Tandem addition/cyclization of alkynylzinc reagents to enantiopure 2-tert-butyl-3,5-dimethyl-2,3-dihydroimidazol-4-one N-oxide: Potential precursors of quaternary α-amino acids

Article abstract of DOI:10.1002/ejoc.200400861

A new enantiopure cyclic nitrone, a potential electrophilic alanine synthon, has been prepared. Its reaction with alkynylzinc reagents led to a tandem addition/cyclization reaction with complete regio- and stereoselectivity. The adducts are potential prec

Full text of DOI:10.1002/ejoc.200400861

FULL PAPER
Tandem Addition/Cyclization of Alkynylzinc Reagents to Enantiopure 2-tert-
Butyl-3,5-dimethyl-2,3-dihydroimidazol-4-one N-Oxide: Potential Precursors of
Quaternary α-Amino Acids
Frédéric Cantagrel,[a] Sandra Pinet,[a] Yves Gimbert,[a] and Pierre Y. Chavant*[a]
Keywords: Nitrones / Alkynylzinc / Amino acids / Imidazolidinone / Dihydroisoxazole / Cyclization
A new enantiopure cyclic nitrone, a potential electrophilic
alanine synthon, has been prepared. Its reaction with alkyn-
ylzinc reagents led to a tandem addition/cyclization reaction
with complete regio- and stereoselectivity. The adducts are
potential precursors of quaternary α-amino acids.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
Introduction
As part of a program involving the use of modified pep-
tides featuring some hydroxamic linkages, we were inter-
ested in the preparation of N-hydroxy-α-amino acid deriva-
tives.[1] For this purpose, we sought to add organometallic
reagents onto properly designed, enantiopure nitrones.
Nitrones are activated imine derivatives that are easily ac-
cessible through the oxidation of secondary amines.[2–11]
Thus, we planned to choose our starting material from
among the secondary amines that have been proposed as
precursors of α-amino acids so that our nitrone precursor
would be produced in a single oxidation step. The imid-
azolidinones 1 designed by Seebach et al.[12] were particu-
larly attractive as starting materials. Enantiopure imid-
azolidinones 1 are easily accessible from simple α-amino ac-
ids. After N-benzoylation, these molecules were used in the
most famous application of the self-reproduction of stereo-
chemistry (SRS) principle.[12] It was easy to imagine that
imidazolidinones 1 could be oxidized to cyclic nitrones 2
(Scheme 1), which could then be used as electrophiles, in
another variant of the SRS principle.
Scheme 1. Nitrone 2 as an electrophilic precursor of amino acids.
reported the first synthesis of 2a (R = H) and its use in the
preparation of nonproteinogenic amino acids by 1,3-dipolar
cycloaddition reactions. All these studies dealt only with
unsubstituted nitrones (R = H, from glycine).
We were mainly interested in reactions of such nitrones
with organometallic reagents. In particular, we treated
nitrones 2 with the alkynylzinc intermediates that we had
developed previously[20,21] (Scheme 1).
In this article we describe the reaction of the new nitrone
2b with 1-alkynes and dimethylzinc. Instead of yielding pro-
pargylic N-hydroxylamines as in the former work, under
mild conditions this reaction proceeded further to 2,3-dihy-
droisoxazoles in high yields and diastereoselectivities.
Analogous cyclic nitrones already exist. Altenbach and
co-workers[13,14] have prepared a menthone-derived spiro
compound in both enantiopure forms. This nitrone was
treated with simple Grignard reagents to produce N-hy-
droxylamines. Other enantiopure cyclic α-acyl nitrones[15–18]
have also been proposed for use as substrates in 1,3-dipolar
cycloaddition reactions. Very recently, Baldwin and Long[19]
Results and Discussion
Preparation of Nitrones 2a,b
We prepared imidazolidinones 1a[22] and 1b[23] according
to the published procedures: the enantiomers of 1a were
obtained from mandelic salts. In the original article,[24]
amines 1 were not purified, but directly N-benzoylated.
However a single recrystallization of the hydrochloride
1b·HCl in ethanol allowed the isolation of 1b in the pure
trans form.[24]
[a] LEDSS, UMR CNRS/UJF 5616, Institut de Chimie Molécu-
laire de Grenoble, Université Joseph Fourier,
BP 53, 38041 Grenoble Cedex 9, France
Fax: +33-4-76-63-59-83
E-mail: Pierre-Yves.Chavant@ujf-grenoble.fr
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
2694
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200400861
Eur. J. Org. Chem. 2005, 2694–2701

Potential Precursors of Quaternary α-Amino Acids
FULL PAPER
In a first attempt to oxidize 1a with 2.1 equiv. of to produce propargylic N-hydroxylamines 4.[31–38] With
MCPBA,[25] we found that nitrone 2a was contaminated simple nitrones in toluene, 0.2 equiv. of diethylzinc is suf-
with small amounts of a more oxidized product.[26] With ficient to allow complete conversion of the nitrone.
less MCPBA, the contaminant was found to be N-hydroxyl-
However the reaction of nitrone 2b failed under the same
amine 3a. In both cases, chromatography led to a low conditions; only traces of the product could be detected.
(Ͻ25%) isolated yield since 2a is unstable on silica gel.[19] Nevertheless, it was possible to achieve a complete transfor-
Note that Baldwin and Long[19] successfully oxidized 1a to mation of 2b if more than 1 equiv. of dimethylzinc was
2a with urea/hydrogen peroxide (UHP) using methyl- used.[39] The reaction of trimethylsilylacetylene (3 equiv.),
trioxorhenium (MTO) in methanol as a catalyst.[8]
dimethylzinc (1.5 equiv.), and 2b led to the propargylic N-
We observed that our samples of 2a decomposed on stor- hydroxylamine 4a in 83% yield (toluene, 20 °C, 6 h).
age. Moreover, an enantiopure solid sample of 2b stored at However, it rapidly became apparent that the reaction
20 °C for 3 months lost optical activity (without decompo- with trimethylsilylacetylene was an exception. With all the
sition). To overcome these problems, we carried out the oxi- other alkynes tested, the nucleophilic addition reaction was
dation in two separate steps: the imidazolidinones 1a,b were followed in situ by a slow cyclization reaction to form the
first treated with 1.2 equiv. of UHP and MTO.[7,8,27] Simple 2,3-dihydroisoxazoles 5.[40] When the reaction mixture was
filtration through a pad of silica produced the correspond- left at room temperature for 18 h, the intermediate 4 was
ing N-hydroxylamines 3; the sole impurity was nitrone 2. totally consumed and high yields of 5 were achieved
At this stage, N-hydroxylamines 3 could be isolated by (Scheme 3).[41]
chromatography (yields 3a: 76%; 3b: 65%) and stored for
months without any loss of optical activity.[28] The second
oxidation step, from N-hydroxylamine 3 to nitrone 2, could
be performed quantitatively in very mild conditions with
[29]
MnO2
in DCM without any side reaction. After fil-
tration, nitrones 2a,b were pure enough to be used as such
in further reactions. Thus, it was very easy to prepare 2 just
before use.
Alternatively, since the impurity in the first step was
nitrone 2, crude 3 could be oxidized at once with MnO2.
