Preparation and Use of Aziridino Alcohols as Promoters for the Enantioselective Addition of Dialkylzinc Reagents to N-(Diphenylphosphinoyl) Imines

Article abstract of DOI:10.1021/jo970918h

A set of chiral aziridino alcohols 2-5 has been synthesized starting from either readily available amino acids (L-serine, L-threonine, and allo-L-threonine) or simple olefins (using Sharpless asymmetric aminohydroxylation and dihydroxylation reactions). Chiral ligands 2-5 have been tested as promoters for the enantioselective addition of dialkylzinc reagents to N-(diphenylphosphinoyl) imines 1. The influence of the substituants on the aziridine ring and the alcohol moiety on the selectivity has been studied, and in the best case, an enantiomeric excess of up to 94% could be obtained. Acidic hydrolysis of the initially formed N-protected amines 6 led to the corresponding free amines 7 without racemization. Although a stoichiometric amount of the ligand was used, about 90% of it could be recovered during the workup and reused without significant loss of chiral induction. The utility of the aziridino alcohols 2-5 as catalysts for the same reaction has also been evaluated and enantiomeric excesses of up to 76% were achieved using 0.25 equiv of the chiral ligand. A possible transition state for the addition reaction is also proposed.

Full text of DOI:10.1021/jo970918h

7364
J . Org. Chem. 1997, 62, 7364-7375
P r ep a r a tion a n d Use of Azir id in o Alcoh ols a s P r om oter s for th e
En a n tioselective Ad d ition of Dia lk ylzin c Rea gen ts to
N-(Dip h en ylp h osp h in oyl) Im in es
Pher G. Andersson,*,† David Guijarro,† and David Tanner*,‡
Department of Organic Chemistry, Uppsala University, Box 531, S-751 21 Uppsala, Sweden, and
Department of Organic Chemistry, The Technical University of Denmark, Building 201,
DK-2800, Lyngby, Denmark
Received May 23, 1997X
A set of chiral aziridino alcohols 2-5 has been synthesized starting from either readily available
amino acids (L-serine, L-threonine, and allo-L-threonine) or simple olefins (using Sharpless
asymmetric aminohydroxylation and dihydroxylation reactions). Chiral ligands 2-5 have been
tested as promoters for the enantioselective addition of dialkylzinc reagents to N-(diphenylphos-
phinoyl) imines 1. The influence of the substituents on the aziridine ring and the alcohol moiety
on the selectivity has been studied, and in the best case, an enantiomeric excess of up to 94% could
be obtained. Acidic hydrolysis of the initially formed N-protected amines 6 led to the corresponding
free amines 7 without racemization. Although a stoichiometric amount of the ligand was used,
about 90% of it could be recovered during the workup and reused without significant loss of chiral
induction. The utility of the aziridino alcohols 2-5 as catalysts for the same reaction has also
been evaluated and enantiomeric excesses of up to 76% were achieved using 0.25 equiv of the chiral
ligand. A possible transition state for the addition reaction is also proposed.
In tr od u ction
In 1990, Tomioka and co-workers reported the addition
of organolithium compounds to N-aryl imines in the
presence of stoichiometric amounts of chiral amino
ethers.7a The corresponding secondary amines were
isolated with enantiomeric purities ranging from 48 to
75%. Shortly after this report, the same authors devel-
oped a catalytic process, using a substoichiometric amount
of the chiral ligand.7b In 1991, Itsuno and co-workers
studied the addition of butyllithium to N-silyl imines in
the presence of chiral promoters such as alcohols, diols,
and amino alcohols.8a The same group has reported on
the addition of butyllithium to N-alumino, N-boryl, and
N-silyl imines in the presence of chiral nitrogen ligands
including (-)-sparteine and proline-derived amino al-
cohols.8b Chiral bis(oxazolines) have been applied as
promoters for the addition of organolithium reagents to
N-aryl imines,9 and ee’s of up to 91% were obtained.
Despite the progress in catalytic enantioselective ad-
dition of dialkylzinc reagents to aldehydes, the applica-
tion of the same methodology to N-aryl or N-silyl imines
failed.8a These substrates are unreactive with diethylzinc
even in the presence of a stoichiometric amount of an
amino alcohol at high temperatures. The use of activated
N-acyl or N-phosphinoyl imines has turned out to be
crucial, and the first example of enantioselective addition
of diethylzinc to the imine function was reported by
Katritzky and co-workers in 1992.10 Diethylzinc was
reacted with N-(amidobenzyl)benzotriazoles (masked N-
acyl imines) to afford the corresponding optically active
amides in ee’s up to 76%. Soon afterward, Soai and co-
workers found that dialkylzinc reagents added to N-
Nucleophilic addition of organometallic reagents to
carbonyl compounds is a very important operation in
organic synthesis, and the asymmetric version of this
reaction is particularly useful. Several highly enantio-
selective additions to prochiral aldehydes leading to
optically active alcohols have been reported,1 and among
the various organometallic compounds, diorganozincs
serve as excellent alkyl nucleophiles. â-Amino alcohols
have proved to be one of the most efficient types of
catalyst for this reaction, and ee’s up to 99% have been
achieved.2
Although enormous progress has been made in the
stereoselective addition of nucleophiles to aldehydes,
considerably less attention has been paid to the corre-
sponding enantioselective conversion of imines to chiral
amines. Optically active amines are important com-
pounds extensively utilized as resolving agents,3 starting
materials for the preparation of biologically active sub-
stances,4 and chiral auxilliaries in asymmetric synthesis.5
However, only a few examples of enantioselective alky-
lation of the imine functionality have been reported,6 and
this is primarily due to the relatively poor electrophilicity
of the CdN function and the tendency of enolizable
imines to undergo deprotonation instead of addition.
† Uppsala University.
‡ The Technical University of Denmark.
X Abstract published in Advance ACS Abstracts, September 15, 1997.
(1) For a review on this subject, see: Solladie´, G. In Asymmetric
Synthesis; Morrison, J . D., Ed.; Academic Press: New York, 1983; Vol.
2A, Chapter 6.
(2) For a review, see: (a) Noyori, R.; Kitamura, M. Angew. Chem.,
Int. Ed. Engl. 1991, 30, 49. (b) Soai, K.; Niwa, S. Chem. Rev.
(Washington, D.C.) 1992, 92, 833.
(3) J acques, J .; Collet, A.; Wilen, S. H. In Enantiomers, Racemates
and Resolution; J ohn Wiley and Sons: New York, 1981.
(4) Moser, H.; Rihs, G.; Santer, H. Z. Naturforsch. 1982, 376, 451.
(5) For a review, see: Whitesell, J . K. Chem. Rev. (Washington, D.C.)
1989, 89, 1581.
(7) (a) Tomioka, K.; Inoue, I.; Shindo, M.; Koga, K. Tetrahedron Lett.
1990, 31, 6681. (b) Tomioka, K.; Inoue, I.; Shindo, M.; Koga, K.
Tetrahedron Lett. 1991, 32, 3095.
(8) (a) Itsuno, S.; Yanaka, H.; Hachisuka, C.; Ito, K. J . Chem. Soc.,
Perkin Trans. 1 1991, 1341. (b) Itsuno, S.; Sasaki, M.; Kuroda, S.; Ito,
K. Tetrahedron: Asymmetry 1995, 6, 1507.
(9) Denmark, S. E.; Nakajima, N.; Nicaise, O. J .-C. J . Am. Chem.
Soc. 1994, 116, 8797.
(10) Katritzky, A. R.; Harris, P. A. Tetrahedron: Asymmetry 1992,
3, 437.
(6) For a review, see: (a) Denmark, S. E.; Nicaise, O. J .-C. J . Chem.
Soc., Chem. Commun. 1996, 999. (b) Enders, D.; Reinhold: U.
Tetrahedron: Asymmetry 1997, 8, 1895.
S0022-3263(97)00918-3 CCC: $14.00 © 1997 American Chemical Society

Aziridino Alcohols as Additions Promoters
J . Org. Chem., Vol. 62, No. 21, 1997 7365
Sch em e 1a
a
Key: (i) R′2Zn, ligand 2-5 (see Figure 1), toluene, 0 °C to rt;
(ii) 1.5 M HCl in MeOH, rt; (iii) 15% NaOH.
(diphenylphosphinoyl) imines in the presence of catalytic
or stoichiometric amounts of chiral â-amino alcohols with
high enantioselectivities (75-91% ee).11 Ukaji and co-
workers have used a nitrone as the acceptor,12 reacting
dialkylzincs with 3,4-dihydroisoquinoline N-oxide deriva-
tives to give the corresponding alkylated hydroxylamines
with ee’s ranging from 25 to 91%.
For the past few years, we have been exploring the use
of chiral aziridines in asymmetric synthesis. Chiral
aziridines are available in enantiomerically pure (or
highly enriched) form by a variety of procedures,13 and
they are suitable building blocks for the synthesis of
biologically active species.14 The utility of C2-symmetric
aziridines as auxilliaries for asymmetric alkylation and
aldol reactions of amide enolates has also been demon-
strated.15 We have evaluated C2-symmetric bis(aziri-
dines) as ligands for a variety of metal-mediated asym-
metric transformations,16 and in the best cases, ee’s of
>99% were obtained. The aziridine ligand is unusual
in that slow pyramidal inversion introduces a stereogenic
nitrogen.13
F igu r e 1.
Sch em e 2a
Recently, we have reported in a preliminary com-
munication the use of some aziridino alcohols as promot-
ers for the addition of diethylzinc to N-(diphenylphos-
phinoyl) imines17 (Scheme 1). In this paper we present
the extension of this study to some new aziridine ligands
2-5 (Figure 1) and substrates. By fine-tuning of the
ligand structure, enantioselectivities of up to 94% ee
could be obtained. A suggestion about a possible transi-
tion state for the addition reaction is proposed. Full
experimental details about the preparation of the ligands
and the addition reaction are given. We have also
evaluated the utility of ligands 2-5 as catalysts in the
same reaction. Acidic hydrolysis of the addition products
6 afforded the free amines 7, without loss of enantiomeric
purity (Scheme 1).
a
Key: (i) 2 M HCl in MeOH, reflux; (ii) Et3N, PhCHO, MeOH,
0 °C; (iii) NaBH4, 70% (three steps); (iv) PPh3, CBr4, Et3N, CH2Cl2/
CHCl3, rt and then reflux, 39%; (v) LiAlH4, THF, -40 °C to rt,
74%.
of the chiral pool, (b) Sharpless asymmetric aminohy-
droxylation,18 and (c) Sharpless asymmetric dihydroxy-
lation.19
(a ) Usin g t h e Ch ir a l P ool: Syn t h esis of t h e
Liga n d s 2, 3a -h , 4a ,b, a n d 5. The starting materials
for the preparation of the aziridino alcohols 2, 3a -h ,
4a ,b, and 5 were the readily available amino acids
L-serine, L-threonine, and allo-L-threonine. The synthesis
of aziridino alcohol 2 is shown in Scheme 2. L-Serine 8
was converted into its methyl ester hydrochloride.20 This
was followed by a reductive alkylation,21 to give the
N-benzyl amino ester 9, and reaction of 9 with Ph3P and
CBr422 afforded the aziridino ester 10. The ligand 2 was
then obtained by reduction of 10 with LiAlH4.
Syn th esis of th e Liga n d s
Figure 1 shows the aziridino alcohols that have been
prepared and tested as promoters of the above-mentioned
reaction. Three different approaches were used to obtain
the ligands in enantiomerically pure form: (a) the use
(11) Soai, K.; Hatanaka, T.; Miyazawa, T. J . Chem. Soc., Chem.
Commun. 1992, 1097.
(12) Ukaji, Y.; Kenmoku, Y.; Inomata, K. Tetrahedron: Asymmetry
1996, 7, 53.
(13) For a review, see: Tanner, D. Angew. Chem., Int. Ed. Engl.
1994, 33, 599.
Starting from L-threonine 11 and following the same
sequence of esterification-reductive alkylation as before,
(14) See, for example: (a) Oppolzer, W.; Flaskamp, E. Helv. Chim.
Acta 1977, 60, 204. (b) Baldwin, J . E.; Adlington, R. M.; Robinson, N.
G. J . Chem. Soc., Chem. Commun. 1987, 153. (c) Tanner, D.; Somfai,
P. Tetrahedron Lett. 1987, 28, 1211. (d) Tanner, D.; He, H. M.
Tetrahedron 1992, 48, 6079.
(15) Tanner, D.; Birgersson, C.; Gogoll, A.; Luthman, K. Tetrahedron
1994, 50, 9797.
(18) Li, G.; Chang, H.-T.; Sharpless, K. B. Angew. Chem., Int. Ed.
Engl. 1996, 35, 451.
(19) (a) Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G.
A.; Hartung, J .; J eong, K.-S.; Kwong, H.-L.; Morikawa, K.; Wang, Z.-
M.; Xu, D.; Zhang, X.-L. J . Org. Chem. 1992, 57, 2768. (b) Kolb, H. C.;
VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev. (Washington, D.C.)
1994, 94, 2483.
(16) (a) Tanner, D.; Andersson, P. G.; Harden, A.; Somfai, P.
Tetrahedron Lett. 1994, 35, 4631. (b) Andersson, P. G.; Harden, A.;
Tanner, D.; Norrby, P.-O. Chem. Eur. J . 1995, 1, 12. (c) Tanner, D.;
Harden, A.; J ohansson, F.; Wyatt, P.; Andersson, P. G. Acta Chem.
Scand. 1996, 50, 361.
(20) Morell, J . L.; Fleckenstein, P.; Gross, E. J . Org. Chem. 1977,
42, 355.
(21) Thompson, C. M.; Frick, J . A.; Green, D. L. C. J . Org. Chem.
1990, 55, 111.
(22) Ha¨ner, R.; Olano, B.; Seebach, D. Helv. Chim. Acta 1987, 70,
1676.
(17) Andersson, P. G.; Guijarro, D.; Tanner, D. Synlett 1996, 727.

