Enantioselective aziridination using copper complexes of biaryl Schiff bases

Article abstract of DOI:10.1021/jo025515i

Racemic 2,2′-diamino-6,6′-dimethylbiphenyl is resolved using simulated moving bed chromatography, and the absolute configuration of the enantiomers is confirmed via the X-ray crystal structure of a derivative. The diamine is condensed with a range of aldehydes to give bidentate aldimine proligands L. Molecular structures of the complexes formed between L and Cu(I) fall into two classes; bimetallic double helices ([Cu₂L₂]²⁺) and monometallic ([CuL]⁺). The latter are strikingly more efficient in the aziridination of alkenes than are the former in terms of rate, turnover, and enantioselection. In particular, the imine ligand formed from the diamine and 2,6-dichlorobenzaldehyde gives, in combination with Cu(I) or Cu(II), up to 99% ee in the aziridination of 6-acyl-2,2-dimethylchromene and 88-98% ee for a range of cinnamate esters. Styrenic and other alkenes are converted with lower selectivities (5-54%). The catalytic system shows a linear response in product ee to catalyst ee, and the product ee does not vary significantly during the reaction. UV spectrophotometric investigations indicate that conversion of Cu(I) to Cu(II) is not essential for catalysis but that Cu(II) is probably also a competent system.

Full text of DOI:10.1021/jo025515i

3450
J . Org. Chem. 2002, 67, 3450-3458
En a n tioselective Azir id in a tion Usin g Cop p er Com p lexes of Bia r yl
Sch iff Ba ses
Kevin M. Gillespie,† Christopher J . Sanders,† Paul O’Shaughnessy,† Ian Westmoreland,†
Christopher P. Thickitt,‡ and Peter Scott*,†
Department of Chemistry, The University of Warwick, Coventry CV4 7AL, U.K., and
GlaxoSmithKline, Harlow, Essex CM19 5AW, U.K.
peter.scott@warwick.ac.uk
Received J anuary 10, 2002
Racemic 2,2′-diamino-6,6′-dimethylbiphenyl is resolved using simulated moving bed chromatog-
raphy, and the absolute configuration of the enantiomers is confirmed via the X-ray crystal structure
of a derivative. The diamine is condensed with a range of aldehydes to give bidentate aldimine
proligands L. Molecular structures of the complexes formed between L and Cu(I) fall into two classes;
bimetallic double helices ([Cu2L2]2+) and monometallic ([CuL]+). The latter are strikingly more
efficient in the aziridination of alkenes than are the former in terms of rate, turnover, and
enantioselection. In particular, the imine ligand formed from the diamine and 2,6-dichlorobenzal-
dehyde gives, in combination with Cu(I) or Cu(II), up to 99% ee in the aziridination of 6-acyl-2,2-
dimethylchromene and 88-98% ee for a range of cinnamate esters. Styrenic and other alkenes are
converted with lower selectivities (5-54%). The catalytic system shows a linear response in product
ee to catalyst ee, and the product ee does not vary significantly during the reaction. UV
spectrophotometric investigations indicate that conversion of Cu(I) to Cu(II) is not essential for
catalysis but that Cu(II) is probably also a competent system.
In tr od u ction
catalytic enantioselective reactions with the Evans group
using chiral nonracemic bis(oxazoline) ligands7 and the
J acobsen group using bifunctional Schiff bases derived
from trans-1,2-diaminocyclohexane.8
Biaryl ligands have been used in the formation of
catalysts for a number of enantioselective transforma-
tions, including hydrogenation,9 Diels-Alder reactions,10
cyclopropanation,11 aldol condensations,12 formation of
chiral nonracemic Mannich bases,13 and ring-closing and
ring-opening olefin metathesis.14 The success of these
ligands is due not least to the fact that the axial chirality
Aziridines are highly versatile synthetic precursors and
have been used as synthons for chiral amines, amino
acids, amino alcohols, alkaloids, and â-lactam antibiot-
ics.1 There are also a number of natural products which
contain this functionality: for example, the mitomycins,
which display potent antitumor and antibiotic activity.2
Aziridines have been synthesized from amino alcohols.3
A number of routes to chiral nonracemic aziridines have
been developed using carbohydrates, hydroxy acids,
epoxides, and 1,2-diols, as well as some enzymatic
methods.1 Catalytic formation of aziridines has been
achieved by the addition of carbenes to imines.4 The first
reported addition of nitrene to olefin was by Kwartz and
Khan, whereby they generated nitrene from tosyl azide
in the presence of copper powder.5 The development by
Evans et al. of the copper-catalyzed aziridination of a
wide range of alkenes using PhINTs (N-(p-toluenesulfon-
yl)iminophenyliodinane) as the nitrene source was a
breakthrough in this area.6 Further developments led to
(7) Evans, D. A.; Faul, M. M.; Bilodeau, M. T.; Anderson, B. A.;
Barnes, D. M. J . Am. Chem. Soc. 1993, 115, 5328.
(8) (a) Li, Z.; Cosner, K. R.; J acobsen, E. N. J . Am. Chem. Soc. 1993,
115, 5326. (b) Li, Z.; Quan, R. W.; J acobsen, E. N. J . Am. Chem. Soc.
1995, 117, 5889. (c) Quan, R. W.; Li, Z.; J acobsen, E. N. J . Am. Chem.
Soc. 1996, 118, 8156.
(9) (a) Takaya, H.; Ohta, T.; Noyori, R. In Catalytic Asymmetric
Synthesis; Ojima, I., Ed.; VCH: Weinheim, Germany, 1993. (b) Noyori,
R. Asymmetric Catalysis in Organic Synthesis; Wiley: New York, 1994.
(10) (a) Corey, E. J .; Imwinkelried, R.; Pikul, S.; Xiang, Y. B. J . Am.
Chem. Soc. 1989, 111, 5493. (b) Chapuis, C.; J urzcak, J . Helv. Chim.
Acta 1987, 70, 436. (c) Narasaka, K. Synthesis 1991, 1. (d) Kagan, H.
B.; Riant, O. Chem. Rev. 1992, 92, 1007.
(11) (a) Gant, T. G.; Noe, M. C.; Corey, E. J . Tetrahedron Lett. 1995,
36, 8745. (b) Uozumi, Y.; Kyota, H.; Kishi, E.; Kitayama, K.; Hayashi,
T. Tetrahedron: Asymmetry 1996, 7, 1603. (c) Suga, H.; Fudo, T.; Ibata,
T. Synlett 1998, 8, 933.
† University of Warwick.
‡ GlaxoSmithKline.
(1) Tanner, D. Angew. Chem., Int. Ed. Engl. 1994, 33, 599.
(2) Kasai, M.; Kono, M. Synlett 1992, 10, 778.
(3) Wenker, H. J . Am. Chem. Soc. 1935, 57, 2328.
(12) Yoshikawa, N.; Yamada, Y. M. A.; Das, J .; Sasai, H.; Shibasaki,
M. J . Am. Chem. Soc. 1999, 121, 4168.
(13) Ishitani, H.; Ueno, M.; Kobayashi, S. J . Am. Chem. Soc. 2000,
122, 8180.
(14) (a) Weatherhead, G. S.; Houser, J . H.; Ford, J . G.; J amieson,
J . Y.; Schrock, R. R.; Hoveyda, A. H. Tetrahedron Lett. 2000, 41, 9553.
(b) Cefalo, D. R.; Kiely, A. F.; Wuchrer, M.; J amieson, J . Y.; Schrock,
R. R.; Hoveyda, A. H. J . Am. Chem. Soc. 2001, 123, 3139. (c) Hoveyda,
A. H.; Schrock, R. R.; Chem. Eur. J . 2001, 7, 945. (d) Alexander, J . B.;
La, D. S.; Cefalo, D. R.; Hoveyda, A. H.; Schrock, R. R. J . Am. Chem.
Soc. 1998, 120, 4041. (e) La, D. S.; Alexander, J . B.; Cefalo, D. R.; Graf,
D. D.; Hoveyda, A. H.; Schrock, R. R. J . Am. Chem. Soc. 1998, 120,
9720. (f) Weatherhead, G. S.; Ford, J . G.; Alexanian, E. J .; Schrock, R.
R.; Hoveyda, A. H. J . Am. Chem. Soc. 2000, 122, 1828.
(4) (a) Hansen, K. B.; Finney, N. S.; J acobsen, E. N. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 676-678. (b) Zhu, Z.; Espenson, J . H. J . Org.
Chem. 1995, 60, 7090. (c) Casarrubios, L.; Pe´rez, J . A.; Brookhart, M.;
Templeton, J . L. J . Org. Chem. 1996, 61, 8358. (d) Aggarwal, V. K.;
Thompson, A.; J ones, R. V. H.; Standen, M. C. H. J . Org. Chem. 1996,
61, 8368. (e) Wang, D.-K.; Dai, L.-X.; Hou, X.-L. Chem. Commun. 1997,
1231.
