Highly stereoselective syntheses of syn- and anti-1,2-amino alcohols

Article abstract of DOI:10.1021/jo010270f

The reduction of N-protected amino ketones can be carried stereoselectively to produce either the syn- or anti-amino alcohol diastereomer. Carbamate-protected amino ketones can be reduced predictably and selectively to anti-amino alcohols with LiAlH(O-t-Bu)₃ in ethanol at -78°C. N-Trityl-protected amino ketones can be reduced selectively to syn-amino alcohols with LiAlH(O-t-Bu)₃ in THF at -5°C.

Full text of DOI:10.1021/jo010270f

VOLUME 67, NUMBER 4
FEBRUARY 22, 2002
© Copyright 2002 by the American Chemical Society
Articles
High ly Ster eoselective Syn th eses of syn - a n d a n ti-1,2-Am in o
Alcoh ols
Robert V. Hoffman,* Najib Maslouh, and Francisco Cervantes-Lee†
Department of Chemistry and Biochemistry, North Horseshoe Drive, New Mexico State University,
Las Cruces, New Mexico 88003-8001
rhoffman@nmsu.edu
Received March 12, 2001
The reduction of N-protected amino ketones can be carried stereoselectively to produce either the
syn- or anti-amino alcohol diastereomer. Carbamate-protected amino ketones can be reduced
predictably and selectively to anti-amino alcohols with LiAlH(O-t-Bu)3 in ethanol at -78 °C. N-Trityl-
protected amino ketones can be reduced selectively to syn-amino alcohols with LiAlH(O-t-Bu)3 in
THF at -5 °C.
In tr od u ction
N-protected amino ketones 1 (R2 ) alkyl, aryl, P )
protecting group, P′ ) alkyl, H) are much more problem-
atic.
1,2-Amino alcohols are important and versatile syn-
thetic intermediates for the preparation of a wide variety
of natural products, drugs, and metal-binding ligands.1
Consequently, the development of synthetic methods for
their preparation in a stereocontrolled manner has
received significant attention for quite some time. In
general, the hydroxyl group of the amino alcohol is
installed either by the addition of an organometallic
reagent to an amino aldehyde 1 (R2 ) H) or by the
reduction of an amino ketone 1 (R2 * H) (Scheme 1).2
Moreover, stereochemical control is good for either strat-
egy when the amino group is a primary, secondary, or
unhindered tertiary amine 1 (P, P′ ) alkyl, H), which is
the most common class of amine used in these reactions.2
In contrast, syntheses of amino alcohols by reduction of
Two modes of stereocontrol determine which diastere-
omer is the major product from the reduction of a
protected amino ketone. Chelation control, in which a
Lewis acid or metal ion coordinates to the carbonyl
oxygen and the amine nitrogen, enforces a syn-periplanar
relationship between the amine and ketone groups and
leads to the anti diastereomer (Figure 1a). Felkin-Anh
control, in which a dihedral angle of about 90° between
the amine and ketone groups maximizes stereoelectronic
interactions in the transition state,3 leads to the syn
diastereomer (Figure 1b).
Carbamate groups are ubiquitous protecting groups for
amines. Stereocontrol in the reduction of carbamate-
protected amino ketones would appear to favor chelation
† To whom inquiries about the X-ray structure should be ad-
dressed: Department of Chemistry, University of Texas El Paso, 500
University Blvd, El Paso, TX 79968-0513.
(1) (a) Ager, D. J .; Prakash, I.; Schaad, D. R. Chem. Rev. 1996, 96,
835. (b) Paleo, M. R.; Cabeza, I.; Sardina, F. J . J . Org. Chem. 2000,
65, 2108.
(3) (a) Cherest,, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968,
18, 2199. (b) Cherest, M.; Felkin, H. Tetrahedron Lett. 1968, 18, 2205.
(c) Anh, N. T.; Eisenstein, O. Nouv. J . Chem. 1977, 1, 61. (d) For an
excellent summary, see: Eliel, E. L.; Wilen, S. H.; Mander, L. N.
Stereochemistry of Organic Compounds; J ohn Wiley and Sons: New
York, 1994; pp 876-880.
(2) Tramontini, M. Synthesis 1982, 605 is a definitive review.
10.1021/jo010270f CCC: $22.00 © 2002 American Chemical Society
Published on Web 11/06/2001

1046 J . Org. Chem., Vol. 67, No. 4, 2002
Hoffman et al.
Sch em e 2
a sphingosine synthesis and screened eight reducing
agents before finding that L-Selectride gives a 19:1 ratio
of the anti product.7 (Normally, L-Selectride gives Fel-
kin-Anh syn selectivity!)
Many other studies where extensive screening of
reducing agents is used to achieve stereocontrolled
reduction8 clearly highlight the need for a reliable,
chelation controlled reduction of carbamate-protected
amino ketones. Such a protocol must be anti selective,
predictable, and structurally tolerant.
F igu r e 1.
Sch em e 1
To achieve Felkin-Anh control in the reduction of
protected amino ketones and thus produce syn-amino
alcohols, the amine group must be modified to make it
sterically bulky. This results in a steric preference for a
Felkin-Anh transition state (Figure 1b) and minimizes
effective chelation involving the amino nitrogen and the
ketone oxygen. Reetz found that the N,N-dibenzylamine
protecting group rendered the amine group sufficiently
bulky to ensure Felkin-Anh control.9 Other groups10
including our own11 utilized N,N-dibenzyl protection for
effective stereocontrol in amino ketone reductions.