7366 J . Org. Chem., Vol. 62, No. 21, 1997
Andersson et al.
Sch em e 3a
Sch em e 5a
a
Key: Tr ) Ph3C. (i) CF3CO2H, CHCl3/MeOH (1/1, v/v), -5 °C,
a
Key: (i) 2 M HCl in MeOH, reflux; (ii) Et3N, PhCHO, MeOH,
77%; (ii) MeI, K2CO3, 18-crown-6 (cat.), THF, rt, 44% (for com-
pound 17f); (iii) R1Br (R1 ) n-Bu, i-Pr), K2CO3, 18-crown-6 (cat.),
NaI (only for R1 ) i-Pr), MeCN, reflux, 70 and 85%, respectively
(for compounds 17g,h ); (iv) LiAlH4, Et2O, 0 °C to rt, 50, 52, 70,
and 65%, respectively.
0 °C; (iii) NaBH4, 69% (three steps); (iv) PPh3, CCl4, Et3N, MeCN,
rt, 67%; (v) LiAlH4, THF, 0 °C to rt, 90% (for compound 3a ); (vi)
MeLi (for compound 3b) or PhMgBr (for compound 3c), THF, -78
°C to rt, 83 and 78%, respectively.
Sch em e 4a
Sch em e 6a
a
Key: Tr ) Ph3C. (i) SO2Cl2, Et3N, toluene, -50 °C, 72%; (ii)
LiAlH4, THF, -40 °C to rt, 78%.
the protected amino ester 12 was obtained (Scheme 3).
It was then cyclized to the aziridino ester 13 by reaction
with Ph3P in the presence of CCl4.23 Reduction with
LiAlH4 led to the aziridino alcohol 3a . Treatment of 13
with excess of MeLi or PhMgBr afforded the ligands 3b
and 3c, respectively.
a
Key: (i) 2 M HCl in MeOH, reflux; (ii) Et3N, MeOH, rt; (iii)
PhCOCl, 0 °C, 84% (three steps); (iv) SOCl2, 0 °C; (v) 6 M HCl,
reflux, 86% (two steps).
According to a literature procedure,24 L-threonine was
transformed into the N-trityl benzyl ester 14 (Scheme
4). This was cyclized to the aziridino ester 15 by reaction
with SO2Cl2,25 and 15 was then reduced with LiAlH4 to
yield the aziridino alcohol 3d .
of NaI was added to the reaction mixture to aid the
substitution reaction by in situ generation of i-PrI. The
aziridino esters 16 and 17f-h were then reduced with
LiAlH4, to yield the desired aziridino alcohols 3e-h .
Each of the ligands 2 and 3a -h were obtained as single
invertomers, due to a relatively high inversion barrier
at nitrogen.13 NOE studies on the ligand 3a showed, as
expected, that the benzyl group on the nitrogen and the
hydroxymethyl group on the aziridine ring had a trans
relationship to each other (see experimental part).
For the preparation of ligands 4a ,b, with opposite
absolute configuration at C3 in the aziridine ring, the
required allo-L-threonine was prepared from L-threonine
in a five-step sequence.27 First, L-threonine 11 was
converted into its methyl ester hydrochloride.20 After
treatment with Et3N, it was reacted in situ with PhCOCl,
to give the N-benzoyl amino ester 18 (Scheme 6). Treat-
The N-H aziridino ester 16 (Scheme 5) was thought
to be a suitable precursor for the synthesis of the ligands
3e-h . The detritylation of 15 was first attempted with
HCO2H in MeOH,25 but this was unsuccessful. The
deblocking was finally achieved by treatment of 15 with
CF3CO2H.24a Reduction of 16 with LiAlH4 then yielded
ligand 3e. The N-methyl aziridino ester 17f was pre-
pared by treatment with K2CO3 and MeI in THF in the
presence of 18-crown-6.26 The alkylation procedure used
for the preparation of 17f was also applied to the
synthesis of both 17g and 17h , but no alkylated product
was formed after several days in refluxing THF. How-
ever, changing the solvent to MeCN afforded the expected
alkylated products 17g and 17h in 70 and 85% yield,
respectively. In the case of 17h , a stoichiometric amount
27
ment of 18 with SOCl2 gave the expected oxazoline 19,
and hydrolysis in 6 M HCl at reflux27 yielded allo-L-
threonine 20 in good overall yield.
allo-L-Threonine was transformed into the aziridino
ester 21 according to Scheme 7. Reduction of 21 with
LiAlH4 afforded ligand 4a , which was obtained as a
mixture of invertomers at nitrogen in a 7:1 ratio, the
major isomer being the one having the benzyl and the
methyl groups in a cis relationship (see the Experimental
Section). Treatment of 21 with excess of EtMgBr led to
(23) For a similar process, see: Shaw, K. J .; Luly, J . R.; Rapoport,
H. J . Org. Chem. 1985, 50, 4515.
(24) (a) Nakajima, K.; Takai, F.; Tanaka, T.; Okawa, K. Bull. Chem.
Soc. J pn. 1978, 51, 1577. (b) Tanaka, T.; Nakajima, K.; Okawa, K.
Bull. Chem. Soc. J pn. 1980, 53, 1352. (c) Wakamiya, T.; Shimbo, K.;
Shiba, T.; Nakajima, K.; Neya, M.; Okawa, K. Bull. Chem. Soc. J pn.
1982, 55, 3878.
(25) Kuyl-Yeheskiely, E.; Lodder, M.; van der Marel, G. A.; van
Boom, J . H. Tetrahedron Lett. 1992, 33, 3013.
(26) For a similar process, see: Åhman, J .; Somfai, P. Synth. Comm.
1994, 24, 1121.
(27) Elliott, D. F. J . Chem. Soc. 1950, 62.

Aziridino Alcohols as Additions Promoters
J . Org. Chem., Vol. 62, No. 21, 1997 7367
Sch em e 7a
Sch em e 9a
a
Key: (i) 2 M HCl in MeOH, reflux; (ii) Et3N, PhCHO, MeOH,
0 °C; (iii) NaBH4, 84% (three steps); (iv) PPh3, CCl4, Et3N, MeCN,
rt, 66%; (v) LiAlH4, THF, 0 °C to rt, 74% (for compound 4a ) or
EtMgBr, THF, -78 °C to rt, 82% (for compound 4b).
Sch em e 8a
a
Key: (i) K2OsO2(OH)4 (4 mol %), (DHQ)2-PHAL (5 mol %),
TsN(Cl)Na‚3H2O, t-BuOH/H2O (v/v ) 1/1), rt, 51%; (ii) 45% HBr
in MeCO2H, PhOH, 75 °C; (iii) 2 M HCl in MeOH, reflux; (iv) Et3N,
PhCHO, MeOH, 0 °C; (v) NaBH4, 41% (four steps); (vi) PPh3, CCl4,
Et3N, MeCN, rt, 86%; (vii) LiAlH4, Et2O, 0 °C to rt, 94% (for
compound 3i) or MeMgBr, THF, -78 to -5 °C, 89% (for compound
3j).
Sch em e 10a
a
Key: (i) TBDMSCl, imidazole, DMF, rt, 94%; (ii) DIBALH,
CH2Cl2, 0 °C to rt, 58%; (iii) PPh3, DEAD, THF, 0 °C to rt, 82%;
(iv) (n-Bu)4N+F-, THF, rt, 69%.
the aziridino alcohol 4b. Only one isomer was detected
in this case, with the same stereochemistry as the major
invertomer of 4a .
The ligand 5 was also prepared from L-threonine. The
N-benzyl amino ester 12 was O-protected with t-BuMe2-
SiCl in the presence of imidazole16c (Scheme 8). Reduc-
tion with DIBALH afforded the amino alcohol 23, which
was then cyclized to the aziridine 24 under Mitsunobu
conditions.16c Finally, 24 was O-deprotected by reaction
with (n-Bu)4NF,28 to yield the desired aziridino alcohol
5.
a
Key: (i) K2OsO2(OH)4 (0.2 mol %), (DHQ)2-PHAL (0.5 mol
%), NMO, t-BuOH, rt, 72%; (ii) Me(OMe)3, p-TsOH, CH2Cl2, rt;
(iii) MeCOBr, CH2Cl2, -20 °C to rt; (iv) K2CO3, MeOH, -20 °C,
76% (three steps); (v) NaN3, NH4Cl, MeOH, reflux; 95%; (vi) H2,
5% Pd-C, MeOH, rt; (vii) 2 M HCl in MeOH, reflux; (viii) Et3N,
PhCHO, MeOH, 0 °C; (ix) NaBH4, 56% (four steps); (x) PPh3, CCl4,
Et3N, MeCN, rt, 88%; (xi) LiAlH4, THF, 0 °C to rt, 59%.
(b) Th e Am in oh yd r oxyla tion Ap p r oa ch : Syn th e-
sis of th e Liga n d s 3i,j. To investigate the effect of the
substituent at C3 in the aziridine ring, ligands 3i,j were
prepared. The amino acid 27 was thought to be a good
precursor for their synthesis and it was obtained as
shown in Scheme 9. First, methyl cinnamate 25 was
submitted to Sharpless asymmetric aminohydroxyla-
tion,18 which gave 26 in 97% ee after recrystallization.
26 was transformed into the free amino acid 27 by
deprotection with HBr in AcOH.29 The latter was then
converted into the N-benzyl amino ester 28 following the
sequence esterification-reductive alkylation at nitrogen21
(vide supra). It was possible to improve the optical purity
at this stage, and after recrystallization of 28 from
pentane/CH2Cl2, only one enantiomer could be detected
by HPLC analysis (see the Experimental Section). The
method that had been used for the preparation of ligand
3a was applied for the cyclization step,23 and the aziridino
ester 29 was obtained in 35% overall yield (from 26). The
reduction of the aziridino ester 29 with LiAlH4 afforded
the ligand 3i. Treatment of 29 with excess of MeMgBr
then led to the aziridino alcohol 3j.
(c) Th e Dih yd r oxyla tion Ap p r oa ch : Syn th esis of
th e Liga n d 4c. Sharpless asymmetric dihydroxylation
protocol19 was used to establish the required stereochem-
istry of ligand 4c. Methyl cinnamate 25 was converted
into the diol 3030 (Scheme 10), which was obtained in 98%
ee after recrystallization. Compound 30 was stereospe-
cifically transformed into the epoxide 31.31 Ring opening
of the epoxide with NaN3,32 followed by hydrogenation
and reductive alkylation21 of the resultant amine, led to
the N-protected amino ester 33. Cyclization as previ-
ously described for 12 (Scheme 3) and reduction of the
ester with LiAlH4 afforded the aziridino alcohol 4c.
Resu lts a n d Discu ssion
The aziridino alcohols 2-5 were tested as promoters
for the addition of dialkylzinc reagents to several N-
(diphenylphosphinoyl) imines. The required imines 1
(30) Wang, Z.-M.; Kolb, H. C.; Sharpless, K. B. J . Org. Chem. 1994,
59, 5104.
(28) Corey, E. J .; Venkateswarlu, A. J . Am. Chem. Soc. 1972, 94,
6190.
(29) Li, G.; Sharpless, K. B. Acta Chem. Scand. 1996, 50, 649.
(31) Kolb, H. C.; Sharpless, K. B. Tetrahedron 1992, 48, 10515.
(32) Legters, J .; Thijs, L.; Zwanenburg, B. Tetrahedron Lett. 1989,
30, 4881.