(5) Kwart, H.; Kahn, A. A. J . Am. Chem. Soc. 1967, 89, 1950-1953.
(6) (a) Evans, D. A.; Woerpel, K. A.; Hinman, M. M.; Faul, M. M. J .
Am. Chem. Soc. 1991, 113, 726. (b) Evans, D. A.; Faul, M. M.; Bilodeau,
M. T. J . Org. Chem. 1991, 56, 6744. (c) Evans, D. A.; Woerpel, K. A.;
Scott, M. J . Angew. Chem., Int. Ed. Engl. 1992, 31, 430. (d) Evans, D.
A.; Faul, M. M.; Bilodeau, M. T. J . Am. Chem. Soc. 1994, 116, 2742.
10.1021/jo025515i CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/17/2002

Aziridination Using Copper Complexes
J . Org. Chem., Vol. 67, No. 10, 2002 3451
Sch em e 1
of the ligand is well expressed in the steric environment
of the active site and because of the structural rigidity
provided by the biaryl backbone.
We have described the synthesis of a number of chiral
metal complexes based around the diaminobiphenyl
backbone 115 and have found that the salicylaldiminate
complexes exhibit nonplanar structures in which coli-
gands (e.g. Cl, alkyls) are forced to occupy mutually cis
coordination sites.15a,b In contrast, salen-derived Schiff-
base ligands such as those used in J acobsen’s catalyst
almost exclusively have trans structures.16 We have
recently reported a ruthenium Schiff-base system which,
as a result of the nonplanar structure, catalyses alkene
cyclopropanation with remarkable enantio- and diaste-
reoselectivity.17 We have previously communicated prin-
ciples for the design of biaryldiimine ligands for azirid-
ination18 and an investigation into their mode of action.19
In this paper we describe synthetic routes to this pre-
catalyst system and detail the methods for its application
to the aziridination of chromenes, cinnamate esters,
styrenes, and other alkenes.
compound was finally solved effectively by the use of
simulated moving bed (SMB) chromatography.24
A suitable analytical HPLC method was modified to
allow maximum throughput without loss of separation.
Initial operating conditions for the SMB were obtained
after determination of the adsorption isotherm. When the
product streams were monitored and the internal con-
centration profile of the system was determined, good
product quality was achieved. It was possible to separate
56 g of racemic 1 in a 24 h period to give the two
enantiomers using this method. Enantiomeric excesses
of 93.3% for the raffinate and 92.5% for the extract were
obtained from an initial pass. Reprocessing of these
fractions under the same conditions gave 99.8% ee for
the raffinate and 99.1% ee for the extract. Confirmation
of the absolute configurations of the diamines produced
in this way is described in the following sections.
Sch iff Ba se Liga n d s. The racemic ligands L1-L6
(Scheme 1) were readily synthesized by condensation of
the appropriate aldehyde with the diamine in refluxing
methanol or ethanol until precipitation was complete.
However, the more soluble chiral nonracemic diimines
condensed more slowly and occasionally required some
method of dehydration to force the reaction to completion.
For example, (R)-(+)-1 and 2,6-dichlorobenzaldehyde
were mixed with triethylamine in dichloromethane with
dropwise addition of TiCl4 resulting in the formation of
(R)-L1.25 This latter compound allowed us to confirm the
absolute configuration of the diamine by the method
described below.
Resu lts a n d Discu ssion
Ch oice, Syn t h esis, a n d R esolu t ion of Bia r yl-
d ia m in e. While 1,1′-binaphthyl-2,2′-diamine is com-
mercially available in chiral nonracemic form, it is
restrictively expensive for use in exploratory chemistry
involving catalyst synthesis and characterization. Also,
the potent carcinogen 2-naphthylamine is either used in
its synthesis or is a byproduct thereof. In our work on
biaryl-bridged chiral diimino ligands we have therefore
chosen to use 2,2′-diamino-6,6′-dimethylbiphenyl (1).
The diamine was synthesized on a scale of up to 50 g
by use of modified literature procedures.20,21 In particular,
the capricious melt-phase Ullmann coupling step was
replaced by an adaptation of the Cu/DMF procedure.22
The subsequent hydrogenation was found to be quantita-
tive in the presence of 5% Pd/C (see the Supporting
Information).
The resolution of racemic 1 was attempted using the
literature procedure: i.e., repeated crystallization of the
tartrate salt.23 However, while the reported properties
of the chiral nonracemic diamines (e.g., melting points
and optical rotatory power) were reproduced, we were
able to determine by chiral HPLC that chiral nonracemic
1 thus formed had an enantiomeric excess of 82% at best.
While exhaustive recrystallization of this material led
eventually to an optically pure sample, we found this
method inconvenient. The problem of resolution of this
Assign m en t of Absolu te Con figu r a tion of (-)-L1
a n d Th u s (+)-1. Single crystals of L1 were grown from
ether/hexane solution, and the molecular structure was
determined. The presence of the relatively heavy Cl
atoms in the structure allowed refinement of Flack’s
parameter x to -0.01(8) for the absolute structure (R)-
(-)-L1 (see the Supporting Information).26 The raffinate
diamine is thus (R)-(+)-1, in agreement with Mislow’s
work, in which S-(-)-1 was synthesized from (S)-(-)-6,6′-
dinitro-2,2′-bis(bromomethyl)biphenyl.27
(15) (a) Woodman, P. R.; Alcock, N. W.; Munslow, I. J .; Sanders, C.
J .; Scott, P. J . Chem. Soc., Dalton Trans. 2000, 3340. (b) Woodman, P.
R.; Munslow, I. J .; Hitchcock, P. B.; Scott, P. J . Chem. Soc., Dalton
Trans. 1999, 4069. (c) Woodman, P. R.; Sanders, C. J .; Alcock, N. W.;
Hitchcock, P. B.; Scott, P. New J . Chem. 1999, 23, 815. (d) Woodman,
P. R.; Hitchcock, P. B.; Scott, P. Chem. Commun. 1996, 2735.
(16) Pospisil, P. J .; Carsten, D. H.; J acobsen, E. N. Chem. Eur. J .
1996, 2, 974.
(17) Munslow, I. J .; Gillespie, K. M.; Deeth, R. J .; Scott, P. Chem.
Commun. 2001, 1638.
(18) Sanders, C. J .; Gillespie, K. M.; Bell, D.; Scott, P. J . Am. Chem.
Soc. 2000, 122, 7132.
(19) Gillespie, K. M.; Crust, E. J .; Deeth, R. J .; Scott, P. Chem.
Commun. 2001, 785.
(20) Meisenheimer, J .; Ho¨ring, M. Chem. Ber. 1927, 60, 1425.
(21) Carlin, R. B.; Foltz, G. E. J . Am. Chem. Soc. 1956, 78, 7.
(22) Wilcox, C. F.; Lassila, K. R.; Kang, S. J . Org. Chem. 1988, 53,
4333.
(24) (a) Migliorini, C.; Mazzotti, M.; Morbidelli, M. J . Chromatogr.
A. 1998, 827, 161. (b) Heuer, C.; Ku¨sters, E.; Plattner, T.; Seidel-
Morgenstern, A. J . Chromatogr. A. 1998, 827, 175. (c) Miller, L.;
Orihuela, C.; Fronek, R.; Honda, D.; Dapremont, O. J . Chromatogr.
A. 1999, 849, 309 and references therein.
(25) (a) Weingarten, H.; Chupp, J . P.; White, W. A. J . Org. Chem.
1967, 32, 3246. (b) Moretti, I.; Torre, G. Synthesis 1970, 141.
(26) Bernardinelli, G.; Flack, H. D. Acta Crystallogr. 1985, A41, 500.
(27) (a) McGinn, F. A.; Lazarus, A. K.; Siegel, M.; Ricci, J . E.; Mislow,
K. J . Am. Chem. Soc. 1958, 80, 476. (b) Fitts, D. D.; Siegel, M.; Mislow,
K. J . Am. Chem. Soc. 1958, 80, 480. (c) O’Connor, M. J .; Ernst, R. E.;
Holm, R. H. J . Am. Chem. Soc. 1968, 90, 4561. (d) Seno, K.; Hagishita,
S.; Sato, T.; Kuriyama, K. J . Chem. Soc., Perkin Trans. 1 1984, 2013.
(23) Dethloff, W.; Mix, H. Chem. Ber. 1949, 534.

3452 J . Org. Chem., Vol. 67, No. 10, 2002
Gillespie et al.
Syn t h esis a n d Ch a r a ct er iza t ion of Cu Com -
p lexes. The choice of metal source has played an
important role in much of the chemistry of copper-
catalyzed aziridination and cyclopropanation reactions.