Despite these successes, we found that the installation
of the N,N-dibenzyl protecting group is very problematic,11b
and its hydrogenolytic removal is not compatible with
unsaturated groups.11a In the course of a synthesis of
sphingosines, we found that an N-trityl protecting group
gave good Felkin-Anh control of an amino ketone reduc-
tion, and it was much easier to attach and remove.12 That
work provides the basis for a general amino ketone
reduction protocol with Felkin-Anh stereocontrol if it is
found to be general.
We had previously developed a very simple and direct
method for preparing N-protected amino ketones with
widely varying structures (Scheme 2). This methodolgy
is very effective in the production of peptidomimetics13
and sphingosines,12 but it also is an excellent method for
almost any amino ketone. Consequently, we undertook
a study of the reductions of protected amino ketones in
order to elucidate factors that contribute to effective
stereocontrol and, hence, extend our amino ketone syn-
thesis into a stereocontrolled synthesis of amino alcohols.
control since both the ketone oxygen and the carbamate
nitrogen are sterically accessible for coordination to a
Lewis acid. Thus, chelation stereocontrol in the reduction
of carbamate-protected amino ketones provides a method
for the stereoselective synthesis of anti carbamate-
protected amino alcohols.
Examination of the literature reveals, however, that
there is no reagent that reduces carbamate-protected
amino ketones with predictable chelation control. In fact,
contradictory results are the rule rather than the excep-
tion. For example, Benedetti et al. reported on the
reduction of Boc-protected amino ketones with sodium
borohydride in methanol, which had anti diastereoselec-
tivities (chelation control) ranging from 3:1 to 32:1. There
was no obvious correlation between the substrate struc-
ture and the diastereoselectivity. These and other experi-
ments led the authors to conclude that the reduction is
“not markedly affected by the reducing agent or solvent”.4
In contrast, Koskinen studied stereoselectivity in the
reductions of a series of Boc-protected unsaturated amino
ketones. Anti-selectivity ranged from 6:1 to 1:4 for various
reducing agents. In this case, sodium borohydride was
found to give the syn isomer selectively (3:1) (Felkin-
Anh control)! These authors concluded that “solvent
effects and reagent size are very important” influences
on the stereoselectivity.5
Good levels of chelation control in reductions of car-
bamate-protected amino ketones are often achieved by
empiricle screening of a variety of reducing agents. For
example, in the elegant synthesis of statine from leucine
reported by J oullie, potassium borohydride gave a 10:1
anti diastereoselectivity for the reduction of a Cbz-
protected γ-amino-â-keto ester and was clearly superior
to other reductants examined.6 More recently, Dondoni
prepared a Boc-protected acetylenic amino ketone during
(7) Dondoni, A.; Perrone, D.; Turturici, E. J . Org. Chem. 1999, 64,
5557.
(8) See, for example: (a) Tramontini, M. Synthesis 1982, 605. (b)
Nishida, A.; Sorimachi, H.; Iwaida, M.; Matsumizu, M.; Kawate, T.;
Nakagawa, M. Synth. Lett. 1998, 389. (c) Chung, S.-K.; Lee, J .-M.
Tetrahedron: Asymmetry 1999, 10, 1441. (d) Rotella, D. P. Tetrahedron
Lett. 1995, 36, 5453.
(9) (a) Reetz, M. T.; Drewes, M. W.; Matthews, B. R.; Lennick, K.
Chem. Commun. 1989, 1474. (b) Reetz, M. T. Angew. Chem., Int. Ed.
Engl. 1991, 30, 1531.
(10) Lagu, B. R.; Liotta, D. C. Tetrahedron Lett. 1994, 35, 547.
(11) (a) Tao, J .; Hoffman, R. V. J . Org. Chem. 1997, 62, 6240. (b)
Hoffman, R. V.; Tao, J . J . Org. Chem. 1997, 62, 2292.
(12) (a) Hoffman, R. V.; Tao, J . Tetrahedron Lett. 1998, 39, 3953.
(b) Hoffman, R. V.; Tao, J . J . Org. Chem. 1998, 63, 3979.
(13) (a) Hoffman, R. V.; Kim, H.-O. Tetrahedron Lett. 1992, 33, 3579.
(b) Hoffman, R. V.; Tao, J . Tetrahedron 1997, 53, 7119. (c) Hoffman,
R. V.; Tao, J . Tetrahedron Lett. 1998, 39, 4195. (d) Hoffman, R. V.;
Tao, J . J . Org. Chem. 1999, 64, 126.
(4) Benedetti, F.; Miertus, S.; Norbedo, S.; Tossi, A.; Zlatoidzky, P.
J . Org. Chem. 1997, 62, 9348.
(5) Koskinen, A. M. P.; Koskinen, P. M. Tetrahedron Lett. 1993, 34,
6765.
(6) Harris, B. D.; J oullie, M. M. Tetrahedron 1988, 44, 3489.

Syntheses of syn- and anti-1,2-Amino Alcohols
J . Org. Chem., Vol. 67, No. 4, 2002 1047
Sch em e 3
Sch em e 4
signals in the proton-decoupled 13C NMR spectrum.
These spectral differences could not be used to assign the
stereochemistry, but they did indicate that a mixture of
diastereomers was present. It was very important to be
able to assign the stereochemistry of the reduction readily
and unambiguously. Thus, syn- and anti-3a were sepa-
rated chromatographically and converted to their cyclic
oxazolidinone derivatives 4a with sodium hydride in
DMF. anti-3a cyclizes to syn-4a while syn-3a cyclizes to
anti-4a (Scheme 4). Irradiation of the benzylic protons
or the R-methylene protons allows the vicinal coupling
constants between the methine protons of the oxazolidi-
none ring to be determined accurately. syn-4a had the
larger coupling constant J vic ) 7.5 Hz, while anti-4a had
a smaller coupling constant J vic ) 5.5 Hz. This agrees
with previous studies that established that syn-oxazoli-
dinones have larger vicinal coupling constants than the
anti diastereomers.14 Ab initio calculations on a truncated
version of syn-4a and anti-4a show that the dihedral
angle between the methine hydrogens is 25.5° in the most
stable conformations of syn-4a and 137° in the most
stable conformations of anti-4a .15 These dihedral angles
also predict that syn isomers of oxazolidinones have
larger coupling constants (6.6 Hz calcd) than the anti
isomers (4.9 Hz calcd).
