Regio- and stereoselectivity of diethylaluminum azide opening of trisubstituted epoxides and conversion of the 3° azidohydrin adducts to isoprenoid aziridines

Article abstract of DOI:10.1021/jo026506c

The regioisomer ratios (3°,2°/2°,3°), and in some cases the stereochemistry, of vicinal azidohydrins formed in reactions of 11 trisubstituted terpene epoxides with Et₂AlN₃ in toluene are reported. The more highly substituted azide usually predominated (3°,2°/2°,3° ratios ≥ 40:1 to 2.5-1) in accord with a Markovnikov orientation and an S_N1-like transition state. Reversed regioisomer ratios were observed with 6,7-epoxygeranyl acetate (1:2.5) and cis-1,2-epoxylimonene (1:3.3 to 1:10). The tertiary azido diols from 2,3-epoxygeraniol, 2,3-epoxyfarnesol, and 2,3-epoxynerol were formed as single isomers with inversion of configuration at C3 (≥ 35-40:1 for the C₁₀ azido diols). The regioselectivity was affected by the presence and proximity of oxy functional groups on the epoxide substrate (OH, OAc, and OSi-tBuMe₂), the equivalents of Et₂AlN₃, and additives (EtOAc or EtOH). The results and trends are rationalized by consideration of the structural and stereoelectronic characteristics of proposed diethylaluminum epoxonium ion intermediates and transition states, together with the nucleophilicity of the azide donor. Six of the 3°,2° azidohydrins were converted to the corresponding aziridines by primary-selective silylations of four azido diols, mesylations, and reductive cyclizations with LiAlH₄.

Full text of DOI:10.1021/jo026506c

Regio- a n d Ster eoselectivity of Dieth yla lu m in u m Azid e Op en in g of
Tr isu bstitu ted Ep oxid es a n d Con ver sion of th e 3° Azid oh yd r in
Ad d u cts to Isop r en oid Azir id in es
Chad E. Davis, J essica L. Bailey, J onathan W. Lockner, and Robert M. Coates*
Department of Chemistry, University of Illinois, 600 South Mathews Avenue, Urbana, Illinois 61801
coates@scs.uiuc.edu
Received October 1, 2002
The regioisomer ratios (3°,2°/2°,3°), and in some cases the stereochemistry, of vicinal azidohydrins
formed in reactions of 11 trisubstituted terpene epoxides with Et2AlN3 in toluene are reported.
The more highly substituted azide usually predominated (3°,2°/2°,3° ratios g 40:1 to 2.5-1) in
accord with a Markovnikov orientation and an SN1-like transition state. Reversed regioisomer ratios
were observed with 6,7-epoxygeranyl acetate (1:2.5) and cis-1,2-epoxylimonene (1:3.3 to 1:10). The
tertiary azido diols from 2,3-epoxygeraniol, 2,3-epoxyfarnesol, and 2,3-epoxynerol were formed as
single isomers with inversion of configuration at C3 (g 35-40:1 for the C10 azido diols). The
regioselectivity was affected by the presence and proximity of oxy functional groups on the epoxide
substrate (OH, OAc, and OSi-tBuMe2), the equivalents of Et2AlN3, and additives (EtOAc or EtOH).
The results and trends are rationalized by consideration of the structural and stereoelectronic
characteristics of proposed diethylaluminum epoxonium ion intermediates and transition states,
together with the nucleophilicity of the azide donor. Six of the 3°,2° azidohydrins were converted
to the corresponding aziridines by primary-selective silylations of four azido diols, mesylations,
and reductive cyclizations with LiAlH4.
In tr od u ction
Aziridine analogues (eq 1) of presqualene pyrophos-
phate1 and many polyene epoxides2,3 in the protonated
aziridinium forms are potential transition state inhibitors
of enzymes associated with terpene and sterol biosyn-
thesis.4,5 However, in most cases the aziridines were
prepared and assayed as racemic or diasteromeric mix-
tures. To understand and to characterize better the
3-dimensional interactions of these inhibitors with the
respective enzyme active sites, we became interested in
developing approaches for synthesis of these and related
terpenyl aziridines as single enantiomers.
Although considerable research on enantioselective
synthesis of aziridines is evident in the recent literature,6
most methods are limited to, or investigated for, access
to mono- and disubstituted types, and appear unsuitable
for unactivated trisubstituted aziridines. The isoprenoid
aziridine inhibitors have most commonly been obtained
from the corresponding olefins by electrophilic addition
of halo azide reagents followed by reductive cyclizations,7
or by epoxide formation, opening to the azidohydrin, and
reductive cyclization of the azido sulfonate.2,8 Since
epoxides of trisubstituted olefins can be secured readily
by asymmetric epoxidation,9 or by asymmetric dihy-
droxylation,10 and subsequent conversion to the epoxide,
(6) (a) Coull, W. M. Synthesis 2000, 1347. (b) Tanner, D. Angew.
Chem., Int. Ed. 1994, 33, 599. (c) Osborn, H. M. I.; Sweeney, J .
Tetrahedron: Asymmetry 1997, 8, 1693. (d) Atkinson, R. S. Tetrahedron
1999, 55, 1519. (e) Pearson, W. H.; Lian, B. W.; Bergmeier, S. C. In
Comprehensive Heterocyclic Chemistry II; Padwa, A., Ed.; Pergamon:
Oxford, UK, 1996; p 1. (f) Rai, K. M.; Hassner, A. In Comprehensive
Heterocyclic Chemistry II; Padwa, A., Ed.; Pergamon: Oxford, UK,
1996; p 61. (g) Davis, F. A.; Liu, H.; Zhou, P.; Fang, T.; Reddy, G. V.;
Zhang, Y. J . Org. Chem. 1999, 64, 7559. Sweeney, J . B. Chem. Soc.
Rev. 2002, 31, 247.
(7) (a) Avruch, L.; Oehlschlager, A. C. Synthesis 1973, 622. (b) Van
Ende, D.; Krief, A. Angew. Chem., Int. Ed. Engl. 1974, 13, 279. (c)
Parrish, E. J .; Nes, W. D. Synth. Commun. 1988, 18, 221. (d) Nes, W.
D.; Parrish, E. J . Lipids 1988, 23, 375.
(1) Koohang, A.; Coates, R. M.; Owen, D.; Poulter, C. D. J . Org.
Chem. 1999, 64, 6.
(2) Corey, E. J .; Ortiz de Montellano, P. R.; Lin, K.; Dean, P. D. G.
J . Am. Chem. Soc. 1967, 89, 2797.
(3) J annssen, G. G.; Nes, W. D. J . Biol. Chem. 1992, 267, 25856.
(4) For reviews on oxidosqualene cyclase inhibitors, see: (a) Cattel,
L.; Ceruti, M.; Viola, F.; Delprino, L.; Balliano, G.; Duriatti, A.; Bouvier-
Nave, P. Lipids 1986, 21, 31. (b) Abe, I.; Rohmer, M.; Prestwich, G. D
Chem. Rev. 1993, 93, 2189. (c) Abe, I.; Tomesch, J . C.; Wattanasin, S.;
Prestwich, G. D Nat. Prod. Rep. 1994, 11, 279.
(5) For other examples and references for isoprenoid aziridine
inhibitors, see: Koohang, A.; Stanchina, C. L.; Coates, R. M. Tetra-
hedron 1999, 55, 9669.
(8) Corey, E. J .; Riddiford, L. M.; Ajami, A. M.; Yamamoto, H.;
Anderson, J . E. J . Am. Chem. Soc. 1971, 93, 1815.