In the latter it was found that Cu(I) was the oxidation
state of choice.6a,b Evans found that both Cu(I) and
Cu(II) were efficient precatalysts for the aziridination of
alkenes. For example, both [Cu(CH3CN)4][ClO4] and
[Cu(acac)2] gave a high turnover number.6b It was later
found that copper(I) triflate was also an excellent cata-
lyst, particularly in combination with chiral ligands such
as those of Evans.6a Norrby has very recently described
detailed calculations on the J acobsen system which
indicate a route by which a Cu(II) starting complex may
enter the Cu(I) manifold.28 Since Cu(I) complexes are
diamagnetic, we decided to synthesize precatalysts in this
oxidation state.
[Cu(CH3CN)4][BF4] and [(CuOTf)2C6H6] are commer-
cially available but are frequently contaminated with
paramagnetic Cu(II). Hence, [Cu(CH3CN)4][BF4] was
synthesized by the reaction of copper(I) oxide with
tetrafluoroboric acid in refluxing acetonitrile, an adaption
of Kubas’ procedure.29 The compound is relatively air and
water stable and can be stored in a sealed container for
up to 1 month without any appreciable hydrolysis or
indefinitely under an inert atmosphere. [(CuOTf)2C6H6]
was synthesized by the reaction of copper(I) oxide with
trifluoromethanesulfonic anhydride in refluxing benzene,
in an adaption of a procedure published by Kochi.30 The
off-white solid can be stored indefinitely under an inert
atmosphere.
F igu r e 1. Views (a, top; b, bottom) of the molecular structure
of [CuL1(CH3CN)2][BF4].
The copper diimine complexes were synthesized by
mixing a slight excess of 1 equiv of the racemic diimine
ligand with either of the copper(I) salts [Cu(CH3CN)4]-
[BF4] and [(CuOTf)2C6H6] in dichloromethane, followed
by crystallization. The chiral nonracemic complexes
crystallized less readily, if at all. The 1H and 13C{1H}
NMR spectra in solution of the free ligand L1 and the
diamagnetic complexes [CuL1(CH3CN)2][BF4] or [CuL1]-
[OTf] are all very similar. For example, the singlet
resonance for the imine protons appears at ca. 8.5 ppm
in all cases. The spectrum obtained by mixing the ligand
with substoichiometric amounts of CuI has essentially the
same appearance; separate ligand/complex peaks are not
resolved. This arises from rapid exchange of coordinated
and uncoordinated ligands on the 1H NMR chemical shift
time scale, as might be expected for a d10 metal center.
Single crystals of the complexes [CuL1(CH3CN)2][BF4]
suitable for X-ray structural analysis were grown by
concentrating the reaction mixtures in dichloromethane
and cooling. Representations of the molecular structure
of the cationic unit in [CuL1(CH3CN)2][BF4] are shown
in Figure 1. The structure of [CuL2(CH3CN)2][BF4] is
similar (see the Supporting Information). The systems
are monometallic, with a pseudo-tetrahedral arrange-
ment of the imine nitrogen atoms of the ligand and
nitrogen atoms of the two acetonitrile ligands around
copper. It is apparent from Figure 1b that the chirality
of the diamine is expressed efficiently in the region of
the labile acetonitrile ligands. There is a relatively short
dative contact between Cl(2) and the metal center (ca.
3.54 Å).
F igu r e 2. Molecular structure of the dicationic unit in
[Cu2L32][OTf]2.
The complexes between L3 and copper sources were
synthesized in the same manner and gave NMR spectra
similar to those above. However, FAB mass spectra of
dichloromethane solutions exhibited ions arising from
+
[Cu2L32][OTf]+ (m/z 1277) and [Cu2L3
]
(m/z 1128) as
2
well as the monometallic [CuL3]+ (m/z 563). A molecular
ion for the monomer unit [CuL3][OTf]+ was not detected
as it was for the complexes [CuL1(OTf)] and [CuL1(OTf)].
Subsequently, single crystals of the complexes [Cu2L3 ]-
2
[OTf]2 were grown from a concentrated solution in
dichloromethane. A structural representation is shown
in Figure 2, The ligands L3 bridge between two two-
coordinate copper(I) centers. Many linearly coordinated
Cu d10 compounds have been structurally characterized,
including Evans’ polymeric bis(oxazoline)copper(I) triflate
complex.6c In [Cu2L32][OTf]2 the biaryl backbone in the
ligand dictates that the complex comprises a double helix,
both ligands having the same absolute configuration. It
(28) Brandt, P.; So¨dergren, M. J .; Andersson, P. G.; Norrby, P. O.
J . Am. Chem. Soc. 2000, 122, 8013.
(29) Kubas, G. J . Inorg. Synth. 1979, 19, 90.
(30) Salomon, R. G.; Kochi, J . K. J . Am. Chem. Soc. 1973, 3300.

Aziridination Using Copper Complexes
J . Org. Chem., Vol. 67, No. 10, 2002 3453
Sch em e 2
Sch em e 3
may be seen from Figure 2 that edge-face contacts
between tert-butylphenyl rings are present.31,32
Hence, there are two distinct structural types present
in our precatalysts. The monometallics [CuL(CH3CN)2]+
arise where ortho substituents are present on the ligand:
i.e., for L1, L2, and (according to mass spectra) L5.
Bimetallics [Cu2L2]2+ arise where ortho substituents are
absent: i.e., L3, L4, and L6.
Ta ble 1. En a n tioselective Azir id in a tion of Ch r om en e 3a
entry
ligand
temp/°C
yield/%
t/minb
ee/%
1
2
3
4
L3
L4
L1
L2
L5
L6
L1
L1
L1
room temp
room temp
room temp
room temp
room temp
room temp
room temp
-40
53
75
83
79
56
0
85
80
87
900
900
<10
<10
<10
900
<10
300d
300d
13
16c
86
55
65
5
To investigate the origins of this behavior, we have
briefly studied models based on the crystallographically
6
7e
8
85
94
99
determined structure of the ion [Cu2L3 2+. Adding 2,6-
]
2
9e
-40
Cl atoms prevents edge-face orientation of the arenes
as in Figure 2, and the face-face conformation is also
destabilized by interligand steric compression (between
a Cl atom and the biaryl unit of the second ligand). We
thus consider that the [Cu2L2] structure is the default
type and its formation is prevented by the use of ortho-
substituted aromatic rings. This has important implica-
tions for turnover, rate, and enantioselectivity in the
subsequent catalytic reaction.
Op tim iza tion of th e Azir id in a tion Rea ction . With
reference to Evans’ procedure, the reaction was optimized
for the aziridination of styrene (Scheme 2).6 A slight
excess of the ligand was used to ensure complete com-
plexation of copper ions. The side product p-toluene-
sulfonamide (H2NTs) was minimized by use of a 5-fold
excess of alkene, although high yields of aziridine 2 were
obtained when only 1 equiv of the alkene was used. The
reaction proceeds very rapidly, and for several reactions
all the PhINTs dissolved in less than 1 min.
A wide range of solvents was examined, but only CH2-
Cl2 and MeCN gave reasonable yields. In preliminary
enantioselective reactions it was found that CH2Cl2 was
far superior; it is probable that acetonitrile competes
successfully with the diimine ligand, leading to catalysis
by achiral derivatives of [Cu(CH3CN)4]+.
A number of nitrene sources were examined briefly,
and iodinane-type materials such as PhINNs (N-(p-
nitrophenylsulfonyl)iminophenyliodinane),33 PhINSs (N-
(2-trimethylsilyl)ethanesulfonylphenyliodinane),34 and
2-(ter t-bu t ylsu lfon yl)(p-t olu en esu lfon ylim in oiodo)-
benzene35 were effective. The more commonly used
PhINTs (N-(p-toluenesulfonyl)iminophenyliodinane) was
chosen for detailed study because of its ease of synthesis
and so that results could be compared directly with those
of others. Chloramine-T did not operate as a nitrene
source, although this has been achieved by Taylor in
another system.36
a
[CuI(CH3CN)4]BF4 was used as the CuI source, unless other-
wise stated. Substitution of [CuI2(OTf)2]‚C6H6 did not affect the
results significantly. Time until all PhINTs had dissolved. c Op-
b
d
posite enantiomer to other ligand systems. The rate of the low-
temperature reactions is probably limited by the rate of dissolution
of PhINTs at -40 °C. e [CuII(CH3CN)4][BF4]2 was used as the
source of copper.38
The performances of the different ligands were as-
sessed in a test reaction involving chromene 3 (Scheme
3). In the case of the systems which form bimetallic
[Cu2L2] precatalysts, dissolution of the sparingly soluble
trivalent iodinane nitrene source PhINTs was not com-
plete after 900 min and the enantiomeric excess of the
product was very poor (Table 1, entries 1, 2, and 6). The
systems having monometallic precatalysts were strik-
ingly more efficient. Dissolution of PhINTs was complete
in minutes, isolated yields were good, and the product
ee was dramatically improved (entries 3-5).37
It is apparent also from Table 1 that the nature of the
ortho substituents on the ligand has a pronounced effect
on the ee, with dichloro giving the best results. We
propose that this arises from the function of Cl as a dative
ligand for the copper center, as noted in the structure of
[CuL1(CH3CN)2][BF4] and also in our DFT calculations.19
Such interactions at the metal will help to shape the
chiral pocket at the active site, without taking up one of
the key tetrahedral positions.