Almost quantitative yields were produced from
1
(Scheme 2). However, although the oxidation of 3a with
MnO2 produced 2a in an enantiomerically impure form
(61% ee in DCM, 40% ee in MeOH),[30] the more stable 2b
was obtained from 3b in an enantiopure form.
Scheme 3. Tandem addition/cyclization reaction of 2b with alkynes:
a) Me2Zn (1.3 equiv.), toluene, 20 °C, 6 h, then NH4Cl; b) Me2Zn
(1.3 equiv.), toluene, 20 °C, 18 h, then NH4Cl.
This reaction was compatible with various functional
groups on the alkyne, even in the propargylic position. In
all cases, a single diastereoisomer was detected by NMR
spectroscopy. The relative stereochemistry of the new ste-
reogenic center was proven by NOE experiments in the case
of 5e.[42]
Although the global outcome was identical to that of a
1,3-dipolar cycloaddition reaction (with the same regioch-
emistry), the reaction clearly followed a two-step pathway.
TLC analysis of hydrolyzed aliquots of the reaction mixture
showed the formation and disappearance of the intermedi-
ate 4. The second step was relatively slow with 1-ethynylcy-
clohexene (Scheme 4) and the reaction could be stopped
Scheme 2. Preparation of N-hydroxylamines 3a,b and nitrones 2a,b:
a) UHP (1.2 equiv.), MTO (2–5%), DCM, 20 °C, 10 h; b) MnO2,
DCM, 20 °C, 2 h.
Reaction of Nitrone 2b with 1-Alkynes and Dimethylzinc
We have previously shown[20,21] that 1-alkynes react with when 4d had accumulated (4d/5d Ϸ 50:50). Chromatog-
nitrones in the presence of diethylzinc in toluene or DCM raphy gave pure 4d in 40% yield. A sample of 4d was then
Eur. J. Org. Chem. 2005, 2694–2701
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2695

F. Cantagrel, S. Pinet, Y. Gimbert, P. Y. Chavant
FULL PAPER
treated in toluene with fresh dimethylzinc to regenerate the
Zn–O bond. After 18 h at 20 °C and hydrolysis, conversion
into 5d was complete (Scheme 4).
Scheme 4. Cyclization of the intermediate N-hydroxylamine 4. a)
Me2Zn (1.5 equiv.), toluene, 20 °C, 6 h, then H2O; b) Me2Zn
(1.5 equiv.), toluene, 20 °C, 18 h, then H2O.
Scheme 5. Accumulation of the vinylzinc intermediate 7.
Thus, the latter process (with dimethylzinc in excess over
alkyne) opens the way to a three-component reaction be-
An efficient cyclization of propargylic N-hydroxylamines tween a nitrone, a 1-alkyne, and an organozinc reagent to
in the presence of zinc iodide and DMAP (both in catalytic yield a new organozinc reagent. We are currently investiga-
amounts) has been published.[43] The mechanism proposed ting the reactivity of 7.
by the authors involves the formation of RRЈN–O–ZnI
The exact mechanism of such a trans oxyzincation of a
(and DMAP iodohydrate). This intermediate adds to the triple bond (from 6 to 7) is not obvious, and any explana-
triple bond to form a vinylzinc species, which is immediately tion should involve the intervention of a second zinc atom
protolyzed by the amine iodohydrate.
that would both activate the alkyne and take part in the
We repeated one of our reactions of Scheme 3 (reaction final C–Zn bond. One could alternatively consider a 1,3-
b) in [D6]benzene [2b (1 equiv.), phenylacetylene (3 equiv.), dipolar cycloaddition of the nitrone to the alkynylzinc triple
and dimethylzinc (1.5 equiv.)]. NMR signals for 2b, 5b, the bond (from 2 to 7 direct; an analogous concerted cycload-
adduct 6 (Scheme 5), acetylenic hydrogen, and Me–ZnY dition of a nitrone and an alkynylcopper species has been
could be distinguished.[44] We observed again that the first postulated in the mechanism for the β-lactam synthesis de-
alkyne addition was faster than the cyclization reaction scribed by Miura et al.[46]). In this hypothesis, the nucleo-
(50% of 2b was consumed after approx. 40 min, 50% of 5b philic addition (from 2 to 6) should be reversible (since in
was formed after approx. 4 h). Noticeably, after 60 min, the the experiment 7 is formed from 6). To address this point
methylzinc signals had shifted from –0.59 to –0.10 ppm. we used DFT calculations[47] to compare the postulated nu-
From then on, the intensities of this methylzinc signal and cleophilic addition reaction (from 2 to 6) with the hypothet-
those of 6 decreased synchronously, in agreement with the ical concerted 1,3-dipolar cycloaddition of the alkynylzinc
structure proposed below for 6. The bicyclic vinylzinc inter- reagent to 2. In the latter reaction, both orientations of the
mediate 7 was not observed in this experiment: the vinylic cycloaddition were checked. Our results are summarized in
proton of 5 was present in quantitative amounts. We con- Scheme 6.
cluded that 7 was rapidly protonated by the excess alkyne.
Two transition states (TS1 and TS2) were located with
Nevertheless, we could accumulate the vinylzinc species activation barriers of 9.7 and 12.5 kcalmol–1, respectively.
7 by changing the reaction protocol. Phenylacetylene The transition state with the higher energy led to 7 with the
(2 equiv.) was treated with excess dimethylzinc (4 equiv.) for experimentally observed regiochemistry. The pathway to the
1 h at 20 °C and then 2b was added to the preformed alkyn- addition product 6 is much more favorable. The nitrone and
ylzinc. After 12 h at 20 °C, the reaction mixture was treated the alkynylzinc first form a complex (–6 kcalmol–1). The
with DCl in D2O. NMR analysis of 5b indicated 87% deu- energy barrier from this complex to adduct 6, via TS3, is
terium incorporation.[45] This reaction was repeated and very low (4 kcalmol–1). Unfortunately, we have so far been
iodolysis (NIS, 20 °C) produced the vinyl iodide 8 in 62% unable to find a solution, in this model, for the reaction
yield.
pathway of the second cyclization step (6 to 7).
2696
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2694–2701

Potential Precursors of Quaternary α-Amino Acids
FULL PAPER
It is evident that the catalytic cycle was effective with
reactive nitrones like 11[48] (Scheme 7), but was blocked if
the nitrone carried a carboxy group (nitrones 9[49] and
10[50]). If a 0.2:1 ratio of dimethylzinc to nitrone was used,
approximately 20% conversion of the nitrone was achieved;
the adducts (approximately 20% yields) were recovered as
dihydroisoxazoles 5h or 5i after 12 h.
The presence of an ester in the reaction mixture is not
the cause of the cyclization: in the presence of 1.5 equiv. of
Me2Zn, the nitrone 11 was totally consumed in less than
15 min producing only 4j. When this reaction was repeated
in the presence of 10 equiv. of ethyl acetate, complete con-
version of 11 required 5 h and also yielded 4j.
The cyclization reaction took place whenever the nitrone
carried an ester (9 and 10) or amide (2b) group. It never
took place with the simple nitrone 11 (or with other exam-
ples reported in ref.[20]), except when the alkyne contained
an acetyl group (entries k and l). It appears that the driving
force for the cyclization reaction is the presence of a car-
bonyl group that can intramolecularly bind with the vinylic
zinc atom. NMR evidence for analogous intramolecular
chelations between an organozinc and a neighbouring car-
boxylic oxygen atom have been published.[51,52]
Scheme 6. Optimized structures at the B3LYP/6-31+G* level of
theory. Distances (in italics) in Å, energies in kcalmol–1.