7368 J . Org. Chem., Vol. 62, No. 21, 1997
Andersson et al.
Sch em e 11a
ee was obtained when the methyl group on the aziridine
ring, R3 (Figure 1), was exchanged for a phenyl group
(Table 1, entry 13). When the steric bulk of the side chain
bearing the hydroxyl group was varied, a decrease in
enantioselectivity was found upon increasing the size of
the R2 substituent in ligands 3 (compare entry 3 with 6
and 7 and entry 13 with 15). Surprisingly, ligand 3c
induced the opposite configuration in the product (Table
1, entry 7) for reasons not yet understood. In the case of
the trans ligands 4, a preference for the S enantiomer
was observed when R2 (Figure 1) was changed to a larger
group (compare entries 17 and 21).
a
Key: (i) R′2Zn (3 equiv), ligand 2-5 (1 equiv), toluene, 0 °C to
rt.
The size of the substituent on nitrogen is also impor-
tant. The simplest ligand, 3e, proved to be unsuitable
for promotion of the addition reaction. No reaction
product was detected after 7 days at rt (Table 1, entry
9). When R1 was an alkyl group, however, good selectivi-
ties were observed. The ee increased upon going from a
primary (Table 1, entries 10 and 11) to a secondary (Table
1, entry 12) alkyl group. According to the result obtained
with the ligand 3a , it seems that the presence of an
aromatic ring in R1 is very important. Ligand 3a , in
which R1 is a benzyl group, gave the highest enantiose-
lectivity (Table 1, entry 3). Increasing the steric bulk of
R1 resulted in an unexpected inversion of the stereo-
chemical outcome of the reaction. When ligand 3d ,
bearing a trityl group on the nitrogen, was used (Table
1, entry 8), the reaction was very sluggish, and we
presume that the very large trityl group inhibits coordi-
nation between the zinc and the nitrogen.
Comparison of entries 3 and 17 clearly shows the
importance of the stereochemistry at C3. Almost racemic
product was obtained when ligand 4a , derived from allo-
L-threonine, was tested. A similar effect was observed
when phenyl was the substituent at C3. The ee dropped
from 86% for ligand 3i (entry 13) to 38% for ligand 4c
(entry 22).
Introduction of a stereocenter at the carbinol site
(ligand 5, entry 23) improved matters somewhat as
compared to ligand 2 (Table 1, entry 1), but we have as
yet no convincing explanation for the fact that both of
these ligands, although having different absolute config-
uration of the aziridine backbone, induce the same sense
of chirality in the product.
Concerning the substrates, imines 1b and 1c were
shown to be less effective than 1a in the enantioselective
addition reaction, although good selectivities were ob-
served with the ligand 3a (Table 1, entries 4 and 5).
However, imine 1d , derived from acetophenone, turned
out to be inert under the standard reaction conditions.
No addition product was detected in the reaction mixture
after 4 days at rt (Table 1, entry 20).
Me2Zn turned out to be much less reactive than Et2-
Zn, and a larger excess of the former as well as much
longer reaction times were necessary to achieve moderate
conversions and low to moderate ee’s (Table 1, entries 2
and 16).
In all cases, although a stoichiometric amount of the
chiral ligand was used, up to 90% of it could be recovered
during the workup in a typical experiment. The recov-
ered ligand could be used in further experiments without
loss of chiral induction.34
Based on the results in Table 1, the stereochemical
outcome of the reaction could be rationalized by assuming
that the addition takes place through the transition state
depicted in Figure 2 (for ligand 3a and imine 1a ). The
steric bulk of the substituent at C3 on the aziridine ring
Ta ble 1. Ad d ition Rea ction of R′2Zn to 1 P r om oted by
th e Liga n d s 2-5
entry imine ligand R′ product yield (%)a ee (%)b configc
1
2
3
4
5
6
7
8
9
1a
1a
1a
1b
1c
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1c
1d
1a
1a
1a
2
Et
Med
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Meh
Et
Et
Et
Et
Et
Et
Et
6a b
6a a
6a b
6b
18
68
63
59
48
92
56
33
12
75
94
80
82
59
16
3
R
R
R
R
Re
R
S
3a
3a
3a
3a
3b
3c
3d
3e
3f
3g
3h
3i
6c
6a b
6a b
6a b
f
6a b
6a b
6a b
6a b
6a b
6a b
6a a
6a b
6b
S
10
11
12
13
14
15
16
17
18
19
20
21
22
23
78
66
72
57
57
60
36
83
21
59
80
83
91
86
50
52
4
5
6
11
R
R
R
R
R
R
R
R
R
Re
3ig
3j
4a
4a
4a
4a
4a
4b
4c
5
6c
i
6a b
6a b
6a b
63
63
45
44
38
55
S
R
R
a
Isolated yield after flash chromatography (silica gel, pentane/
acetone). bDetermined by HPLC analysis on
(ChiralCel OD-H). Determined by comparison of the optical
rotation of the free amine with the data given in the literature
(see the Experimental Section). dSix equivalents of dimethylzinc
a chiral column
c
e
was used and the reaction was run for 11 days at rt. Tentatively
assigned by the retention times of the enantiomers on the HPLC
analysis, according to our observations with the products 6a ,b.
fAfter 7 days at rt, no product was detected (1H NMR). gThe optical
h
purity of the ligand was 50%. Six equivalents of dimethylzinc
i
was used and the reaction was running for 8 days at rt. After 4
days at rt, no product was detected (1H NMR).
(Scheme 11) were prepared from the corresponding
aldehydes or ketones, according to the literature.33 As a
general procedure, the dialkylzinc reagent (3 equiv) was
added dropwise to a stirred solution of the imine 1 and
the corresponding aziridino alcohol 2-5 (1 equiv) in dry
toluene, under nitrogen, at 0 °C. The temperature was
allowed to rise slowly to rt and then the reaction was
stirred for 4 days. After quenching with saturated
aqueous NH4Cl and usual workup, the expected products
6 were obtained (Scheme 11, Table 1).
Table 1 shows the yields and ee’s obtained with all the
ligands 2-5. The imine 1a proved to be the best
substrate for the addition reaction. Ligand 2, derived
from serine, gave only poor conversion and low ee (Table
1, entry 1). However, introduction of a methyl group as
in 3a resulted in a dramatic increase in both yield and
enantioselectivity (Table 1, entry 3). A somewhat lower
(33) (a) J ennings, W. B.; Lovely, C. J . Tetrahedron 1991, 47, 5561.
(b) Krzyzanowska, B.; Stec, W. J . Synthesis 1982, 270.

Aziridino Alcohols as Additions Promoters
J . Org. Chem., Vol. 62, No. 21, 1997 7369
Ta ble 2. Ad d ition Rea ction of Et2Zn to 1a Ca ta lyzed by
th e Liga n d s 3-4
ee cat./
entry ligand equiv yield (%)a ee (%)b configc stoichd
1
2
3
4
5
6
7
8
9
10
11
12
13
3a
3a
3a
3b
3c
3f
3g
3h
3i
3i
3j
4b
4c
0.50
0.25
0.10
0.25
0.25
0.25
0.25
0.25
0.50
0.25
0.25
0.25
0.25
76
68
37
24
50
52
63
60
64
50
50
47
48
87
75
49
33
9
45
60
65
81
76
22
17
28
R
R
R
R
S
R
R
R
R
R
R
S
0.93
0.80
0.52
0.56
0.56
0.56
0.72
0.71
0.94
0.88
0.42
0.39
0.73
F igu r e 2.
R
Sch em e 12a
a
Isolated yield after flash chromatography (silica gel, pentane/
acetone). bDetermined by HPLC analysis on
(ChiralCel OD-H). Determined by comparison of the optical
rotation of the free amine with the data given in the literature
a chiral column
c
d
(see the Experimental Section). Ratio between the ee obtained
in the catalytic and in the corresponding stoichiometric reaction
(see Table 1).
a
Key: (i) Et2Zn (3 equiv), ligand 3a (1 equiv), toluene, 0 °C,
45%, racemic.
substrate for the study, because this pair gave the best
ee in the stoichiometric reaction. Several addition reac-
tions of Et2Zn (3 equiv) to the imine 1a were set up with
different amounts of the ligand 3a (0.50, 0.25, and 0.10
equiv) under the same conditions as for the stoichiometric
reaction. The results are shown in Table 2 (entries 1-3).
As can be seen, the ee decreased upon reducing the
amount of ligand. When 0.10 equiv of the ligand was
used, the reaction was much slower than in the other
cases, indicating a ligand acceleration effect.36 After the
normal reaction time (4 days at rt), only 37% of the
addition product was isolated after workup, together with
some unreacted starting material. However, moderate
enantioselectivity was still obtained (Table 2, entry 3).
When 0.50 and 0.25 equiv of the ligand were used,
conversions obtained were similar to the one observed
in the stoichiometric reaction (Table 1, entry 3) and with
the same reaction times, although the products were
obtained with a slight decrease in the enantioselectivity.
Since the ee obtained with 0.25 equiv of the ligand was
still relatively good, aziridino alcohols 3b,c,f-j and 4b,c
were also tested in the catalytic reaction (0.25 equiv of
the ligand) in order to get some information about the
optimum structure for the catalyst. The rest of the
ligands were not evaluated because of their inefficiency
to induce enantioselectivity in the stoichiometric reaction.
The results are summarized in Table 2. To determine
which is the ligand structure that leads to the smallest
loss of enantioselectivity, the ratio between the ee in the
catalytic and in the stoichiometric reactions was calcu-
lated (Table 2, last column). As can be seen, that ratio
varied from 0.39 to 0.88. Again, ligands with a primary
alcohol on the side chain and an R-branched substituent
at nitrogen gave less loss of selectivity (Table 2, entries
2, 7, 8, 10, and 13). The presence of an aromatic ring on
the substituent at nitrogen seems to give an additional
beneficial effect (Table 2, entries 2 and 10), possibly by
π-π stacking in the transition state.37 Ligand 3i, with
forces ZnB to be in the position shown, determining in
this way the stereochemistry at ZnA. The extra coordina-
tion between the oxygen of the imine and ZnB would lead
to the formation of a bicyclic transition state. The
stereochemistry observed in the product would be ob-
tained if an ethyl group from ZnB was transferred to the
imine carbon. For steric reasons, it appears advanta-
geous to have a small group on the hydroxylic carbon,
and this assumption was borne out experimentally. π-π
Interactions between the phenyl group of R1 and the
aromatic group on the imine carbon are expected to
stabilize the depicted transition state to some extent,
leading to the improved selectivity observed. The loss
of enantioselectivity found with ligands 2 and 4, having
no substituent or inverted configuration at C3, might be
due to the possibility of forming two different transition
states with opposite configuration at ZnA.
Another imine with a different electron-withdrawing
group on the nitrogen was also tested. Et2Zn (3 equiv)
was added dropwise to a stirred solution of the N-tosyl
imine 3533a and the aziridino alcohol 3a (1 equiv) in dry
toluene, under nitrogen, at 0 °C (Scheme 12). The
reaction was complete after 38 h at 0 °C. Workup as for
the addition to imines 1 afforded the expected product
3635 in 45% isolated yield, but in racemic form. Assuming
that the reaction takes place through a transition state
similar to the one depicted in Figure 2, the possibility of
coordination of both oxygens of the sulfonyl group to ZnB-
would allow the imine to aproach the complex between
the aziridino alcohol moiety and Et2Zn in two different
ways: the one shown in Figure 2 and another one in
which the imine is turned 180° around the N-ZnA axis.
This would explain the obtention of the racemic product
in this case.
The good enantioselectivities obtained with some of the
ligands prompted us to test them in a catalytic reaction.
Ligand 3a was chosen as catalyst and imine 1a as
(34) A sample of ligand 3a was recovered from two successive
stoichiometric reactions and it was used to set up a catalytic (0.50 equiv
of 3a ) addition reaction to the imine 1a . The expected product was
isolated in 79% yield and 81% ee, a value that is only slightly lower
than the one obtained in the same catalytic reaction but using fresh
ligand (Table 2, entry 1).
(36) Berrisford, D. J .; Bolm, C.; Sharpless, K. B. Angew. Chem., Int.
Ed. Engl. 1995, 34, 1059.
(37) For a similar discussion concerning the AD reaction, see: Kolb,
H. C.; Andersson, P. G.; Sharpless, K. B. J . Am. Chem. Soc. 1994, 116,
1278.
(38) Bedekar, A. V.; Koroleva, E. B.; Andersson, P. G. J . Org. Chem.
1997, 62, 2518.
(35) Sisko, J .; Weinreb, S. M. J . Org. Chem. 1990, 55, 393.