The mass spectrometric data presented above indicate
that the structures obtained from X-ray crystallography
appear also to be present in solution. Nevertheless, given
the rapid ligand exchange possible in Cu systems, it
remained that a nonlinear effect19 might be responsible
for the variation in catalyst performance. This possibility
is almost certainly excluded by the fact that essentially
linear plots were obtained for eeligand vs eeproduct for L1,
L3, and L4 (Figure 3).
Further in support of a simple, “single-site” catalyst,
we have found that the enantiomeric excess of the
product obtained does not vary significantly with the
percent conversion (Figure 4); it is very unlikely that the
nature of the active catalyst changes during the course
of the reaction.
(31) Nishio, M.; Umezawa, Y.; Hirota, M.; Takeuchi, Y. Tetrahedron
1995, 51, 8665.
(32) The phenomenon of noncovalent interactions between arene
rings is well-documented, and for edge-face interactions the distances
between the two ring centroids are found in the range 4.5-7.0 Å, with
a
theoretical optimum distance of 5.2 Å. In [Cu2L32}2][OTf]2 the
centroid-centroid distances are ca. 4.99 Å. (a) Burley, S. K.; Petsko,
G. A. Science 1985, 229, 23. (b) J orgensen, W. L.; Severance, D. L. J .
Am. Chem. Soc. 1990, 112, 4768.
The degree of enantiomeric excess obtained, the rate
of reaction, and the yield thus relate very strongly to the
(33) (a) Fukuyama, T.; J ow, C.-K.; Cheung, M. Tetrahedron Lett.
1995, 36, 6373. (b) Fukuyama, T.; Cheung, M.; J ow, C.-K.; Hidai, Y.;
Kan, T. Tetrahedron Lett. 1997, 38, 5831.
(34) Dauban, P.; Dodd, R. H. J . Org. Chem. 1999, 64, 5304.
(35) Macikenas, D.; Skrzypczak-J ankun, E.; Protasiewicz, J . D. J .
Am. Chem. Soc. 1999, 121, 7164.
(36) Albone, D. P.; Aujla, P. S.; Taylor, P. C.; Challenger, S.; Derrick,
A. M. J . Org. Chem. 1998, 63, 9569.
(37) Strenuous attempts to determine the absolute configuration of
the chromene aziridine by crystallographic and synthetic methods were
unsuccessful.

3454 J . Org. Chem., Vol. 67, No. 10, 2002
Gillespie et al.
Ta ble 2
F igu r e 3. Variation of the enantiomeric excess of the product
with that of the precatalyst for the aziridination of chromene
3 with [Cu(CH3CN)4]BF4 in the presence of chiral nonracemic
L: (0) L1; ()) L3; (4) L4.
F igu r e 5. UV spectra of Cu(I) and Cu(II) complexes reacted
with PhINTs in CH2Cl2: (a) [CuI(CH3CN)4][BF4] + L1, 10 mM
in CH2Cl2; (b) [CuI(CH3CN)4][BF4] + L1 + PhINTs, 10 mM in
CH2Cl2; (c) [CuII(CH3CN)4][BF4]2 + L1, 10 mM in CH2Cl2; (d)
[CuII(CH3CN)4][BF4]2 + L1 + PhINTs, 10 mM in CH2Cl2.
F igu r e 4. Variation of product ee with conversion for the
aziridination of 3 with PhINTs using 5% [Cu(CH3CN)4][BF4]
and 5.5% (R)-L1 in CH2Cl2.
Sch em e 4
structure of the precatalyst determined by X-ray crystal-
lography. This could arise for a number of reasons. Since
the bimetallic Cu2L2 complexes do not have sterically
available coordination sites, the aziridination reaction in
these cases may be being mediated by an equilibrium
amount of an achiral copper complex, e.g. [Cu(CH3CN)n]+.
The low rates of reaction obtained would thus be due to
very low concentrations of this catalyst at equilibrium.
The low ee’s obtained could arise from catalysis by a
small equilibrium amount of the monomeric [CuL]+. It
seems less likely that the Cu2L2 catalyzes the reaction
with low enantioselection.
These observations indicate that in the present system
the conversion of Cu(I) to Cu(II) is not essential for
catalysis. Nevertheless, it seems either that Cu(II) is also
a competent catalyst or that a small amount of Cu(II) is
reduced to a Cu(I) catalyst in situ, perhaps by the Norrby
mechanism.28
To ascertain whether this effect holds for other alkene
substrate types, catalyst systems of both mono- and
bimetallic types were used in the aziridination of styrene
and ethyl cinnamate. The results in Table 2 show that,
as with chromene, the time for completion of the reaction,
yield, and ee are significantly better for the monometallic
[CuL] systems than for [Cu2L2].
En a n tioselective Azir id in a tion of Cin n a m a te Es-
ter s. We set out to optimize the aziridination of cin-
namate esters, the products of which may allow access
to phenylalanine analogues.39 The ligand L1 (Scheme 4)
was used throughout the experiments, whose results are
shown in Table 3. In most cases a significant improve-
ment in yield and ee was observed for reactions per-
formed at -40 °C compared to those at room tempera-
ture. At still lower temperatures the rate was made
UV Stu d ies. Evans proposes oxidation of Cu(I) to a
Cu(II) active catalyst in the bis(oxazoline) ligated sys-
tem.7 The identical performance of the two catalysts in
terms of yield and ee and indistinguishable UV spectra
for the complexes treated with PhINTs strongly support
this. In our system it can be seen from Figure 5 that the
spectrum (a) of precatalyst [CuL1(CH3CN)2][BF4] is little
affected by treatment with PhINTs (b). The analogous
Cu(II) spectra behave similarly ((c) and (d)).38 Comparison
of (b) with (d) shows that the addition of PhINTs has
apparently not caused oxidation of the Cu(I) complex.
(38) (a) Hathaway, B. J .; Underhill, A. E. J . Chem. Soc. 1960, 3705.
(b) Hathaway, B. J .; Holah, D. G.; Underhill, A. E. J . Chem. Soc. 1962,
2444.
(39) (a) Dauban, P.; Dodd, R. H. Tetrahedron Lett. 1998, 39, 5739.
(b) Rousseau, J .-F.; Dodd, R. H. J . Org. Chem. 1998, 63, 2731.

Aziridination Using Copper Complexes
J . Org. Chem., Vol. 67, No. 10, 2002 3455
Ta ble 3. Azir id in a tion of Cin n a m a te Ester s
Ta ble 4. Azir id in a tion of Olefin s Usin g Cu I/L1 a s
Ca ta lyst
a
Absolute configuration 2R,3R-(-) determined by comparison
with data from ref 7. ee for trans product. c ee for cis product.
Absolute configuration 2S,3R-(-) determined by comparison with
data from ref 8b.
b
d
namate esters, significant levels of trans/cis isomerization
are apparent in some cases. Similarly to Evans,7 we found
that the absolute configuration of styrene-derived aziri-
dine was opposite to that for the trans substrates
â-methylstyrene and methyl cinnamate.40
a
b
Recovered yield. Enantiomeric excess determined by HPLC
using Chiralpak Chiralcel OD column 15 cm × 0.46 cm i.d. with
mobile phase hexane/propan-2-ol (various ratios) unless otherwise
stated. c Enantiomeric excess determined by 1H NMR in the
presence of the chiral shift reagent Eu(hfc)3.
Con clu sion s
inconveniently slow, probably limited by the rate of
dissolution of PhINTs. It was first noted that the size of
the ester alkyl substituent has a small but significant
influence on enantioselection; the tert-butyl compound
(entry 2) gave a higher ee than the methyl (entry 1) and
ethyl compounds (entry 4, Table 2). For this reason, the
tert-butyl esters 5, synthesized readily from the com-
mercially available acids (see the Supporting Informa-
tion) were used in subsequent studies. Arene substitution
with various groups appears to have little effect on
enantioselection, although the 3-NO2 compound gave low
yields (entries 2-8). Significantly, no cis/trans isomerism
was observed; only anti isomers of the aziridines were
detected. The absolute configuration of the unsubstituted
aziridine (entry 2) was determined by comparison of the
observed optical rotation with the literature value7 and
confirmed by X-ray crystallography (see the Supporting
Information). The 2S,3R configuration is as predicted by
our previously communicated model.19
Efficient, and in some cases highly enantioselective,
catalysts for the aziridination of alkenes have been
produced using a biaryldiimine ligand set Ln at copper.
The performance of the catalysts is highly dependent on
the substitution pattern of the ligand arylimine unit,
which determines catalyst speciation; ortho disubstitu-
tion is required to furnish monometallic, active systems.