These results led to several conclusions about the
reduction. First, the reduction is very efficient for all the
reducing agents as the crude yields are very high and
the crude product was nearly pure in all cases. Only trace
amounts of contaminants could be detected in the NMR
of the crude products. The syn/anti ratios in Table 1 were
determined on the crude products to avoid fractionation
during purification. Second, borohydride reagents in
alcohol solvents give chelation-controlled reduction, but
the chelating atom is not the metal counterion. Identical
results are found (Table 1, entries 1-5) irrespective of
the metal (or none, Table 1, entry 4) present in solution.
Third, solvent appears to play an important role in the
diastereoselection. Our hypothesis is that hydroxylic
solvents (EtOH and other alcohols) promote the exchange
and/or disproportionation of ligands attached to boron so
that the substrate can become bound to boron in a
chelated fashion needed for effective chelation stereo-
Ta ble 1. Th e Red u ction of 1a w ith Va r iou s Red u cin g
Agen ts/Con d ition s
yield
entry
reagent
NaBH4
conditions
(% crude) anti/syna
1
2
3
4
5
6
7
8
9
EtOH, -78 °C
EtOH, -78 °C
EtOH, -78 °C
EtOH, -78 °C
EtOH, -78 °C
THF, -78-0 °C
ether, -78-0 °C
THF, -78 °C
98
97
98
97
97
95
94
91
93
95
98
7:1
7:1
7:1
7:1
7:1
2:1
1:2
1:9
1:9
LiBH4
KBH4
-
Me4N+BH4
NaBH4/CeCl3
Zn(BH4)2
Zn(BH4)2
L-Selectride
N-Selectride
LiAl(O-t-Bu)3H THF, -78- -20 °C
LiAl(O-t-Bu)3H EtOH, -78 °C
THF, -78 °C
10
11
1:1
>95:5b
a
The anti/syn ratio was determined on the crude product.
Only one diastereomer could be observed in the NMR spectrum
b
of the crude product.
Resu lts
N-Ca r ba m a te-P r otected Am in o Keton es. A series
of eight carbamate-protected amino ketones 1a -h were
prepared via â-keto esters 2a -h (Scheme 3). Noteworthy
is that the method is compatible with a variety of
functional groups and carbamate protecting groups.
Somewhat surprisingly, even the Fmoc group of 1f
appears to pose no problem even though strong bases are
used during its preparation.
Ch ela tion Con tr ol. To examine chelation control in
the reductions of N-carbamate-protected amino ketones,
compound 1a was reduced to mixtures of syn- and anti-
3a with a variety of reducing agents and conditions that
have been reported in the literature. These data appear
in Table 1.
(14) Harris, B. D.; Bhat, K. L.; J oullie, M. M. Tetrahedron Lett. 1987,
28, 2837.
(15) Calculations were carried out on syn- and anti-i at the HF
6-31G* level using the Titan suite of programs from Wavefunction. A
solvation model was not included in the calculations.
The diastereomers produced from reductions of 1a
could be distinguished by both 1H and 13C NMR. For
example, the benzylic protons of syn- and anti-3a ap-
peared as distinct singlets in the 1H NMR spectrum, and
the carbinol carbons of syn- and anti-3a gave distinct

1048 J . Org. Chem., Vol. 67, No. 4, 2002
Hoffman et al.
Ta ble 2. Red u ction of Ca r ba m a te-P r otected Am in o
Keton es 1a -h w ith LiAl(O-t-Bu )3H in Eth a n ol a t -78 °C
control (vide infra). In non-hydroxylic solvents (Table 1,
entries 6 and 7) diastereoselection is nearly absent.
In support of this mechanistic rationale is the observa-
tion that Selectride reagents, which do not have ex-
changeable ligands on boron and thus should not readily
chelate, give the Felkin-Anh product. These results
should be viewed with caution, however, as we, in this
work, and others5,7 have observed that relatively minor
changes in the substrate structure can result in a switch
to the anti product with Selectride reducing agents.
Since alkoxy ligands bound to aluminum are also
readily exchangeable,22 lithium tri-tert-butoxyaluminum
hydride was tested. In THF, the diastereoselection was
poor. This was presumably due to inefficient ligand
exchange in the non-hydroxylic medium according to the
mechanistic hypothesis derived for borohydride reduc-
tion. These considerations culminated in the experiment
shown in entry 11, Table 1. If solid LiAl(O-t-Bu)3H is
added to -78 °C ethanol, very little hydrogen formation
is observed upon dissolution, and smooth reduction of 1a
to anti-3a in very high yield is observed. Temperature
control is critical since at -50 °C reaction between the
aluminum hydride and ethanol begins to compete with
the reduction.
yield (%)
compd
crude (recrystallized)
anti/syn
1a
1b
1c
1d
1e
1f
97 (88)
96 (80)
97 (89)
97 (82)
97 (82)
95 (81)
96 (82)
99 (85)
>95:5
>95:5
>95:5
>95:5
>95:5
>95:5
>95:5
>95:5
1g
1h
Sch em e 5
As a further confirmation of the stereochemical as-
signments, amino ketone 1c was reduced with sodium
borohydride in ethanol to give a mixture of syn-3c and
1
anti-3c. Both diastereomers could be detected in the H
of 7.5 Hz demonstrating the anti geometry of 3b,h .