Conclusions
Finally, we undertook a series of tests (Scheme 7 and
Table 1) in order to explain the changes in reactivity (stoi-
We have shown that the self-regenerating stereocenter
chiometric vs. catalytic ZnII) and in the outcome (cyclic di- principle[12] can be applied to the preparation of enantio-
hydroisoxazole vs. open-chain propargylic N-hydroxyl- pure nitrones 2a,b from natural α-amino acids. We have also
amine) when 2b was used instead of other nitrones.[20]
found that 2b and other α-carboxy-nitrones react with 1-
Scheme 7. Factors influencing the cyclization reaction.
Table 1. Factors influencing the cyclization reaction.
Entry
Nitrone
R3
Me2Zn [equiv.]
Nitrone[a] [%]
4[a][b] [%]
5[a][b] [%]
h
h
i
9
Ph
Ph
0.2
1.5
0.2
1.5
0.2
1.5
0.2
1.5
0.2[c]
82
0
75
0
0
0
0
0
0
0
0
0
0
18
100 (91)
25
100 (95)
0
0
0
9
10
10
11
11
11
11
12
CH2OtBu
CH2OtBu
CH2OtBu
CH2OtBu
CH2OAc
CH2OAc
CH2OAc
i
j
j
k
k
l
100 (82)[20]
100 (88–92)
100 (70)[20]
0
0
100 (82)
100[20]
[a] Ratio of products in the crude hydrolyzed material. [b] The yields of the isolated pure materials are given in parentheses. [c] Diethylzinc
(0.2 equiv.), 19 h, see ref.[20]
.
Eur. J. Org. Chem. 2005, 2694–2701
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2697

F. Cantagrel, S. Pinet, Y. Gimbert, P. Y. Chavant
FULL PAPER
necessary to add another 50 mg of MTO after several hours). After
filtration through Celite and concentration, the crude N-hydroxyl-
amine 3 (contaminated with nitrone 2) was either directly oxidized
with MnO2 (see below) or purified by chromatography on silica.
alkynylzinc derivatives in a tandem addition/cyclization
process, leading regio- and diastereoselectively to a series of
highly functionalized bicyclic imidazoisoxazoles 5 in excel-
lent yields. α,α-Disubstituted α-amino acids are masked in
5. We are currently studying further transformations of
these molecules into new quaternary amino acids that fea-
ture complex side-chains.
(R)-(+)-2-tert-Butyl-1-hydroxy-3-methylimidazolidin-4-one
(3a):
Chromatography (cyclohexane/ethyl acetate, 1:1, Rf = 0.18) yielded
1
76% (1.48 g) of 3a, m.p. 68 °C. H NMR (300 MHz, CDCl3): δ =
4.22 (s, 1 H), 3.73, (d, J = 17.8 Hz, 1 H), 3.52 (d, J = 17.8 Hz, 1
H), 3.01 (s, 3 H), 0.97 (s, 9 H) ppm. 13C NMR (75 MHz, CDCl3):
δ = 172.9 (C), 97.7 (CH), 61.8 (CH2), 37.2 (C), 31.8 (CH3), 26.2
Experimental Section
(CH ) ppm. FTIR: ν = 3302, 2961, 2874, 1677, 1484, 1476, 1453,
˜
3
1414, 1293, 1262, 1247, 1205, 1124, 1084, 907, 826 cm–1. HRMS:
found: m/z = 195.1107 [M + Na]; calculated: m/z = 195.1109 [M +
Na]. [α]2D5 = +40.6 (c = 2.4, CH2Cl2).
General Remarks: 1H (200 or 300 MHz) and 13C NMR (50 or
75 MHz) spectra were recorded with Bruker AC200 or Advance300
spectrometers in CDCl3 with tetramethylsilane as the internal stan-
dard. Chemical shifts are reported in parts per million (ppm) and
coupling constants (J) are given in Hertz (Hz). The number of hy-
drogen atoms on each carbon atom was obtained from DEPT ex-
periments. IR spectra were recorded with a Nicolet Impact-400
FTIR spectrometer from neat films or sintered KBr discs. HRMS
(chemical ionisation) and elemental analyses were performed at the
Service Central dЈAnalyse du CNRS, Vernaison, France. Thin-layer
chromatography (TLC) was carried out on Merck precoated silica
gel 60 F-254 plates. Spots were visualized with UV light or by
dipping the TLC plate into a basic potassium permanganate solu-
tion and heating, or by dipping the plate in a 1% triphenyl tetrazol-
ium chloride (TTC) solution and then in a 5% NaOH solution of
ethanol/water and heating; a strong permanent red colour is char-
acteristic of N-hydroxylamines. Forced-flow column chromatog-
raphy was performed by using Macherey–Nagel silica gel 60, 230–
400 mesh. Melting points were determined on a Büchi B-545 appa-
ratus and are uncorrected. All reactions were performed by using
conventional Schlenk techniques under dry nitrogen in oven-dried
glassware and with magnetic stirring. All reagents were purchased
from Aldrich, Acros, or Fluka and used as received. Dimethylzinc
was used as a commercial (Aldrich) 2 m solution in toluene. Dichlo-
romethane (DCM) was distilled from CaH2. Toluene was distilled
from sodium. Dimethylformamide (DMF) and N-methylpyrroli-
din-2-one were distilled and stored on 4- molecular sieves.
NMR measurement of enantiomeric purity: A sample of enantio-
enriched 3a and approximately 2 mol-equiv. of (S)-O-acetylman-
delic acid were mixed in CDCl3. Lines for (S)-3a in this solution
are 4.25 (s, 1 H), 3.73 (d, J = 18 Hz, 1 H), 3.50 (d, J = 18 Hz, 1
H), 2.99 (s, 1 H), 0.95 (s, 1 H) ppm; for (R)-3a 4.24 (s, 1 H), 3.76
(d, J = 18 Hz, 1 H), 3.53 (d, J = 18 Hz, 1 H), 2.99 (s, 1 H), 0.91
(s, 1 H) ppm.
(2S,5S)-(–)-2-tert-Butyl-1-hydroxy-3,5-dimethylimidazolidin-4-one
(3b): Chromatography (cyclohexane/ethyl acetate, 1:1, Rf = 0.56)
yielded 65 % (1.21 g) of 3b, m.p. 83 °C. 1H NMR (300 MHz,
CDCl3): δ = 6.28 (s, 1 H), 4.17 (s, 1 H), 3.65 (q4, J = 7.2 Hz, 1 H),
3.00 (s, 3 H), 1.36 (d, J = 7.2 Hz, 3 H), 1.00 (s, 9 H) ppm. 13C
NMR (75 MHz, CDCl3): δ = 175.5 (C), 96.4 (CH), 62.5 (CH), 36.9
(C), 32.3 (CH ), 26.8 (CH ), 10.8 (CH ) ppm. FTIR: ν = 3349,
˜
3
3
3
3317, 2960, 2872, 1680 cm–1. HRMS: found: m/z = 209.1266; calcu-
lated: m/z = 209.1266 [M + Na]. [α]2D5 = –32 (c = 2.0, CH2Cl2).
