e p rox i m a l P and distal D isomers are defined as shown.
Results and discussion
1. Reaction of sorbate esters: an efficient synthesis of
5-methylornithine
In general terms, the more electron-withdrawing the substituent
on the nitroso group, the more efficient is the cycloaddition, by
virtue of lowering the LUMO energy. In early work in this area,
a-chloro nitroso derivatives were frequently used.6 Acyl nitroso
Our first investigations were conducted with commercially avail-
able ethyl sorbate 3a and either Z or Boc hydroxylamine 4a,b
using periodate oxidants. According to Houk’s prediction4 this
combination is expected to give entirely (Scheme 2) the proximal
adduct, viz. 5a–f, and indeed this has been observed between
sorbate esters and various nitroso dienophiles15 (a literature
correction16 should also be noted).
aRobert Robinson Laboratories, Department of Chemistry, University of
Liverpool, Liverpool, L69 7ZD, UK. E-mail: stachuls@liv.ac.uk; Fax: 44
(0) 151 794 3588; Tel: 44 (0) 151 794 3542
bSchool of Chemistry, University of Manchester, Oxford Rd., Manchester,
M13 9PL, UK
We studied in detail and optimised the reaction of ethyl sorbate 3
† Electronic supplementary information (ESI) available: Additional exper-
imental information; 1H and 13C NMR spectra. CCDC reference numbers
725211 and 738892. For ESI and crystallographic data in CIF or other
electronic format see DOI: 10.1039/b912963d
t
=
=
with PhCH2OCON O and Bu OCON O, Scheme 2: full details
are in the ESI†. In brief, benzyloxycarbonyl hydroxylamine 4a
gave a higher yield than the t-butoxycarbonyl analogue 4b, and
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4531–4538 | 4531
©

View Article Online
Scheme 2 Oxidant: Na+ IO4-, or Bun4N+ IO4-. See ESI† for more details.
NaIO4 in aq. MeOH was a fully satisfactory oxidant. In this series
organic-soluble periodates (we used mainly Bun N+ IO4 ) gave
-
4
lower yields, in striking contrast to the sorbic alcohol series (vide
Fig. 1 Single-crystal X-ray structure of phthalimide 7 showing the 2,5-cis
stereochemistry.
infra), and indeed Bun N+ IO4 performed better in a more polar
-
4
solvent, MeOH. A two-phase variant17 (EtOAc–aq. buffer, pH 5)
was less successful here. Interestingly, ethyl ester 3 gave better
yields than other simple alkyl esters from Me to But: we believe
this detail has not◦been noted by others. The reaction was best
performed at -10 C, and using 2 equivalents of oxidant a very
good yield of 85% of 5a was obtained: a similarly improved yield
of 5b was found under these conditions. We suspect that this
particular ANDA is close to the reactivity limit, with the additional
a-NH2 epimers resulted, reflecting the high acidity of the a-proton.
Instead, aqueous acid hydrolysis achieved clean deprotection and
ring-opening: the desired amino acid 8 was obtained as a single
diastereoisomer in 100% yield. Alkylated analogues of ornithine
are of biological interest, e. g. 5-fluoromethylornithine is an
inhibitor of ornithine aminotransferase.20
=
steric hindrance of BocN O an additional negative factor. Steric
2. Diene alcohols and derived esters
hindrance in the diene (though without a conjugating substituent,
as here) is known greatly to favour the proximal isomer in a related
ANDA, using the dimethyl acetal of sorbic aldehyde.18
Sorbic alcohol 9a is known to give a mixture of proximal and
distal adducts on reaction with acylnitroso compounds, as studied
by Kouklovsky12 et al. One solution to the regiochemical issue (and
obtain exclusively the distal product) is to employ an intramolecu-
lar tether, as effectively demonstrated by Sheradsky15b and Russell
et al.21 We nevertheless probed further the regioselectivity of the
intermolecular ANDA of sorbic alcohol esters, noting in particular
Houk’s comment4 that ‘a very sensitive balance between FMO
interactions, electrostatics and steric effects’ can profoundly affect
the outcome. We have also obtained what we believe is the first
X-ray structure of an acyl nitroso ANDA product of this type, and
evaluated a practical NMR determinant for the regioselectivity of
such reactions. Here again we have screened a large number of
reactions (see ESI†) and restrict ourselves to the main conclusions
here.
Adduct 5a proved to be an ideal a-amino acid precursor.
Thus (Scheme 3) exhaustive hydrogenation of 5a gave the
a-hydroxylactam 6 in very good yield, by cyclisation of the
intermediate amino-ester: this product was obtained by Belleau
et al.6 (where the precursor was accessed using an a-Cl nitroso
dienophile). Conversion of OH to NH2 might be achieved via
the a-bromide, followed by azide displacement and reduction.19
We found that, much more conveniently, 6 was an excellent
Mitsunobu substrate, and on reaction with phthalimide, Ph3P
and (Pr O2CN )2 the highly crystalline phthalimide 7 resulted
in 66% yield. Single-crystal X-ray analysis of this product (Fig. 1)
confirmed both the clean SN2 inversion in the Mitsunobu step and
the regiochemistry of precursor 5a.
i
=
Thus (Scheme 4) sorbic alcohol 9a itself gives a 3:2 proximal:
distal (viz. 10a + 11a) mixture on reaction with ZNHOH and
Bun N+ IO4 in CH2Cl2, in excellent (87%) yield.12 In contrast to
-
4
sorbate esters (vide supra) the ‘organic’ oxidant, viz. Bun N+ IO4- in
4
CH2Cl2, was appreciably more effective than NaIO4 in this series.
=
The steric bulk of the nitroso compound has less effect, ZN O
=
and BocN O giving similar yields. The relatively high amount of
distal product seen with 9a suggests that in the non-polar solvent
(CH2Cl2) an H-bonding interaction between the alcohol OH and
Scheme 3 Synthesis of a single diastereoisomer of 5-Me ornithine.
Reagents and conditions: i) Pd, H2, THF, 65%; ii) phathalimide, Ph3P,
(Pr O2CN )2, THF, 66%; iii) N2H4, CHCl3, 20 ◦C; iv) aq. HCl, reflux,
i
=
100%.