Cyclization of the remaining examples was problematic
because of reactions of the side-chain protecting groups.
Since only one diastereomer was observed in the NMR,
it is reasonable to conclude that the geometry of these
products is anti as well. Noteworthy is that the method
appears to be insensitive to the nature of the carbamate
group and tolerates a range of functionality in the R1 and
R2 substituents in amino ketones 1a -h . This generality
suggests that many other functional groups should be
tolerated as well.
F elk in -An h Ster eocon tr ol. We had used the trityl
protecting group on nitrogen as an easily manipulated,
bulky protecting group to achieve Felkin-Anh control of
an amino ketone reduction during a sphingosine synthe-
sis.12 In that work, sodium borohydride was used as the
reductant. We since had reason to question the stereo-
chemical assignment16 and hoped to establish a general
protocol for achieving Felkin-Anh control in amino
ketone reductions.
To that end, a series of four N-trityl amino ketones
9a ,c,g,h were prepared from amino acids 5 in good
overall yields by the same general sequence used for
carbamate-protected examples (Scheme 5). The bulky
trityl group causes a significant increase in reaction time
for conversion of the tritylated amino acid to the â-keto
ester (3 h at -40 °C) compared to carbamate protecting
groups (30 min at -78 °C).
NMR by distinct methyl doublets at δ 1.05 and 1.13 and
in the 13C NMR by separate signals for the carbinol
carbons at δ 74.4 and 74.9. These diastereomers were
not separable by silica gel chromatography. Only a single
diasteromer was observed from the reduction of 1c with
LiAl(O-t-Bu)3H, which had a methyl doublet at δ 1.05 and
a carbinol carbon signal at δ 74.4. It was converted to
oxazolidinone 4c, which had a coupling constant of 7.5
Hz between the methine protons. Thus, 4c is the syn
diastereomer. This was further confirmed by X-ray
analysis of a crystal of 4c that showed the syn geometry.
The syn stereochemistry of 4c derives from the anti
isomer of 3c, which results from chelation controlled
reduction of 1c by LiAlH(OtBu)3.
With these results in hand, amino ketones 1a -h were
reduced with LiAl(O-t-Bu)3H in ethanol at -78 °C. For
the entire series, the yields were very high and only one
1
diastereomer could be detected by both H and 13C NMR
in the crystalline crude products (Table 2). Recrystalli-
zation gave analytically pure products (75-80%); how-
ever, the NMR spectra of the crude material and the
recrystallized products were virtually indistinguishable.
Amino alcohols 3b and 3h were also cyclized to oxa-
zolidinones 4b,h , which had methine coupling constants
(16) NMR and optical rotation were used to determine the reduction
diastereoselection. These methods may not be reliable for the accurate
determination of diastereomeric excess in sphingosines. Li, C. Brigham
Young University, personal communication.
Amino ketone 9c was reduced with sodium borohydride
in ethanol and with L-Selectride. The syn-amino alcohol
10c was the major product (Felkin-Anh control), but
only poor stereoselectivities of 1.5:1 and 1.6:1, respec-
tively, were obtained. These reductions required 12 h at
room temperature to be completed, which shows the rate-
retarding effect of the trityl group. We reasoned that
increased stereoselection would require a reducing agent
that was bulkier and more reactive. We therefore exam-
ined LiAlH(O-t-Bu)3 in THF (Table 3). High yields and
(17) (a) Koga, K.; Yamada, S. Chem. Pharm. Bull. 1972, 20, 526.
(b) Yamada, S.; Koga, K. Tetrahedron Lett. 1967, 1711. (c) Portoghese,
P. S.; Williams, D. A. Tetrahedron Lett. 1966, 6299.
(18) (a) Pierre, J . L.; Handel, H.; Baret, P. Tetrahedron 1974, 30,
3213. (b) Handel, H.; Pierre, J . L. Tetrahedron 1975, 31, 997. (c) Costes,
E.; Be´nard, C.; Lattes, A. Tetrahedron Lett. 1976, 1185.
(19) Nogradi, M. Stereoselective Synthesis: A Practical Approach,
2nd ed.; VCH Publishers: New York, 1995; p 111.
(20) Brown, H. C.; Mead, E. J . J . Am. Chem. Soc. 1956, 78, 3616.
(21) (a) J ohnson, M. R.; Rickborn, B. J . Org. Chem. 1970, 35, 1041.
(b) Wuesthoff, M. T.; Rickborn, B. J . Am. Chem. Soc. 1970, 92, 6894.
(22) (a) Haubenstock, H.; Eliel, E. L. J . Am. Chem. Soc. 1962, 84,
2363. (b) Brown, H. C.; Shoaf, C. J . J . Am. Chem. Soc. 1964, 86, 1079.

Syntheses of syn- and anti-1,2-Amino Alcohols
J . Org. Chem., Vol. 67, No. 4, 2002 1049
Ta ble 3. Red u ction s of Am in o Keton es 9a ,c,g,h in THF
It also appears that hydrogen bonding between the NH
proton and the carbonyl oxygen plays a minor role in
controlling the conformation. In THF or ether where
intramolecular H-bonding should be the greatest, only
low diastereoselectivity is found. Much higher diastereo-
selectivity is found in ethanol, a solvent that should
minimize intramolecular H-bonding by competetive H-
bonding with the solvent. Moreover, in a given solvent
at a given temperature the conformational control ex-
erted by hydrogen bonding should remain relatively
constant and should be relatively independent of the
hydride donor. Consequently, one should at least observe
the same major isomer irrespective of the reducing agent.