10.1021/jo026506c CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/23/2002
J . Org. Chem. 2003, 68, 75-82
75

Davis et al.
the latter approach appears quite attractive for this
purpose. However, the methods for converting a chiral
epoxide to the enantiomeric aziridine depend critically
upon a Markovnikov orientation in the epoxide opening
reaction with nitrogen nucleophiles such as azide to form
the 3°,2° azido alcohol (eq 2) owing to the difficulty of
activating the tertiary hydroxyl group in the regioiso-
meric 2°,3° azidohydrins. Unfortunately the low acidity
of HN3 together with the high nucleophilicity of azide
ion and covalent azide donors usually favor the anti-
Markovnikov adduct in reactions with trisubstituted
epoxides.11-13
selected 3°,2° azidohydrins to the corresponding trisub-
stituted aziridines.
Ep oxid es, Dieth yla lu m in u m Azid e Rea gen t, a n d
Hyd r oa zid a tion P r oced u r e
A series of twelve trisubstituted epoxides with different
electronic and structural environments was used in this
study. 2,3-Epoxides of (E,E)-farnesol, geraniol, and nerol
(1, 2, and 5) were chosen to establish the stereochemistry
of azide addition and to prepare the way for synthesis
of enantio pure aziridine inihibitors. Other epoxides
selected were six 6,7-epoxygeranyl and 6,7-epoxycitronel-
lyl derivatives (7-OH, 7-OAc, 7-OTBS, 9-OH, 9-OAc,
and 9-OTBS), trans- and cis-1,2-epoxylimonene isomers
(11 and 12), and 24,25-epoxy lanosterol tert-butyl-
dimethylsilyl ether (15-OTBS). The 2,3-epoxides were
prepared by hydroxyl-directed epoxidations of (E,E)-
farnesol, geraniol, and nerol (VO(acac)2, t-BuOOH).18
Most of the others were obtained by selective epoxida-
tions of the precursor terpenols with m-ClC6H4CO3H
followed by acetylation or silylation. Terminal epoxida-
tion of geranyl acetate afforded 7-OAc, saponification of
which provided 7-OH. A trans-cis mixture of limonene-
1,2-epoxides is commercially available. The epoxides
derived from citronellol and lanosterol were 1:1 mixtures
of diastereomers.
Solutions of diethylaluminum azide in toluene were
generated before use by reaction of sodium azide with
commercial diethylaluminum chloride in toluene (room
temperature, 6 h) according to a literature procedure.15
Azide addition reactions were carried out by adding
solutions of the epoxide in toluene to usually 2 molar
equiv of Et2AlN3 in toluene at -78 °C, 0 °C, or room
temperature for the times shown in Table 1. In some
cases 1 equiv of ethyl acetate or ethanol was added to
evaluate the effect of a polar coordinating carbonyl group
or a proton donor. Reactions were stopped by dilution
with ethyl acetate. Hydrolysis of the aluminoxide inter-
mediates and isolation of the crude azidohydrin products
were accomplished by adding solid NaF and water, and
filtering through anhydrous Na2SO4.15 Reaction progress
was monitored by the disappearance of the epoxide
starting material by TLC analysis of aliquots.
The use of oxyphilic organometallic azide reagents
such as organoaluminum and organotitanium azides for
Markovnikov opening of epoxides affords an appealing
alternative. Thus the reaction of diethylaluminum azide
(generated in solution from Et3Al and HN3) with 1-meth-
ylcyclohexene epoxide resulted in exclusive formation of
the trans 3°,2° azidohydrin.14 Similar electrophilic open-
ing of 2,3-epoxy-3-methylbutenol with diethylaluminum
azide (generated from Et2AlCl and NaN3) afforded a
mixture of the Markovnikov adducts as azido- and
chorohydrins.15 Exclusive Markovnikov orienation was
also observed in the reaction of 2,3-epoxygeraniol with
(iPrO)2Ti(N3)2,16,17 although a substantial amount of the
methylene diol arising from competing elimination was
also formed.16 In this paper we report the results of
an investigation on the regio- and stereoselectivity of
azide openings of representative trisubstituted isoprenoid
epoxides with diethylaluminum azide and conversion of
(9) (a) Sharpless, K. B.; Gao, Y.; Hanson, R. M.; Klunder, J . M.; Ko,
S. Y.; Masamune, H. J . Am. Chem. Soc. 1987, 109, 5765. (b) Katsuki,
T.; Martin, V. S. Org. React. 1996, 48, 1.
(10) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem.
Rev. 1994, 94, 248.
(11) Anderson, R. J .; Henrick, C. A.; Siddall, J . B. J . Org. Chem.
1972, 37, 1266.
(12) For reviews see: (a) Scriven, E. F. V.; Turnbull, K. Chem. Rev.
1988, 88, 297. (b) The Chemistry of the Azide Group; Patai, S., Ed.;
Wiley: New York, 1971.
(13) For recent articles and leading references, see: (a) aq NaN3
and polymeric phase transfer catalyst: Tamami, B.; Mahdavi, H.
Tetrahedron Lett. 2001, 42, 8721. (b) NaN3 and metal salt catalysts:
Chini, M.; Crotti, P.; Macchia, F. Tetrahedron Lett. 1990, 31, 5641. (c)
Me3SiN3 and Yb(OiPr)3 catalyst: Meguro, M.; Asao, N.; Yamamoto,
Y. J . Chem. Soc., Chem. Commun. 1995, 1021. (d) aq NaN3 and
catalysts: Fringuelli, F.; Piermatti, O.; Pizzo, F.; Vaccaro, L. J . Org.
Chem. 1999, 64, 6094. (e) Me3SiN3 and Al(OiPr)3 or Ti(OiPr)4 cata-
lyst: Sutowardoyo, K. I.; Emziane, M.; Lhoste, P.; Sinou, D. Tetra-
hedron 1991, 47, 1435. (f) NaN3/PhB(OH)2 with 2,3-epoxy alcohols:
Hayakawa, H.; Okada, N.; Miyazawa, M.; Miyashita, M. Tetrahedron
Lett. 1999, 40, 4589. (g) Chiral Cr(III) complexes and meso epoxides:
Schaus, S. E.; Larrow, J . F.; J acobsen, E. N. J . Org. Chem. 1997, 62,
4197. (h) N3- or Me3SiN3 and transition metal catalysts: Crotti, P.; Di
Bussolo, V.; Favero, L.; Macchia, L. E.; Pineschi, M. Tetrahedron Lett.
1996, 37, 1675. (i) NaN3/NH4Cl and 2,3-epoxyalcohols: Behrens, C.
H.; Sharpless, K. B. J . Org. Chem. 1985, 50, 5696.
Ra tios
a n d
Id en tifica tion
of
Azid oh yd r in
Ad d u cts
The principal results of reactions of Et2AlN3 with the
isoprenoid epoxide substrates are summarized in Table
1 and eqs 3-8. The tertiary azido diols 3, 4, and 6 from
2,3-epoxides of farnesol, geraniol, and nerol (1, 2, and 5)
were formed as single isomers and obtained in 66, 54,
and 63% yields, respectively, after purification by flash
chromatography on silica gel (eqs 3 and 4). Integration
(C3-CH3 signals) of the proton NMR spectra of the crude
products from 2 and 5 spiked with small amounts of 6
and 4 demonstrated that azide capture at C3 was
essentially stereospecific (g97%), i.e. isomer ratios g 35-
40:1. Furthermore no other isomers such as the regio-
(14) Mereyala, H. B.; Frei, B. Helv. Chim. Acta 1986, 69, 415.
(15) Benedetti, F.; Berti, F.; Norbedo, S. Tetrahedron Lett. 1998, 39,
7971.