Scheme 4 Acyl nitroso Diels–Alder reaction of sorbic alcohol and derived
esters. Reagents and conditions: as Scheme 2. See ESI† for all analogues
studied.
Interestingly, standard hydrazinolysis conditions proved unsat-
isfactory for deprotection of 7: in a slow reaction, a mixture of
4532 | Org. Biomol. Chem., 2009, 7, 4531–4538
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
carbonyl O of the dienophile is partly overriding the inherent
preference for the proximal isomer. In support of this, use of NaIO4
in aq. MeOH raises the P:D ratio to 3:1 (see ESI†).
convenient method depending on long-range 1H–15N interaction,
and show its value by assigning unambiguously the P:D mixtures
of isomers 10d/11d and 12/13. We were initially assisted by a
supply of 15NH2OH·HCl, which was converted into the 15N form
of 4b by the usual procedure.23
Kouklovsky reported that the TBS ether of 9a strongly favoured
the proximal isomer (ratio 7:1).12 We studied electron-withdrawing
substituents, namely the acetate 9b and 4-nitrobenzoate 9c. Using
We first studied 10e whose structure had been unambiguously
determined by X-ray analysis. It was necessary first to assign
unambiguously all the protons of 10e: this was complicated by
the large magnitude of the allylic coupling. COSY experiments
with a low flip-angle read pulse24 were used to distinguish between
the olefinic protons H-4 and H-5 [see Fig. 3 for numbering] and
in particular to determine the signs of the coupling constants,
assuming the large vicinal coupling JH-4,H-5 to be positive. We found
that H-4/H-5 were defined as d 6.02 [1H, ddd, J = +10.5, +4.6 and
-2.3 Hz (H-4)] and d 5.75 [1H, ddd, J = +10.5, -1.8 and +1.5 Hz
(H-4)] on the basis of the relative signs of their coupling constants,
the vicinal coupling JH-3,H-4 being positive and the allylic JH-4,H-6
negative. A complete listing of the ring coupling constants, with
signs, is given in the ESI†. A number of attempts were made to find
13C NMR evidence for or against the proximal structure for 10e.
13C NOE difference experiments showed significant Overhauser
effects between the ester carbonyl resonance and a number of ring
protons, but none of these effects were diagnostic. 1H–13C HMBC
experiments showed no coupling between the carbonyl and H-3
(d 4.5); subsequent selective decoupling experiments showed that
all the couplings to the ring protons are less than 0.1 Hz, the only
significant coupling (ca. 1 Hz) being to the methyl protons.
BocNHOH and Bun N+ IO4 in CH2Cl2, the P:D ratios were
-
4
between 4:1 and 5:1 for 9b and 7:1 for 9c, strongly suggesting
an electrostatic effect. In general P and D isomers such as 10d
and 11d were very close-running in this series, but the proximal
adduct 10e from 9c was readily purified by crystallisation and
gave crystals suitable for a further X-ray structure determination,
Fig. 2. Increasing the steric bulk of the ester further, by using the
pivalate of 9a, had little further effect on the P:D ratio: see ESI†.
Fig. 2 Single crystal X-ray structure of p-nitrobenzoate 10e.
The p-nitrobenzoate ester 10e was clearly the proximal isomer
and displayed a half-chair structure, Fig. 2. We believe this is
the first X-ray structural determination for an intermolecular acyl
nitroso–diene adduct; a recent report22 detailed X-ray structures
for a regioisomeric pair of intramolecular ANDA adducts.
1
We therefore turned to H–15N HSQC and HMBC measure-
ments. Using the 15N-enriched sample of 4b, we prepared the 15
N
version of 5b, as we were confident this was a single regioisomer.25
The long-range HSQC spectrum of this product, run with a
coherence transfer delay of 31.25 ms, is shown in Fig. 4. The
optimum delay for detecting a given coupling J is nominally
1/(4J), corresponding here to a J of 8 Hz. In practice the
relationship between cross-peak intensity, J, and the delay is
complicated by relaxation, homonuclear multiplet structure and
field inhomogeneity, and the optimum delay for a given cross-
peak is usually significantly less than 1/(4J). Fig. 4 shows strong
responses for the three-bond 1H–15N interaction with the terminal
3. NMR Studies: a new criterion for distinguishing proximal and
distal adducts
It has been stated12 that, for a number of ANDA adducts derived
from unsymmetrical dienes, the regiochemistry was unambigu-
ously determined using HMBC and HSQC NMR data. Earlier
literature15c,18 relied on correlations of 13C NMR chemical shift
data to determine which ring carbon in the dihydrooxazine adducts
(see Fig. 3) was attached to N and which to O. In addition, it was
postulated that the dihydrooxazines adopted half-chair conforma-
tions such that H-6 (proximal isomer) and H-3 (distal isomer) were
axially oriented. One important consequence was that a relatively
large four-bond coupling, H-6 to H-4, was observed in P adducts
(D adducts were not considered). Nevertheless, this still requires a
well-defined conformational preference for certainty, particularly
where the relative signs of coupling constants are not determined,
and we sought long-range heteronuclear interactions, which could
provide a more general criterion that might encompass a wider
range of unsymmetrical ANDA adducts. We now report a new and
Fig. 4 1H–15N long-range HSQC NMR of 15N-enriched 5b (R = But),
Fig. 3 Numbering system of dihydrooxazines (both regioisomers).
using a coherence transfer delay of 31.25 ms.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4531–4538 | 4533
©

View Article Online
methyl group protons and the two-bond interaction with H-3,
and somewhat weaker cross-peaks for the three- and four-bond
interactions with H4 and H5 respectively. The interaction with
H-6 (5-bond through C and 3-bond through O) is extremely weak.