Comparison of entries 6 and 8 of Table 1 shows this is
not the case. These results cast serious doubt on intra-
molecular H-bonding to explain the diastereoselectivity.
There is also some earlier evidence suggesting that
hydrogen bonding is not an important stereocontrol
element in amino ketone reductions. Reductions of
tertiary amino ketones generally give syn products
because of Felkin-Ahn control by bulky tertiary amines.
Reduction of protonated tertiary ammonium ketones by
sodium borohydride in ethanol also gives the syn product.
The ammonium ion NH proton should provide even
greater hydrogen-bonding capability than an amide NH
proton but fails to influence the diastereoselection
significantly.17b
Our results suggest that chelation of the amino ketone
to boron or aluminum is the key to stereochemical
control. Both borohydride and aluminum hydride re-
agents can have multiple alkoxide ligands which can be
exchanged to produce a chelated intermediate. The
exchange/disproportionation of alkoxy ligands on boro-
hydrides is rapid20 and occurs by a dissociation/recom-
bination mechanism.21 Alkoxy ligands on aluminum
hydrides also exchange rapidly22 and by a dissociative
process.22,23 Thus it is possible for chelate formation to
occur by dissociative ligand exchange for both boron and
aluminum.
a t -5 °C
substrate
time (h)
yielda (%)
syn/antib
9a
9c
9g
9h
48
8
60
60
88 (76)
91d
>18:1c
4:1
mixturee
12:1
90 (71)
a
Yields are crude yields of the syn-3/anti-3 mixture. Yields in
b
parentheses are purified yields of syn-3. The syn/anti ratio was
determined by NMR of the crude product mixture. c Only one
d
isomer was detected by NMR. The mixture was not purified.
e Products of both ketone and ester reduction were obtained.
good to excellent syn-selectivity was found for the reduc-
tions of 9a ,c,h with this reagent. The reaction time was
lengthened considerably. Glutamate derivative 9g gave
a mixture of products. It appeared that reduction of the
ester and some cyclization accompanied reduction of the
ketone. This mixture was not pursued further.
The isomer distributions were determined by exchang-
ing the trityl protecting group for a carbamate protecting
group (Boc or Cbz) and cyclizing to the corresponding
oxazolidinone. In these cases, the methine coupling
constants indicated the anti-oxazolidinone was the major
isomer; thus, amino alcohols 10a ,c,h were predominantly
the syn stereoisomers. A simple recrystallization gave
pure carbamate-protected syn-amino alcohols.
Discu ssion
Prior to this study, the reagent and conditions needed
to produce anti-amino alcohols from carbamate-protected
amino ketones were largely determined by trial and error.
While chelation control is needed to ensure the proper
stereochemical outcome, the literature provides no gen-
eral guidelines on how to achieve it.
Boron24 and aluminum25 form only weak complexes
with neutral species but much stronger complexes with
anionic species. It is thus reasonable to postulate that
the N-carbamate-protected amino ketone must be con-
verted to an anion prior to complexation. The most likely
anion is the deprotonated carbamate, which could bind
through nitrogen to the boron or aluminum.26 Chelate
formation to the carbonyl oxygen follows (Scheme 6). This
process is directly analogous to chelate formation to
aluminum(III) in hydroxy ketones.27 It is also related to
the chelation of amino acids to aluminum.25 Reduction
of the chelated amino ketone occurs stereoselectivly to
give the amino alcohol product. Although Scheme 6 is
shown for borohydrides, a similar process is likely for
aluminum hydride reagents with alkoxy ligands.
The results also show that the use of an alcohol as
solvent increases the diastereoselectivity significantly for
borohydride reducing agents. It is possible that small
Several rationales have been proposed to account for
the anti diastereoselectivity seen in the reductions of
R-amino ketones. It is generally accepted that there must
be some type of conformational control to keep the amino
group coplanar (or nearly so) with the ketone (Figure
1a).8a It has been proposed that the aluminum or boron
atom of the aluminum hydride or borohydride reductant
is the control element in binding to both the amino
nitrogen and the ketone oxygen.17 Other workers have
argued that the sodium or lithium (or other metal)
counterion provides the conformational control by che-
lation to the nitrogen and oxygen.18 Still others have
suggested that hydrogen bonding between the NH proton
of primary or secondary amines or the amide NH proton
of N-acylated amines and the ketone oxygen provides the
needed conformational bias.19
The data from Table 1 provide additional insight into
the origins of diastereoselectivity in these reductions. In
the first place, it appears that the metal ions examined
do not bind sufficiently strongly to carbamate-protected
amino ketones to produce chelation control. Identical
stereoselection is found for borohydride regardless of
the counterion. It is expected that lithium or zinc ions
would provide increased conformational control and
hence higher diastereoselectivity. This is not the case;
thus, the counterion does not appear to be involved in
the diastereoselection.
(23) Brown, A. J .; Howarth, O. W.; Moore, P.; Parr, W. J . E. J . Chem.
Soc., Dalton Trans. 1978, 1776.
(24) Mathur, V.; Kanoongo, N.; Mathur, R.; Narang, C. K.; Mathur,
N. K. J . Chromatogr. A 1994, 685, 360.
(25) Pohlmeier, A.; Knoche, W. Int. J . Chem. Kinet. 1996, 28, 125.
(26) Delocalized amide anion complexes of aluminum are known to
be quite stable. Healy, M. D.; Power, M. B.; Barron, A. R. Coord. Chem.
Rev. 1994, 130(1-2), 63.
(27) Ishihara, K.; Funahashi, S.; Tanaka, M. Inorg. Chem. 1986,
25, 2898.

1050 J . Org. Chem., Vol. 67, No. 4, 2002
Hoffman et al.