(16) Caron, M.; Carlier, P. R.; Sharpless, K. B. J . Org. Chem. 1988,
53, 5185.
(17) (a)Aguilar, N.; Moyana, A.; Perica´s, M. A.; Riera, A. J . Org.
Chem. 1998, 63, 3560. (b) Reddy, K. S.; Sola´, L.; Moyano, A.; Perica´s,
M. A.; Riera, A. J . Org. Chem. 1999, 64, 3969.
(18) (a) Sharpless, K. B.; Michaelson, R. C. J . Am. Chem. Soc. 1973,
95, 6136. (b) Sharpless, K. B.; Verhoeven, T. R. Aldrichim. Acta 1979,
12, 63. (c) Tanaka, S.; Yamamoto, H.; Nozaki, H.; Sharpless, K. B.;
Michaelson, R. C.; Cutting, J . D. J . Am. Chem. Soc. 1974, 96, 5254.
76 J . Org. Chem., Vol. 68, No. 1, 2003

Formation of Vicinal Azidohydrins
TABLE 1. Regioisom er Ra tios a n d Yield s of Azid oh yd r in s fr om Rea ction of Et 2AlN3 w ith Ep oxid es (eqs 3-8)
epoxide
name
conditionsa
azidohydrin product(s)
d
entry
1
no.
variationb temp (°C) time (h) conversion (%)c
no.
3°N3/2°N3
yield (%)e
2,3-epoxy-farnesol
1
-78
rt
1
16
3
66
2
3
4
5
2,3-epoxy-geraniol
2,3-epoxy-nerol
6,7-epoxy-geraniol
6,7-epoxy-geranyl
acetate
2
5
4
6
g40:1
g35:1
13:1
54
63
69
81
7-OH
7-OAc
0
-78
0.50
22
8a ,b-OH
8a ,b-OAc
94
70
1:2.5
6
7
6,7-epoxy-citronellolf
6,7-epoxy-citronellyl
acetatef
9-OH
9-OAc
0
-78
0.50
22
10,a b-OH
10a ,b-OAc
11:1
3:1
72
68
8
9
10
0
0
-78
0.50
0.50
22
4.5:1
5.9:1
5.8:1
89
78
78
5 equivg
EtOAch
6,7-epoxy-citronellyl
silyl etherf
9-OTBS
77
75
10a ,b-OTBS
11
12
13
0
0
0
0.50
0.50
0.17
5:1
2.5:1
5:1
68
75
89
83
73
14
69
52
83
75
51
31
1,2-epoxy limonenesi
11 (trans)
12 (cis)
13
14a ,b
1:5
14
15
16
17
-78
0
0.17
0.25
0.25
6
96
62
>10:1
1:6
EtOHj
g5:1
1:3.3
g10:1
1:10
g6:1
1:4
EtOAch
1 equivk
EtOHl
0
rt
18
19
-78
0.50
1
0
24,25-epoxy lanosterol 15-OTBS
0
16a ,b-OTBS
3:1
55
silyl etherf
a
b
2 equiv of Et2AlN3 in toluene unless noted otherwise. rt ) room temperature. Equivalents of Et2AlN3 or ethyl acetate added or
ethanol added. c Conversion based on the amount of recovered epoxide estimated by 1H NMR analysis and/or isolated yield. Conversion
was ∼100% unless specified otherwise. Isomer ratio was determined by 1H NMR analysis of crude product. The azidohydrins obtained
d
from the citronellyl derivatives were ∼1:1 mixtures except 10b-OAc was a ∼2:1 mixture of diastereomers. e Yields are based on the
amount of unrecovered epoxides in incomplete reactions. f An approximate 1:1 mixture of diastereomers. 5 equiv of Et2AlN3. 1 equiv
g
h
i
j
k
l
of ethyl acetate added. A 1:1.3 mixture of trans and cis epoxides. 1 equiv of ethanol added. 1 equiv of Et2AlN3. 2 equiv of ethanol
added.
and secondary hydroxyl structures shown were confirmed
by conversion to the corresponding silyl azido mesylates
(t-BuMe2SiCl; MsCl) and reductive cyclizations (LiAlH4)
to the known 2,3-aziridino alcohols described in more
detail below (eqs 9 and 10). The C2 CHOMs protons (δH
4.56-4.57) of the silyl azido mesylates intermediates are
shifted downfield considerably by the methanesulfonyl
substituent, thus verifying that the hydroxyl group had
been secondary. Furthermore, it seems unlikely that the
tertiary mesylates derived from the 2°,3° azido diols
would have been stable on silica gel during chromato-
graphic purifications.
isomeric 2° azidohydrins could be detected although small
amounts (5-6%) of the known 3-methylene-1,2-diol elimi-
nation byproduct16,19 were isolated from reactions of the
2,3-epoxy monoterpene alcohols. Preliminary attempts to
effect azide addition to 2,3-epoxyfarnesyl benzyl and tert-
butyldimethylsilyl ethers with NaN3,/NH4Cl,13i Et3Al/
The reactions of the 6,7-epoxides of geraniol and
citronellol and their OH derivatives with Et2AlN3 af-
forded mixtures of tertiary and secondary azidohydrins
with ratios varying from 13:1 to 1:2.5 (eqs 5 and eq 6,
1
and Table 1). Product ratios were estimated by H NMR
16
HN3,14 Et2AlN3,15 and Ti(OiPr)2(N3)2 were largely un-
successful or proceeded slowly in low yields.
The racemic azidohydrin products were characterized
by appropriate 1H NMR, 13C NMR, IR (NdNdN and
O-H stretch), and mass spectral data. However the
distinction of the two isomeric possibilities was somewhat
subtle owing to the similarity of the chemical shifts for
the CHOH and CHN3 protons and the complex ABX spin
systems for the CHXCH2OH group. The tertiary azide
analysis of the product mixtures prior to purification.
High ratios favoring tertiary azido diols 8a -OH (13:1) and
10a -OH (11:1) were obtained from the epoxy alcohols
whereas the acetates and silyl ethers gave considerably
(19) Morgans, D. J .; Sharpless, K. B.; Traynor, S. G. J . Am. Chem.
Soc. 1981, 103, 462.
J . Org. Chem, Vol. 68, No. 1, 2003 77

Davis et al.
lower selectivities. Remarkably the product distribution
for epoxy acetate 7-OAc favored the opposite isomers
(8a -OAc/8b-OAc 1:2.5) and the ratio for epoxy acetate
9-OAc was markedly decreased (10a -OAc/10b-OAc 3:1)
under comparable conditions (2 equiv of Et2AlN3, -78 °C).
The selectivity in the latter case increased to 5.9:1 when
5 equiv of Et2AlN3 at 0 °C was used instead of 2. All of
the citronellol azidohydrins were 1:1mixtures of distere-
omers with the exception of the secondary azidoacetate
(10b-OAc) that was 2:1.
oxiranes. Since each isomer gives a discrete set of readily
identifiable azidohydrin products that could be analyzed
by H NMR analysis of the mixture, there seemed to be
1
no compelling reason to study the pure isomers individu-
ally. As expected the reactions of the trans epoxide (11)
afforded tertiary trans diaxial azidohydrin 13 as the only
identified product with selectivities g 5-10:1. In contrast
the cis isomer 12 gave mixtures in which the secondary
diaxial azidohydrin 14b was the major product (14a /14b
ratios 1:3.3-10). The higher reactivity of the trans
epoxide was evident from the reaction at -78 °C, which
resulted in 96% conversion of 11 to 13 (g 10:1 selectivity)
as opposed to 62% conversion of 12.