Armed with this information, we re-examined 10e using natural
abundance 1H–15N HSQC and were pleased to find clear-cut results
using experiments with a total data acquisition time of about
20 h and setting a range of coherence transfer delays (83.3, 31.25
in a gradient HMBC (gHMBC) experiment, which allowed the
NH coupling constants to be determined. Using a delay of 125 ms
(corresponding to a 4 Hz coupling) the only significant interactions
seen were with the Me protons and the H-4 olefinic H. A cross-
section of this experiment at the 15N frequency is shown (Fig. 6).
There is a clear splitting of the Me doublet into a double doublet
and a further doublet splitting of the olefinic proton multiplet:
both heteronuclear coupling constants JNH were approximately
3 Hz.
1
and 20.8 ms) corresponding to H–15N couplings of J = 3, 8
and 12 Hz (Fig. 5; the spectrum shown was run at a delay of
31.25 ms, nominally J = 8 Hz). There was a clear splitting of the
Me doublet into a double doublet and further interactions with the
olefinic protons. There was some variation in response for cross-
peaks with the olefinic Hs: at 83.3 ms the strongest response was
from H-5 whereas at 31.25 ms (as shown) the stronger response
was from H-4 and a large coupling to methine proton H-3 was
observed. At 20.8 ms a cross-peak for the N coupling to H-4 was
seen but none for H-3. Nevertheless, a consistent coupling to the
Me protons was seen in all three experiments, with no evidence of
coupling to H-6 or the CH2 group. We obtained a clearer result
We applied these methods to a more stringent test, namely
the assignment of the proximal/distal mixture of 10d and 11d.
Here, as 11d comprises just 20% of the total product (vide supra),
long acquisition was essential but after 60 h using a coherence
transfer delay of 20.8 ms the minor isomer peaks were sufficiently
strong and the 1H–15N HSQC spectrum showed a firm distinction
between the major and minor isomer peaks (Fig. 7). The clearest
indication is given by the strong interaction between 15N and
the Me protons in the proximal isomer 10d (this corresponds
to the situation in 4b and 10e), whereas the distal isomer 11d
showed no interaction between 15N and the Me protons but a
strong interaction, again presumably 3-bond, between 15N and the
non-equivalent CH2 protons. We had already confidently assigned
the CH2 signal of the minor isomer 11d to the region d 4.10–
4.30, overlapping with the major isomer CH2. It is also clear
that the coupling from N to both olefinic Hs is significant here
for each regioisomer, with the minor/major pairs being correctly
distinguished (e. g. d 5.80, 5.90 for the minor isomer olefinic Hs).
In the preceding example, the HSQC interaction differentiated
between nearby CH2 and CH3 groups. We now demonstrate that,
in a case where a near-symmetrical diene is used and the two
ANDA adducts are formed in almost equal amounts, two CH2
groups can be clearly distinguished. Thus the distal/proximal pair
12 (37% of mixture) and 13 (42%) were prepared and separated.26
Fig. 5 1H–15N long-range HSQC NMR of natural 15N abundance 10e
(aryl hydrogens excluded) using a coherence transfer delay of 31.25 ms,
equivalent to 8 Hz.
Fig. 6 (Top) F2 cross-section through the 500 MHz gHMBC spectrum of 10e at the 15N chemical shift; (bottom) full proton spectrum of 10e.
4534 | Org. Biomol. Chem., 2009, 7, 4531–4538 This journal is The Royal Society of Chemistry 2009
©

View Article Online
1H–15N HMBC/HSQC is an excellent method: differing 3-bond
connectivities in the regioisomers are clearly distinguished. Finally,
we have demonstrated the synthetic value of the sorbate ester
adduct 5a by conversion to a single diastereoisomer of an alkylated
ornithine analogue 8 in three steps: a highly efficient Mitsunobu
reaction of phthalimide with an a-hydroxylactam is a notable
feature of this sequence.
Experimental
1H NMR spectra were recorded for CDCl3 solutions (unless
otherwise stated) on a Bruker 500 Avance II+ (500 MHz), a Bruker
400 Avance (400 MHz), a Bruker AMX 200 (200 MHz) or a
Bruker AMX 250 (250 MHz) spectrometer; 13C NMR spectra
were recorded on a Bruker 400 Avance (at 100 MHz) and a
Bruker Avance 500 MHz instrument was used for some of the
1H–15N correlations, notably Fig. 6. Chemical shifts are recorded
(dH, dC) in ppm and coupling constants (J) are recorded in hertz
(Hz). Chemical shifts were referenced to residual non-deuterated
solvent present in the sample, e.g. CHCl3 in CDCl3, or to a
calibration reference of tetramethylsilane (TMS) incorporated in
the sample. Proton (1H) spectra were assigned using COSY data
(where necessary) while carbon (13C) spectra were assigned using
HMQC data. Proximal and distal regioisomers of ANDA reaction
products were assigned using a combination of various NMR
techniques, including 1H–15N HMBC, and X-ray crystallography;
vide supra. Elemental analyses were performed in the University of
Liverpool microanalysis laboratory by Mr. S. Apter. Mass spectra
were recorded on a VG analytical 7070E machine or a Fison
TRIO spectrometer using electron ionisation (EI) or chemical
ionisation (CI). Infrared spectra were recorded on a JASCO
4100 type A FT/IR spectrometer in the range of 4000–600 cm-1.
Samples were either a liquid film or solid sample. All absorptions
are reported as wave numbers, n (cm-1). Melting points were
measured using a Stuart melting point apparatus (SMP3 model)
and are uncorrected. Thin layer chromatography was carried out
on Merck 5 ¥ 2 cm aluminium-backed plates with a 0.2 mm layer
of Kieselgel 60 F254with visualisation by KMnO4 stain, ninhydrin
or anisaldehyde. Preparative column chromatography was carried
out using ICN (230–400 mesh) silica gel under bellows pressure.
Fig. 7 The 1H–15N long-range HSQC spectrum of a mixture of 10d and
11d (4:1) acquired over 60 h with a coherence transfer delay of 42 ms,
equivalent to 12 Hz.