Oxidation of N-Hydroxylamines 3 to Nitrones 2: Crude 3a,b from
the preceding preparation was added to DCM (20 mL) and stirred
with activated MnO2 (Fluka, ref. 63548, 1.74 g, 20 mmol) for 2 h
at 20 °C. Filtration of the slurry through Celite and repeated wash-
ing with EtOAc yielded nitrone 2a,b in sufficient purity for direct
use. When crude 3a,b were oxidized, the yields of 2a,b from 1a,b
were 95 and 99% respectively. When the same MnO2 oxidation
reaction was performed with purified samples of 3a and 3b, quanti-
tative yields were obtained.
DFT calculations were carried out using the Gaussian 98[53,54] sys-
tem of programs. Critical points (reactants, transition structures,
precomplexes, and products) were fully characterized as minima or
first-order saddle points by diagonalizing the Hessian matrices of
the optimized structures at the B3LYP/6-31+G* level of theory.[47]
All critical points were further characterized by analytic computa-
tion of the harmonic frequencies at the same level of calculation.
Transition structures were found to have only one negative eigen-
value with the corresponding eigenvector involving the formation
of the newly created C–C and C–O bonds. Vibrational frequencies
were calculated (1 atm, 298.15 K) for all optimized structures and
used to compute both ZPVE and activation energies. The intrinsic
reaction coordinates[55,56] were also calculated in order to analyze
in detail the mechanism for all transition structures obtained.
The ee values of 2a and 2b were determined by reversed-phase
HPLC on a Daicel CHIRALPAK AD-RH column (length:
150 mm, flow rate: 0.3 mLmin–1, mobile phase: acetonitrile/water
(40:60) for 3a, 25:75 for 3b. Retention times: (R)-2a 20.4 min, (S)-
2a 24.8 min, (S)-3a 17.4 min, and (R)-3a 18.6 min.
(R)-2-tert-Butyl-3-methyl-2,3-dihydroimidazol-4-one N-Oxide (2a):
Yield = 95% from 3a (crude material, oil). 1H NMR (300 MHz,
CDCl3): δ = 7.07 (s, 1 H), 4.64 (s, 1 H), 3.10 (s, 3 H), 1.14 (s, 9
H) ppm. 13C NMR (75 MHz, CDCl3): δ = 165.9 (C), 126.4 (CH),
94.9 (CH), 37.3 (C), 32.3 (CH ), 25.5 (CH ) ppm. FTIR: ν = 3080,
˜
3
3
2967, 2936, 2869, 1717, 1574, 1559, 1288, 647, 557 cm–1. HRMS:
found: m/z = 171.1137; calculated: m/z = 171.1134 [M + H]. The
crude sample had 61% ee. [α]2D5 = –73.3 (c = 1.34, CHCl3) [lit.[19]
+218 (c = 0.15, CHCl3 for (S)-2a)].
Preparation of Imidazolidinones 1a,b: Imidazolidinone 1a was pre-
pared as its mandelic salt[22] on a 50-mmol scale. Imidazolidinone
1b, obtained following the procedure described in ref.[23], was
recrystallized once in ethanol to yield the pure trans isomer.
Oxidation with UHP/MTO: UHP (77 mg, 0.82 mmol) and then
MTO (5 mg, 0.03 mmol) were added to enantiopure (R)-3a (54 mg,
0.35 mmol) in methanol (1 mL) at 0 °C . The reaction was stirred
for 18 h at 20 °C. Work up according to ref.[19] led to 2a (quant.)
in 95% ee. This crude sample had [α]2D5 = –142.6 (c = 2.06, CHCl3).
Preparation of N-Hydroxylamines 3: Urea/hydrogen peroxide
(UHP, 1.13 g, 12 mmol) followed by methyltrioxorhenium (MTO,
50 mg, 0.2 mmol) were added to a solution of imidazolidinone (S)-
1a or (S,S)-1b (10 mmol) in dichloromethane (20 mL). The reaction
was monitored by TLC (silica, eluent cyclohexane/ethyl acetate,
1:1) until the conversion of 1a was complete (in some runs it was
(2S)-2-tert-Butyl-3,5-dimethyl-2,3-dihydroimidazol-4-one N-Oxide
(2b): Yield 99% from 2b (crude material, solid), m.p. 67 °C. 1H
2698
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2694–2701

Potential Precursors of Quaternary α-Amino Acids
FULL PAPER
NMR (300 MHz, CDCl3): δ = 4.57 (s, 1 H), 3.11 (s, 3 H), 2.08 (s, 1071 cm–1. HRMS: found: m/z = 267.2066; calculated: m/z =
3 H), 1.12 (s, 9 H) ppm. 13C NMR (75 MHz, CDCl3): δ = 166.5 267.2073 [M + H]. [α]2D5 = –27 (c = 2.9, CH2Cl2).
(C), 135.9 (C), 93.0 (CH), 37.5 (C), 32.4 (CH3), 25.5 (CH3), 7.6
(6S,3aS)-6-tert-Butyl-2-(cyclohex-1-enyl)-3a,5-dimethyl-5,6-dihydro-
(CH ) ppm. FTIR: ν = 2964, 2925, 2874, 1692, 1594, 1468, 1448,
˜
3
3aH-imidazo[1,5-b]isoxazol-4-one (5d): Prepared according to the
general procedure from 2b (172 mg, 0.93 mmol), 1-ethynylcyclohex-
ene (300 mg, 2.84 mmol), and dimethylzinc (0.75 mL, 1.5 mmol).
Chromatography (EtOAc/cyclohexane, 4:6) yielded 255 mg (94%)
of 5d as an oil (Rf = 0.55 in EtOAc/cyclohexane, 1:1). 1H NMR
(300 MHz, CDCl3): δ = 6.13 (m, 1 H), 4.83 (s, 1 H), 4.05 (s, 1 H),
2.95 (s, 3 H), 2.14–2.12 (m, 4 H), 1.69–1.59 (m, 4 H), 1.44 (s, 3 H),
1.07 (s, 9 H) ppm. 13C NMR (75 MHz, CDCl3): δ = 173.2 (C),
154.5 (C), 128.5 (C), 125.4 (C), 96.7 (CH), 92.5 (CH), 75.1 (C),
35.7 (C), 31.0 (CH3), 27.1 (CH2), 25.9 (CH3), 25.4 (CH2), 23.3
1380, 1335, 1268, 1228, 1197, 1056, 1003, 734, 577, 545 cm–1
.
HRMS: found: m/z = 185.1282; calculated: m/z = 185.1290 [M +
H]. [α]2D5 = +158.5 (c = 4.46, CH2Cl2). Chiral HPLC: one enantio-
mer was detected. A sample was prepared according to ref.[19]
:
[α]2D5 = +152.1 (c = 1.46, CH2Cl2); one enantiomer was detected by
HPLC.
General Procedure: Reaction of Nitrone 2b with Excess Alkyne and
Dimethylzinc: Nitrone 2b (184 mg, 1 mmol) and the alkyne
(3 mmol) in toluene (2 mL) were introduced into a 30 mL Schlenk
vessel under N2. A 2 m solution of dimethylzinc in toluene
(1.5 mmol) was added at 20 °C, and the reaction mixture was
stirred at 20 °C overnight. Hydrolysis was performed with saturated
aqueous NH4Cl (2 mL) at 20 °C for 30 min. After extraction with
EtOAc, the organic phase was washed with brine, dried with
Na2SO4, and concentrated to give a crude oil which was purified
by chromatography.