Sch em e 6
F igu r e 2.
amounts of alcohol are converted to ethoxide by reaction
with borohydride and the ethoxide deprotonates the
carbamate to initiate complex formation and ultimately
chelate formation.
tures compared. From these structures, it is clear that
the trityl group not only blocks one face of the carbonyl
group itself, but it also plays a dominant role in deter-
mining conformational preferences of the other groups
in the molecule by forcing them toward the opposite face
of the carbonyl group. In essence, it causes both faces of
the carbonyl group to become sterically congested.
In the normal depiction of the Felkin-Anh transition
state, the large group is pictured as blocking one face of
the carbonyl group, whereas the opposite face is pictured
as relatively unencumbered. Stereoselection derives from
the difference between the incoming nucleophile eclipsing
a proton or the R1group at the R-position (Figure 1b).
The size difference between R1 and the proton is the basis
for the stereoselectivity. In addition, the size of R2 does
not exert a large influence on the stereoselectivity when
the R′-carbon is unbranched.29 It is usually assumed that
R1or R2 (when the R-carbon is unbranched) can rotate
out of the way of the approaching nucleophile.
In N-tritylated amino ketones, a much different picture
emerges. In 9c, which gives only modest stereoselectivity,
only two low energy conformers 9c-1 and 9c-2 have
accessible faces of the carbonyl group (Figure 2). The rest
(e.g., 9c-3) have the 3-phenylpropyl substituent wrapping
away from the tritylamino group and blocking the op-
posite face of the ketone. Thus, 9c-1 and 9c-2 appear to
be the reactive conformers and are similar in energy. As
a result of their similarity, there are only small differ-
ences in the nucleophilic approach to the carbonyl face,
and stereoselectivity is modest. However, since there are
many other low energy conformers of the type 9c-3
(which are all within 1-2 kcal of 9c-1), the concentration
of the reactive conformers is low and the rate of reduction
is relatively slow as well (8 h at -5 °C)
The choice of ethanol as a solvent for LiAlH(OtBu)3 is
unprecedented but effective. The ability of a hydroxylic
solvent to increase diastereoselectivity was clear from the
results with borohydride reductants. When it was found
that no stereocontrol was observed for LiAlH(O-t-Bu)3 in
THF, it was logical to try an alcohol solvent. There was
a good chance that the experiment would fail, since
experience suggested that reaction with the solvent to
produce hydrogen would be much faster than the reduc-
tion of the ketone, particularly at -78 °C. To our surprise
and delight, LiAlH(O-t-Bu)3 dissolves in ethanol at -78
°C without evidence of hydrogen evolution. Warming to
-50 °C causes gas evolution to begin. Nevertheless,
LiAlH(O-t-Bu)3 is sufficiently reactive at -78 °C to reduce
ketones, so that high yields of anti-amino alcohols are
obtained. The reaction appears to be broadly applicable
and thus provides a predictable and general means to
achieve chelation control. This reagent-solvent combina-
tion might be useful in other situations where ligand
exchange is needed prior to reduction.
This study also highlights the mixed blessing of the
trityl protecting group as a stereocontrol element. To
achieve Felkin-Anh stereocontrol, a large nitrogen pro-
tecting group is needed. Certainly the trityl group is
large. Moreover, it is much more easily attached and
removed than the N,N-dibenzyl protecting group, which
is presently the only alternative for enforcing Felkin-
Anh control. However, while the stereoselectivities shown
in Table 3 are usable, they are not as high as expected
on the basis of the bulk of the trityl group. In addition,
the reaction rate dropped significantly. This lowered
reactivity led to complications in the reduction of 9g
where reduction of the ester group competes with ketone
reduction to produce product mixtures.
When the R1 group is larger than methyl, steric
congestion about the carbonyl group increases dramati-
cally. For example, conformational analysis of 9a shows
that the lowest energy conformers (>10 within ∼1 kcal
of each other) all have the tritylamino group forming a
dihedral angle of about 47° with the carbonyl group
(instead of 90°). One phenyl ring of the trityl group is
parallel to the plane of the carbonyl and completely hides
that face of the carbonyl group. In addition, the low lying
To understand the origins of these trends, we carried
out a conformational search on amino ketones 9a ,c,g,h .28
The conformers were ranked by energy and their struc-
(28) Calculations were carried out using the Titan suite of programs
from Wavefunction. Conformational searching used the MMFF (Merck)
force field to identify the conformational minima which were then
ranked according to their energies. The equilibrium geometries of the
lowest several conformers were calculated by ab initio calculations at
the 6-31G* level. It was found that the confomational energy differences
between the low energy conformers were very similar for the two
methods. A solvation model was not included in the calculations.
(29) (a) Eliel. E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of
Organic Compounds; Wiley-Interscience: New York; 1994, p 876. (b)
Nogradi, M. Stereoselective Synthesis: A Practical Approach, 2nd ed.;
VCH Publishers: New York, 1995; p 84, Table 3-1; p 86, Table 3-2. (c)
Che´rest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968, 2199.

Syntheses of syn- and anti-1,2-Amino Alcohols
J . Org. Chem., Vol. 67, No. 4, 2002 1051
indicated, CDCl3 was used as the NMR solvent. Thin-layer
chromatography was performed on silica gel 60 F254 plates from
EM reagents and visualized by UV irradiation or by spraying
Ce(SO4)2 solution and heating on a hot plate. Flash chroma-
tography was performed using silica gel 60 (230-400 mesh).
Tetrahydrofuran was distilled from benzophenone ketyl. Other
solvents were HPLC grade and were used without further
purification. Starting materials were purchased from Acros,
Aldrich, or Novabiochem and used as received. Elemental
analyses were performed by MHW Laboratories. N-Tritylated
amino aicds were prepared from amino acids by a procedure
modified slightly from the standard preparation.30
F igu r e 3.