1
Here a H–15N HSQC experiment essentially the same as that
used for 10d/11d (Fig. 7), but with an acquisition time of 16 h,
showed a clear three-bond interaction between the ring nitrogen
N-2 and the C-8 methylene protons (viz. CH2OH), but no inter-
action to the C-7 methylene protons (CH2NHBoc) in compound
12. (See ESI†; note that N-2 is easily distinguished from the other
nitrogen by its chemical shift of -220 ppm). This is consistent with
Fig. 7 where only the distal isomer 11d showed such a 3-bond
interaction; no interaction between N-2 and CH2OH is expected
in compound 13, cf. 10d. This result has also been independently
confirmed by X-ray analysis of the bis-Boc analogue of 12. In the
1D proton spectrum of 12, the olefinic protons are overlapping (H-
3 and H-6 are also partially overlapping) and we do not believe a
confident assignment could be made by the older criterion,18 viz.
lower-field olefinic H diametrically opposite ring oxygen.
General procedure for synthesis of N-alkoxycarbonyl
hydroxylamines
Conclusions
A suspension of hydroxylamine hydrochloride (1.5N mmol) and
potassium carbonate (0.75N mmol, 1.5 equiv.) in a 1:1 ether–water
mixture was stirred for approx. 1 h with the evolution of CO2.
Benzyl chloroformate (1.0N mmol) or di-tert-butyl dicarbonate
We have conducted a study of the acyl nitroso Diels–Alder reaction
between a series of unsymmetrical diene esters and alcohols and
alkoxycarbonyl nitroso compounds. High yields were obtained
straightforwardly using appropriate periodate oxidants, though
careful attention to detail was needed, with NaIO4/aq. MeOH
◦
(1.0N mmol) was added portionwise at 0 C and the suspension
was stirred at r.t. for 5–16 h. The solids were filtered from the
organic layer and the aqueous layer washed with ether. The
combined organic layers were then dried (MgSO4), filtered and the
organic layer evaporated to give a colourless liquid that crystallised
overnight. If necessary, recrystallisation (cyclohexane–toluene)
gave a pure product.
giving significantly better yields on sorbate esters whereas Bun N+
4
IO4- in CH2Cl2 was the reagent of choice for sorbic alcohol and its
acyl derivatives. Sorbate esters gave essentially complete preference
for the proximal regioisomeric product, but sorbic alcohol and its
acyl derivatives gave significant variation, though the proximal
adduct was always favoured. We have additionally obtained a
single-crystal X-ray structure for a proximal adduct (10e). For
definitive assignment of regioisomers, especially in cases where a
significant mixture results, we have found that natural abundance
N-Benzyloxycarbonylhydroxylamine 4a. Yield, 91% on a scale
of 129 mmol hydroxylamine hydrochloride, mp (from toluene–
◦
cyclohexane) 68–70 C (lit.27 mp 62–64 ◦C from ether–hexane);
dH (200 MHz): 5.19 (2H, s, CH2Ph), 5.30 (1H, s, OH), 6.82 (1H,
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4531–4538 | 4535
©

View Article Online
bs, NH) and 7.20–7.40 (5H, m, ArH). (Lit.27 dH 5.19 (2 H, s,
CH2Ph), 5.89 (1H, s), 7.21 (1H, s) and 7.36 (5 H, s).
CHCHCO2Et), 6.00 (1H, m, CHCHCH3), 7.40 (5H, m, ArH);
=
=
dC (100 MHz, CDCl3): 14.5 (CH3CH2, 1C), 18.4 (CH3CH, 1C),
50.9 (CHCH3, 1C), 62.2 (CH2CH3, 1C), 68.1 (CH2Ph, 1C), 76.4
(CHCO2Et, 1C), 122.4 (CHCHCO2Et, 1C), 127.4 (CHCHCH3,
1C), 128.0 (Ph, 1C), 128.5, 128.7 (¥2), 129.0 (Ph, 4C), 136.3
(Ph, 1C), 155.1 (NC(O)OCH2Ph, 1C) and 167.6 (C(O)OEt, 1C);
N-tert-Butoxycarbonyl hydroxylamine 4b. Yield, 84% on a
scale of 136 mmol hydroxylamine hydrochloride, mp 52–54 ◦C
◦
(from ether) (lit.23 mp 58–59 C, pet. ether); dH (200 MHz): 1.50
(9 H, s, CMe3), 6.35 (1H, s, OH) and 7.00 (1H, s, NH). (Lit.16 dH
1.48 (9 H, s, CMe3) and 6.55, 7.01 (2 H, 2bs, NHOH).
n
max.(film) 3033, 2979, 1731, 1706, and 698 cm-1; m/z (CI, NH3):
323 ([M + NH4]+, 13%), 288 (100%), 172 ([M - PhCH2OC(O) +
H]+, 44%), 169 (34%), 108 (29%). Found (CI, NH3) m/z 323.16040,
C16H23N2O5 ([M + NH4]+) requires 323.16071.
15N-tert-Butoxycarbonyl hydroxylamine 15N-4b. This was ob-
tained on a 5 mmol scale by the above procedure but using
1.25 eq of both hydroxylamine and base as a colourless liquid
that crystallised overnight. Recrystallisat◦ion gave pure product
as a clear solid (0.56 g, 73%), mp 57–58 C (from cyclohexane–
toluene); dH (400 MHz): 1.50 (9 H, s, CMe3), 6.80 (1H, s, OH) and
7.00 (1H, s, NH). [Lit.16 dH (CDCl3): 1.48 (9 H, s, CMe3,), 6.55,
7.01 (2 H, 2 br s, NHOH)]; dC (100 MHz): 28.6 (CMe3, 3C), 82.6
2-tert-Butyl 6-ethyl 3-methyl-3,6-dihydro-2H-1,2-oxazine-2,6-
dicarboxylate 5b
Yield, 47% (method 1). Found: C, 57.20; H, 7.86; N, 5.09.