(CH ), 22.3 (CH ), 22.0 (CH ) ppm. FTIR: ν = 2930, 2866, 1715,
˜
3
2
2
1659, 1620, 1482, 1398, 1368, 1280, 1266, 1248, 1076, 998, 959,
926 cm–1. HRMS: found: m/z = 291.2093; calculated: m/z =
291.2073 [M + H]. [α]2D5 = –146 (c = 2.1, CH2Cl2).
(6S,3aS)-2-tert-Butoxymethyl-6-tert-butyl-3a,5-dimethyl-5,6-dihy-
dro-3aH-imidazo[1,5-b]isoxazol-4-one (5e): Prepared according to
the general procedure from 2b (213 mg, 1.15 mmol), 3-tert-butoxy-
prop-1-yne (390 mg, 3.5 mmol), and dimethylzinc (0.87 mL,
1.75 mmol). Chromatography (EtOAc/cyclohexane, 4:6) yielded
319 mg (90%) of 5e as an oil (Rf = 0.66 in EtOAc/cyclohexane,
(2S,5S)-2-tert-Butyl-1-hydroxy-3,5-dimethyl-5-(trimethylsilylethy-
nyl)imidazolidin-4-one (4a): Prepared according to general pro-
cedure from 2b (173 mg, 0.94 mmol), trimethylsilylacetylene
(0.45 mL, 3.0 mmol), and dimethylzinc (0.75 mL, 1.5 mmol).
Chromatography (EtOAc/cyclohexane, 4:6) yielded 220 mg (83%)
of 4a (Rf = 0.75 in EtOAc/cyclohexane, 1:1), m.p. 113 °C. 1H NMR
(300 MHz, CDCl3): δ = 4.67 (s, 1 H), 3.75 (s, 1 H), 2.94 (s, 3 H),
1.52 (s, 3 H), 1.01 (s, 9 H), 0.14 (s, 9 H) ppm. 13C NMR (75 MHz,
CDCl3): δ = 169.3 (C), 98.0 (C), 95.2 (C), 88.3 (CH), 65.5 (C), 35.2
1
1:1). H NMR (300 MHz, CDCl3): δ = 4.90 (s, 1 H), 4.06 (s, 1 H),
3.95 (s, 1 H), 2.93 (s, 3 H), 1.43 (s, 3 H), 1.20 (s, 9 H), 1.03 (s, 9
H) ppm. 13C NMR (75 MHz, CDCl3): δ = 173.0 (C), 154.5 (C),
98.9 (CH), 92.6 (CH), 74.7 (C), 74.4 (C), 55.8 (CH2), 35.8 (C), 31.8
(CH ), 27.5 (CH ), 25.9 (CH ), 23.3 (CH ) ppm. FTIR: ν = 3116,
˜
3
3
3
3
2974, 2930, 2871, 1710, 1396, 1366, 1194, 1071 cm–1. HRMS:
found: m/z = 297.2184; calculated: m/z = 297.2178 [M + H]. [α]2D5
= –122 (c = 1.2, CH2Cl2). For details of the NOE experiment see
Scheme 8.
(C), 30.9 (CH ), 26.5 (CH ), 23.3 (CH ), 0.1 (CH ) ppm. FTIR: ν
˜
3
3
3
3
= 3370, 2996, 2962, 2900, 2871, 2174, 1706, 1404, 1396, 1345, 1254,
1176, 1094, 1086, 896, 848, 767 cm–1. HRMS: found: m/z =
305.1682; calculated: m/z = 305.1661 [M + Na]. [α]2D5 = –64 (c =
2.4, CH2Cl2).
(6S,3aS)-6-tert-Butyl-3a,5-dimethyl-2-phenyl-5,6-dihydro-3aH-imid-
azo[1,5-b]isoxazol-4-one (5b): Prepared according to the general
procedure from 2b (346 mg, 1.88 mmol), phenylacetylene (602 mg,
6.0 mmol), and dimethylzinc (1.5 mL, 3 mmol). Chromatography
(EtOAc/cyclohexane, 4:6) yielded 495 mg (92%) of 5b as an oil (Rf
1
= 0.55 in EtOAc/cyclohexane, 1:1). H NMR (300 MHz, CDCl3):
Scheme 8.
δ = 7.48–7.45 (m. 2 H), 7.34–7.29 (m. 3 H), 5.30 (s. 1 H), 4.13 (s.
1 H), 2.93 (s. 3 H), 1.51 (s. 3 H), 1.08 (s. 9 H) ppm. 13C NMR
(75 MHz, CDCl3): δ = 173.3 (C), 153.8 (C), 129.8 (CH), 128.8
(CH), 128.1 (C), 126.1 (CH), 97.4 (CH), 92.9 (CH), 75.8 (C), 36.0
(6S,3aS)-2-(Acetoxymethyl)-6-tert-butyl-3a,5-dimethyl-5,6-dihydro-
3aH-imidazo[1,5-b]isoxazol-4-one (5f): Prepared according to the
general procedure from 2b (92 mg, 0.5 mmol), prop-2-ynyl acetate
(150 mg, 1.5 mmol), and dimethylzinc (0.38 mL, 0.75 mmol).
Chromatography (EtOAc/cyclohexane, 4:6) yielded 107 mg (76%)
of 5f as an oil (Rf = 0.50 in EtOAc/cyclohexane, 1:1). 1H NMR
(300 MHz, CDCl3): δ = 4.99 (s, 1 H), 4.63 (s, 2 H), 4.08 (s, 1 H),
2.95 (s, 3 H), 2.10 (s, 3 H), 1.44 (s, 3 H), 1.04 (s, 9 H) ppm. 13C
NMR (75 MHz, CDCl3): δ = 172.3 (C), 169.9 (C), 150.4 (C), 101.0
(CH), 92.5 (CH), 74.6 (C), 56.4 (CH2), 35.6 (CH3), 30.9 (C), 25.6
(C), 31.3 (CH ), 26.2 (CH ), 23.5 (CH ) ppm. FTIR: ν = 3107,
˜
3
3
3
3061, 2961, 2925, 2869, 2851, 1707, 1495, 1481, 1449, 1396, 1282,
1241, 1071, 1004, 997, 765, 740, 716, 691 cm–1. HRMS: found:
m/z = 287.1755; calculated: m/z = 287.1760 [M + H]. [α]2D5 = +28
(c = 3.1, MeOH).
(6S,3aS)-2-Butyl-6-tert-butyl-3a,5-dimethyl-5,6-dihydro-3aH-imid-
azo[1,5-b]isoxazol-4-one (5c): Prepared according to the general
procedure from 2b (364 mg, 1.98 mmol), hex-1-yne (0.91 mL,
8 mmol), and dimethylzinc (1.5 mL, 3 mmol). Chromatography
(EtOAc/cyclohexane, 3:7) yielded 318 mg (60%) of 5c as an oil (Rf
(CH ), 22.9 (CH ), 20.6 (CH ) ppm. FTIR: ν = 3115, 2963, 2873,
˜
3
3
3
1752, 1706, 1482, 1450, 1429, 1397, 1367, 1232, 1062, 1032,
990 cm–1. HRMS: found: m/z = 305.1464; calculated: m/z =
305.1477 [M + Na]. [α]2D5 = –34.4 (c = 1.7, CH2Cl2).