Gen er a l P r oced u r e for th e P r ep a r a tion Ca r ba m ic
N-P r otected r-Am in o-â-k eto Allyl Ester s (2).12a,13b CDI
(carbonyl-1,1′-diimidazole) (1.70 g, 10.5 mmol) was added at
room temperature to a stirred solution of a carbamic N-
protected R-amino acid (10 mmol) in dry THF (20 mL) under
a N2 atmosphere. The resulting solution was stirred for 1 h at
the same temperature and used for the next step without
further purification. Meanwhile, a solution of lithium enolate
of allyl acetate was made from BuLi (2.5 M, 14 mL, 35 mmol,
3 equiv are required), diisopropylamine (4.9 mL, 35 mmol),
and allyl acetate (3.8 mL, 35 mmol). The above imidazole
solution was added dropwise to the pale yellow solution of
lithium enolate at -78 °C under a nitrogen atmosphere. The
resulting mixture was stirred for 30 min and then quenched
at -78 °C with 10% citric acid and extracted with ethyl acetate
(3 × 50 mL). The organic extracts were combined, washed with
saturated bicarbonate (2 × 50 mL) and brine (50 mL), dried
(MgSO4), passed through a short pad of silica gel, and con-
centrated to provide the crude product, which could be purified
by chromatography or recrystallization: yield (90-94%).
conformers all have the R-benzyl group curving away
from the trityl such that its phenyl ring partially blocks
the opposite face of the carbonyl group. The octyl ketone
substituent is largely in the extended conformation. This
is depicted schematically in Figure 3. Thus, the “trityl
face” of the carbonyl group is more sterically blocked than
the “benzyl face”, and reduction occurs from the “benzyl
face” preferentially to give the syn isomer.
Even though the preferred product is syn, this is not
typical Felkin-Anh stereoselection because the stereo-
selectivity does not arise from differences between two
reactive conformers. Instead, the low-lying conformers
are all similar. Stereoselection results from steric differ-
ences between the two faces of the carbonyl group. In
the case of 9a , the benzyl face is only partially blocked,
and thus, there is good facial selectivity. However both
faces are sterically congested to some degree and this
results in a much slower rate of reduction (48 h).
Similar factors are operative in 9g and 9h . The low
energy conformers are similar with respect to the dihe-
dral angle between the tritylamino group and the car-
bonyl group. Thus, the choice is not between different
Felkin-Anh conformers; rather, the choice is between the
two faces of the carbonyl group. The trityl group provides
steric congestion on one face and the R-substituent bends
away from the trityl group and thus provides steric
congestion to the other face. It is the relative steric
crowding on each face which determines the stereoselec-
tivity. Moreover, as both faces are fairly hindered, the
overall reduction rate decreases significantly.
This analysis suggests that the trityl protecting group
is actually too large for efficient Felkin-Anh control. Its
large size restricts rotation about the bond between the
carbonyl group and the R-carbon and restricts the con-
formational mobility of the remaining ketone substitu-
ents in the molecule forcing them to bend toward the
carbonyl group. This results in steric crowding at both
faces of the carbonyl group. Stereoselection is dependent
on the relative crowding at each face, which may account
for the unexpectedly low stereoselectivities found for 9h
and the competitive ester reduction in 9g (Table 3).
We are evaluating other nitrogen protecting groups
smaller than trityl that may give Felkin-Anh stereo-
selection, but with higher reactivity and stereoselectivity.
In the meantime, the N-trityl protecting group can be
used as an effective stereocontrol element to give syn-
amino alcohols in high yields and with very good stereo-
control for substrates without other reducible groups.
Much longer reaction times are required, however.
4(S)-Ben zyloxycar bon ylam in o-3-oxo-5-ph en ylpen tan o-
ic Acid Allyl Ester (2a ). Compound 2a (3.50 g, 92%) was
obtained as a colorless solid after purification on silica gel
eluting with 90:10 hexanes/EtOAc. Pure material could be
obtained by recrystallization from hexanes: mp 57 °C; [R]21
D
+30.4 (c 1.00, CH2Cl2); FTIR (KBr) 1755, 1730, 1701, 1536,
cm-1; 1H NMR δ 3.00 (dd, 1H, J ) 7.4, 14.2 Hz), 3.15 (dd, 1H,
J ) 6.1, 14.1 Hz), 3.48 (d, 1H, J ) 17.2 Hz), 3.52 (d, 1H, J )
16.2 Hz), 4.60 (d, 2H, J ) 5.8 Hz), 4.62 (m, 1H), 5.04 (s, 2H),
5.27 (m, 3H), 5.87 (m, 1H), 7.12-7.32 (m, 10H); 13C NMR δ
37.1, 46.9, 60.9, 66.1, 67.2, 118.9, 127.2, 128.1, 128.2, 128.6,
128.8, 129.3, 135.8, 155.8, 166.4, 201.2.
4(S)-ter t-Bu toxycar bon ylam in o-3-oxo-5-ph en ylpen tan o-
ic Acid Allyl Ester (2b). Compound 2b (3.10 g, 89%) was
obtained as a colorless oil after purification on silica gel eluting
with 90:10 hexanes/EtOAc: [R]21D +10.5 (c 1.00, CH2Cl2); FTIR
1
(neat) 1751, 1715, 1691, 1526 cm-1; H NMR δ 1.40 (s, 9H),
3.15 (dd, 1H, J ) 5.9, 14.2 Hz), 2.99 (dd, 1H, J ) 7.4, 14.2
Hz), 3.48 (d, 1H, J ) 16.2 Hz), 3.54 (d, 1H, J ) 16.1 Hz), 4.54
(m, 1H), 4.61 (d, 2H, J ) 5.6 Hz), 5.00 (d, 1H, J ) 7.0 Hz),
5.28 (m, 2H), 5.90 (m, 1H), 7.15-7.28 (m, 5H); 13C NMR δ 28.3,
37.1, 46.8, 60.6, 66.1, 80.4, 118.9, 127.1, 128.8, 129.4, 131.7,
136.2, 155.3, 166.6, 201.7.