C13H21NO5 requires C, 57.55; H, 7.80; N 5.16; dH (200 MHz,): 1.30
(3H, t, J = 7.1 Hz, CH3CH2), 1.37 (3H, d, J = 6.0 Hz, CH3CH,),
1.50 (9 H, s, CMe3), 4.26 (2 H, q, J = 7.1 Hz, CH2CH3), 4.45 (1H,
m, CHCH3), 5.15 (1H, d, J = 1.6 Hz, CHCO2Et,), 5.90 (1H, m,
=
(CMe3, 1C), 159.2 and 160.0 (d, J = 8 Hz, C O); m/z (CI, NH3):
152 ([M + NH4]+, 6%), 134 ([M]+, 100%), 117 ([M - OH]+, 98%), 74
([M - C(O)NHOH + H]+, 31%), 58 ([M - OC(O)NHOH]+, 10%).
Found (CI, NH3): m/z, 152.10578; C5H1515NNO3 ([M + NH4]+)
requires 152.10530.
=
=
CHCHCO2Et) and 5.98 (1H, m, CHCHCH3); dC (100 MHz):
14.5 (CH3CH2, 1C), 18.3 (CH3CH, 1C), 28.7 (CMe3, 3C), 50.8
(CHCH3, 1C), 62.1 (CH2CH3, 1C), 75.9 (CHCO2Et, 1C), 82.3
=
=
(CMe3, 1C), 122.1 ( CHCHCO2Et, 1C), 131.1 ( CHCHCH3,
1C), 154.5 (NC(O)OC(Me)3, 1C) and 167.8 (CHC(O)OEt, 1C).
Found (CI, NH3): m/z, 289.17633; C13H25N2O5 ([M + NH4]+)
requires m/z, 289.17635.
General methods for the acyl nitroso Diels–Alder reactions
Method 1. To a mixture of the diene (1 equiv.) and hy-
droxylamine (1 equiv.) in 1:1 MeOH–H2O was slowly added
drop◦wise, with stirring, a solution of NaIO4 (1 equiv.) in H2O
at 0 C. After addition the reaction was left to stir for 30 min at
the same temperature during which time a heavy precipitate of
NaIO3 appeared. The reaction was allowed to warm to 20 ◦C
and stirred for a further 40 min. Water and ether were then
added and the layers separated. The aqueous layer was washed
again with ether and the combined ethereal layers were washed
with satd. aq. NaHCO3 (¥ 3) then brine. After drying (MgSO4),
filtering and concentration in vacuo, the product was purified by
flash chromatography (ethyl acetate–hexane) to give pure product.
For additional variants, see notes with Tables 1 and 2 in the
ESI†.
2-tert-Butyl 6-ethyl 3-methyl-3,6-dihydro-2H-1,2-oxazine-2,6-
dicarboxylate 15N-5b
Yield, 27% (method 1). Found: C, 57.20; H, 7.86; N, 5.09; m/z
290.17271. C13H21NO5 requires C, 57.55; H, 7.80; N, 5.16%;
C13H2515NO5N ([M + NH4]+) requires 290.17228; dH (200 MHz):
1.30 (3H, t, J = 7.1 Hz, CH3CH2), 1.37 (3H, dd, J = 6.0 and
1 Hz, CH3CH15N), 1.50 (9 H, s, CMe3), 4.26 (2 H, q, J = 7.1 Hz,
CH2CH3), 4.45 (1H, m, CHCH3), 5.15 (1H, d, J = 1.6 Hz,
=
CHCO2Et), 5.90 (1H, m, CHCHCO2Et) and 5.98 (1H, m,
=
CHCHCH3); dC (100 MHz): 14.5 (CH3CH2, 1C), 18.3 (CH3CH,
1C), 28.7 (CMe3, 3C), 50.7 and 50.8 (CHCH3, 1C, J = 7), 61.6
(CH2CH3, 1C), 76.0 (CHCO2Et, 1C), 82.3 (CMe3, 1C), 122.4,
Method 2. To a solution of the diene (1 equiv.) and hydroxy-
◦
=
=
122.5 ( CHCHCO2Et, 1C) and 131.1 ( CHCHCH3, J = 2,
1C), 154.4 and 154.7 (NC(O)OC(Me)3, J = 21, 1C) and 167.9
(CHC(O)OEt, 1C); m/z (CI, NH3): 290 ([M + NH4]+, 66%), 273
([M + H]+, 15%), 216 ([M - But + H]+, 18%), 199 ([M - CO2Et]+,
12%), 172 ([M - Boc + H]+, 36%) and 157 ([M - O15NBoc + NH3]+,
36%).
lamine (1 equiv.) in CH2Cl2 at 0 C was slowly added dropwise,
with stirring, Bun N+ IO4 (0.5 equiv.) in CH2Cl2. After addition,
-
4
the reaction was purged with N2 and left to stir for 2.5 h at 20 ◦C
under N2 during which time a yellow-orange colour appeared.
Ethyl acetate and water were added and the layers separated.
The aqueous layer was washed again with ethyl acetate and
the combined organic layers washed with satd. aq. Na2S2O3,
satd. aq. NaHCO3 (¥ 2) then brine. After drying over MgSO4,
filtering and concentration in vacuo, the product was purified by
flash chromatography (ethyl acetate–hexane) to give pure product.
For additional variants, see notes with Tables 1 and 2 in the
ESI†.