1
= 0.70 in EtOAc/cyclohexane, 1:1). H NMR (300 MHz, CDCl3):
δ = 4.53 (s, 1 H), 3.93 (s, 1 H), 2.86 (s, 3 H), 2.05 (t, J = 8.2 Hz, 2 (6S,3aS)-6-tert-Butyl-2-(diethoxymethyl)-3a,5-dimethyl-5,6-dihydro-
H), 1.33 (s, 3 H), 1.34 (m, 4 H), 0.97 (s, 9 H), 0.83 (t, J = 7.2 Hz, 3aH-imidazo[1,5-b]isoxazol-4-one (5g): Prepared according to the
3 H) ppm. 13C NMR (75 MHz, CDCl3): δ = 172.4 (C), 155.4 (C),
general procedure from 2b (296 mg, 1.61 mmol), 1,1-diethoxyprop-
95.6 (CH), 91.7 (CH), 73.6 (C), 34.6 (C), 29.9 (CH3), 27.8 (CH2), 2-yne (0.69 mL, 4.8 mmol), and dimethylzinc (1.2 mL, 2.4 mmol).
24.8 (CH3), 24.2 (CH2), 22.3 (CH3), 21.2 (CH2), 12.7 (CH3) ppm. Chromatography (EtOAc/cyclohexane, 2:8) yielded 390 mg (78%)
FTIR: ν = 2957, 2926, 2855, 1712, 1482, 1467, 1425, 1396, 1365, of 5g as an oil (Rf = 0.82 in EtOAc/cyclohexane, 1:1). 1H NMR
˜
Eur. J. Org. Chem. 2005, 2694–2701
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2699

F. Cantagrel, S. Pinet, Y. Gimbert, P. Y. Chavant
FULL PAPER
(300 MHz, CDCl3): δ = 5.07 (d, J = 1.0 Hz, 1 H), 5.03 (d, J =
1.0 Hz, 1 H), 4.07 (s, 1 H), 3.70–3.49 (m, 4 H), 2.94 (s, 3 H), 1.45
(s, 3 H), 1.21 (t, J = 7.2 Hz, 6 H), 1.03 (s, 9 H) ppm. 13C NMR
(75 MHz, CDCl3): δ = 173.0 (C), 152.8 (C), 101.1 (CH), 95.3 (CH),
93.0 (CH), 74.9 (C), 62.3 (CH2), 61.8 (CH2), 36.1 (C), 31.4 (CH3),
Treatment of a sample of 5b with NIS in THF (20 °C, 2 h) did not
produce 8.
Ethyl 2-Benzyl-2,3-dihydro-5-phenylisoxazole-3-carboxylate (5h):
Prepared according to the general procedure from nitrone 9 (52 mg,
0.25 mmol), phenylacetylene (77 mg, 0.75 mmol), and dimethylzinc
(0.18 mL, 0.36 mmol). Chromatography (EtOAc/cyclohexane, 2:8)
yielded 70 mg (91%) of 5h as an oil (Rf = 0.82 in EtOAc/cyclohex-
15.5 (CH ) ppm. FTIR: ν = 2976, 2930, 2877, 1709, 1396, 1121,
˜
3
1065 cm–1. HRMS: found: m/z = 267.1716; calculated for
C14H23N2O3: m/z = 267.1709. [α]2D5 = –4 (c = 2,3, CH2Cl2).
1
ane, 2:8). H NMR (300 MHz, CDCl3): δ = 7.54–7.28 (m, 10 H),
1H NMR Observation of the Tandem Reaction: Signals observed
in the reaction mixture ([D8]toluene, 300 MHz, reference toluene
2.09 ppm, all singlets). Nitrone 2b: 0.76 (9 H), 1.91 (3 H), 2.50 (3
H), 3.86 (1 H) ppm. Adduct 6: 0.82 (9 H), 1.84 (3 H), 2.65 (3 H),
3.84 (1 H) ppm. Cycloadduct 5b: 0.88 (9 H), 1.60 (3 H), 2.57 (3 H),
3.84 (1 H), 5.43 (1 H) ppm.
5.28 (d, J = 3.0 Hz, 1 H), 4.62 (d, J = 3.0 Hz, 1 H), 4.40 (d, J =
12.8 Hz, 1 H), 4.17 (quad, J = 7.1 Hz, 2 H), 4.05 (d, J = 12.8 Hz,
1 H), 1.23 (t, J = 7.1 Hz, 3 H) ppm. 13C NMR (75 MHz, CDCl3):
δ = 170.8 (C), 155.0 (C), 135.8 (C), 133.8 (C), 130.0 (CH), 129.8
(CH), 128.7 (CH), 129.0 (CH), 128.9 (CH), 128.8 (CH), 90.4 (CH),
72.0 (CH), 63.8 (CH ), 61.7 (CH ), 14.5 (CH ) ppm. FTIR: ν =
˜
2
2
3
3114, 3088, 3061, 3036, 2960, 2935, 2898, 2871, 1764, 1455, 1006,
763, 740, 701, 532 cm–1. HRMS: found: m/z = 310.1451; calculated:
m/z = 310.1443 [M + H].
Preparation of (2S,5S)-2-tert-Butyl-5-(cyclohex-1-enylethynyl)-1-hy-
droxy-3,5-dimethylimidazolidin-4-one (4d): Nitrone 2b (184 mg,
1 mmol) in toluene (2 mL), 1-ethynylcyclohexene (300 mg,
2.84 mmol), and dimethylzinc (1.5 mmol) in toluene (0.75 mL) were
introduced into a 30 mL Schlenk vessel under N2. The reaction
mixture was stirred at 20 °C for 6 h. Hydrolysis was performed with
saturated aqueous NH4Cl (2 mL) at 20 °C for 30 min. After extrac-
tion of the reaction mixture with EtOAc, the organic phase was
washed with brine, dried with Na2SO4, and concentred to give a
crude oil which was purified chromatography (EtOAc/cyclohexane,
1:9 to 4:6). Compound 4d was eluted first as an oil in 40% yield
(116 mg). Rf = 0.76 in EtOAc/cyclohexane, 1:1. 1H NMR
(300 MHz, CDCl3): δ = 6.19–6.15 (m, 1 H), 4.72 (s, OH), 3.83 (s,
1 H), 2.98 (s, 3 H), 2.11–2.08 (m, 4 H), 1.61–1.57 (m, 4 H), 1.58 (s,
3 H), 1.05 (s, 9 H) ppm. 13C NMR (75 MHz, CDCl3): δ = 169.7
(C), 136.9 (CH), 119.6 (C), 91.6 (C), 88.2 (CH), 79.5 (C), 65.3 (C),
35.2 (C), 30.8 (CH3), 29.3 (CH2), 26.6 (CH3), 25.8 (CH2), 23.5
Methyl 2-(tert-Butoxymethyl)-5,6-dihydro-4H-pyrrolo[1,2-b]isox-
azole-3a-carboxylate (5i): Prepared according to the general pro-
cedure from nitrone 10 (86 mg, 0.60 mmol), 3-tert-butoxyprop-1-
yne (200 mg, 1.8 mmol), and dimethylzinc (0.45 mL, 0.9 mmol).