By some way the highest enantioselectivities are obtained
for L1, which contains o-dichlorophenyl groups. Interest-
ingly, the most successful ligands in J acobsen’s diimi-
nocyclohexane8 system and Suga and co-workers’ binaph-
thyldiimine catalysts11c for aziridination and cyclopropana-
tion repectively, also incorporate this unit. In the former
case this may be due to the presence of Cl‚‚‚Cu dative
interactions.19 In the latter, catalyst speciation as de-
scribed above is likely to be the determining factor.
Exp er im en ta l Section
Azir id in a tion of Styr en ic Alk en es. The results for
aziridination of a number of styrenic and other alkenes
are given in Table 4. Although yields are reasonable,
enantiomeric excesses for these substrates are low in
comparison to those with the functionalized systems
above. The contrast with cinnamates is significant and
can be attributed to a lack of secondary binding of the
substrate.19 Evans observed similar selectivities for the
cinnamate and styrenic substrates.7 Unlike the cin-
Gen er a l Exp er im en ta l P r oced u r es. Catalytic procedures
were carried out under an inert atmosphere of argon by using
a dual-manifold vacuum/argon line and standard Schlenk
techniques. Solvents were dried by refluxing over the ap-
propriate drying agent (calcium hydride for dichloromethane,
potassium for THF) for 3 days under dinitrogen. All glassware,
cannulas, and Celite were stored in an oven (>100 °C) and
(40) This refers to the configuration of the benzylic carbon.

3456 J . Org. Chem., Vol. 67, No. 10, 2002
Gillespie et al.
flame-dried immediately prior to use. Deuterated chloroform
was dried in the bottle in air over molecular sieves (4 Å).
Infrared spectra were obtained either as Nujol mulls or by
evaporation of dichloromethane solutions onto IR plates. SMB
separation was performed on a Novasep Licoseplab with 8 ×
Chiralpak AD columns, 10 cm bed length × 26 mm id. The
eluent was hexane/IPA in the ratio 60:40 maintained at 28
°C. HPLC experiments were performed on a Kontron instru-
ment: Model 332 detector, Model 320 system. The columns
used were Chiracel OD and AD, 15.0 cm × 0.46 cm i.d. Column
chromatography was performed using a selection of column
widths and Merck silica gel 60. Thin-layer chromatography
was performed using Merck 0.25 mm silica layer foil-backed
plates.
J ) 7.5 Hz), 7.23-7.28 (m, 3H), 7.45-7.65 (m, 4H), 7.80-7.83
(m, 10H), 8.25 (s, 2H). 13C NMR (CDCl3; δ): 20.0 (Me). 115.5
(Ar), 123.8, 126.4, 126.8, 127.2, 127.8, 128.3, 128.6, 130.8,
132.1, 132.9, 134.2, 134.8, 136.9, 151.4, 159.6. MS (EI+; m/z):
488 (M+), 473 (M+- CH3). IR (CD2Cl2; ν, cm-1): 3054, 2917,
2859, 1925, 1620, 1571, 1453, 819, 764, 745. Anal. Calcd for
C
36H28N2: C, 88.49; H, 5.77; N, 5.73. Found: C, 87.85; H, 5.72;
N, 5.72.
(R)-(-)-L4. Yield: 1.8 g, 80%. [R]D25 ) -832.4° (c 0.250, CH2-
Cl2). HPLC (Chiracel OD, hexane/IPA 90:10, 0.5 mL/min): tR
) 9.8 (0.4%), 11.5 (99.6%) min; ee ) 99.2%.
(R)-(-)-N,N′-Bis(a n th r a cen -9-ylid en e)-6,6′-d im eth ylbi-
p h en yl-2,2′-d ia m in e ((R)-(-)-L5). Yellow powder. Yield: 2.56
1
g, 93%. H NMR (CDCl3; δ): 2.24 (s, 6H), 7.18-7.25 (m, 6H),
Syn th esis of Liga n d s L1-L6. (()-N,N′-Bis(2,6-d ich lo-
r oben zylid en e)-6,6′-d im eth ylbip h en yl-2,2′-d ia m in e (L1).
The following represents the general procedure for ligand
synthesis. (()-2,2′-Diamino-6,6′-dimethylbiphenyl (1; 1.06 g,
5.0 mmol) and 2,6-dichlorobenzaldehyde (0.88 g, 5.0 mmol)
were dissolved in methanol (10 mL) and stirred at room
temperature for 3 h to produce a yellow crystalline solid. The
reaction mixture was cooled to -30 °C, and the resultant
crystals of the Schiff base were isolated by vacuum filtration,
washed with cold methanol, and dried in vacuo. Yield: 2.2 g,
7.33-7.39 (m, 6H), 7.54 (t, 2H, J ) 7.7 Hz), 7.89 (d, 4H, J )
8.5 Hz), 8.18 (d, 4H, J ) 9.0 Hz), 8.38 (s, 2H), 9.48 (s, 2H). 13
C
NMR (CD2Cl2; δ): 20.7, 117.1, 125.5, 126.0, 127.7, 128.0, 128.3,
129.4, 129.5, 131.0, 131.1, 131.9, 133.7, 138.1, 153.4, 160.3.
MS (EI+; m/z): 588 (M+). MS (CI+; m/z): 607 (MNH4+), 589
(MH+). IR (CH2Cl2; ν, cm-1): 3051, 2918, 2862, 1671, 1626,
1572, 1520, 1451, 1263, 1239, 890, 784, 732. Anal. Calcd for
C
44H32N2: C, 89.76; H, 5.48; N, 4.76. Found: C, 89.33; H, 5.34;
25
N, 4.84. [R]D ) -6.1° (c 0.114, CH2Cl2).
(()-N,N′-Bis(4-n itr oben zyliden e)-6,6′-dim eth ylbiph en yl-
1
89%. Mp: 166 °C. H NMR (CDCl3; δ):, 2.05 (s, 6H), 6.86 (d,
2,2′-d ia m in e (L6). Yellow crystalline solid. Yield: 2.0 g, 88%.
2H, J ) 7.3 Hz), 7.07-7.29 (m, 10H), 8.51 (s, 2H). 13C NMR
(CDCl3; δ): 19.8, 115.7, 127.3, 128.5, 130.1, 131.8, 132.4, 135.0,
137.1, 151.0, 155.7. MS (EI+): m/z 527 (M+), 511 (M+ - CH3),
353 (M+ - NC7H4Cl2). IR (CD2Cl2; ν, cm-1): 3056, 2915, 2358,
1635, 1575, 1557, 1455, 1430, 775, 749. Anal. Calcd for
1
Mp: 278 °C. H NMR (CDCl3; δ): 2.04 (s, 6H), 6.88 (d, 2H, J
) 7.7 Hz), 7.16 (d, 2H, J ) 7.5 Hz), 7.28 (t, 2H, J ) 7.7 Hz),
7.62 (d, 4H, J ) 8.7 Hz), 8.17 (d, 4H, J ) 8.9 Hz), 8.28 (s, 2H).
13C NMR (CDCl3; δ): 19.9, 114.4, 123.8, 127.9, 128.1, 128.9,
132.5, 137.3, 141.7, 149.0, 149.6, 156.3. MS (EI+; m/z): 478
(M+). IR (CD2Cl2; ν, cm-1): 3010, 2359, 1638, 1600, 1523, 1343,
855, 841, 744, 689. Anal. Calcd for C28H22N4O4: C, 70.28; H,
4.63; N, 11.71. Found: C, 70.19; H, 4.66; N, 11.63.
C
28H20N2Cl4: C, 63.90; H, 3.83; N, 5.32. Found: C, 63.89; H,
3.80; N, 5.36.
(R)-(-)-L1. (R)-1 (1.06 g, 5.0 mmol) and 2,6-dichlorobenz-
aldehyde (0.88 g, 5 mmol) were placed in a 100 mL round-
bottom flask, and dichloromethane (20 mL) and Et3N (2.5 mL)
were added. The solution was cooled to 0 °C in an ice/salt bath.
A solution of TiCl4 in dichloromethane (2.5 mmol) was added
over 5 min. The solution was stirred overnight and then
filtered through Celite before evaporation in vacuo. The
residue was redissolved in toluene, and the solution was stirred
for 30 min. After filtration the toluene was removed in vacuo
and the crude residue was recrystallized from ether/hexane.
Cop p er Com p lexes. [Cu L1(CH3CN)2][BF 4]‚CH2Cl2. A
Schlenk vessel was charged with [Cu(CH3CN)4][BF4] (72.4 mg,
230 µmol) and (()-L1 (133 mg, 253 µmol). The two solids were
dissolved in dichloromethane with stirring, and an equal
volume of pentane was added to the yellow solution. When
the mixture was cooled, yellow needlelike crystals formed. The
crystals were of sufficient quality for X-ray analysis. Drying
in vacuo caused loss of 1 mol of dichloromethane of crystal-
lization. 1H NMR (CD2Cl2; δ): 1.95 (s, 6H), 2.05 (s, 6H), 7.00-
7.44 (m, 12H), 8.54 (s, 2H). 13C NMR (CD2Cl2; δ): 2.5, 20.2,
118.2, 129.5, 129.96, 130.04, 132.1, 133.0, 135.0, 139.7, 162.6.