(3R,6R)/(3S,6S)-3-Hydroxy-6-methylpiperidin-2-one 66
10% Pd/C (2 g) was added to a solution of dihydrooxazine 5a
(5.72 g, 18.7 mmol) in anhydrous THF (20 mL) and the mixture
was hydrogenated under H2 (30 atm). The catalyst was filtered
and washed with THF and PriOH, then the combined filtrate
and washings were evaporated and the residue recrystallised from
THF to afford the product 6 as a solid (1.57 g, 65%), mp 151–
153 ◦C. Found: C, 55.7; H, 8.6; N, 10.9; m/z 130.08660; C6H11NO2
requires C, 55.8; H, 8.6; N, 10.8%; C6H12NO2 (MH+) requires m/z,
130.08681; nmax.(cm-1) 3315, 3178, 2929 and 1650; dH (CD3OD,
400 MHz): 1.15 (3H, d, J = 6.5 Hz, CH3CH), 1.35, 1.60, 1.90 and
2.05 (4H, 4 m, CH2CH2), 3.40 (1H, m, CH3CHN) and 3.90 (1H,
2-Benzyl 6-ethyl 3-methyl-3,6-dihydro-2H-1,2-oxazine-2,6-
dicarboxylate 5a
Yield, 85% (method 1, 2 eq. of ZNHOH). dH (200 MHz): 1.31
(3H, t, J = 7.2, CH3CH2), 1.38 (3H, d, J = 6.9, CH3CH),
4.27 (2 H, q, J = 7.2, CH2CH3), 4.54 (1H, m, CHCH3), 5.15
(1H, m, CHCO2Et), 5.20 (2 H, ABq, OCH2Ph), 5.90 (1H, m,
4536 | Org. Biomol. Chem., 2009, 7, 4531–4538
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
m, CH2CHO); dC (CD3OD, 100 MHz): 23.2, 30.7 (¥2), 51.0, 69.1
Benzyl 6-(hydroxymethyl)-3-methyl-3,6-dihydro-2H-1,2-oxazine-
and 175.9; m/z (CI, NH3) 147 (MNH4 , 100%), 130 (MH+, 93%)
2-carboxylate (proximal and distal isomers) 10a, 11a28
+
and 114 (29%).
Yield, 87% (method 2). For 10a: Found: C, 64.0; H, 6.55; N,
5.3. C14H17NO4 requires C, 63.9; H, 6.5; N, 5.3%; dH (400 MHz):
1.35 (3H, d, J = 6.7 Hz, CHCH3), 3.65 and 3.77 (2H, dd, J =
12.4 and 3.0 Hz, J = 12.4 and 6.5 Hz, CH2OH), 4.52 (1H, m,
CHCH3), 4.70 (1H, m, CHCH2OH), 5.20 (2H, ABq, CH2Ph),
2-[(3R,6S)/(3S,6R)-6-Methyl-2-oxopiperidin-3-yl]isoindoline-1,3-
dione 7
i
=
(Pr O2CN )2 (1.07 g, 5.30 mmol) in THF (5 mL) was added to
=
5.70 (1H, dt, J = 10.3 and 1.6 Hz, CHCHCH2OH), 5.91 (1H,
a solution of lactam 6 (0.65 g, 5.03 mmol), phthalimide (0.82 g,
5.60 mmol) and Ph3P (1.45 g, 5.53 mmol) in THF (10 mL) with
stirring under N2 at 20 ◦C. After 5 days the solid was filtered off,
then the mother liquors were concentrated to yield a further crop;
recrystallisation of the combined solids from THF afforded pure
product 7 (0.85 g, 66%), mp 55–56 ◦C. Found: C, 64.9; H, 5.4;
N, 10.8; m/z, 259.10737; C14H14N2O3 requires C, 65.1; H, 5.5; N,
10.9%; C14H15N2O3 (MH+) requires m/z, 259.10827; nmax.(cm-1)
3182, 2962, 1709, 1666 and 1387; dH (400 MHz): 1.40 (3H, d, J =
6.6 Hz, CH3CH), 1.85, 2.00, 2.10 and 2.55 (4H, 4 m, CH2CH2),
3.75 (1H, m, CH3CHN), 4.75 (1H, m, CH2CHNPhth), 5.90 (1H,
br, NH), 7.70 and 7.85 (4H, 2 m, ArH); dC (100 MHz): 22.2, 22.4,
27.8, 47.8, 48.8, 123.5, 132.1, 134.1, 167.7 and 167.8; m/z (CI,
=
ddd, J = 10.3, 4.5 and 2.3 Hz, CHCHCH3) and 7.35 (5H,
m, Ph); dC (100 MHz): 18.7 (CHCH3, 1C), 51.0 (CHCH3, 1C),
64.0 (CH2OH, 1C), 68.0 (CH2Ph, 1C), 79.4 (CHCH2OH, 1C),
=
124.2 ( CHCHCH2OH, 1C), 127.4, 128.0, 128.4, 128.7, 128.9,
=
129.0 and 131.7 (ArC and CHCH CHCH, 7C; both proximal
=
and distal), 130.6 ( CHCHCH3, 1C), 136.4 (ArC, 1C) and
155.3 (NC(O)OCH2Ph, 1C); MS (m/z, CI): 281 ([M + NH4]+,
40%), 264 ([M + H]+, 13%), 173 ([M - CH2Ph + H]+, 8%), 114
([M - NC(O)CH2Ph]+, 27%), 98 ([M - ONC(O)CH2Ph]+, 66%).
Found (CI, NH3): m/z, 281.14958; C14H21N2O4 ([M + NH4]+)
requires 281.15013. The distal isomer 11a was distinguished by
dH (400 MHz): 1.25 (3H, d, J = 6.7 Hz, CH3CH) and 5.75–5.85
=
(2H, m, CHCH CHCH); dC (100 MHz): 19.2 (CH3CH, 1C),
+
NH3) 259 (MH+, 100%), 276 (MNH4 , 51%) and 518 [2(MH+),
25%].
63.7 (HOCH2CH, 1C) and 68.2 (CH2Ph, 1C); as noted above, the
olefinic carbons of the distal isomer cannot be distinguished with
certainty from the Ar carbons.
(2S,5R)/(2R,5S)-2,5-Diaminohexanoic acid dihydrochloride 8
(5-methylornithine dihydrochloride)
tert-Butyl 6-(acetoxymethyl)-3-methyl-3,6-dihydro-2H-1,2-
oxazine-2-carboxylate (proximal and distal isomers) 10d, 11d
Phthalimide 7 (0.102 g, 0.395 mmol) was heated at reflux in
6 M HCl (10 mL) for 5 h. The solution was cooled, washed
with ether (2 ¥ 5 mL) and EtOAc (1 ¥ 5 mL) and the aqueous
layer was concentrated to dryness, affording the dihydrochloride
8 as a yellow powder (0.086 g, quant.). Found: m/z, 293.2202;
C12H29N4O4 [(2M + H)+] requires m/z, 293.2189; nmax.(cm-1) 3398,
2897, 1720, 1589, 1384 and 1223; dH (CD3OD, 400 MHz): 1.35
(3H, d, J = 6.6 Hz, CH3CH), 1.85, 2.05 (4H, 2 m, CH2CH2), 3.40
(1H, m, CHCH3) and 4.05 (1H, t, J = 6.1 Hz, CH2CHN); dC
(CD3OD, 100 MHz): 18.9, 28.0, 31.7, 49.0, 53.9 and 171.7; m/z
(ES +ve mode) 147 (MH+) and 293 [(2M + H)+].