Chromatography (EtOAc/cyclohexane, 2:8) yielded 145 mg (95%)
of 5i as an oil (Rf = 0.85 in AcOEt/cyclohexane, 1:1). 1H NMR
(300 MHz, CDCl3): δ = 4.79 (s, 1 H), 3.95 (s, 2 H), 3.75 (s, 3 H),
3.75–3.32 (m, 2 H), 2.25–2.14 (m, 1 H), 2.10–2.02 (m, 1 H), 1.95–
1.79 (m, 2 H), 1.21 (s, 9 H) ppm. 13C NMR (75 MHz, CDCl3): δ
= 173.8 (C), 155.6 (C), 96.5 (CH), 82.2 (C), 74.3 (C), 60.2 (CH2),
56.0 (CH2), 52.6 (CH3), 36.8 (CH2), 29.8 (C), 27.5 (CH3), 23.2
(CH ) ppm. FTIR: ν = 2954, 2950, 2860, 1747, 1561, 1458, 1388,
˜
2
1259, 1182, 1112 cm–1. HRMS: found: m/z = 256.1538; calculated:
m/z = 256.1549 [M + H].
(CH ), 22.3 (CH ), 21.5 (CH ) ppm. FTIR: ν = 3391, 2934, 2863,
˜
3
2
2
2934, 2863, 2230 (w), 1705, 1488, 1402, 1365, 1260, 1204, 1137,
1068 cm–1. HRMS: found: m/z = 313.1909; calculated: m/z =
313.1879 [M + Na]. [α]2D5 = –29 (c = 2.0, CH2Cl2).
(2-Benzyl-2,3-dihydro-3-phenylisoxazol-5-yl)methyl Acetate (5k):
Prepared according to the general procedure from nitrone 11
(106 mg, 0.50 mmol), 3-acetoxyprop-1-yne (147 mg, 1.5 mmol),
and dimethylzinc (0.40 mL, 0.8 mmol). Chromatography (EtOAc/
cyclohexane, 2:8) yielded 126 mg (82%) of 5i as an oil (Rf = 0.75
in AcOEt/cyclohexane, 1:1). 1H NMR (300 MHz, CDCl3): δ =
7.40 –7.10 (m, 10 H), 4.98 (d, J = 2.5 Hz, 1 H), 4.87 (d, J = 2.5 Hz,
1 H), 4.67 (d, J = 13.4 Hz, 1 H), 4.61 (d, J = 13.4 Hz, 1 H), 4.27
(d, J = 13.0 Hz, 1 H), 2.04 (s, 3 H) ppm. 13C NMR (75 MHz,
CDCl3): δ = 170.35 (C), 149.78 (C), 141.53 (C), 136.11 (C), 129.50
(CH), 128.49 (CH), 128.34 (CH), 128.13 (CH), 127.62 (CH), 126.94
(CH), 99.91 (CH), 72.98 (CH), 63.34 (CH2), 57.18 (CH2), 20.75
Cyclization of 4d: A 2 m solution of dimethylzinc (0.35 mmol) in
toluene (0.17 mL) was added to pure 4d (50 mg, 0.17 mmol) in tol-
uene (0.5 mL). After 18 h at 20 °C, the reaction mixture was hy-
drolyzed with sat. NH4Cl (1 mL). After the usual work up and
filtration through silica, a 93% yield (47 mg) of 5d was obtained.
The reaction was repeated identically and deuterolysis was per-
formed with a 1 m solution of DCl in D2O (0.5 mL) for 30 min,
1
followed by the usual EtOAc extraction. The H NMR signal at δ
= 4.83 ppm (vinylic) was less intense (intensity 0.13 H) in the de-
uteriated sample.
(CH ) ppm. FTIR: ν = 3090, 3062, 3030, 2927, 1746, 1669, 1230,
˜
3
932, 761, 698 cm–1. HRMS: found: m/z = 310.1455; calculated:
(rac)-6-tert-Butyl-3-iodo-3a,5-dimethyl-2-phenyl-5,6-dihydro-3aH-im-
idazo[1,5-b]isoxazol-4-one (8): Toluene (2 mL), phenylacetylene
(0.25 mL, 2.2 mmol), and a 2 m solution of dimethylzinc
m/z = 310.1443 [M + H].
(4.8 mmol) in toluene (2.4 mL) were introduced into a Schlenk tube Acknowledgments
under N2. After stirring for 1.5 h at 20 °C, a solution of (rac)-2b
(216 mg, 1.17 mmol) in toluene (2 mL) was added. The mixture was
stirred at 20 °C for 12 h. Then, N-iodosuccinimide (2.5 g, 6 mmol)
in THF (10 mL) was introduced. After 1 h, the usual aqueous work
up and chromatography (EtOAc/cyclohexane, 1:9) yielded 299 mg
(62%) of 8 as an oil (Rf = 0.82 in EtOAc/cyclohexane, 1:1). 1H
NMR (300 MHz, CDCl3): δ = 7.75–7.72 (m, 2 H), 7.30–7.28 (m, 3
F.C. thanks LEDSS (UMR, 5616) for financial support. We thank
Dr. C. Philouze for carrying out the X-ray structure determination.
[1] H. C. J. Ottenheijm, J. D. M. Herscheid, Chem. Rev. 1986, 86,
697–707.
[2] For oxidation with SeO2/H2O2, see: S.-I. Murahashi, T. Shiota,
Tetrahedron Lett. 1987, 28, 2383–2386.
[3] For oxidation with Na2WO4/H2O2, see: S.-I. Murahashi, H.
Mitsui, T. Shiota, T. Tsuda, S. Watanabe, J. Org. Chem. 1990,
55, 1736–1744.
H), 3.99 (s, 1 H), 2.87 (s, 3 H), 1.40 (s, 3 H), 0.99 (s, 9 H) ppm. 13
C
NMR (75 MHz, CDCl3): δ = 170.1 (C), 151.5 (C), 130.0 (CH),
128.2 (CH), 127.9 (CH), 127.3 (C), 91.3 (CH), 76.5 (C), 35.4 (C),
30.5 (CH3), 25.7 (CH3), 21.9 (CH3) ppm; C–I was not detected.
FTIR: ν = 2976, 2930, 2877, 1709, 1396, 1121, 1065 cm–1. HRMS:
˜
[4] S.-I. Murahashi, T. Shiota, Y. Imada, Org. Synth. 1991, 70,
found: m/z = 435.0520; calculated: m/z = 435.0545 [M + Na].
265–271.
2700
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2694–2701

Potential Precursors of Quaternary α-Amino Acids
FULL PAPER
[5]
[32] S. K. Patel, K. Murat, S. Py, Y. Vallee, Org. Lett. 2003, 5, 4081–
4084.
[33] S. Hanessian, M. Bayrakdarian, X. Luo, J. Am. Chem. Soc.
2002, 124, 4716–4721.
[34] For recent examples of C=N bond alkynylation reactions, see:
D. A. Black, B. A. Arndtsen, Org. Lett. 2004, 6, 1107–1110.
[35] C. Wei, C.-J. Li, J. Am. Chem. Soc. 2002, 124, 5638–5639.