The product ratio from cis epoxide 12 proved to
be dependent on the additives and the equivalents of
Et2AlN3 reagent used. Lower anti-Markovnikov selectiv-
ity was observed in the presence of 1 equiv of EtOH
(1:3.3) and when only 1 equiv of the aluminum reagent
was employed. However, the anti-Markovnikov selectivity
increased to 1:10 in the presence of 1 equiv of EtOAc.
Although the EtOH additive appears to enhance the
Markovnikov selectivity somewhat (1:3.3), the presence
of the proton donor clearly depresses the rate of opening
of the trans epoxide at -78 °C.
Pure samples of the azidohydrin products 13, 14a , and
14b were isolated by flash chromatography. The stereo-
chemistry of the axial 2-OH and 2-N3 substituents in the
major azidohydrins 13 and 14b was evident from the lack
of vicinal 1H NMR couplings (13: δH 3.64, s; 14b: δH 3.52,
s), and the equatorial position of secondary OH in 14a
could be assigned in the same manner (δH 3.59, dd, J )
11.8, 4.5 Hz). The configurations of the tertiary azide
substituents in 13 and 14a were assumed to be trans to
the adjacent secondary OH groups as a consequence of
the expectation of trans openings. However, small amounts
(e 5-10%) of other minor unidentified products isolated
might have contained an azide functionality. The struc-
ture and trans stereochemistry of 13 were verified by
conversion to the known cis-1,2-aziridino limonene
by reductive cyclization of the corresponding mesylate
(eq 12).20
Although similar 3°/2° azide adducts (∼6.5:1) seemed
to be formed in reactions of 7-OTBS, full characterization
of the products was not carried out in this case owing to
the multicomponent product mixtures formed, extensive
purification needed, and low recoveries obtained. How-
ever, the corresponding reactions of 9-OTBS gave sat-
isfactory yields and consistent ratios (-78 °C, 5.8:1, 78%;
0 °C, 5:1, 68%). The reactions of both 7-OAc and 9-OTBS
were incomplete at -78 °C after 22 h (∼70-77% conver-
sions). The presence of 1 equiv of ethyl acetate in the
latter case resulted in a much slower rate (0 °C, 0.5 h,
75% conversion) and a lower regioisomer ratio (2.5:1).
The isomers were readily separated by flash chroma-
tography owing to the increased polarity of the 2°
azidohydrins (tertiary OH). The rather small chemical
1
shift differences between the C6 protons in the H NMR
spectra of less polar isomers (CHOH δH 3.30) compared
to those of the more polar isomers (CHN3 δH 3.11) were
insufficient for conclusive assignments. However, as in
the case of the 2,3 azidohydrins above, the structures
were confirmed by the downfield shifts for the CHOMs
protons in the derived silyl azido mesylate (eq 11). The
azidohydrin acetates 8b-OAc and 10b-OAc were cor-
related with the corresponding diols 8b-OH and 10b-OH
by methanolysis of the esters, and tertiary azido diol
10a -OH was interrelated with 10a -OTBS by selective
silylation of the primary hydroxyl group.
Diethylaluminum azide-mediated opening of the 1:1
diastereomeric mixture of 24,25-epoxylanosterol tert-
butyldimethylsilyl ethers 15-OTBS by the usual proce-
dure gave rise to a 3:1 mixture of tertiary and secondary
azidohydrins 16a -OTBS and 16b-OTBS (eq 8) both of
The reactions of Et2AlN3 with a 1:1.3 mixture of
trans- and cis-limonene-1,2-epoxides (eq 7 and Table 1)
which were ∼1:1 mixtures of C-24 diastereomers, which
could be separated by chromatography. The less polar
isomer was determined to be the 3° isomer 16a -OTBS
1
based on the H NMR shift upon conversion to the azido
mesylate. Reactions of lanosterol 24,25-epoxide directly
with Et2AlN3 afforded the azidohydrin adducts in much
lower yield (24 and 32%) in two preliminary trials,
were investigated to evaluate the stereoelectronic effects
in addition to these conformationally fixed trisubstituted
(20) Ferrero, L.; Geribaldi, S.; Rouillard, M.; Azzaro, M. Can. J .
Chem. 1975, 53, 3227.
78 J . Org. Chem., Vol. 68, No. 1, 2003

Formation of Vicinal Azidohydrins
possibly owing to side products arising from hydrazoic
acid generated by OH-Et2Al exchange (see eq 14 below).
the silyl derivative of 24,25-aziridino lanosterol (75%),
respectively.
Con ver sion of Azid oh yd r in s to Azir id in es
Five of the monoterpene tertiary azidohydrins (eqs
9-12) and the diastereomeric lanosteryl azidohydrin
mixture were converted to the corresponding aziridines.
Azido diols 3, 4, 6, and 8a -OH were selectively silylated
at the primary hydroxyl groups (t-BuMe2SiCl, imidazole,
DMAP, DMF, 0 °C).21 The 1,2-diols afforded mixtures of
primary (62-65%) and secondary isomers (14%) that
were separated by flash chromatography. In contrast
silylation of 1,6-diol 8a -OH proceeded cleanly to the
primary ether 8a -OTBS (91%). The azidohydrin silyl
ethers including 16a -OTBS and limonene-derived
azidohydrin 13 were converted to the corresponding
azido mesylates with methanesulfonyl chloride (Et3N,
CH2Cl2, 0 °C).22 The large downfield chemical shifts
(∆δH(CHfCHOMs) ) 0.83-1.17) for the carbinyl protons
(CHOMs)23 confirmed the tertiary-secondary structures
(17, 18, 21, 23, 25, and the diastereomeric mesylates of
16a -OTBS).
Discu ssion
The reactions of Et2AlN3 with the trisubstituted ep-
oxides afforded the tertiary azido alcohols as the major
product in most cases, consistent with an SN1-like mech-
anism.14,15 However, the isomer ratios varied considerably
with the character of the substituents on the epoxide,
functionality present, molar ratio, and absence or pres-
ence of a proton donor or ester additive.
The 2,3-epoxy isoprenoid alcohols underwent exclusive
Markovnikov additions to form the tertiary azido diols
3, 4, and 6 with complete inversion of configuration at
C3 (eq 13). The high regioselectivity is attributed to the
proximal hydroxyl group that either coordinates to the
Et2AlN3 reagent or possibly exchanges proton for alumi-
num to form an aluminoxide intermediate and HN3 (eq
14). Nucleophilic capture at C3 would be favored by
displacement of an exocyclic leaving group (path a), as
opposed to the endocyclic alternative (path b). Bias
toward an SN1-like mechanism with increased positive
charge at carbon may also be caused by attack of a
neutral azide nucleophile (HN3 and/or excess Et2AlN3)
upon the diethylaluminum epoxonium intermediate.
Reductive cyclizations of the azido mesylates 17, 18,
and 21 proceeded with cleavage of the proximal silyl
ethers to provide the known 2,3-aziridino farnesol, ge-
raniol, and nerol (19, 20, and 22)5 in moderate yields (43,
57, and 45% yields). The low yields are attributed to
inefficient recovery of the potentially chelating azirdino
alcohols from the aluminum salts formed in the hydroly-
sis step following the relatively small-scale LiAlH4 reduc-
tions.24 Reaction of azido mesylate 23 with LiAlH4 gave
aziridino ether 24-OTBS (79%), which was desilated
(Bu4F, THF, 0 °C) to the known 6,7-aziridino geraniol
(82%).5,7b,c Similar reductive cyclizations of limonene-
derived azido mesylate 25 and the mesylate of 16a -OTBS
gave the known cis-1,2-aziridino limonene (50%)20 and
The significant effect of oxygen functionality present
in the epoxide substrate on the regioselectivity is dra-
matically illustrated by comparison of the outcome of
reactions with 6,7-epoxygeraniol (13:1 3° N3/2° N3 ratio)
and the corresponding acetate (1:2.5 3° N3/2° N3 ratio).