Yield, 89% (method 2). For 10d, found: dH (400 MHz): 1.33
(3H, d, CH3CH, J = 6.7 Hz), 1.50 (9 H, s, CMe3), 2.10 (3H, s,
CH3CO), 4.13 and 4.25 (2H, m, CH2OAc), 4.46 (1H, m, CHCH3),
4.79 (1H, m, CHCH2OAc), 5.66 (1H, dt, J = 10.3 and 1.6 Hz,
=
=
CHCHCH2OAc), 5.95 (1H, ddd, J = 10.3, 4.6 and 2.3 Hz,
CHCHCH3); dC (100 MHz): 18.4 (CH3CH, 1C), 21.2 (CH3CO,
1C), 28.7 (Me3C), 50.6 (CHCH3, 1C), 64.7 (CH2OAc), 75.6
=
(CHCH2OAc, 1C), 81.9 (CMe3, 1C), 123.3 ( CHCHCH2OAc,
=
1C), 131.5 ( CHCHCH3, 1C), 154.6 (NC(O)OCH2Ph, 1C) and
171.3 (CH2OC(O)CH3, 1C); m/z (CI, NH3): 289 ([M + NH4]+,
3%), 272 ([M + H]+, 1%), 172 ([M - Boc + H]+, 11%), 156 ([M -
OAc - But]+, 26%), 140 ([M - OAc - OBut]+, 15%), 112 ([M - OAc -
Boc]+, 7%). Found (CI+) 289.17578, C13H25N2O5 ([M + NH4]+)
requires 289.17635. The distal isomer 11d was distinguished by
dH (400 MHz): 1.27 (3H, d, CH3CH, J = 6.7 Hz), 2.05 (3H, s,
CH3CO), 4.65 (1H, m, CHCH3), 5.78 (1H, ddd, J = 10.4, 4.3 and
(2E,4E)-Hexa-2,4-dienyl-4-nitrobenzoate 9c
4-Nitrobenzoyl chloride (2.04 g, 11 mmol) was added in one
portion to a solution of (2E,4E)-hexa-2,4-dienol (0.98 g, 10 mmol)
and pyridine (1.2 mL) in CH2Cl2 (15 mL), which was stirred under
N2 at 0 ◦C and protected from light. After 1 h the mixture was
diluted with EtOAc (70 mL) and washed sequentially with 3 M
HCl (3¥), water, satd. aq. NaHCO3 soln. and brine. After drying
over MgSO4 the solution was evaporated to give crude product
as a solid (2.50 g, approx. quant.), which was sufficiently pure for
further use; an analytical sample was obtained by chromatography,
eluting with 30% EtOAc–hexane, giving 9c as a rather light-
sensitive white solid (1.29 g, 52%), mp 57–59 ◦C. Found: C,
63.1; H, 5.35; N, 5.6; m/z, 270.0741. C13H13NO4 requires C,
63.2; H, 5.3; N, 5.7%; C13H13NO4Na requires m/z, 270.0742;
dH (200 MHz): 1.78 (3H, d, J = 6.4 Hz, CH3CH), 4.85 (2H,
d, J = 6.7 Hz, CH2CH), 5.80 (2H, m, 2-H + 5-H), 6.10 (1H,
m, 4-H), 6.36 (1H, dd, J = 15.1 and 10.2 Hz, 3-H) and 8.25
(4H, dd, ArH); m/z (ES +ve mode) 270 (MNa+, 100%), 517
(2M + Na+, 72%).
=
2.1 Hz, CHCHCH3), and 5.86 (1H, dt, J = 10.3 and 1.6 Hz,
=
CHCHCH2OAc); dC (100 MHz, CDCl3): 19.1 (CH3CH, 1C),
63.9 (CH2OAc, 1C) and 73.4 (CHCH2OAc, 1C) with other signals
overlapping.
tert-Butyl 6-[(4-nitrobenzoyloxy)methyl]-3-methyl-3,6-dihydro-
2H-1,2-oxazine-2-carboxylate (proximal and distal isomers)
10e, 11e
Yield, 76% (method 2). The major (proximal) isomer was
isolated by crystallization from EtOAc–hexane: 10e, mp 118–
119 ◦C. Found: C, 57.05; H, 5.85; N, 7.4. C18H22N2O7 requires
C, 57.1; H, 5.8; N, 7.4%; dH (400 MHz): 1.35 (3H, d, J =
6.7 Hz, CH3CH), 1.50 (9 H, s, Me3C), 4.43 and 4.57 (2H, 2dd,
J = 12.2 and 6.9 Hz; 12.2 and 2.7 Hz; CHCH2OCOAr), 4.50
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4531–4538 | 4537
©

View Article Online
(1H, m, CHCH3), 4.95 (1H, m, CHCH2OCOAr), 5.75 (1H, dt,
11 (a) G. Calvet, M. Dussausois, N. Blanchard and C. Kouklovsky, Org.
Lett., 2004, 6, 2449–2451; (b) G. Calvet, R. Guillot, N. Blanchard and
C. Kouklovsky, Org. Biomol. Chem., 2005, 3, 4395–4401; in the latter
example Kouklovsky eventually concluded that, although there was
genuine Lewis acid catalysis especially by Zn(OTf)2, a better product
profile was obtained using added water only.
=
J = 10.5 and 1.5 Hz, CHCHCH2OCOAr), 6.02 (1H, ddd, J =
=
10.5, 4.6 and 2.3 Hz, CHCHCH3) and 8.03 (4H, dd, ArH); dC
(100 MHz): 18.4 (CH3CH, 1C), 28.7 (CMe3, 3C), 50.7 (CH3CH,
1C), 65.8 (CH2OAc, 1C), 75.6 (CHCH2OAc, 1C), 82.1 (CMe3),
12 G. Calvet, N. Blanchard and C. Kouklovsky, Synthesis, 2005, 3346–
=
123.0 ( CHCHCH2OCOR), 123.9, 131.3 (ArC, each 2C), 131.9
3354.