[36] C. Fischer, E. M. Carreira, Org. Lett. 2004, 6, 1497–1499.
[37] J. F. Traverse, A. H. Hoveyda, M. L. Snapper, Org. Lett. 2003,
5, 3273–3275.
For oxidation with SeO2 or Na2WO4/UHP, see: E. Marcantoni,
M. Petrini, O. Polimanti, Tetrahedron Lett. 1995, 36, 3561–
3562.
For oxidation with MTO/H2O2, see: R. W. Murray, K. Iyanar,
J. Chen, J. T. Wearing, J. Org. Chem. 1996, 61, 8099–8102.
For oxidation with MTO/UHP, see: A. Goti, L. Nannelli, Tet-
rahedron Lett. 1996, 37, 6025–6028.
A. Goti, F. Cardona, G. Soldaini, Org. Synth. 2004, 81, 204–
212.
[6]
[7]
[8]
[9]
[10]
S. Yamazaki, Bull. Chem. Soc. Jpn. 1997, 70, 877–883.
[38] N. Gommermann, C. Koradin, K. Polborn, P. Knochel, Angew.
Chem. Int. Ed. 2003, 42, 5763–5766.
For N-phenylsulphonyl-C-phenyloxaziridine, see: W. W. Za-
jac, Jr., T. R. Walters, M. G. Darcy, J. Org. Chem. 1988, 53,
5856–5860.
For dimethyldioxirane, see: R. W. Murray, M. Singh, J. Org.
Chem. 1990, 55, 2954–2957.
D. Seebach, A. R. Sting, M. Hoffmann, Angew. Chem. Int. Ed.
Engl. 1996, 35, 2708–2748.
A. Vogt, H.-J. Altenbach, M. Kirschbaum, M. G. Hahn,
M. S. P. Matthaeus, A. R. Hermann, Eur. Pat. Appl. 99-114037,
2000.
B. Westermann, A. Walter, U. Floerke, H.-J. Altenbach, Org.
Lett. 2001, 3, 1375–1378.
A. Long, S. W. Baldwin, Tetrahedron Lett. 2001, 42, 5343–5346.
S. W. Baldwin, B. G. Young, A. T. McPhail, Tetrahedron Lett.
1998, 39, 6819–6822.
O. Tamura, K. Gotanda, J. Yoshino, Y. Morita, R. Terashima,
M. Kikuchi, T. Miyawaki, N. Mita, M. Yamashita, H. Ishiba-
shi, M. Sakamoto, J. Org. Chem. 2000, 65, 8544–8551.
[39] We observed that when stoichiometric amounts of diethylzinc
were present in the reaction of nitrones, trace amounts of the
ethyl adduct were sometimes observed. Replacing diethylzinc
with dimethylzinc suppressed this side reaction.
[40] We previously observed one occurrence of this cyclization reac-
tion (see ref.[20]): see the last entry of Table 1.
[41] Under the same conditions, the reaction of 2a led mainly to
decomposition products.
[42] See the Exp. Sect.
[11]
[12]
[13]
[43] P. Aschwanden, D. E. Frantz, E. M. Carreira, Org. Lett. 2000,
2, 2331–2333.
[44] See the Exp. Sect.
[14]
[15]
[16]
[45] Nevertheless, mass spectrometry indicated only 22% deute-
rium.
[46] M. Miura, M. Enna, K. Okuro, M. Nomura, J. Org. Chem.
1995, 60, 4999–5004.
[47] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[48] M. Murahashi, H. Mitsui, T. Shiota, T. Tsuda, S. Watanabe, J.
Org. Chem. 1990, 55, 1736–1744.
[49] Y. Inouye, Y. Watanabe, S. Takahashi, H. Kakisawa, Bull.
Chem. Soc., Jpn. 1979, 52, 3763–3764.
[50] A. Goti, L. Nannelli, Tetrahedron Lett. 1996, 37, 6025–6028.
[51] F. Lambert, B. Kirschleger, J. Villiéras, J. Organomet. Chem.
1991, 405, 273–286.
[52] R. F. W. Jackson, R. J. Moore, C. S. Dexter, J. Org. Chem.
1998, 63, 7875–7884.
[53] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R.Cheeseman, V. G. Zakrzewski, J. A. Mont-
gomery, Jr., R. E. Burant, J. C. Stratmann, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas,
J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson,
P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rab-
uck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Or-
tiz, A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Pis-
korz, P. Komaromi, I. Gomperts, R. R. Martin, R. L. Fox, D. J.
Keith, T. Al Laham, M. A. Peng, C. Y. A. Nanayakkara, C.
Gonzalez, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head Gor-
don, E. S. Replogle, J. A. Pople, Gaussian 98, Revision A6,
Gaussian, Inc.: Pittsburgh PA, 1998.
[17]
[18] N. Katagiri, M. Okada, Y. Morishita, C. Kaneko, Tetrahedron
1997, 53, 5725–5746.
[19] S. W. Baldwin, A. Long, Org. Lett. 2004, 6, 1653–1656.
[20] S. Pinet, S. U. Pandya, P. Y. Chavant, A. Ayling, Y. Vallee, Org.
Lett. 2002, 4, 1463–1466.
[21] S. K. Patel, S. Py, S. U. Pandya, P. Y. Chavant, Y. Vallee, Tetra-
hedron: Asymmetry 2003, 14, 525–528.
[22] R. Fitzi, D. Seebach, Tetrahedron 1988, 44, 5277–5292.
[23] R. Naef, D. Seebach, Helv. Chim. Acta 1985, 68, 135–143.
[24] R. Fitzi, D. Seebach, Angew. Chem. Int. Ed. Engl. 1986, 25,
345–346; a corrigendum to this publication has been published
in Angew. Chem. Int. Ed. Engl. 1986, 25, 766.
[25] R. W. Feenstra, E. H. M. Stokkingreef, A. M. Reichwein,
W. B. H. Lousberg, H. C. J. Ottenheijm, Tetrahedron 1990, 46,
1745–1756.
[26] Not isolated. We propose the structure (2S)-2-tert-butyl-2-hy-
droxy-3-methyl-2,3-dihydroimidazol-4-one 1-oxide on the basis
of NMR evidence.
[27] R. W. Murray, K. Iyanar, J. Chen, J. T. Wearing, Tetrahedron
Lett. 1996, 37, 805–808.
[28] For the NMR determination with O-acetylmandelic acid as
chiral shift reagent, see the Exp. Sect.
[29] S. Cicchi, M. Marradi, A. Goti, A. Brandi, Tetrahedron Lett.
2001, 42, 6503–6505.
[54] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[55] For IRC analysis, see: K. Fukui, Acc. Chem. Res. 1981, 14,
363–368.
[30] Determined by HPLC. In our hands Baldwin and Long’s pro-
cedure (see ref.[19]) produced 2a in 95% ee from enantiopure
3a (see Exp. Sect.).
[31] For nitrone alkynylation, see: D. E. Frantz, R. Faessler, E. M.
Carreira, J. Am. Chem. Soc. 1999, 121, 11245–11246.
[56] M. Head-Gordon, J. A. Pople, J. Chem. Phys. 1988, 89, 5777–
5786.
Received: December 3, 2004
Eur. J. Org. Chem. 2005, 2694–2701
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2701