The tendency toward Markovnikov orientation with the
epoxy alcohols may be a consequence of OH-AlEt2
exchange (eq 14), proton catalysis, and a neutral, rela-
tively weakly polarized HN3 nucleophile. However, co-
ordination of aluminum to the carbonyl group of the
allylic acetate would presumably promote ionization (eq
15). The increased concentration of the highly nucleo-
philic azide anion would favor an SN2-like transition state
(21) (a) Chaudhary, S. K.; Hernandez, O. Tetrahedron Lett. 1979,
20, 99. (b) Fujii, N.; Nakai, K.; Habashita, Y.; Hotta, Y.; Tamamura,
H.; Otaka, A.; Ibuka, T. Chem. Pharm. Bull. 1994, 42 (11), 2241.
(22) Crossland, R. K.; Servis, K. L. J . Org. Chem. 1970, 35, 3195.
(23) Pretsch, E.; Clerc, T.; Seibel, J .; Simon, W. Tables of Spectral
Data for Structure Determination of Organic Compounds; Springer-
Verlag: Berlin, Germany, 1989; p H145.
(24) Much higher yields (70 and 83%) were obtained in reductions
of the enantiopure azido mesylates 18 conducted on gram scale (2.95
and 2.45 g): Bailey, J . L. MS Thesis, University of Illinois at Urbana-
Champaign, 2000.
J . Org. Chem, Vol. 68, No. 1, 2003 79

Davis et al.
leading to the secondary azidohydrin. This inference is
supported by the effect of the ethyl acetate additive,
which resulted in decreased 3°/2° ratios (entries 12 and
16).
OAc and 8b-OAc. The relative rate of π-cyclization must
be somewhat depressed by the C1-OH (or C1-OAlEt2)
and C1-OAc substituents, which accounts for the higher
yields of azide adducts in these cases.
The regio-reversal between 6,7-epoxygeranyl acetate
(1:2.5 3° N3/2° N3 ratio) and its 2,3-dihydro derivative
6,7-epoxycitronellyl acetate (3:1 3°/2° ratio) seems un-
likely to be caused by a long-range electronic effect. The
nonpolar medium and the tendency for the oxyphilic
aluminum to become tetrahedral suggest the possibility
of intramolecular coordination (eq 16). It seems plausible
The formation of a single trans azidohydrin from trans-
1,2-epoxy limonene 11 is attributable to the reinforcing
stereoelectronic and electronic effects (eq 18). Thus a
Markovnikov orientation is favored by trans diaxial
opening of the trans epoxonium ion through a chairlike
transition state that gives rise to tertiary azido alcohol
13. However, in the cis isomer the stereoelectronic and
electronic effects are opposed to each other, and the major
product was invariably the secondary azidohydrin 14b
arising from the opposite diaxial addition (eq 19). The
minor tertiary azidohydrin 14a is presumably formed
through a strained twist boat intermediate that relaxes
to the more stable trans diequatorial conformer.
that the different conformation(s) of the 11-membered
dioxyaluminate rings from the rigid, planar 2,3-E double
bond in the geranyl intermediate as opposed to the more
flexible 2,3-ethano linkage of the citronellyl case would
affect the regioselectivity of nucleophilic attack. Perhaps
the 12-membered ring would better accommodate the E
double bond, which would favor path a and formation of
the 2°,3° adduct. On the other hand, the more confor-
mationally mobile citronellyl moiety in the ring would
permit azide capture at C7 and lead to Markovnikov
orientation (path b). The deviation of the diastereomer
ratio of citronellyl azidohydrin acetates 10b (2:1 ratio by
NMR analysis) from the expected 1:1 proportion may be
explained by the influence of the C-3 configuration on
the conformational properties of the distereomeric ep-
oxonium-aluminate macrocyles.
Several attempts to isolate the azido hydrin adducts
from 6,7-epoxygeranyl tert-butyldimethylsilyl ether (7-
OTBS) were frustrated by low yields and the presence
of numerous side products. The bicyclic ether (27-OTBS,
eq 17) was obtained in pure form (9%) and the spectral
properties of the corresponding alcohol 27-OH obtained
by fluoride desilation correlate with literature data for
this known product of Lewis acid-induced cyclizations of
6,7-epoxygeranyl acetate.25 Evidently 2,3 double bond
participation leading to cyclization onto the epoxonium
ion (path b) competes with nucleophilic attack by an azide
donor (eq 17). Nevertheless, the ratio of tertiary and
In summary all but one of the trisubstituted epoxides
underwent regioselective addition of the elements of
HN3 in the presence of Et2AlN3 to form azidohydrins
bearing the tertiary azide as major products in accord
with a Markovnikov orientation. The conversion of
representative azidohydrins to the corresponding aziri-
dines indicates enantiomerically or diasteromerically
pure isoprenoid aziridines should be accessible from the
corresponding epoxides by this means for evaluation of
their inhibitory properties.
Exp er im en ta l Section
A representative complete procedure for the epoxide opening
reactions using Et2AlN3 in toluene is presented below in the
case of 2,3-epoxyfarnesol (1). Amounts, conditions, and pro-
cedural variations for the epoxide reactions of 2,3-epoxyge-
raniol (2) and 2,3-epoxynerol (3) are given in the following
abbreviated format: epoxide, reagents, temperature(s), time-
(s), crude product analysis. The procedures for synthesis of
aziridines 19, 20, and 22 from the corresponding azidohydrins
are detailed below. Other experimental procedures and data
are presented in the Supporting Information.
(()-(2R*,3R*,6E)-3-Azid o-3,7,11-tr im eth yl-6,10-d od eca -
d ien -1,2-d iol (3). The following procedure is based on that
reported by Benedetti.15 A suspension of NaN3 (1.23 g, 18.9
mmol) in toluene (14 mL) was stirred and cooled at 0 °C as a
solution of diethylaluminum chloride (9.44 mL, 17.1 mmol, 1.8
M in toluene) was added dropwise over 10 min via syringe.
After 6 h at room temperature, the resulting suspension was
secondary azidohydrins 8a -OTBS and 8b-OTBS ap-
peared to be similar to that of the related acetates 8a -
(25) Barrero, A. F.; Alvarez-Manzaneda, E. J .; Palomino, P. L.
Tetrahedron 1994, 50, 13239.