7370 J . Org. Chem., Vol. 62, No. 21, 1997
Andersson et al.
Ta ble 3. Stu d y on th e Va r ia tion of th e ee w ith th e
Rea ction Tim e
Con clu sion s
A set of new chiral aziridino alcohols 2-5 has been
synthesized and tested as promoters for the enantiose-
lective addition of dialkylzinc reagents to N-(diphen-
ylphosphinoyl) imines 1. In the best case, an enantiose-
lectivity of up to 94% could be obtained. Acidic hydrolysis
of the initially formed N-protected amines 6 led to the
corresponding free amines 7 without racemization. Al-
though a stoichiometric amount of the ligand was used,
about 90% of it could be recovered during the workup
and reused without significant loss of chiral induction.
The utility of the aziridino alcohols 2-5 as catalysts for
the same reaction has also been evaluated, and enantio-
selectivities of up to 76% were achieved using 0.25 equiv
of the chiral ligand. To the best of our knowledge, this
is the first time that aziridino alcohols have been applied
as ligands in the above-mentioned reaction.
reaction
time (h)
yield
(%)a
ee
reaction
time (h)
yield
(%)a
ee
(%)b
(%)b
5.0
9.0
22.5
32.5
35
44
58
61
75
79
76
75
47.0
61.0
102.0
64
66
69
76
75
75
a
Estimated by 1H NMR on the crude sample before purification
by flash chromatography. bDetermined by HPLC analysis on a
chiral column (ChiralCel OD-H).
a phenyl group on the aziridine ring, proved to be the
most efficient, giving a loss of ee of only 12% (Table 2,
entry 10).
A kinetic study was performed in order to check if there
was any degradation of the catalytic system during the
reaction time. Imine 1a was used as substrate and
ligand 3a (0.25 equiv) as catalyst. Aliquots were ex-
tracted from the reaction during the whole reaction time.
After hydrolysis with saturated aqueous NH4Cl and
extraction with ether, the organic phase was dried over
MgSO4. Evaporation of the solvent afforded a residue
that was purified by flash chromatography. The product
was analyzed by chiral HPLC (see the Experimental
Section) and the results obtained are presented in Table
3. Two interesting features of the reaction are apparent.
The rate is fast at the beginning and slows down after
the first day. The yield of the addition product reaches
58% after 22.5 h and increases only very slightly from
that point till the end. On the other hand, the ee is
practically constant during the whole reaction time, so
the efficiency of the catalytic system in inducing chirality
is not impaired during the reaction.
Exp er im en ta l Section
For general experimental information, see ref 38. Unless
otherwise noted, final product solutions were dried over
MgSO4, filtered, and evaporated. Flash chromatography was
performed on silica gel (Matrex 60A, 37-70 µm). TLC
analyses were performed on precoated TLC plates, SIL G-60
UV254, which were purchased from Macherey-Nagel. The
following chromatography solvent systems were used: A,
pentane/Et2O; B, pentane/Et2O/MeOH; C, pentane/ethyl ac-
etate; D, pentane/ethyl acetate/MeOH; E, pentane/acetone; F,
pentane/CHCl3/MeOH; G, pentane/CHCl3/MeOH/Et3N; H,
CHCl3/MeOH; I, CHCl3/MeOH/Et3N. Unless otherwise men-
tioned, Rf values were taken in pentane/Et2O 1/1, and [R]D
values were measured in CH2Cl2. HPLC analysis was carried
out on a chiral column (ChiralCel OD-H), using a 254 nm UV
detector, 10% i-PrOH in hexane as eluent, and a flow rate of
0.5 mL/min. 1H and 13C NMR spectra were recorded at 400
and 100.4 MHz, respectively. Et2Zn and Me2Zn were pur-
chased from Aldrich Co. Imines 1 were prepared according
to a literature procedure.33
The presence of an excess of Et2Zn in the reaction
medium seems to be beneficial in order to get good
selectivities in the catalytic process. The addition reac-
tion indicated in Table 2, entry 2, was repeated, but
adding Et2Zn very slowly (addition time, 26 h) with the
aid of a syringe pump. The reaction was quenched and
worked up in the usual way 3 h after the addition was
complete. Product 6a b was isolated in 48% yield and
62% ee, which is slightly lower than the one obtained
under the standard conditions (Table 2, entry 2). Ac-
cording to previous observations made in the case of the
addition to aldehydes,2b the ethylzinc amide that is
formed after the addition of Et2Zn to the imine could also
form a complex with the chiral catalyst, which might
reduce the enantioselectivity of the latter. When an
excess of Et2Zn is present, the chiral complex may be
more easily formed between Et2Zn and the chiral cata-
lyst, securing a stereoselective pathway.
[(2S)-N-Ben zyl-2-a zir id in yl]m eth a n ol (2).17,39 The aziri-
dino ester 10 was synthesized following a literature proce-
dure.22 10 (1.339 g, 7.0 mmol) was dissolved in dry THF (20
mL) under N2 and cooled to -40 °C. LiAlH4 (7.0 mL 1.0 M
solution in THF, 7.0 mmol) was added dropwise and the
reaction was stirred for 4 h, allowing the temperature to rise
slowly to rt. Then, the reaction was quenched following a
literature procedure.40 After evaporation of the solvent, the
resultant residue was purified by flash chromatography
(solvent G, 90/10/1/1), yielding 850 mg (74%) of the expected
aziridino alcohol 2: Rf 0.30 (solvent I, 95/5/1); mp 84-85 °C;
[R]25 ) -11.3 (c ) 1.01); IR (KBr, cm-1) 3125; 1H NMR δ
D
1.45-1.49 (1 H, m), 1.80-1.86 (2 H, m), 3.25 (1 H, br s), 3.36
(1 H, dd, J ) 11.8, 5.4 Hz), 3.43, 3.47 (1 H each, 2 d, J ) 13.4
Hz each), 3.76 (1 H, dd, J ) 11.8, 2.9 Hz), and 7.24-7.37 (5
H, m); 13C NMR δ 31.1, 40.4, 62.5, 64.0, 127.2, 128.0, 128.4,
and 138.8; MS (EI) m/z (rel intensity) 91 (M+ - 72, 100), 72
(82), 65 (20), and 44 (56).
N-Ben zyl-L-th r eon in e Meth yl Ester (12). MeOH (20
mL) was cooled in an ice-NaCl bath and SOCl2 (1.46 mL, 20.0
mmol) was added dropwise. To the resultant solution of HCl
in MeOH, L-threonine (2.431 g, 20.0 mmol) was added and the
reaction mixture was refluxed for 1 h.20 After that time, MeOH
was evaporated and another 20 mL of a 2 M solution of HCl
in MeOH, prepared in the same way as before, was added and
the reaction was refluxed for another hour. MeOH was
evaporated again and the resultant L-threonine methyl ester
hydrochloride was submitted to a reductive alkylation, accord-
ing to a previously described procedure,21 to yield 3.084 g (69%
from threonine) of the title compound: Rf 0.19; [R]24D ) -61.3
(c ) 1.05); IR (neat, cm-1) 3444, 1732, 1199; 1H NMR (300
One experiment was performed in order to study the
correlation between the optical purity of the ligand and
the ee obtained in the addition product. A sample of
ligand 3i of 50% optical purity was prepared from
optically pure and racemic 3i. A stoichiometric addition
of Et2Zn to 1a was set up with the optically enriched
ligand as promoter, giving 57% yield and 50% ee (Table
1, entry 14). If there had been a linear correlation
between the ee’s of the product and the ligand, a value
of 43% [corresponding to 50% of the ee obtained in the
stoichiometric reaction with optically pure 3i (Table 1,
entry 13)] would have been expected. Therefore, it can
be concluded that there is a small chiral amplification
in this case.
(39) Schwan, A. L.; Refvik, M. D. Tetrahedron Lett. 1993, 34, 4901.
(40) Micovic, V. M.; Mihailovic, M. L. J . Org. Chem. 1953, 18, 1190.