Anal. Calcd for C33H28BCl6CuF4N4: C, 46.96; H, 3.35; N, 6.64.
Found: C, 47.44; H, 3.37; N, 6.35.
25
Yield: 1.4 g, 56%. [R]D ) -28.1° (c 0.360, CH2Cl2). A crystal
suitable for X-ray structural determination was chosen and
the molecular structure solved. Refinement of the Flack
parameter showed that the correct absolute structure was the
R isomer.
[Cu L1][OTf]. (()-L1 (13.4 mg, 25.4 µmol) and (CuOTf)2‚
C6H6 (6.0 mg, 12.0 µmol) were dissolved in d2-dichloromethane.
1H NMR (CD2Cl2; δ): 1.89 (s, 6H). 7.20-7.30 (m, 15H), 8.46
(s, 2H). 13C NMR (CD2Cl2; δ): 20.31, 118.75, 129.07, 129.75,
130.08, 131.25, 132.97, 135.10, 162.56. IR (CD2Cl2; ν, cm-1):
3339, 3121, 2302, 2196, 1750, 1628, 1579, 1560, 1433, 1385,
1313, 1270, 1234, 1211, 1170. MS (EI; m/z): 738 [CuL1][OTf]+,
(()-N,N′-Bis(2,4,6-tr im eth ylben zylid en e)-6,6′-d im eth -
ylbip h en yl-2,2′-d ia m in e (L2). Yellow crystalline solid.
1
Yield: 4.9 g, 87%. Mp: 138 °C. H NMR (CDCl3; δ): 2.03 (s,
6H). 2.09 (s, 12H), 2.22 (s, 6H), 6.75 (s, 4H), 6.75 (d, 2H, J )
7.6 Hz), 7.06 (d, 2H, J ) 7.6 Hz), 7.22 (t, 2H, J ) 7.6 Hz), 8.55
(s, 2H). 13C NMR (CDCl3; δ): 19.9, 20.5, 21.0,. 115.4, 126.4,
127.7, 129.5, 130.1, 132.4, 136.7, 138.7, 139.2, 152.3, 159.3.
MS (EI+; m/z): 472 (M+), 457 (M+ - CH3). IR (CD2Cl2; ν, cm-1):
3057, 2917, 2860, 2358, 1925, 1631, 1607, 1573, 1453, 1377,
845, 778, 743. Anal. Calcd for C34H36N2: C, 86.40; H, 7.68; N,
5.93. Found: C, 86.47; H, 7.70; N, 5.91.
589 [CuL1]+, 527 [L1] +. MS (CI; m/z): 589 [CuL1]+, 527 [L1] +
.
[Cu L2(CH3CN)2][BF 4]. (()-L2 (12.0 mg, 25.4 µmol) and
[Cu(CH3CN)4][BF4] (7.3 mg, 23.0 µmol) were dissolved in d2-
dichloromethane with warming. The sample was placed in the
freezer (-30 °C). The yellow crystals obtained were suitable
for X-ray structural determination. Drying in vacuo caused
complete loss of the solvent of crystallization. 1H NMR (CDCl3;
δ): 1.86 (s, 6H). 1.94 (s, 12H), 2.02 (s, 6H), 2.23 (s, 6H), 6.81
(s, 4H), 6.94 (d, 2H, J ) 7.6 Hz), 7.25 (d, 2H, J ) 7.6 Hz), 7.39
(t, 2H, J ) 7.6 Hz), 8.55 (s, 2H). IR (CH2Cl2; ν, cm-1): 2360,
2342, 1573, 1385, 1299, 1242, 1210, 1057, 872, 850, 784, 741.
Anal. Calcd for C38H42BCuF4N4: C, 64.73; H, 6.00; N, 7.95.
Found: C, 64.44; H, 6.06; N, 8.05.
(R)-(-)-L2. Yield: 52%. [R]D ) -578.2° (c 0.188, CH2Cl2).
25
(()-N,N′-Bis(4-ter t-bu tylben zylid en e)-6,6′-d im eth ylbi-
p h en yl-2,2′-d ia m in e (L3). White crystalline solid. Yield: 0.76
1
g, 65%. Mp: 165 °C. H NMR (CDCl3; δ): 1.31 (s, 18H), 2.05
(s, 6H,), 6.79 (d, 2H, J ) 8.7 Hz), 7.07 (d, 2H, J ) 8.7 Hz),
7.21 (t, 2H, J ) 9.3 Hz), 7.35 (d, 4H, J ) 9.9 Hz), 7.48 (d, 4H,
J ) 9.9 Hz), 8.18 (s, 2H). 13C NMR (CDCl3; δ): 20.4, 31.6, 116.0,
125.8, 126.9, 128.0, 128.7, 132.4, 134.4, 137.3, 159.6 151.7,
154.6. MS (EI+; m/z) 500 (M+), 485 (M+ - Me). MS (CI+; m/z)
501 (MH+). IR (CH2Cl2; ν, cm-1): 3056, 2961, 2865, 1630, 1566,
1456, 830, 746. Anal. Calcd for C36H40N2: C, 86.35; H, 8.05;
N, 5.60. Found: C, 86.44; H, 8.05; N, 5.57.
[Cu L2][OTf]. (()-L2 (13.0 mg, 27.5 µmol) and (CuOTf)2‚
C6H6 (6.3 mg, 12.5 µmol) were dissolved in d2-dichloromethane.
1H NMR (CD2Cl2; δ): 1.91 (s, 6H), 1.94 (s, 12H), 2.14 (s, 6H),
6.73 (s, 4H), 6.93 (d, 2H, J ) 7.7 Hz), 7.18 (d, 2H, J ) 7.5 Hz),
7.22 (s, 3H, C6H6), 7.30 (t, 2H, J ) 7.7 Hz), 8.54 (s, 2H). 13C
NMR (CD2Cl2; δ): 20.2, 20.3, 21.7, 129.8 (C6H6), 119.2, 128.9,
1230.0, 130.3, 130.8, 131.0, 137.5, 139.1, 141.5, 149.2, 169.4.
MS (FAB+; m/z): 684 ([CuL2][OTf]+), 535 ([CuL2]+), 471 [L2]+.
(R)-(-)-L3. Yield: 1.1 g, 90%. [R]D25 ) -765.4° (c 0.156, CH2-
Cl2).
(()-[N,N′-Bis(2-n a p h th ylid en e)-6,6′-d im eth ylbip h en yl-
2,2′-d ia m in e (L4). Yield: 2.1 g, 85%. Mp: 168 °C. 1H NMR
(CDCl3; δ): 2.12 (s, 6H). 6.84 (d, 2H, J ) 7.5 Hz), 7.12 (d, 2H,

Aziridination Using Copper Complexes
J . Org. Chem., Vol. 67, No. 10, 2002 3457
[Cu 2L32][OTf]2. (()-L3 (12.5 mg, 25.0 µmol) and (CuOTf)2‚
C6H6 (6.0 mg, 12 µmol) were dissolved in d2-dichloromethane.
After NMR analysis, concentration of the solution resulted in
formation of crystals suitable for X-ray diffraction. Drying in
vacuo caused complete loss of the solvent of crystallization.
1H NMR (CD2Cl2; δ): 1.12 (s, 36H), 1.73 (s, 12H), 7.00 (d, 8H,
J ) 8.3 Hz), 7.15 (d, 8H, J ) 7.5 Hz), 7.19 (t, 4H, J ) 7.5 Hz),
7.42 (d, 8H, J ) 8.3 Hz), 7.26 (s, C6H6), 8.62 (s, 4H). 13C NMR
(CD2Cl2; δ): 20.4, 31.4, 129.1 (C6H6), 122.0, 126.8, 129.7, 129.9,
130.8, 132.0, 133.1, 140.3, 160.7, 149.6, 173.0. MS (EI+; m/z):
712 [{CuIL3}][OTf]+, 500 [L3]+, 485 [L3 - CH3+]. MS (CI+;
m/z): 713 [CuIL3][OTf][H] +, 649 [(M+) - Cu], 563 [CuIL3]+,
485 [L3 - CH3+], 340 [L3 - NdCHC6H4C(CH3)3+]. MS (FAB+;
N-(p-Tolylsu lfon yl)-2-ca r bo-ter t-bu toxy-3-p h en yla zir i-
d in e (6a ). Pale, off-white solid. Yield: 144 mg, 77%. Mp: 55-
57 °C. IR (NaCl plates; νmax, cm-1): 3026, 2981, 2937, 1739,
1600, 1350, 1160, 1084, 909. 1H NMR (CDCl3; δ): 1.47 (s, 9H),
2.34 (s, 3H), 3.34 (d, 1H, J ) 4.0 Hz), 4.31 (d, 1H, J ) 4.0 Hz),
7.19 (m, 5H), 7.73 (d, 2H, J ) 8.3 Hz). 13C NMR (CDCl3; δ):
20.6, 26.8, 46.6, 47.7, 82.5, 126.2, 126.4, 127.5, 127.7, 128.5,
132.1, 136.6, 143.1, 163.6. MS (CI; m/z): 391 (M + NH4+), 374
1
(M + H+). Enantiomeric excess determined by H NMR using
Eu(hfc)3; ee ) 89%. [R]D25 ) -49.2° (c 1.00, CH2Cl2).7 Crystals
obtained were suitable for X-ray structural determination (see
Supporting Information).