80 J . Org. Chem., Vol. 68, No. 1, 2003

Formation of Vicinal Azidohydrins
cooled to -78 °C under N2, and a solution of (()-2,3-epoxy-
farnesol (1, 2.03 g, 8.53 mmol) in toluene (5.5 mL) was added
dropwise over 10 min. The mixture was stirred for 1 h at -78
°C and for 16 h at room temperature, cooled to 0 °C, and
diluted with EtOAc (40 mL). NaF (16.2 g) and H2O (2.1 mL)
were added sequentially, and the heterogeneous mixture was
allowed to stir for 4 h at room temperature. Filtration through
a pad of anhydrous Na2SO4 (ca. 100 g) to remove salts and
water followed by solvent evaporation gave 1.83 g of a pale
yellow oil. Purification by flash column chromatography (50%
EtOAc/hexanes) afforded 1.58 g (66%) of azido diol 3 as a
colorless oil: 1H NMR (500 MHz, CDCl3) δ 1.37 (s, 3H), 1.48
(ddd, J ) 14.0, 10.8, 6.1 Hz, 1H), 1.60 (s, 3H), 1.62 (s, 3H),
1.64 (ddd, J ) 14.1, 10.6, 5.9 Hz, 1H), 1.68 (s, 3H), 1.97 (3
peak m, 2H), 2.08 (br 8 peak m, 4H), 2.59 (s, 1H, exch with
D2O), 3.58 (br d, J ) 7.0 Hz; ABX upon D2O exch, J AX ) 3.2
Hz, J BX ) 7.7 Hz; 1H), 3.65 (br dd, J ) 11.2, 7.7 Hz; ABX upon
D2O exch, J AB ) 11.2 Hz, J BX ) 7.7 Hz, 1H), 3.74 (br dd, J )
11.2, 3.0 Hz; ABX upon D2O exch, J AB ) 11.2 Hz, J AX ) 3.2
Hz, 1H), 5.09 (m, 2H); 13C NMR (126 MHz, CDCl3) δ 15.79,
17.47, 19.16, 22.02, 25.47, 26.39, 36.32, 39.42, 62.31, 65.4,
75.94, 122.82, 123.96, 131.32, 136.01; IR 3386 (OH), 2103 (N3).
Anal. Calcd for C15H27N3O2 (MW 281.39): C, 64.02; H, 9.67;
N, 14.93. Found: C, 64.35; H, 9.52; N, 14.94.
g of a colorless oil. Purification by flash column chromatogra-
phy (5% EtOAc/hexanes) afforded 892 mg (65%) of the primary
silyl ether and 198 mg (14%) of the secondary silyl ether in
elution order. Data for primary silyl ether 3-OTBS: 1H NMR
(500 MHz, CDCl3) δ 0.09 (s, 3H), 0.10 (s, 3H), 0.91 (s, 9H),
1.32 (s, 3H), 1.46 (ddd, J ) 13.9, 10.2, 5.9 Hz, 1H), 1.60 (s,
3H), 1.62 (s, 3H), 1.63 (m, 1H), 1.68 (s, 3H), 1.98 (3 peak m,
2H), 2.08 (br 8 peak m, 4H), 2.78 (d, J ) 2.93 Hz, 1H, exch
with D2O), 3.56 (ABX upon D2O exch J AX ) 3.2 Hz, J BX ) 8.1
Hz, 1H), 3.61 (ABX upon D2O exch, J AB ) 9.7 Hz, J BX ) 8.1
Hz, 1H), 3.74 (ABX upon D2O exch, J AB ) 9.7 Hz, J AX ) 3.2
Hz, 1H), 5.10 (m, 2H); 13C NMR (126 MHz, CDCl3) δ -5.62,
15.74, 17.46, 17.99, 18.93, 21.91, 25.47, 25.62, 26.42, 36.40,
39.44, 62.60, 64.69, 75.49, 123.12, 124.04, 131.23, 135.66; IR
3568 (OH), 2103 (N3). Anal. Calcd for C21H41N3O2Si (MW
395.65): C, 63.75; H, 10.44; N, 10.62. Found: C, 64.14; H,
10.49; N, 9.99. MS m/z 395.6. Data for secondary silyl ether:
1H NMR (500 MHz, CDCl3) δ 0.07 (s, 3H), 0.08 (s, 3H), 0.91
(s, 9H), 1.46 (ddd, J ) 13.9, 10.2, 5.9 Hz, 1H), 1.58 (s, 3H),
1.60 (s, 3H), 1.62 (s, 3H), 1.62 (m, 1H), 1.68 (s, 3H), 1.98 (3
peak m, 2H), 2.08 (br 8 peak m, 4H), 3.55 (dd, J ) 8.3, 3.2 Hz,
1H), 3.61 (dd, J ) 9.4, 7.9 Hz, 1H), 3.75 (dd, J ) 9.5, 3.3 Hz,
1H), 5.10 (m, 2H).
(()-(2R*,3R*,6E)-3-Azido-1-(ter t-bu tyldim eth ylsilyloxy)-
3,7,11-tr im eth yl-6,10-d od eca d ien -2-yl Meth a n esu lfon a te
(17). The following procedure is based on that reported by
Crossland.22 A solution of the primary silyl ether (3-OTBS,
0.366 g, 0.926 mmol) and Et3N (0.291 g, 2.87 mmol) in CH2-
Cl2 (4 mL) was stirred and cooled at 0 °C under N2 as neat
methanesulfonyl chloride (0.402 g, 2.78 mmol) was added
dropwise over 10 min. After 1 h, ice-cold H2O (2 mL) was
added, and the aqueous phase was extracted with Et2O (3 ×
5 mL). The organic extracts were combined, washed with satd
NaCl (2 × 5 mL), dried (MgSO4), and concentrated under
reduced pressure to give 400 mg of a yellow oil. Purification
by flash column chromatography (5% EtOAc/hexanes) afforded
336 mg (78%) of the azido mesylate 17 as a colorless oil: 1H
NMR (500 MHz, CDCl3) δ 0.09 (s, 3H), 0.11 (s, 3H), 0.91 (s,
9H), 1.40 (s, 3H), 1.58 (m, 1H), 1.60 (s, 3H), 1.62 (s, 3H), 1.64
(m, 1H), 1.68 (s, 3H), 1.97 (m, 2H), 2.10 (m, 4H), 3.15 (s, 3H),
3.86 (ABX, J AB ) 11.7 Hz, J BX ) 7.7 Hz, 1H), 3.94 (ABX, J AB
(()-(2R*,3R*,6E)-3-Azid o-3,7-d im et h yl-6-oct en -1,2-d i-
ol (4). (()-2,3-Epoxygeraniol (2, 1.00 g, 5.90 mmol), NaN3 (0.85
g, 13.02 mmol), Et2AlCl (6.6 mL, 11.83 mmol, 1.8 M in toluene),
PhMe (10 mL); rt; 16 h; 1.05 g. Column purification of a 700-
mg sample in the same manner as that described for 3 gave
0.436 g (54% based on recovered epoxide) of azido diol 4 as a
pale yellow oil and 48 mg (6%) of the known enediol byprod-
uct.16 4: 1H NMR (500 MHz, CDCl3) δ 1.36 (s, 3H), 1.47 (ddd,
J ) 13.9, 10.9, 5.8 Hz, 1H), 1.62 (s, 3H), 1.69 (d, J ) 1.1 Hz,
3H), 1.65 (ddd, J ) 14.1, 10.7, 4.7 Hz, 1H), 2.02-2.14 (10 peak
m, 2H), 2.40-2.70 (br s, 2H, exch with D2O), 3.57 (ABX, J AX
) 3.2 Hz, J BX ) 7.7 Hz, 1H), 3.66 (ABX, J AB ) 11.1 Hz, J BX
)
7.7 Hz, 1H), 3.75 (ABX, J AB ) 11.1 Hz, J AX ) 3.2 Hz, 1H), 5.08
(t sept, J ) 7.1, 1.5 Hz, 1H); 13C (126 MHz, CDCl3) δ 17.42,
19.14, 22.11, 25.44, 36.32, 62.29, 65.41, 75.94, 122.96, 132.40;
IR 3389 (OH), 2105 (N3); MS (FI) m/z 214.2. Some peaks of
the enediol (∼11%) were also present [δ 4.99 (s, 0.1H), 5.11
(s, 0.1H)] and the data correspond to the literature values.16
) 11.7 Hz, J AX ) 2.6 Hz, 1H), 4.57 (ABX, J AX ) 2.6 Hz, J BX
)
7.7 Hz, 1H), 5.08 (m, 2H); 13C NMR (126 MHz, CDCl3) δ -5.66,
15.79, 17.46, 18.15, 19.53, 21.90, 25.47, 25.66, 26.37, 36.80,
38.84, 39.43, 62.34, 64.12, 87.68, 122.43, 123.96, 131.28,
136.17; IR 2109 (N3).