Aziridino Alcohols as Additions Promoters
J . Org. Chem., Vol. 62, No. 21, 1997 7371
MHz) δ 1.20 (3 H, d, J ) 6.2 Hz), 2.05 (1 H, br s), 3.04 (1 H,
d, J ) 7.6 Hz), 3.63-3.70 (2 H, m), 3.72 (3 H, s), 3.84 (1 H, d,
J ) 12.9 Hz), and 7.22-7.37 (5 H, m); 13C NMR (75.5 MHz) δ
19.3, 51.9, 52.6, 67.3, 68.0, 127.3, 128.3, 128.4, 139.1, and
174.1; MS (EI) m/ z (rel intensity) 208 (M+ - 15, <1), 179 (17),
178 (18), 91 (100), and 88 (20). Anal. Calcd for C12H17NO3:
C, 64.54; H, 7.69; N, 6.27. Found: C, 64.30; H, 7.73; N, 6.12.
Meth yl (2S,3S)-N-Ben zyl-3-m eth yl-2-a zir id in eca r boxy-
la te (13).17,41 The N-protected amino ester 12 was cyclized to
the aziridino ester 13 following a previously described proce-
dure.23 The isolated yield of 13 was 67%: Rf 0.41; (lit.41 bp2
The reaction mixture was then refluxed for 2 h. After that
time, the reaction was cooled to -78 °C and the solution of
the aziridino ester 13 (205 mg, 1.0 mmol) in dry THF (2 mL)
was added. The reaction mixture was stirred overnight,
allowing the temperature to rise to rt. Water (10 mL) was
added and the resultant mixture was extracted with ether (2
× 20 mL). The combined extracts were washed with water
and brine and dried. The solvent was evaporated and the
residue was chromatographed (solvent A, 9/1), to give 257 mg
(78%) of the title compound: Rf 0.65; mp 117-118 °C; [R]25
D
) +60.2 (c ) 1.10); IR (KBr, cm-1) 3366; H NMR (300 MHz)
1
120-122 °C); [R]25 ) -77.1 (c ) 0.96); IR (neat, cm-1) 1750,
δ 1.07 (3 H, d, J ) 6.1 Hz), 1.94 (1 H, quintet, J ) 6.1 Hz),
2.47 (1 H, d, J ) 6.1 Hz), 3.51, 3.78 (1 H each, 2 d, J ) 13.4
Hz each), 4.62 (1 H, s), 7.10-7.37, and 7.48-7.54 (13 and 2
H, respectively, 2 m); 13C NMR (75.5 MHz) δ 13.5, 40.3, 51.2,
63.6, 73.6, 125.7, 126.3, 126.4, 126.8, 127.1, 127.8, 128.0, 128.3,
138.5, 145.1, and 149.5; MS (EI) m/z (rel intensity) 329 (M+,
1), 224 (92), 167 (64), 105 (65), 91 (100), and 77 (63). Anal.
Calcd for C23H23NO: C, 83.84; H, 7.05; N, 4.25. Found: C,
83.64; H, 7.15; N, 4.20.
D
1718, 1195, and 1124; 1H NMR (300 MHz) δ 1.29 (3 H, d, J )
5.7 Hz), 1.97-2.06 (1 H, m), 2.24 (1 H, d, J ) 6.0 Hz), 3.56,
3.63 (1 H each, 2 d, J ) 13.8 Hz each), 3.73 (3 H, s), and 7.20-
7.40 (5 H, m); 13C NMR (75.5 MHz) δ 13.2, 41.8, 42.6, 52.0,
63.6, 127.1, 127.8, 128.3, 137.9, and 170.2; MS (EI) m/z (rel
intensity) 205 (M+, 2), 114 (97), 91 (100), 59 (79), and 54 (26).
[(2S ,3S )-N -Be n zyl-3-m e t h yl-2-a zir id in yl]m e t h a n ol
(3a ).17,42 To a stirred suspension of LiAlH4 (593 mg, 15.0
mmol) in dry THF (20 mL), under N2, cooled to 0 °C, a solution
of the aziridino ester 13 (1.035 g, 5.0 mmol) in dry THF (10
mL) was added dropwise during ca. 10 min. The reaction
mixture was stirred for 3 h, allowing the temperature to rise
to rt. Then the reaction was quenched following a literature
procedure.40 After evaporation of the solvent, the residue was
chromatographed (solvent A, 6/1, 4/1, 1/1 and then solvent B,
50/50/2), giving 798 mg (90%) of the aziridino alcohol 3a : Rf
[(2S,3S)-N-Tr ityl-3-m eth yl-2-a zir id in yl]m eth a n ol (3d ).
The aziridino ester 15 was prepared as previously reported.25
All the physical and spectroscopic data of the product were in
complete agreement with the reported data,25 except for the
melting point (140-142 °C; lit.25 mp 99-101 °C) and the optical
rotation {[R]24 ) -114.3 (c ) 1.55, CHCl3); lit.25 [R]20
)
D
D
-108.9 (c ) 1.54, CHCl3)}. 15 was reduced in the same way
as for the synthesis of 2. The crude residue was purified by
flash chromatography (solvent A, 95/5, 9/1, 6/1), giving 78%
0.17 (Et2O); (lit.42 bp0.005 94-96 °C); [R]24 ) +17.3 (c ) 1.04);
D
of the expected aziridino alcohol 3d : Rf 0.36; mp 113 °C; [R]24
IR (neat, cm-1) 3383; H NMR (300 MHz) δ 1.18 (3 H, d, J )
1
D
) +13.5 (c ) 1.05); IR (KBr, cm-1) 3436; 1H NMR δ 1.31-1.37
(4 H, m), 1.44-1.49 (1 H, m), 1.76 (1 H, t, J ) 5.6 Hz), 3.81 (1
H, dt, J ) 11.2, 5.6 Hz), 3.88-3.94 (1 H, m), 7.20-7.32 and
7.47-7.51 (9 and 6 H, respectively, 2 m); 13C NMR δ 13.5, 30.9,
36.3, 61.2, 74.5, 126.7, 127.5, 129.4, and 144.7; MS (EI) m/z
(rel intensity) 243 (M+ - 86, 100), 241 (13), 165 (63), 77 (10),
and 58 (16). Anal. Calcd for C23H23NO: C, 83.84; H, 7.05; N,
4.25. Found: C, 84.03; H, 7.19; N, 4.20.
5.7 Hz), 1.71-1.85 (2 H, m), 2.38 (1 H, br s), 3.47-3.57 (3 H,
m), 3.73 (1 H, dd, J ) 11.6, 5.0 Hz), and 7.20-7.40 (5 H, m);
13C NMR (75.5 MHz) δ 13.3, 39.3, 44.3, 60.0, 64.2, 127.1, 127.9,
128.4, and 139.1; MS (EI) m/z (rel intensity) 178 (M+ + 1, <1),
177 (M+, 1), 91 (87), 86 (85), and 58 (100).
3a was obtained as a single invertomer at nitrogen, as
shown by the NMR data, and a NOESY experiment43 was
performed in order to establish the absolute configuration at
nitrogen. An enhancement of the signal corresponding to the
benzylic protons was observed upon irradiation of both aziri-
dine protons. An NOE was also obtained from the hydrogens
of the CH2O group when the methyl group was irradiated.
These results confirm that the benzyl group on the nitrogen
has a trans relationship to both the methyl and the hydroxy-
methyl groups on the aziridine ring.
Ben zyl (2S,3S)-3-Meth yl-2-a zir id in eca r boxyla te (16).25
The detritylation of the aziridino ester 15 was carried out
according to a previously described procedure.24a The expected
N-H aziridino ester was obtained in 77% yield: Rf 0.42
(solvent H, 95/5); mp 55-56 °C; [R]24 ) -0.6 (c ) 1.08); IR
D
(KBr, cm-1) 3185, 1733, and 1184; 1H NMR δ 1.02 (1 H, br s),
1.29 (3 H, d, J ) 5.7 Hz), 2.27-2.35 (1 H, m), 2.68 (1 H, d, J
) 6.3 Hz), 5.19, 5.22 (1 H each, 2 d, J ) 12.2 Hz each), and
7.31-7.41 (5 H, m); 13C NMR δ 13.1, 34.0, 34.9, 67.1, 128.3,
128.4, 128.6, 135.5, and 170.8; MS (EI) m/z (rel intensity) 100
(M+ - 91, 85), 91 (100), 69 (10), 65 (19), and 45 (46).
2-[(2S,3S)-N-Ben zyl-3-m et h yl-2-a zir id in yl]-2-p r op a -
n ol (3b). The aziridino ester 13 (205 mg, 1.0 mmol) was
dissolved in dry THF (5 mL) under N2 and cooled to -78 °C.
Methyllithium (2.5 mL 1.2 M solution in THF, 3.0 mmol) was
added dropwise and the reaction was stirred overnight allow-
ing the temperature to rise to rt. Then, water (10 mL) was
added and the mixture was extracted with ether (2 × 20 mL).
The combined organic layers were washed with brine and
dried. Flash chromatography (solvent A, 9/1, 6/1), afforded
170 mg (83%) of the aziridino alcohol 3b as an oil: Rf 0.40;
Ben zyl (2S,3S)-N,3-Dim et h yl-2-a zir id in eca r b oxyla t e
(17f). The N-methylation of 16 was performed following a
literature procedure.26 The expected product 17f was obtained
in 44% yield: Rf 0.32 (Et2O); [R]24D ) -92.2 (c ) 1.27); IR (neat,
cm-1) 1746 and 1167; 1H NMR δ 1.22 (3 H, d, J ) 5.6 Hz),
1.79 (1 H, dq, J ) 6.7, 5.6 Hz), 2.09 (1 H, d, J ) 6.7 Hz), 2.40
(3 H, s), 5.14, 5.21 (1 H each, 2 d, J ) 12.4 Hz each), and 7.29-
7.39 (5 H, m); 13C NMR δ 12.9, 43.0, 43.7, 46.9, 66.6, 128.3,
128.4, 128.5, 135.7, and 169.7; MS (EI) m/z (rel intensity) 114
[R]25 ) +11.3 (c ) 1.09); IR (neat, cm-1) 3447; 1H NMR (300
D
MHz) δ 1.03, 1.20 (3 H each, 2 s), 1.36 (3 H, d, J ) 5.9 Hz),
1.45 (1 H, d, J ) 6.6 Hz), 1.66-1.72 (1 H, m), 2.97 (1 H, br s),
3.44, 3.71 (1 H each, 2 d, J ) 13.2 Hz each), and 7.21-7.39 (5
H, m); 13C NMR (75.5 MHz) δ 13.7, 25.6, 31.7, 40.3, 50.3, 64.5,
67.1, 127.1, 128.27, 128.32, and 139.1; MS (EI) m/z (rel
intensity) 205 (M+, 22), 190 (49), 178 (37), 147 (52), and 91
(100). Anal. Calcd for C13H19NO: C, 76.04; H, 9.35; N, 6.82.
Found: C, 75.82; H, 9.43; N, 6.71.
(M+ - 91, 100), 91 (54), and 65 (15). Anal. Calcd for C12H15
-
NO2‚0.1H2O:44 C, 69.60; H, 7.41; N, 6.77. Found: C, 69.33;
H, 7.30; N, 6.57.
Ben zyl (2S,3S)-N-(n -Bu t yl)-3-m et h yl-2-a zir id in eca r -
boxyla te (17g). The N-H aziridino ester 16 (511 mg, 2.7
mmol), n-BuBr (0.45 mL, 4.1 mmol), and 18-crown-6 (72 mg,
0.3 mmol) were dissolved in dry MeCN (15 mL). K2CO3 (560
mg, 4.1 mmol) was added and the reaction mixture was
refluxed for 3 days. The solvent was evaporated, water was
added, and the resultant mixture was extracted with CH2Cl2
(3 × 20 mL). The combined organic layers were dried. Flash
chromatography (solvent A, 95/5, 9/1) afforded the desired
r-[(2S,3S)-N-Ben zyl-3-m et h yl-2-a zir id in yl]-r-p h en yl-
ben zyl Alcoh ol (3c). PhMgBr was prepared by adding a
solution of PhBr (0.24 mL, 2.2 mmol) in dry THF (2 mL) to a
suspension of Mg (55 mg, 2.3 mmol) in the same solvent (1
mL) under N2, at such a rate to allow a gentle reflux of THF.
product 17g in 70% yield: Rf 0.46; [R]25 ) -77.5 (c ) 1.10);
D
(41) Bouteville, G.; Gelas-Mialhe, Y.; Vessie`re, R. Bull. Soc. Chim.
Fr. 1971, 3264.
(42) Capeller, R. V.; Griot, R.; Ha¨ring, M.; Wagner-J auregg, T. Helv.
Chim. Acta 1957, 40, 1652.
IR (neat, cm-1) 1747 and 1169; 1H NMR δ 0.89 (3 H, t, J ) 7.3
Hz), 1.24 (3 H, d, J ) 5.6 Hz), 1.26-1.41, 1.52-1.61 (2 H each,
(43) We are grateful to Dr. Adolf Gogoll for helping to set up the
experiment and to interpret the data.
(44) The product was very hygroscopic.