=
( CHCHCH3), 135.5, 151.0 (ArC), 154.5 (NC(O)OCH2Ph, 1C)
13 A. Hall, P. D. Bailey, D. C. Rees and R. H. Wightman, Chem. Commun.,
1998, 2251–2252.
+
and 164.8 (CH2OC(O)Ar, 1C); m/z (CI, NH3) 396 (MNH4 ,
14 S. F. Martin, M. Hartmann and J. A. Josey, Tetrahedron Lett., 1992,
100%). The distal isomer 11e was distinguished by dH (400 MHz):
33, 3583–3586.
=
1.26 (3H, d, J = 6.7 Hz, CH3CH), 5.80–5.95 (2H, m, 2 ¥ CH)
15 (a) H. Felber, G. Kresze, R. Prewo and A. Vasella, Helv. Chim. Acta,
1986, 69, 1137–1146(a-Cl nitroso dienophile used); (b) T. Sheradsky
and E. R. Silcoff, Molecules, 1998, 3, 80–87; (c) Sorbic acid itself reacts
with an a-Cl nitroso dienophile to give only the proximal adduct: A.
Defoin, H. Sarazin, T. Sifferlen, C. Strehlerr and J. Streith, Helv. Chim.
Acta, 1998, 81, 1417–1428.
and 8.30 (4H, dd, ArH, partially overlapping); no 13C NMR was
determined for this isomer in view of the very small amount.
Acknowledgements
16 S. F. Martin, M. Hartmann and J. A. Josey, Tetrahedron Lett., 1993,
34, 2852.
17 P. Bach and M. Bols, Tetrahedron Lett., 1999, 40, 3461–3464.
18 A. Defoin, H. Fritz, G. Geffroy and J. Streith, Helv. Chim. Acta, 1988,
71, 1642–1658.
19 As employed on a similar lactam: (a) T. Beisswenger and F. Effenberger,
Chem. Ber., 1984, 117, 1513–1522; (b) U. Kraatz, W. Hasenbrink, H.
Wamhoff and F. Korte, Chem. Ber., 1971, 104, 2458–2466.
20 M. X. Li, T. Nakajima, T. Fukushige, K. Kobayashi, N. Seiler and T.
Saheki, Biochim. Biophys. Acta, 1999, 1455, 1–11.
We are grateful to the EPSRC for funding (grants EP/D05592X,
EP/E057888 to GAM, EP/C005643/1 to JAI, DTA award to LB)
and to Drs. John Harding and Ryan Bragg of Astra Zeneca for a
kind donation of 15NH2OH·HCl.
References and notes
1 O. Wichterle, Collect. Czech. Chem. Commun., 1947, 12, 292–304.
2 (a) B. T. Shireman and M. J. Miller, J. Org. Chem., 2001, 66, 4809–4813;
(b) B. T. Shireman, M. J. Miller, M. Jonas and O. Wiest, J. Org. Chem.,
2001, 66, 6046–6056.
3 (a) A. G. Pepper, G. Procter and M. Voyle, Chem. Commun., 2002,
1066–1067; (b) S. B. King and B. Ganem, J. Am. Chem. Soc., 1994, 116,
562–570.
4 A. G. Leach and K. N. Houk, J. Org. Chem., 2001, 66, 5192–5200.
5 D. L. Boger, M. Patel and F. Takusagawa, J. Org. Chem., 1985, 50,
1911–1916.
6 B. Belleau and Y.-K. Au-Young, J. Am. Chem. Soc., 1963, 85, 64–71.
7 G. W. Kirby and J. G. Sweeny, J. Chem. Soc., Perkin Trans. 1, 1981,
3250–3254.
8 J. E. T. Corrie, G. W. Kirby and J. W. M. Mackinnon, J. Chem. Soc.,
Perkin Trans. 1, 1985, 883–886.
9 J. E. Baldwin, D. J. Aldous, C. Chan, L. M. Harwood, I. A. O’Neil and
J. M. Peach, Synlett, 1989, 9–14.
10 (a) T. Kawabata, K. Itoh and T. Hiyama, Tetrahedron Lett., 1989, 30,
4837–40; (b) M. Joubert, A. Defoin, C. Tarnus and J. Streith, Synlett,
2000, 1366–1368.
21 G. Davies, A. T. Russell, A. J. Sanderson and S. J. Simpson, Tetrahedron
Lett., 1999, 40, 4391–4394.
22 S. M. Sparks, C. P. Chow, L. Zhu and K. J. Shea, J. Org. Chem., 2004,
69, 3025–3035.
23 L. A. Carpino, C. A. Griza and B. A. Carpino, J. Am. Chem. Soc.,
1959, 81, 955–957.
24 A. Bax and R. Freeman, J. Magn. Reson., 1981, 44, 542–561.
25 None of the sorbate ester series adducts could be suitably crystallized:
unambiguous proof of the regiochemistry of 5a was obtained by
chemical correlation, namely the conversion of 5a to 7 (Scheme 3).
26 Details of the synthesis of compounds 12 and 13 will be published
elsewhere, together with the X-ray crystal structure of the bis-Boc
analogue of 12. For characterization of 12 and 13, in particular NMR
data, see the ESI†.
27 R. W. Darbeau, G. A. Trahan and L. M. Siso, Org. Biomol. Chem.,
2004, 2, 695–700.
28 Only the P isomer is named, for simplicity, where P/D pairs were
obtained. The D isomer 11a is: Benzyl 3-(hydroxymethyl)-6-methyl-
3,6-dihydro-2H-1,2-oxazine-2-carboxylate—similarly for 11d, 11e and
examples in the ESI†.
4538 | Org. Biomol. Chem., 2009, 7, 4531–4538
This journal is
The Royal Society of Chemistry 2009
©