(()-(2S*,3R*,6E)-3-Azid o-3,7-d im et h yl-6-oct en -1,2-d i-
ol (6). (()-2,3-Epoxynerol (5, 1.01 g, 5.95 mmol), NaN3 (0.85
g, 13.03 mmol), Et2AlCl (6.6 mL, 11.85 mmol, 1.8 M in PhMe),
PhMe (10 mL); rt; 16 h; 0.97 g. Column purification of an 850-
mg portion was carried out as described for 3 to give 0.690 g
(63% based on recovered epoxide) of azido diol 6 as a pale
yellow oil and 60 mg (5%) of the enediol. Data for azido diol 6:
1H NMR (400 MHz, CDCl3) δ 1.25 (s, 3H), 1.60 (ddd, J ) 13.7,
10.9, 5.4 Hz, 1H, H4), 1.63 (s, 3H), 1.69 (d, J ) 1.0 Hz, 3H),
1.73 (ddd, J ) 13.9, 10.9, 5.8 Hz, 1H), 2.0-2.20 (br 10 peak
m, 3H), 2.52-2.61 (br s, 1H, exch with D2O), 3.63 (br 3 peak
m, 2H), 3.74 (br 4 peak m, 1H), 5.10 (t sept, J ) 7.1, 1.5 Hz,
1H); 13C NMR (101 MHz, CDCl3) δ 17.41, 18.85, 22.14, 25.44,
36.44, 62.29, 65.60, 75.81, 122.90, 132.36; IR 3389 (OH), 2111
(N3); MS m/z 214.2; HRMS (CI+) m/z calcd for C10H19N3O2:
214.1809. Found: 214.1807. Some peaks of the enediol were
also present (∼6%): δ 4.99 (s, 0.06H), 5.10 (s, 0.06H).
(()-tr a n s-3-((E)-4,8-Dim eth yl-3,7-n on adien yl)-3-m eth yl-
a zir id in e-2-m eth a n ol (19). The following procedure is based
on that reported by Guedj.26 A suspension of LiAlH4 (28 mg,
0.520 mmol) in Et2O (2 mL) was stirred and cooled at 0 °C
under N2 as a solution of the azido mesylate (17, 100 mg, 0.210
mmol) in Et2O (1 mL) was added dropwise over 2 min via
syringe. After 4 h, the gray suspension was cooled to 0 °C and
stirred as H2O (28 µL), 15% NaOH (28 µL), and H2O (84 µL)
were added sequentially with occasional additions of Et2O to
maintain the volume at ca. 3 mL. After 2 h, the white solid
was filtered and washed with EtOAc (4 × 3 mL). The filtrates
were combined, dried (MgSO4), and concentrated under re-
duced pressure to give 30 mg of a pale yellow oil. Purification
by flash column chromatography (10% MeOH/CH2Cl2) afforded
(()-(2R*,3R*,6E)-3-Azid o-3,7,11-tr im eth yl-6,10-d od eca -
d ien -1,2-d iol 1-ter t-Bu tyld im eth ylsilyl Eth er (3-OTBS).
The following procedure is based on that reported by Fujii and
Ibuka.21 A solution of the azido diol (3, 1.00 g, 3.57 mmol),
imidazole (1.44 g, 21.2 mmol), and 4-DMAP (44 mg, 0.359
mmol) in DMF (5 mL) was stirred and cooled at 0 °C under
N2 as TBDMSCl (1.59 g, 10.6 mmol) in DMF (6 mL) was added
dropwise over 5 min. The solution was warmed to room
temperature and after 30 min ethanolamine (0.647 mL, 10.7
mmol) was added to scavenge excess TBDMSCl. After 2 h, H2O
(3 mL) was added, and the product was extracted with benzene
(3 × 10 mL). The organic extracts were combined, dried
(MgSO4), and concentrated under reduced pressure to give 1.10
1
22 mg (43%) of aziridine 19 as a colorless oil. The H and 13C
NMR spectral data and assignments correspond to those
reported in the literature.1,5,27
(()-(2R*,3R*,6E)-3-Azid o-3,7-d im et h yl-6-oct en -1,2-d i-
ol 1-ter t-Bu tyld im eth ylsilyl Eth er (4-OTBS). The silyl
ether was prepared as described previously for 3 using azido
diol 4 (478 mg, 2.24 mmol), imidazole (918 mg, 13.50 mmol),
(26) Forestier, M. A.; Ayi, A. Y.; Condom, R.; Boyode, B. P.; Conlin,
J . N.; Selway, J .; Challand, R.; Guedj, R. Nucleosides Nucleotides 1993,
12 (9), 915.
(27) Koohang, A., Ph.D. Thesis. University of Illinois at Urbana-
Champaign, 1999.
J . Org. Chem, Vol. 68, No. 1, 2003 81

Davis et al.
4-DMAP (28 mg, 0.23 mmol), and TBDMSCl (1.016 g, 6.74
mmol) in DMF (7 mL). After 0.5 h, ethanolamine (4.30 g, 6.95
mmol) was added. Isolation of the crude product (925 mg) and
purification in the same manner as that described for com-
pound 3 gave 457 mg (62%) of the primary silyl ether and 105
mg (14%) of the secondary silyl ether. Data for primary silyl
ether (4-OTBS): 1H NMR (500 MHz, CDCl3) δ 0.09 (s, 3H),
0.10 (s, 3H), 0.91 (s, 9H), 1.32 (s, 3H), 1.45 (ddd, J ) 14.1,
11.1, 5.8 Hz, 1H), 1.60 (m, 1H), 1.63 (s, 3H), 1.69 (d, J ) 1.1
Hz, 3H), 2.04-2.12 (6 peak m, 2H), 2.79 (br s, 1H, exch with
D2O), 3.56 (ABX, J AX ) 3.4 Hz, J BX ) 8.2 Hz, 1H), 3.61 (ABX,
J AB ) 9.4 Hz, J BX ) 8.2 Hz, 1H), 3.74 (ABX, J AB ) 9.4 Hz, J AX
) 3.4 Hz, 1H), 5.10 (t sept, J ) 7.1, 1.5 Hz, 1H); 13C NMR
(126 MHz, CDCl3) δ -5.39, 19.17, 22.20, 25.68, 25.84, 36.61,
53.07, 62.78, 64.86, 75.65, 123.44, 123.49, 132.28; IR 3569
(OH), 2105 (N3). Data for secondary silyl ether: 1H NMR (500
MHz, CDCl3) δ 0.15 (s, 3H), 0.16 (s, 3H), 0.95 (s, 9H), 1.49 (m,
1H), 1.57 (s, 3H), 1.64 (s, 3H), 1.64 (m, 1H), 1.70 (d, J ) 1.2
Hz, 3H), 1.97 (dd, J ) 6.9, 5.3 Hz, 1H), 2.08 (m, 2H), 3.60 (app
t, J ) 4.5 Hz, 1H), 3.70 (app dd, J ) 9.5, 4.5 Hz, 1H), 3.79
(app dd, J ) 9.4, 5.0 Hz, 1H), 5.09 (m, 1H).