7372 J . Org. Chem., Vol. 62, No. 21, 1997
Andersson et al.
2 m), 1.82 (1 H, dq, J ) 6.9, 5.6 Hz), 2.10 (1 H, d, J ) 6.9 Hz),
2.27 (1 H, dt, J ) 11.4, 7.5 Hz), 2.35-2.43 (1 H, m), 5.16, 5.21
of 26,29 was then submitted to the sequence esterification-
reductive alkylation at nitrogen used for the synthesis of 12,
giving the N-benzyl amino ester 28 in 41% yield (from 26): Rf
0.32 (solvent A, 1/2); mp 107-108 °C ; [R]24D ) +73.9 (c ) 1.04);
(1 H each, 2 d, J ) 12.4 Hz each), and 7.29-7.39 (5 H, m); 13
C
NMR δ 13.2, 14.0, 20.4, 31.5, 42.0, 42.6, 60.6, 66.5, 128.2, 128.3,
128.5, 135.8, and 169.8; MS (EI) m/z (rel intensity) 156 (M+
91, 100), 91 (70), 84 (46), and 57 (47). Anal. Calcd for C15H21
1
-
-
IR (KBr, cm-1) 3509, 3371, 1725, 1256, and 1105; H NMR δ
3.51, 3.77 (1H each, 2 d, J ) 13.3 Hz each), 3.71 (3 H, s), 3.96
(1 H, d, J ) 3.9 Hz), 4.26 (1 H, d, J ) 3.9 Hz), and 7.22-7.41
(10 H, m); 13C NMR δ 50.7, 52.4, 63.5, 74.9, 127.0, 127.7,
127.76, 128.20, 128.3, 128.6, 139.6, 140.0, and 173.8; MS (EI)
NO2: C, 72.83; H, 8.57; N, 5.66. Found: C, 72.85; H, 8.55; N,
5.66.
Ben zyl (2S,3S)-3-Met h yl-N-isop r op yl-2-a zir id in eca r -
boxyla te (17h ). The N-isopropylaziridino ester 17h was
prepared in the same way as 17g, but 10 equiv of i-PrBr was
used and NaI (2 equiv) was also added to the reaction mixture
before reflux. The expected product 17h was isolated in 85%
m/z (rel intensity) 196 (M+
and 65 (13). Anal. Calcd for C17H19NO3: C, 71.55; H, 6.72; N,
4.91. Found: C, 71.29; H, 6.58; N, 4.80.
- 89, 54), 104 (7), 91 (100), 77 (9),
Recrystallization from pentane/CH2Cl2 gave the amino ester
28 optically pure, according to HPLC. Analysis by HPLC of
racemic 28 (prepared in the same way as the optically active
product) showed the retention times 16.5 and 25.5 min for the
two enantiomers. In the analysis of the recrystallized product
under the same conditions, only the enantiomer at 16.5 min
could be detected.
Meth yl (2S,3S)-N-Ben zyl-3-p h en yl-2-a zir id in eca r boxy-
la te (29). The N-protected amino ester 28 was cyclized to the
aziridino ester 29 following a literature procedure.23 The
yield: Rf 0.42; [R]25 ) -64.2 (c ) 1.00); IR (neat, cm-1) 1747
D
and 1166; 1H NMR δ 1.13, 1.15 (3 H each, 2 d, J ) 6.4 Hz
each), 1.26 (3 H, d, J ) 5.6 Hz), 1.57 (1 H, septet, J ) 6.4 Hz),
1.87 (1 H, dq, J ) 6.7, 5.6 Hz), 2.12 (1 H, d, J ) 6.7 Hz), 5.17,
5.20 (1 H each, 2 d, J ) 12.3 Hz each), and 7.28-7.38 (5 H,
m); 13C NMR δ 13.5, 21.5, 21.7, 41.3, 42.3, 61.3, 66.5, 128.16,
128.22, 128.5, 135.9, and 169.7; MS (EI) m/z (rel intensity) 142
(M+ - 91, 70), 91 (93), 84 (38), 73 (100), and 70 (52). Anal.
Calcd for C14H19NO2‚0.3H2O:44 C, 70.43; H, 8.29; N, 5.87.
Found: C, 70.21; H, 8.11; N, 5.70.
isolated yield of 29 was 86%: Rf 0.50; mp 73-74 °C; [R]24
)
D
+5.8 (c ) 1.20); IR (Nujol, cm-1) 1739 and 1458; H NMR δ
2.66 (1 H, d, J ) 6.8 Hz), 3.07 (1 H, d, J ) 6.8 Hz), 3.48 (3 H,
s), 3.67, 3.93 (1 H each, 2 d, J ) 13.7 Hz each), 7.20-7.35 and
7.37-7.44 (6 and 4 H, respectively, 2 m); 13C NMR δ 45.9, 47.8,
51.7, 63.6, 127.3, 127.4, 127.7, 127.9, 128.0, 128.4, 135.0, 137.6,
and 168.4; MS (EI) m/z (rel intensity) 267 (M+, 3), 176 (51),
117 (53), 116 (89), and 91 (100). Anal. Calcd for C17H17NO2:
C, 76.37; H, 6.42; N, 5.24. Found: C, 76.37; H, 6.41; N, 5.24.
(2S,3S)-N-Ben zyl-3-p h en yl-2-a zir id in em et h a n ol (3i).
Compound 29 was reduced in the same way as 13, but in Et2O
instead of THF. The crude residue was purified by flash
chromatography (solvent A, 6/1, 4/1 and then solvent B, 50/
1
Red u ction of Ester s 16 a n d 17f-h . Gen er a l P r oce-
d u r e. The reduction of 16 and 17f-h was carried out in the
same way as for 13, but in Et2O instead of THF. After careful
evaporation of the solvent, the residue was chromatographed.
[(2S,3S)-3-Meth yl-2-azir idin yl]m eth an ol (3e). Flash chro-
matography (neutral aluminum oxide; solvent F, 70/30/1, 60/
40/2, 50/50/3, 50/50/5) afforded the ligand 3e in 50% yield: Rf
0.17 (solvent I, 95/5/1); mp 83-85 °C; [R]25D ) +1.3 (c ) 1.10);
IR (KBr, cm-1) 3260; 1H NMR δ 1.16 (3 H, d, J ) 5.8 Hz), 2.22
(1 H, dq, J ) 6.6, 5.8 Hz), 2.28 (1 H, ddd, J ) 7.4, 6.6, 4.7 Hz),
2.50 (2 H, br s), 3.47 (1 H, dd, J ) 11.8, 7.4 Hz), and 3.72 (1
H, dd, J ) 11.8, 4.7 Hz); 13C NMR δ 13.7, 29.8, 35.9, and 60.8;
MS (EI) m/z (rel intensity) 86 (M+ - 1, 8), 70 (57), 69 (78),
and 54 (100). Anal. Calcd for C4H9NO‚0.1H2O:44 C, 54.01; H,
10.45; N, 15.75. Found: C, 53.84; H, 10.21; N, 15.48.
50/1), yielding 94% of the aziridino alcohol 3i: Rf 0.20; [R]24
D
1
) +123.8 (c ) 1.00); IR (neat, cm-1) 3383; H NMR δ 1.36 (1
H, dd, J ) 7.3, 4.9 Hz), 2.16-2.22 (1 H, m), 2.91 (1 H, d, J )
6.6 Hz), 3.28-3.35 (1 H, m), 3.46 (1 H, ddd, J ) 11.7, 7.3, 5.8
Hz), 3.69, 3.74 (1 H each, 2 d, J ) 13.3 Hz each), and 7.19-
7.44 (10 H, m); 13C NMR δ 46.2, 47.0, 60.6, 64.4, 126.9, 127.2,
127.5, 128.09, 128.13, 128.5, 136.6, and 138.9; MS (EI) m/z
(rel intensity) 239 (M+, <1), 148 (62), 91 (100), and 65 (11).
Anal. Calcd for C16H17NO: C, 80.30; H, 7.16; N, 5.85. Found:
C, 80.67; H, 7.26; N, 5.76.
2-[(2S,3S)-N-Be n zyl-3-p h e n yl-2-a zir id in yl]-2-p r op a -
n ol (3j). A solution of the aziridino ester 29 (278 mg, 1.0
mmol) in dry THF (5 mL), under N2, was cooled to -78 °C
and MeMgBr (1.4 mL 3.0 M solution in THF, 4.2 mmol) was
added dropwise. The reaction was stirred for 2 h at -78 °C
and then it was allowed to warm slowly to -5 °C during ca. 6
h. Then, it was quenched with saturated aqueous NH4Cl and
extracted with Et2O (2 × 20 mL). The combined organic layers
were washed with brine and dried. Flash chromatography
(solvent A, 9/1) yielded 248 mg (89%) of the aziridino alcohol
3j: Rf 0.43; mp 62-63 °C; [R]24D ) +106.1 (c ) 1.11); IR (KBr,
cm-1) 3451; 1H NMR δ 0.97, 1.03 (3 H each, 2 s), 1.87 (1 H, d,
J ) 6.7 Hz), 2.08 (1 H, s), 2.89 (1 H, d, J ) 6.7 Hz), 3.64, 3.88
(1 H each, 2 d, J ) 12.7 Hz each), and 7.16-7.46 (10 H, m);
13C NMR δ 25.9, 30.4, 46.6, 53.9, 65.0, 68.8, 126.7, 127.5, 127.7,
128.0, 128.5, 128.8, 137.0, and 138.8; MS (EI) m/z (rel
intensity) 252 (M+ - 15, <1), 176 (82), 91 (100), and 65 (27).
Anal. Calcd for C18H21NO: C, 80.86; H, 7.92; N, 5.24. Found:
C, 80.74; H, 8.02; N, 5.24.
[(2S,3S)-N,3-Dim et h yl-2-a zir id in yl]m et h a n ol
(3f).45
Flash chromatography (neutral aluminum oxide; solvent D,
80/20/1, 70/30/1) gave 52% yield: Rf 0.29 (solvent I, 95/5/1);
[R]25 ) -3.0 (c ) 1.07); IR (neat, cm-1) 3358; 1H NMR δ 1.13
D
(3 H, d, J ) 5.8 Hz), 1.46-1.54, 1.56-1.62 (1 H each, 2 m),
2.38 (3 H, s), 3.44 (1 H, dd, J ) 11.9, 7.4 Hz), 3.69 (1 H, dd, J
) 11.9, 4.5 Hz), and 4.15 (1 H, br s); 13C NMR δ 12.9, 40.1,
46.0, 47.2, and 60.2; MS (EI) m/z (rel intensity) 101 (M+, <1),
86 (35), 84 (100), and 70 (36).
[(2S,3S)-N-(n -Bu t yl)-3-Met h yl-2-a zir id in yl]m et h a n ol
(3g). Flash chromatography (solvent A, 3/1 and then solvent
B, 75/25/2, 75/25/4). The isolated yield was 70%: Rf 0.10
(solvent H, 95/5); [R]24D ) +8.1 (c ) 1.21); IR (neat, cm-1) 3356;
1H NMR δ 0.90 (3 H, t, J ) 7.3 Hz), 1.15 (3 H, d, J ) 5.7 Hz),
1.27-1.40 (2 H, m), 1.49-1.63 (4 H, m), 2.27, 2.34 (1 H each,
2 dt, J ) 11.6, 7.4 Hz each), 3.00 (1 H, br s), 3.49 (1 H, dd, J
) 11.6, 6.6 Hz), and 3.71 (1 H, dd, J ) 11.6, 5.1 Hz); 13C NMR
δ 13.4, 14.0, 20.5, 32.0, 39.0, 44.3, 59.9, and 60.7; MS (EI) m/z
(rel intensity) 128 (M+ - 15, 14), 112 (100), 58 (38), 57 (49),
and 56 (83). Found: M, 143.1288. C8H17NO requires M,
143.1310.
[(2S,3S)-3-Met h yl-N-isop r op yl-2-a zir id in yl]m et h a n ol
(3h ).45 Flash chromatography (solvent A, 3/1 and then solvent
B, 75/25/2, 75/25/4), yielded 65% of the aziridino alcohol 3h :
Rf 0.16 (solvent H, 95/5); [R]24 ) +12.7 (c ) 1.19); IR (neat,
D
cm-1) 3355; H NMR δ 1.09, 1.10 (3 H each, 2 d, J ) 6.3 Hz
1
N-Ben zoyl-(L)-th r eon in e Meth yl Ester (18).27 L-Threo-
nine was converted into its methyl ester hydrochloride follow-
ing the same procedure as for the preparation of 12 (first
step).20 The amino ester hydrochloride (210 mmol) was
dissolved in 130 mL of MeOH, and Et3N (88 mL, 630 mmol)
was added. After 15 min, the reaction mixture was cooled to
0 °C and PhCOCl (27 mL, 231 mmol) was added dropwise.
The reaction was stirred for 2 h at 0 °C. Then, MeOH and
the excess Et3N were evaporated. Water (100 mL) was added
and the mixture was extracted with ethyl acetate (3 × 100
mL). The combined organic layers were washed with water
(50 mL) and brine (50 mL) and dried. The solvent was
each), 1.16 (3 H, d, J ) 5.7 Hz), 1.50-1.65 (3 H, m), 2.55 (1 H,
br s), 3.51 (1 H, dd, J ) 11.5, 6.5 Hz), and 3.68 (1 H, dd, J )
11.5, 5.3 Hz); 13C NMR δ 13.7, 21.9, 22.0, 38.2, 43.9, 60.1, and
60.7; MS (EI) m/z (rel intensity) 129 (M+, 4), 86 (37), 70 (29),
58 (100), and 56 (42).
Meth yl (2R,3S)-3-(Ben zyla m in o)-2-h yd r oxy-3-p h en yl-
p r op ion a te (28). The N-tosyl amino ester 26 was prepared
from methyl cinnamate via Sharpless asymmetric aminohy-
droxylation.18 The â-amino acid 27, obtained by deprotection
(45) Wartski, L.; Sierra-Escudero, A. Bull. Soc. Chim. Fr. 1975,
1663.