N-(p -Tolylsu lfon yl)-2-ca r b o-ter t-b u t oxy-3-((4-m et h yl-
oxy)p h en yl)a zir id in e (6b). Yellow oil. Yield: 136.7 mg, 67%.
IR (NaCl plates; νmax, cm-1): 2977, 2927, 1746, 1614, 1516,
m/z): 1277 ([Cu2L32][OTf]+), 1128 [Cu2L3 +], 1063 [CuL32].
2
High-resolution MS (EI+): calcd mass (M+) 712.200 774, found
712.193 039. Anal. Calcd for C74H80Cu2F6N4O6S2: C, 62.30; H,
5.65; N, 3.93. Found: C, 62.30; H, 6.00; N, 4.15.
1
1368, 1332, 1253, 1161, 1091, 1032, 816. H NMR (CDCl3; δ):
1.15 (s, 9H), 2.42 (s, 3H), 3.56 (d, 1H, J ) 7.6 Hz), 3.74 (s,
3H), 4.29 (d, 1H, J ) 7.6 Hz), 6.78 (dd, 2H, J ) 6.7, 1.8 Hz),
7.22 (dd, 2H, J ) 6.7, 1.8 Hz), 7.33 (d, 2H, J ) 8.2 Hz), 7.90
(d, 2H, J ) 8.2 Hz). 13C NMR (CDCl3; δ): 22.1, 28.1, 44.7, 45.2,
55.7, 83.7, 113.9, 128.4, 129.2, 130.2, 164.0. MS (EI; m/z): 403
(M+), 347(M - C4H8+), 330 (M - OC4H9). HRMS (CI; m/z):
[Cu 2L42][OTf]2. (()-L4 (12.0 mg, 24.6 µmol) and (CuOTf)2‚
C6H6 (6.0 mg, 12 µmol) were dissolved in d2-dichloromethane.
Concentration of the solution resulted in the formation of
crystals suitable for X-ray diffraction. 1H NMR (CD2Cl2; δ):
1.96 (s, 12H), 6.64 (d, 4H, J ) 8.6 Hz), 6.84 (d, 4H, J ) 8.7
Hz), 7.06 (d, 4H, J ) 7.9 Hz), 7.26 (s, C6H6), 7.74 (s, 4H), 7.38-
7.14 (m, 16H), 8.77 (s, 4H). 13C NMR (CD2Cl2; δ): 20.7, 121.7,
122.3, 128.0, 128.4, 129.1, 129.2, 129.4, 129.6, 130.5, 132.4,
132.6, 133.1, 136.3, 149.9, 140.5, 164.9, 174.0. MS (FAB+;
m/z): 1253 ([{CuL4}2][OTf]+), 1104 ([{CuL4}2]+), 1040
([(CuL4)2]+), 552 ([CuL4]+). Anal. Calcd for C74H56Cu2F6-
N4O6S2: C, 63.37; H, 4.02; N, 3.99. Found: C, 63.30; H, 4.10;
N, 3.95.
404.1534, calcd [M+ + 1] 404.1550. Anal. Calcd for C21H25
-
NO5S: C, 62.51; H, 6.25; N, 3.47. Found: C, 62.49; H, 6.32;
N, 3.37. Enantiomeric excess determined by HPLC: Chiralcel
OD, 15 cm × 0.46 cm i.d.; mobile phase hexane/propan-2-ol
(95:5), tr(major) ) 16.41 min, tr(minor) ) 20.17 min, ee ) 93%.
26
[R]D ) +30.4° (c 1.06, CH2Cl2).
N-(p-Tolylsulfonyl)-2-carbo-tert-butoxy-3-(4-methylphen-
yl)a zir id in e (6c). Yellow oil. Yield: 114.3 mg, 59%. IR (NaCl
plates; νmax, cm-1): 2978, 2931, 1738, 1597, 1336, 1162, 1088,
[Cu L5][OTf]. (R)-L5 (15.0 mg, 25.5 µmol) and (CuOTf)2‚C6H6
(6.0 mg, 12 µmol) were dissolved in d2-dichloromethane. 1H
NMR (CD2Cl2; δ): 2.06 (s, 6H), 7.23 (s, C6H6), 6.79-8.11 (m,
26H), 9.38 (s, 2H). 13C NMR (CD2Cl2; δ): 20.5, 119.9, 123.0,
124.7, 125.0, 128.4, 129.0, 129.1 (C6H6), 129.2, 130.5, 130.8,
131.1, 131.3, 131.4, 132.8, 135.9, 139.9, 149.6, 164.4, 169.2.
MS (FAB+; m/z): 651 ([CuL5]+). IR (CD2Cl2; ν, cm-1): 3485,
3054, 1666, 1623, 1606, 1553, 1446, 1261, 1235, 1173, 1030,
954, 896, 736.
Gen er a l P r oced u r e for Azir id in a tion . A round-bottom
flask with a sidearm incorporating a PTFE stopcock was
charged with the metal salt (5 mol %) and the diimine ligand
(6 mol %) and was flushed with argon. Solvent (6 mL), usually
dichloromethane, was added, and the resulting mixture was
stirred for 15 min. Reactions were performed at room tem-
perature or in a bath of 2-propanol cooled to -40 °C using an
immersion cooler. The olefin (1-5 equiv) was added to the
stirred solution either via syringe or by solid addition against
a slow positive flow of argon. The solid nitrene source (1 equiv)
was added in a similar manner in a single portion to the stirred
solution. The reaction mixture was quenched by dilution with
hexane and was filtered through a small plug of silica. The
mixture was then concentrated to dryness on silica and
separated by flash chromatography. The product was identified
by NMR spectroscopy. Cis/ trans ratios were calculated from
1H NMR spectroscopy, and enantiomeric excesses (if ap-
plicable) were determined using (+)-Eu(hfc)3 as the chiral shift
reagent or by chiral HPLC on a Chiracel OD column, as
appropriate.
P r ep a r a tion of Azir id in es. N-(p-Tolylsu lfon yl)-2,3-(6-
a cetyl-2,2-d im eth ylch r om a n -3,4-yl)a zir id in e (4). White
solid. Yield: 79%. Mp (DSC): 160 °C. 1H NMR (CDCl3; δ): 1.25
(s, 3H), 1.31 (s, 3H), 2.43 (s, 3H), 2.54 (s, 3H), 3.37 (d, 1H, J )
7.3 Hz), 3.97 (d, 1H, J ) 7.3 Hz), 6.81 (d, 1H, J ) 8.4 Hz),
7.33 (d, 2H, J ) 8.1 Hz), 7.85-7.82 (m, 3H), 7.93 (s, 1H). 13C
NMR (CDCl3; δ): 19.6, 21.7, 23.8, 24.3, 37.6, 47.6, 70.9, 115.9,
116.3, 126.0, 127.7, 128.0, 128.8, 128.9, 132.7, 142.9, 154.8,
194.2. IR (Nujol; ν, cm-1): 1669, 1456, 1377, 1365, 1328, 1267,
1160, 904, 868, 830, 825. MS (EI+; m/z): 372 (M + H+), 216
(M+ - SO2C6H4CH3), 91 (C6H4CH3+), 43 (CH3CO+). High-
resolution MS (EI+): calcd mass (M+ - Ts) 216.102 454, found
216.101 996. Anal. Calcd for C20H21NO4S: C, 64.67; H, 5.70;
N, 3.77. Found: C, 64.48; H, 5.74; N, 3.61. Chiral HPLC
analysis (hexane/IPA (90:10), 1 mL/min): tr ) 14.9 min (major),
18.8 min (minor).
1
921, 813, 677. H NMR (CDCl3; δ): 1.46 (s, 9H), 2.24 (s, 3H),
2.33 (s, 3H), 3.35 (d, 1H, J ) 4.0 Hz), 4.26 (d, 1H, J ) 4.0 Hz),
7.05 (m, 4H),7.19 (d, 2H, J ) 8.5 Hz), 7.71 (d, 2H, J ) 8.5 Hz).
13C NMR (CDCl3; δ): 21.6, 22.0, 28.2, 48.2, 48.7, 83.8, 127.6,
127.7, 129.6, 129.9, 130.3, 138.1, 139.1, 144.4, 165.1. MS (EI;
m/z): 387 (M+), 332 (M - C4H8). HRMS (CI; m/z): found
388.1589, calcd [M+ + 1] 388.1601. Anal. Calcd for C21H25
-
NO4S: C, 65.09; H, 6.50; N, 3.61. Found: C, 64.98; H, 6.29;
N, 3.53. Enantiomeric excess determined by 1H NMR using
30
Eu(hfc)3, ee ) 88%. [R]D ) -18.9° (c 0.84, CH2Cl2).