and purification in the same manner gave 646 mg (65%) of
primary silyl ether 6-OTBS and 137 mg (14%) of its secondary
silyl ether isomer. Data for 6-OTBS: 1H NMR (500 MHz,
CDCl3) δ 0.09 (s, 3H), 0.10 (s, 3H), 0.91 (s, 9H), 1.22 (s, 3H),
1.60 (m, 1H), 1.63 (s, 3H), 1.66 (m, 1H), 1.69 (s, 3H), 1.98-
2.15 (12 peak m, 2H), 2.76 (d, J ) 2.4 Hz, 1H, exch with D2O),
3.60 (ABX, J AB ) 8.6 Hz, J BX ) 8.2 Hz, 1H), 3.62 (ABX, J AX
)
2.3 Hz, J BX ) 8.2 Hz, 1H), 3.73 (ABX, J AB ) 8.6 Hz, J AX ) 2.4
Hz, 1H), 5.11 (t sept, J ) 5.8, 1.5 Hz, 1H); 13C NMR (126 MHz,
CDCl3) δ -5.63, 17.40, 18.01, 19.03, 22.12, 25.45, 25.61, 36.13,
62.74, 64.65, 76.13, 123.25, 132.07; IR 3571 (OH), 2105 (N3);
MS (CI+) m/z 328.3. Data for secondary silyl ether: 1H NMR
(500 MHz, CDCl3) δ 0.16 (s, 3H), 0.18 (s, 3H), 0.96 (s, 9H),
1.28 (s, 3H), 1.60 (m, 1H), 1.64 (s, 3H), 1.70 (d, J ) 1.0 Hz,
3H), 1.91 (br s, 1H), 2.08 (m, 2H), 3.59 (app dd, J ) 8.6, 4.0
Hz, 1H), 3.65 (br dd, J ) 11.2, 4.1 Hz, 1H), 3.73 (br dd, J )
11.3, 2.4 Hz, 1H), 5.10 (m, 1H).
(()-(2S*,3R*)-3-Azid o-1-(ter t-b u t yld im et h ylsilyloxy)-
3,7-d im eth yl-6-octen -2-yl Meth a n esu lfon a te (21). The azi-
do mesylate was prepared as described previously for 3-OTBS
using silyl ether 6-OTBS (419 mg, 1.28 mmol), Et3N (417 mg,
4.10 mmol), and methanesulfonyl chloride (444 mg, 3.84 mmol)
in CH2Cl2 (5 mL). Purification in the same manner gave 318
mg (69%) of the azido mesylate 21 as a yellow oil and 50 mg
(12%) of recovered starting material. Data for 21: 1H NMR
(500 MHz, CDCl3) δ 0.09 (s, 3H), 0.10 (s, 3H), 0.91 (s, 9H),
1.38 (s, 3H), 1.57 (m, 1H), 1.63 (s, 3H), 1.65 (m, 1H), 1.69 (s,
3H), 2.01-2.18 (14 peak m, 2H), 3.15 (s, 3H), 3.83 (ABX, J AB
) 11.8 Hz, J BX ) 6.9 Hz, 1H), 3.94 (ABX, J AB ) 11.8 Hz, J AX
) 3.0 Hz, 1H), 4.56 (ABX, J AX ) 3.0 Hz, J BX ) 6.9 Hz, 1H),
5.07 (t sept., J ) 7.1, 1.3 Hz, 1H); 13C NMR (126 MHz, CDCl3)
δ -5.73, 17.45, 19.55, 21.91, 25.47, 36.10, 38.70, 52.86, 62.32,
64.24, 87.55, 122.62, 132.53; IR 2106 (N3); MS m/z 406.2;
HRMS (CI+) calcd for C17H35N3O4SSi 406.2432, found 406.2434.
(()-cis-3-(4-Met h yl-3-p en t en yl)-3-m et h yla zir id in e-2-
m eth a n ol (22). The aziridine was prepared as described
previously for (()-2,3-aziridinofarnesol (19) using azido me-
sylate 21 (280 mg, 0.691 mmol) and LiAlH4 (166 mg, 4.37
mmol) in Et2O (2 mL). After 5 h, H2O (166 µL), 15% NaOH
(166 µL), and H2O (498 µL) were added dropwise in succession
to give 66 mg of a colorless oil. Isolation and purification in
the same manner as that described previously gave 53 mg
(45%) of the aziridino alcohol as a white solid. The melting
(()-(2R*,3R*)-3-Azid o-1-(ter t-b u t yld im et h ylsilyloxy)-
3,7-d im eth yl-6-octen -2-yl Meth a n esu lfon a te (18). The azi-
do mesylate was prepared as described previously for 17 using
primary silyl ether 4-OTBS (425 mg, 1.30 mmol), Et3N (421
mg, 4.16 mmol), and methanesulfonyl chloride (447 mg, 3.90
mmol) in CH2Cl2 (5 mL). Purification in the same manner as
that described for compound 17 gave 383 mg (73%) of azido
mesylate 18 as a yellow oil: 1H NMR (500 MHz, CDCl3) δ 0.09
(s, 3H), 0.10 (s, 3H), 0.91 (s, 9H), 1.39 (s, 3H), 1.56 (ddd, J )
14.2, 11.6, 5.6 Hz, 1H), 1.62 (s, 3H), 1.65 (m, 1H), 1.68 (d, J )
0.9 Hz, 3H), 2.02-2.17 (12 peak m, 2H), 3.15 (s, 3H), 3.86
(ABX, J AB ) 11.8 Hz, J BX ) 7.9 Hz, 1H), 3.94 (ABX, J AB
11.8 Hz, J AX ) 2.6 Hz, 1H), 4.57 (ABX, J AX ) 2.6 Hz, J BX
)
)
7.9 Hz, 1H), 5.07 (t sept., J ) 7.1, 1.5 Hz, 1H); 13C NMR (126
MHz, CDCl3) δ -5.51, 18.35, 19.64, 19.72, 22.19, 25.65, 25.93,
37.03, 39.09, 62.52, 64.27, 87.83, 122.72, 132.76; IR 2108 (N3).
(()-tr a n s-3-(4-Meth yl-3-p en ten yl)-3-m eth yla zir id in e-2-
m eth a n ol (20). The aziridine was prepared as described
previously for (()-2,3-aziridinofarnesol (19) using the azido
mesylate (18, 355 mg, 0.876 mmol) and LiAlH4 (94 mg, 2.46
mmol) in Et2O (2 mL). After 5 h, H2O (94 µL), 15% NaOH (94
µL), and H2O (282 µL) were added dropwise in succession to
give 100 mg of a colorless oil. Isolation and purification in the
same manner as that described for compound 19 gave 85 mg
(57%) of the aziridino alcohol 20 as a low-melting white solid.
The aziridine was a clear oil at room temperature and a white
1
point and H NMR spectral data and assignments correspond
to those reported in the literature.5
Ack n ow led gm en t. We thank the National Insti-
tutes of Health for financial support (GM 13952).
1
solid at -20 °C. The H NMR spectral data and assignments
correspond to those reported in the literature.5
(()-(2S*,3R*)-3-Azid o-3,7-d im et h yl-6-oct en -1, 2-d iol
1-ter t-Bu tyld im eth ylsilyl Eth er (6-OTBS). The silyl ether
was prepared as described previously for 3-OTBS using azido
diol 6 (638 mg, 3.00 mmol), imidazole (1.23 g, 18.05 mmol),
4-DMAP (45 mg, 0.36 mmol), and TBDMSCl (1.36 g, 9.02
mmol) in DMF (10 mL). After 0.5 h, ethanolamine (0.607 g,
9.00 mmol) was added. Isolation of the crude product (963 mg)
Su p p or tin g In for m a tion Ava ila ble: Procedural informa-
tion and characterization data for all other substrates, general
experimental details, and copies of 1H NMR spectra for
selected compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
J O026506C
82 J . Org. Chem., Vol. 68, No. 1, 2003