Aziridino Alcohols as Additions Promoters
J . Org. Chem., Vol. 62, No. 21, 1997 7373
evaporated, and the resulting residue was refluxed for 1 h with
stirring in 300 mL of dry ether. After cooling, the white solid
was filtered off. A 40.287 g portion of the expected product
18 was obtained. On cooling the mother liquors in the
refrigerator, 1.500 g more of product crystallized. The yield
of this one-pot procedure was 84%: Rf 0.44 (solvent H, 95/5);
Hz), 0.91 (3 H, t, J ) 7.6 Hz), 1.31 (3 H, d, J ) 6.1 Hz), 1.36
(2 H, q, 7.5 Hz), 1.44 (1 H, d, 3.4 Hz), 1.47-1.54 (2 H, m), 2.20
(1 H, qd, 6.1, 3.4 Hz), 2.71 (1 H, br s), 3.73 (2 H, s), and 7.24-
7.40 (5 H, m); 13C NMR δ 7.8, 8.0, 11.0, 29.3, 32.1, 33.7, 51.0,
54.4, 70.3, 126.8, 128.0, 128.3, and 140.0; MS (EI) m/z (rel
intensity) 233 (M+, 1), 142 (63), 91 (100), and 56 (20). Anal.
Calcd for C15H23NO: C, 77.21; H, 9.93; N, 6.00. Found: C,
77.09; H, 10.02; N, 5.99.
mp 97-98 °C; [R]25D ) +24.0 (c ) 5.43, EtOH); IR (KBr, cm-1
)
3427, 3360, 1746, 1644, 1526, and 1259; 1H NMR δ 1.27 (3 H,
d, J ) 6.4 Hz), 2.76-2.81 (1 H, m), 3.77 (3 H, s), 4.40-4.48 (1
H, m), 4.81 (1 H, dd, J ) 8.9, 2.5 Hz), 7.03 (1 H, d, J ) 8.9
Hz), 7.39-7.45, 7.48-7.53, and 7.82-7.85 (2, 1, and 2 H,
respectively, 3 m); 13C NMR δ 20.0, 52.6, 57.7, 68.2, 127.2,
128.6, 131.9, 133.7, 167.9, and 171.6; MS (EI) m/z (rel
intensity) 193 (M+ - 44, 13), 161 (21), 105 (100), 77 (63), and
51 (21).
Meth yl (2R,3R)-3-(Ben zyla m in o)-2-h yd r oxy-3-p h en yl-
p r op ion a te (33). As previously reported, epoxide 3131 was
synthesized and ring-opened with NaN3.32 The azido alcohol
32 (5.308 g, 24.0 mmol) was dissolved in MeOH (30 mL) and
hydrogenated in the presence of 5% Pd-C (228 mg) under 1
atm of hydrogen at rt for 2 days. The catalyst was filtered off
and washed with MeOH. Solvent was evaporated and the
resultant oily residue was dissolved in a 2 M solution of HCl
in MeOH (25 mL) at rt and stirred for 10 min. Solvent was
evaporated and the resultant hydrochloride was then submit-
ted to the reductive alkylation process described above for the
synthesis of 12. The crude residue was purified by flash
chromatography (solvent A, 1/1, 1/2 and then solvent B, 50/
50/2). The N-benzyl amino ester 33 was obtained in 56% yield
(from 32): Rf 0.14; mp 98-99 °C; [R]24D ) -81.5 (c ) 1.05); IR
(KBr, cm-1) 3467, 3333, 1720, and 1282; 1H NMR δ 2.30, 3.30
(1 H each, 2 br s), 3.60 (3 H, s), 3.62, 3.78 (1 H each, 2 d, J )
13.1 Hz each), 4.07 (1 H, d, J ) 4.2 Hz), 4.54 (1 H, d, J ) 4.2
Hz), and 7.23-7.38 (10 H, m); 13C NMR δ 51.0, 52.1, 63.8, 73.6,
127.1, 127.7, 128.0, 128.2, 128.4, 128.5, 137.7, 139.7, and 173.0;
MS (EI) m/z (rel intensity) 226 (M+ - 59, 4), 197 (19), 196 (80),
91 (100), and 65 (32). Anal. Calcd for C17H19NO3: C, 71.56;
H, 6.71; N, 4.91. Found: C, 71.67; H, 6.71; N, 4.95.
Meth yl (2S,3R)-N-Ben zyl-3-m eth yl-2-a zir id in eca r box-
yla te (21).46 allo-L-Threonine 20 was prepared from 18 in 86%
yield27 and transformed into 21 in the same way as for the
corresponding L-threonine derivative 13. The yield of the
isolated product was 55% (from 20). 21 was obtained as a 1:0.7
mixture of invertomers at nitrogen: Rf 0.42; [R]26 ) -90.3 (c
D
)1.16); IR (neat, cm-1) 1732 and 1200; 1H NMR δ major
invertomer: 1.25 (3 H, d, J ) 5.5 Hz), 2.34 (1 H, qd, J ) 5.5,
2.8 Hz), 2.46 (1 H, d, J ) 2.8 Hz), 3.67 (3 H, s), 3.88, 4.04 (1 H
each, 2 d, J ) 13.8 Hz each), and 7.21-7.39 (5 H, m). Minor
invertomer: 1.38 (3 H, d, J ) 6.0 Hz), 2.04 (1 H, d, J ) 2.8
Hz), 2.62 (1 H, qd, J ) 6.4, 2.8 Hz), 3.71 (3 H, s), 3.68, 3.84 (1
H each, 2 d, J ) 14.1 Hz each), and 7.21-7.39 (5 H, m); 13C
NMR δ major invertomer: 17.9, 41.1, 42.8, 51.9, 55.0, 126.8,
128.0, 128.26, 139.3, and 170.1. Minor invertomer: 10.7, 40.1,
44.1, 52.1, 54.7, 126.9, 127.6, 128.33, 138.7, and 171.5; MS (EI)
m/z (rel intensity) 206 (M+ + 1, 2), 205 (M+, 13), 146 (49), 114
(100), 91 (98), and 59 (63).
Recrystallization from pentane/CH2Cl2 gave the amino ester
33 optically pure, according to HPLC. Analysis by HPLC of
racemic 33 (prepared in the same way as the optically active
product) showed the retention times 19.6 and 21.3 min for the
two enantiomers. In the analysis of the recrystallized product
under the same conditions, only the enantiomer at 21.3 min
could be detected.
[(2S ,3R )-N -Be n zyl-3-m e t h yl-2-a zir id in yl]m e t h a n ol
(4a ).17,42 The title compound was prepared in the same way
as for the corresponding serine derivative 2. The yield of the
isolated product was 74%. The product was obtained as a
1:0.15 mixture of invertomers at nitrogen: Rf 0.22 (solvent H,
95/5); mp 72-73 °C; [R]24 ) +27.7 (c ) 1.02); IR (KBr, cm-1
)
Meth yl (2S,3R)-N-Ben zyl-3-p h en yl-2-a zir id in eca r boxy-
la te (34).47 Compound 33 was cyclized to the aziridino ester
34.23 The isolated yield of 34 was 88%. The product was
obtained as a 1:0.15 mixture of invertomers at nitrogen: Rf
D
1
3357; H NMR δ 1.20 (0.45 H, d, J ) 5.5 Hz), 1.31 (3 H, d, J
) 6.1 Hz), 1.55 (0.15 H, qd, J ) 5.5, 3.6 Hz), 1.63 (1 H, dt, J
) 5.5, 3.4 Hz), 2.01 (0.15 H, dt, J ) 8.6, 3.6 Hz), 2.16 (1 H, qd,
J ) 6.1, 3.4 Hz), 2.63 (0.15 H, br s), 3.10 (1 H, br s), 3.38 (1 H,
m), 3.55 (1 H, d, J ) 13.9 Hz), 3.65 (0.15 H, d, J ) 14.4 Hz),
3.67-3.74 (1 H + 0.15 H, m), 3.74 (1 H, d, J ) 13.9 Hz), 3.84
(0.15 H, d, J ) 14.4 Hz), 3.84-3.92 (0.15 H, m), and 7.23-
7.40 (5 H + 0.75 H, m); 13C NMR δ major invertomer: 11.0,
35.5, 47.0, 54.8, 62.2, 126.9, 127.8, 128.4, and 139.8; MS (EI)
m/z (rel intensity) 176 (M+ - 1, 4), 91 (89), 86 (100), 65 (26),
and 58 (60).
0.57; [R]24 ) +7.2 (c ) 1.16); IR (neat, cm-1) 1730, 1201, and
D
1
1174; H NMR δ major invertomer: 2.81 (1 H, br s), 3.35 (1
H, br s), 3.72 (3 H, br s), 4.11, 4.27 (1 H each, 2 br d, J ) 13.9
Hz each), and 7.15-7.45 (10 H, m); 13C NMR δ major
invertomer: 44.4, 48.6, 52.1, 54.9, 126.3, 126.9, 127.6, 128.0,
128.3, 128.4, 138.1, 139.1, and 169.2; MS (EI) m/z (rel
intensity) 268 (M+ + 1, <1), 267 (M+, 5), 176 (83), 117 (48),
116 (100), and 91 (68).
A NOESY experiment43 was carried out at -50 °C in order
to establish the absolute configuration at nitrogen for the
major invertomer. An enhancement of the signals correspond-
ing to the methyl group and the proton at C2 in the aziridine
ring was observed upon irradiation of the benzylic protons.
An NOE was also obtained from the hydrogens of the CH2O
group when the hydrogen at C3 in the aziridine ring was
irradiated. No enhancement of the signal corresponding to
the benzylic protons was observed in the latter case. These
results confirm the trans relationship between the benzyl
group and the hydroxymethyl groups on the aziridine ring, as
well as the cis relationship between the benzyl and the methyl
groups. The energy barrier for nitrogen inversion was esti-
mated to be ca. 70 kJ /mol, the value being determined by full
line-shape fitting of spectra obtained at various temperatures
using the program gNMR 3.6.5 (Cherwell Scientific).43
3-[(2S,3R)-N-Ben zyl-3-m et h yl-2-a zir id in yl]-3-p en t a -
n ol (4b). The title compound was prepared from 21 and
EtMgBr, following the same procedure as for the synthesis of
ligand 3c. The crude mixture was purified by flash chroma-
tography (solvent A, 9/1, 6/1, 4/1, 2/1, 1/1), to afford the desired
[(2S,3R)-N-Ben zyl-3-ph en yl-2-a zir id in yl]m eth a n ol (4c).
The aziridino ester 34 was reduced with LiAlH4, following the
procedure described above for the synthesis of 3a . The crude
residue was purified by flash chromatography (solvent E, 9/1,
6/1), to afford the ligand 4c in 59% yield. The product was
obtained as a 1:0.36 mixture of invertomers at nitrogen: Rf
0.28 (solvent E, 2/1); [R]24 ) +31.7 (c ) 1.13); IR (neat, cm-1
)
D
3354; 1H NMR δ 1.44 (0.36 H, br s), 2.12 (1 H, t, J ) 6.2 Hz),
2.42-2.60 (1+0.72 H, m), 3.14, 3.48 (1 H each, 2 d, J ) 13.6
Hz each), 3.34 (1 H, d, J ) 3.6 Hz), 3.56-3.64 (1 H, m), 3.85-
4.09 (1+1.44 H, m), and 7.15-7.47 (10+3.60 H, m); 13C NMR
δ 43.9, 44.0, 45.4, 48.2, 55.4, 55.5, 59.0, 62.3, 126.2, 126.9,
127.5, 127.95, 128.02, 128.1, 128.3, 130.2, 133.1, and 139.3;
MS (EI) m/z (rel intensity) 239 (M+, 1), 148 (58), and 91 (100).
Anal. Calcd for C16H17NO: C, 80.30; H, 7.16; N, 5.85. Found:
C, 80.07; H, 7.22; N, 5.87.
Meth yl N-Ben zyl-O-(ter t-bu tyld im eth ylsilyl)-L-th r eo-
n in a te (22). Methyl N-benzyl-L-threoninate 12 (866 mg, 3.9
mmol), t-BuMe2SiCl (1.212 g, 7.8 mmol), and imidazole (1.062
g, 15.6 mmol) were dissolved in DMF (4 mL), and the reaction
mixture was stirred at rt for 3 days. The DMF was evapo-
rated, and water (10 mL) and CH2Cl2 (20 mL) were added to
the resultant residue. The layers were separated, and the
aziridino alcohol 4b in 82% yield: Rf 0.18; [R]24 ) -35.2 (c )
D
1
1.20); IR (neat, cm-1) 3425; H NMR δ 0.82 (3 H, t, J ) 7.5
(46) Wartski, L.; Wakselman, C.; Sierra-Escudero, A. Bull. Soc.
Chim. Fr. 1972, 1478.
(47) Bouayad, Z.; Chanet-Ray, J .; Ducher, S.; Vessiere, R. J . Het-
erocycl. Chem. 1991, 28, 1757.