N-(p-Tolylsulfonyl)-2-carbo-tert-butoxy-3-(3-methylphen-
yl)a zir id in e (6d ). Yellow oil. Yield: 109 mg, 56%. IR (NaCl
plates; νma, cm-1):x 2979, 2926, 1736, 1638, 1337, 1163, 1087,
1
784, 690. H NMR (CDCl3; δ): 1.53 (s, 9H), 2.28 (s, 3H), 2.40
(s, 3H), 3.39 (d, 1H, J ) 4.0 Hz), 4.31 (d, 1H, J ) 4.0 Hz),
6.99-7.28 (m, 6 H), 7.78 (d, 2H, J ) 8.6 Hz). 13C NMR (CDCl3;
δ): 20.8, 20.6, 26.8, 46.6, 47.5, 82.4, 123.2, 126.4, 127.0, 127.4,
128.5, 128.5, 132.0, 136.5, 137.3, 143.0, 163.6. MS (CI; m/z):
405 (M + NH4+), 388 (M + H+). HRMS (CI; m/z): found
388.1581, calcd [M+ + 1] 388.1601. Anal. Calcd for C21H25
-
NO4S: C, 65.09; H, 6.50; N, 3.61. Found: C, 65.11; H, 6.59;
N, 3.67. Enantiomeric excess determined by 1H NMR using
28
Eu(hfc)3, ee ) 96%. [R]D ) -19.9° (c 0.24, CH2Cl2).
N-(p-Tolylsu lfon yl)-2-car bo-ter t-bu toxy-3-(4-flu or oph en -
yl)a zir id in e (6e). Yellow oil. Yield: 130.0 mg, 66%. IR (NaCl
plates; νmax, cm-1): 2982, 1732, 1608, 1515, 1336, 1222, 1088,
1
925, 814, 738, 690. H NMR (CDCl3; δ): 1.52 (s, 9H), 2.40 (s,
3H), 3.39 (d, 1H, J ) 4.0 Hz), 4.33 (d, 1H, J ) 4.0 Hz), 6.97
(dd, 2H, J ) 8.5 Hz), 7.17-7.28 (m, 4H), 7.77 (d, 2H, J ) 8.5
Hz). 13C NMR (CDCl3; δ): 20.6, 26.8, 45.9, 47.5, 82.6, 114.4,
114.7, 126.5, 128.0, 128.1, 128.5, 136.4, 143.2, 163.5. MS (CI;
m/z): 409 (MNH4+), 392 (MH+). HRMS (CI; m/z): found
392.1337, calcd [M+ + 1] 392.1351. Anal. Calcd for C20H22
-
FNO4S: C, 61.36; H, 5.66; N, 3.58. Found: C, 61.48; H, 5.29;
N, 3.73. Enantiomeric excess determined by HPLC: Chiralcel
OD, 15 cm × 0.46 cm i.d.; mobile phase hexane/propan-2-ol
(95:5), tr(major) ) 15.56 min, tr(minor) ) 19.37 min, ee ) 98%,
30
[R]D ) -42.2° (c 0.25, CH2Cl2).
N-(p-Tolylsu lfon yl)-2-car bo-ter t-bu toxy-3-(4-ch lor oph en -
yl)a zir id in e (6f). Yellow oil. Yield: 174 mg, 89%. IR (NaCl
plates; νmax, cm-1): 2980, 2931, 1738, 1598, 1495, 1339, 1164,
1090, 915, 814, 731, 686. 1H NMR (CDCl3; δ): 1.52 (s, 9H),
2.40 (s, 3H), 3.36 (d, 1H, J ) 4.0 Hz), 4.32 (d, 1H, J ) 4.0 Hz),
7.15 (dd, 2H, J ) 8.9 Hz),7.17-7.28 (m, 4H), 7.77 (d, 2H, J )

3458 J . Org. Chem., Vol. 67, No. 10, 2002
Gillespie et al.
8.6 Hz). 13C NMR (CDCl3; δ): 20.6, 26.8, 45.8, 47.7, 82.7, 126.3,
127.6, 128.6, 130. 7, 133.7, 136.4, 143.3, 163.3 (CdO). MS (CI;
m/z): 428 (MNH4+), 408 (MH+). HRMS (CI; m/z): found
plates; νmax, cm-1): 2964, 2929, 1730, 1597, 1532, 1439, 1260,
1162, 108, 803, 737, 679. 1H NMR (CDCl3; δ): 1.54 (s, 9H),
2.41 (s, 3H), 3.41 (d, 1H, J ) 4.0 Hz), 4.42 (d, 1H, J ) 4.0 Hz),
7.30 (d, 2H, J ) 8.2 Hz), 7.45 (dd, 1H, J ) 7.6 Hz), 7.58 (dd,
1H, J ) 7.6 Hz), 7.78 (d, 2H, J ) 8.2 Hz), 8.03 (s, 1H), 8.13
(dd, 1H, J ) 8.3 Hz). 13C NMR (CDCl3; δ): 20.6, 26.8, 44.8,
47.9, 83.1, 121.1, 121.6, 122.7, 122.8, 126.4, 126.6, 128.7, 132.4,
132.6, 134.6, 143.7, 163.4. MS (CI; m/z): 405 (M + NH4+), 388
(M + H+). HRMS (CI; m/z): found 419.1282, calcd [M+ + 1]
419.2288. Anal. Calcd for C20H22N2O6S: C, 57.40; H, 5.30; N,
6.69. Found: C, 57.51; H, 5.29; N, 6.54. Enantiomeric excess
determined by HPLC: Chiralcel OD, 15 cm × 0.46 cm i.d.;
mobile phase hexane/propan-2-ol (98:2), tr(major) 35.65 min,
tr(minor) ) 39.79 min, ee ) 96%. [R]D30 ) -53.8° (c 0.41, CH2-
Cl2).
408.1041, calcd [M+ + 1] 408.1055. Anal. Calcd for C20H22
-
ClNO4S: C, 58.89; H, 5.44; N, 3.43. Found: C, 58.99; H, 5.35;
N, 3.37. Enantiomeric excess determined by HPLC: Chiralcel
OD 15 cm × 0.46 cm i.d; mobile phase hexane/propan-2-ol (95:
26
5), tr(major) 8.99 min, tr(minor) ) 10.43 min, ee ) 93%. [R]D
) -41.0° (c 0.66, CH2Cl2).
N-(p-Tolylsu lfon yl)-2-car bo-ter t-bu toxy-3-(4-br om oph en -
yl)a zir id in e (6g). Colorless oil. Yield: 130 mg, 57%. IR (NaCl
plates; νmax, cm-1): 2980, 2928, 1738, 1597, 1338, 1163, 1087,
1
916, 812. H NMR (CDCl3; δ): 1.53 (s, 9H), 2.41 (s, 3H, 3.36
(d, 1H, J ) 4.0 Hz), 4.31 (d, 1H, J ) 4.0 Hz), 7.09 (dd, 2H, J
) 6.7, 1.8 Hz), 7.27 (d, 2H, J ) 8.2 Hz), 7.41 (dd, 2H, J ) 6.7,
1.8 Hz), 7.79 (d, 2H, J ) 8.2 Hz). 13C NMR (CDCl3; δ): 22.0,
28.2, 47.2, 49.1, 84.1, 123.3, 127.7, 129.2, 130.0, 132.1, 132.6,
137.8, 144.7, 164.7. MS (CI; m/z): 452/454 (MH+). HRMS (CI;
m/z): found 452.0525, calcd [M+ + 1] 452.0550. Anal. Calcd
for C20H22BrNO4S: C, 53.10; H, 4.90; N, 3.10. Found: C, 53.16;
H, 4.79; N, 3.21. Enantiomeric excess determined by HPLC:
Chiralcel OD 15 cm × 0.46 cm i.d.; mobile phase hexane/
propan-2-ol (95:5), tr(major) 15.89 min, tr(minor) ) 23.21 min,
Ack n ow led gm en t. We thank EPSRC for a postdoc-
toral fellowhip to K.M.G.
Su p p or tin g In for m a tion Ava ila ble: Details of synthesis
of 2,2′-diamino-6,6′-dimethylbiphenyl, copper sources, and
butyl cinnamates 5a -j and crystallographic data for (R)-(-)-
L1, four copper complexes, and 6a . This material is available
free of charge via the Internet at http://pubs.acs.org.
30
ee ) 98%. [R]D ) -41.8° (c 0.33, CH2Cl2).
N-(p-Tolylsu lfon yl)-2-car bo-ter t-bu toxy-3-(3-n itr oph en -
yl)a zir id in e (6h ). Yellow oil. Yield: 109 mg, 56%. IR (NaCl
J O025515I