7374 J . Org. Chem., Vol. 62, No. 21, 1997
Andersson et al.
aqueous layer was extracted with CH2Cl2 (2 × 20 mL). The
combined organic phases were washed with water (2 × 10 mL)
and dried. The solvent was evaporated and the residue was
alcohol 2-5 (0.25 mmol) were dissolved in dry toluene (1.5 mL)
under nitrogen, and the mixture was stirred for 10 min at rt.
The solution was cooled to 0 °C, Et2Zn (0.68 mL 1.1 M solution
in toluene, 0.75 mmol) was added, the reaction was stirred,
the temperature was allowed to rise slowly to rt (3-4 h), and
then the mixture was stirred for 4 days more. The reaction
was quenched with saturated aqueous NH4Cl (10 mL) and the
mixture was extracted with CH2Cl2 (4 × 10 mL) and dried.
Solvents were evaporated and the residue was purified by flash
chromatography (solvent E, 6/1, 4/1, 2/1, 1/1). The resultant
solid was analyzed by HPLC. The retention times are as
follows: 6a a , 14.7 (R) and 18.6 min (S); 6a b, 13.6 (R) and 17.7
min (S); 6b, 16.3 (R) and 19.8 min (S); and 6c, 14.7 (R) and
16.7 min (S). Yields and ee’s are given in Table 1. The
absolute configuration of the major isomer was determined by
hydrolysis of the reaction product and comparison of the optical
rotation of the obtained amine with the reported data.48
Gen er a l P r oced u r e for th e Hyd r olysis. The product was
stirred overnight at rt in a 1.5 M solution of hydrogen chloride
in methanol (2 mL). Then, solvent was evaporated and 2 M
aqueous HCl (5 mL) was added. The mixture was extracted
with ethyl acetate (2 × 10 mL), and the organic layers were
discarded. The aqueous layer was basified with 15% NaOH
and extracted with CH2Cl2 (3 × 10 mL). The combined organic
phases were dried. Evaporation of the solvent gave the pure
amine.
purified by flash chromatography (solvent A, 98/2, 95/5).
A
1.238 g (94%) portion of the protected threoninate 22 was ob-
tained: Rf 0.56 (solvent A, 3/1); [R]25 ) -53.7 (c ) 1.02); IR
D
(neat, cm-1) 1745, 1157, 1097, and 836; 1H NMR (300 MHz) δ
-0.01, 0.03 (3 H each, 2 s) 0.86 (9 H, s), 1.25 (3 H, d, J ) 6.2
Hz), 2.13 (1 H, br s), 3.12 (1 H, d, J ) 3.4 Hz), 3.62, 3.98 (1 H
each, 2 d, J ) 13.4 Hz each), 3.72 (3 H, s), 4.18 (1 H, dq, 6.2,
J ) 3.4 Hz), and 7.20-7.39 (5 H, m); 13C NMR (75.5 MHz) δ
-5.2, -4.4, 17.9, 20.8, 25.7, 51.5, 52.0, 66.2, 69.8, 126.8, 128.17,
128.22, 140.2, and 174.0; MS (EI) m/z (rel intensity) 338 (M+
+ 1, <1), 337 (M+, <1), 293 (40), 159 (54), 91 (100), and 73
(49). Anal. Calcd for C18H31NO3Si: C, 64.04; H, 9.27; N, 4.15.
Found: C, 63.80; H, 9.28; N, 4.09.
(2R,3R)-2-(Ben zyla m in o)-3-[(ter t-bu tyld im eth ylsilyl)-
oxy]-1-bu ta n ol (23). Compound 22 (190 mg, 0.6 mmol) was
dissolved in dry CH2Cl2 (2.5 mL) under N2 and cooled to -78
°C. DIBALH (2.0 mL 1.0 M solution in THF, 2.0 mmol) was
added and the reaction mixture was stirred overnight, allowing
the temperature to rise to rt. Then the reaction was quenched
with water (5 mL). KOH (20%, 5 mL) was added and the
mixture was extracted with ethyl acetate (3 × 20 mL). The
combined organics were washed with water and dried. Flash
chromatography (solvent C: 2/1) yielded 101 mg (58%) of the
expected amino alcohol 23: Rf 0.22 (solvent C, 1/1); [R]25
)
The reported optical rotation for 1-phenylpropylamine was
found to be in error. Hydrolysis of 6a b with 61.5% ee gave
D
-10.6 (c ) 1.09); IR (neat, cm-1) 3364, 1098, and 835; 1H NMR
(300 MHz) δ 0.04, 0.07 (3 H each, 2 s), 0.87 (9 H, s), 1.21 (3 H,
d, J ) 6.3 Hz), 2.58 (1 H, dt, J ) 6.3, 4.3 Hz), 3.44 (1 H, dd, J
) 10.9, 4.0 Hz), 3.68 (1 H, dd, J ) 10.9, 4.6 Hz), 3.79, 3.85 (1
H each, 2 d, J ) 13.2 Hz each), 3.91 (1 H, quintet, J ) 6.3
Hz), and 7.20-7.35 (5 H, m); 13C NMR (75.5 MHz) δ -5.0,
-4.2, 17.9, 20.5, 25.8, 51.8, 59.9, 64.0, 68.8, 127.0, 128.0, 128.4,
and 140.3; MS (EI) m/z (rel intensity) 294 (M+ - 15, 2), 278
the corresponding free amine, with [R]25 ) +8.8 (c ) 1.04,
D
CHCl3). According to the literature,48a that value corresponds
to 24% ee for the amine. However, the free amine was
converted into the corresponding benzamide by reaction with
benzoyl chloride in the presence of triethylamine. Analysis
by HPLC gave a value of 61.3% ee, which is in very good
agreement with the optical purity of the starting amide 6a b.
The retention times for the two enantiomers of the benzamide
were 20.5 (R) and 32.2 min (S).
(24), 150 (100), 91 (92), and 73 (24). Anal. Calcd for C17H31
-
NO2Si: C, 65.95; H, 10.11; N, 4.53. Found: C, 65.91; H, 10.02;
N, 4.47.
N-(1-P h en yleth yl)-P ,P -diph en ylph osph in ic am ide (6aa):
11
(R)-N-Ben zyl-2-{(R)-1-[(ter t-b u t yld im et h ylsilyl)oxy]-
eth yl}a zir id in e (24). To the solution of the protected amino
alcohol 23 (181 mg, 0.6 mmol) in dry THF (6 mL) was added
Ph3P (232 mg, 0.9 mmol). The solution was cooled to 0 °C,
and DEAD (0.14 mL, 0.9 mmol) was added dropwise. The ice
bath was removed and the reaction was stirred at rt overnight.
The solvent was evaporated and the residue was purified by
flash chromatography (solvent A, 9/1), affording 139 mg (82%)
Rf 0.48 (solvent H, 95/5); mp 166-170 °C (75% ee); [R]24
)
D
+28.5 [c ) 1.01, 75% ee (R)]; IR (KBr, cm-1) 3167 and 1181;
1H NMR δ 1.57 (3 H, d, J ) 6.7 Hz), 3.20 (1 H, br s), 4.34-
4.46 (1 H, m), 7.20-7.52 and 7.79-7.94 (11 and 4 H, respec-
tively, 2 m); 13C NMR δ 25.9 (d, J ) 3.1 Hz), 51.1, 125.9, 127.1,
128.3, 128.39, 128.42, 128.51, 128.54, 131.69, 131.72, 131.78,
131.80, 131.9, 132.0, 132.4, 132.5, and 145.1 (d, J ) 6.1); MS
(EI) m/z (rel intensity) 321 (M+, 2), 306 (23), 201 (54), 120 (100),
and 77 (45).
of the desired aziridine 24: Rf 0.59; [R]23 ) +44.4 (c ) 1.31);
D
IR (neat, cm-1) 1112, and 836; 1H NMR (300 MHz) δ 0.01, 0.04
(3 H each, 2 s), 0.88 (9 H, s), 1.13 (3 H, d, J ) 6.4 Hz), 1.36 (1
H, d, J ) 6.4 Hz), 1.58-1.69 (2 H, m), 3.21, 3.60 (1 H each, 2
d, J ) 13.0 Hz each), 3.43 (1 H, quintet, J ) 6.4 Hz), and 7.20-
7.38 (5 H, m); 13C NMR (75.5 MHz) δ -4.8, -4.6, 18.2, 21.2,
25.9, 31.0, 46.0, 64.9, 71.3, 127.0, 128.3, 128.4, and 139.2; MS
(EI) m/z (rel intensity) 234 (M+ - 57, 29), 115 (32), 91 (100),
75 (59), and 73 (40). Anal. Calcd for C17H29NOSi: C, 70.03;
H, 10.05; N, 4.81. Found: C, 69.91; H, 10.18; N, 4.73.
N-(1-P h en ylp r op yl)-P ,P -d ip h en ylp h osp h in ic a m id e
(6a b):11 Rf 0.42 (solvent H, 95/5); mp 111-114 °C (45% ee);
[R]24 ) +31.3 [c ) 1.35, 81% ee (R)]; IR (KBr, cm-1) 3151,
D
1
1197, and 1182; H NMR (300 MHz) δ 0.78 (3 H, t, J ) 7.4
Hz), 1.75-2.09 (2 H, m), 3.24 (1 H, dd, J ) 9.7, 6.2 Hz), 4.03-
4.17 (1 H, m), 7.11-7.52 and 7.70-7.91 (11 and 4 H, respec-
tively, 2 m); 13C NMR (75.5 MHz) δ 10.5, 32.5 (d, J ) 3.3 Hz),
57.1, 126.5, 127.0, 128.2, 128.3, 128.4, 128.5, 131.6, 131.67,
131.73, 131.79, 131.84, 132.4, 132.5, 132.6, 132.8 and 143.5
(d, J ) 6.0); MS (EI) m/z (rel intensity) 306 (M+ - 29, 100),
202 (12), 201 (72), and 77 (15).
(R)-1-[(R)-N-Ben zyl-2-a zir id in yl]eth a n ol (5). Compound
24 (114 mg, 0.4 mmol) was dissolved in dry THF (2.5 mL).
(n-Bu)4NF (1.2 mL 1.0 M solution in THF, 1.2 mmol) was
added and the reaction was stirred at rt for 32 h. Water (10
mL) was added and the mixture was extracted with ether (3
× 20 mL). The combined organic extracts were washed with
water and brine and dried. Flash chromatography (solvent
A, 6/1 and then Et2O) gave 48 mg (69%) of the ligand 5: Rf
N -[1-(1-Na p h t h yl)p r op yl]-P ,P -d ip h e n ylp h osp h in ic
a m id e (6b): Rf 0.25 (hexane/acetone: 2/1); mp 167-169 °C
(6% ee); [R]25 ) -23.3 [c ) 1.00, 73% ee (R)]; IR (KBr, cm-1
)
D
3197 and 1190; 1H NMR δ 0.85 (3 H, t, J ) 7.4 Hz), 2.05-2.13
(2 H, m), 3.42 (1 H, dd, J ) 9.2, 6.8 Hz), 4.99 (1 H, tt, J ) 9.8,
6.8 Hz), 7.15-7.21, 7.29-7.55, and 7.66-7.91 (2, 8, and 7 H,
respectively, 3 m); 13C NMR δ 10.6, 32.5, 52.7, 122.9, 123.3,
125.3, 125.4, 125.9, 127.6, 128.1, 128.2, 128.4, 128.5, 128.7,
130.6, 131.5, 131.6, 131.75, 131.78, 131.8, 132.3, 132.4, 133.8,
and 139.7; MS (EI) m/z (rel intensity) 356 (M+ - 29, 75), 201
(100), 184 (38), and 154 (35). Anal. Calcd for C25H24NOP: C,
77.89; H, 6.29; N, 3.63. Found: C, 77.79; H, 6.13; N, 3.84.
N-[1-(4-Ch lor op h en yl)p r op yl]-P ,P -d ip h en ylp h osp h in -
ic a m id e (6c): Rf 0.53 (solvent H, 95/5); mp 162-164 °C (82%
0.18 (ethyl acetate); [R]24 ) -11.5 (c ) 0.99); IR (neat, cm-1
)
D
1
3384; H NMR (300 MHz) δ 1.14 (3 H, d, J ) 6.4 Hz), 1.49 (1
H, d, J ) 6.6 Hz), 1.60-1.66 (1 H, m), 1.82 (1 H, d, J ) 3.6
Hz), 2.59 (1 H, br s), 3.35-3.50 (1 H, m), 3.40, 3.50 (1 H each,
2 d, J ) 13.0 Hz each), and 7.24-7.37 (5 H, m); 13C NMR (75.5
MHz) δ 20.8, 31.8, 45.0, 64.3, 68.0, 127.3, 128.2, 128.5, and
138.9; MS (EI) m/z (rel intensity) 177 (M+, 3), 91 (100), 86 (51),
65 (20), and 58 (19). Anal. Calcd for C11H15NO: C, 74.52; H,
8.55; N, 7.90. Found: C, 74.36; H, 8.65; N, 7.88.
Ad d ition Rea ction of Dia lk ylzin c Rea gen ts to 1 P r o-
m oted by th e Liga n d s 2-5. Gen er a l P r oced u r e. The
phosphinoylimine 1 (76 mg, 0.25 mmol) and the aziridino
(48) (a) Yang, T.-K.; Chen, R.-Y.; Lee, D.-S.; Peng, W.-S.; J iang, Y.-
Z.; Mi, A.-Q.; J ong, T.-T. J . Org. Chem. 1994, 59, 914. (b) Pridgen, L.
N.; Mokhallalati, M. K.; Wu, M.-J . J . Org. Chem. 1992, 57, 1237.

Aziridino Alcohols as Additions Promoters
J . Org. Chem., Vol. 62, No. 21, 1997 7375
ee); [R]25D ) +44.1 [c ) 1.02, 82% ee (R)]; IR (KBr, cm-1) 3222
and 1183; 1H NMR δ 0.79 (3 H, t, J ) 7.4 Hz), 1.73-1.85 (1 H,
m), 1.97 (1 H, dqd, J ) 13.6, 7.4, 6.1 Hz), 3.25 (1 H, dd, J )
9.5, 6.1 Hz), 4.03-4.13 (1 H, m), 7.06-7.10, 7.21-7.25, 7.28-
7.35, 7.38-7.51, 7.69-7.76, and 7.81-7.88 (2 H, 2 H, 2 H, 4,
2, and 2 H, respectively, 6 m); 13C NMR δ 10.5, 32.3 (1 C, d, J
) 4.5 Hz), 56.5, 127.9, 128.2, 128.3, 128.4, 128.5, 131.4, 131.68,
131.71, 131.75, 131.84, 132.38, 132.47, 132.7, 133.7, and 142.1
(2 C, d, J ) 4.6 Hz); MS (EI) m/z (rel intensity) 369 (M+, <1),
Ack n ow led gm en t. We thank the Wenner-Gren
Foundation (postdoctoral stipend to D.G.), the Swedish
Natural Science Research Council, and the Danish
Natural Science Research Council for financial support.
We are also grateful to Dr. Adolf Gogoll (University of
Uppsala) for his very useful assistance with the NMR
experiments.
340 (66), 201 (100), and 77 (29). Anal. Calcd for C21H21
-
ClNOP: C, 68.19; H, 5.73; N, 3.79. Found: C, 67.87; H, 5.65;
N, 3.50.
J O970918H