4(S)-ter t-Bu toxyca r bon yla m in o-3-oxop en ta n oic Acid
Allyl Ester (2c). Compound 2c (2.43 g, 90%) as a colorless
oil after purification on silica gel eluting with 90:10 hexanes/
EtOAc: [R]21D -12.0 (c 1.00, CH2Cl2); FTIR (neat) 1755, 1735,
1
1701, 1526 cm-1; H NMR δ 1.36 (d, 3H, J ) 7.2 Hz), 1.45 (s,
1H), 3.59 (d, 1H, J ) 16.7 Hz), 3.63 (d, 1H, J ) 17.3 Hz), 4.37
(m, 1H), 4.64 (d, 1H, J ) 5.8 Hz), 5.29 (m, 3H), 5.90 (m, 1H);
13C NMR δ 16.9, 28.3, 45.7, 55.4, 66.0, 80.0, 118.5, 131.5, 155.2,
166.6, 202.3.
8-Ben zyloxyca r bon yla m in o-4(S)-ter t-bu toxyca r bon yl-
a m in o-3-oxoocta n oic Acid Allyl Ester (2d ). Compound 2c
(4.06 g, 88%) was obtained as a colorless oil after purification
on silica gel eluting with 85:15 hexanes/EtOAc: [R]21D -3.6 (c
1.00, CH2Cl2); FTIR (neat) 1756, 1716, 1526 cm-1; 1H NMR δ
1.42 (broad, 12H), 1.48 (m, 2H), 1.79 (m, 1H), 3.17 (m, 2H),
3.57 (s, 2H), 4.29 (m, 1H), 4.62 (d, 1H, J ) 5.8 Hz), 5.09 (broad,
3H), 5.27 (m, 3H), 7.19-7.34 (m, 5H); 13C NMR δ 22.3, 28.3,
29.4, 30.3, 40.4, 46.0, 59.6, 66.0, 66.6, 80.1, 118.8, 128.0, 128.5,
131.6, 136.7, 156.6, 166.6.
Exp er im en ta l Section
Infrared spectra were taken as a KBr pellets, as solution in
CHCl3, or as neat liquids. 1H NMR and 13C NMR spectra were
recorded at 200 and 50 MHz, respectively. Unless otherwise

1052 J . Org. Chem., Vol. 67, No. 4, 2002
Hoffman et al.
4(S)-Ben zyloxycar bon ylam in o-6-m eth yl-3-oxoh eptan o-
ic Acid Allyl E st er (2e). Compound 2e (3.10 g, 89%) was
obtained as a colorless oil after purification on silica gel eluting
with 90:10 hexanes/EtOAc: [R]22D -3.6 (c 1.00, CH2Cl2); FTIR
being quenched with 10 mL of 10% citric acid, the reaction
mixture was extracted with ethyl acetate (3 × 50 mL). The
organic extracts were combined, washed with saturated bi-
carbonate and brine, dried (MgSO4), passed through a short
pad of silica gel, and concentrated to provide a pale yellow oil.
Without further purification, this oil was dissolved in THF (20
mL) and added dropwise to a stirred solution of palladium
acetate (14.3 mg, 0.064 mmol), PPh3 (33.8 mg, 0.13 mmol),
formic acid (0.2 mL, 5 mmol), and Et3N (0.9 mL, 6.5 mmol) in
dry THF at room temperature under nitrogen. The mixture
was stirred overnight. After the filtrate was concentrated, the
residue was chromatographed on SiO2 with hexanes-ethyl
acetate (90:10) to afford the ketone in 71-76% yield.
1
(neat) 1761, 1731, 1531 cm-1; H NMR δ 0.89 (d, 3H, J ) 5.8
Hz), 0.92 (d, 3H, J ) 5.1 Hz), 1.42 (m, 1H), 1.60 (m, 2H), 3.53
(d, 1H, J ) 16.2 Hz), 3.57 (d, 1H, J ) 15.7 Hz), 4.40 (m, 1H),
4.57 (d, 2H, J ) 5.7 Hz), 5.06 (s, 2H), 5.25 (m, 2H), 5.36 (d,
1H, J ) 8.5 Hz), 5.84 (m, 1H), 7.18-7.34 (m, 5H); 13C NMR δ
21.5, 23.2, 24.8, 39.8, 46.2, 58.7, 66.0, 67.1, 118.8, 128.0, 128.2,
128.5, 131.5, 136.2, 156.1, 166.5, 202.1.
4(S)-(9H -F lu or en -9-ylm et h oxyca r b on yla m in o)-3-oxo-
5-p h en ylp en ta n oic Acid Allyl Ester (2f). Compound 2f
(4.28 g, 91%) was obtained as a colorless solid after purification
on silica gel eluting with 85:15 hexanes/EtOAc. Pure material
could be obtained by recrystallization from EtOH: mp 103 °C;
[R]20D -2.0 (c 1.00, CH2Cl2); FTIR (KBr) 1745, 1720, 1686, 1541
cm-1; 1H NMR δ 3.00 (dd, 1H, J ) 6.1, 14.1 Hz), 3.16 (dd, 1H,
J ) 7.2, 14.1 Hz), 3.45 (d, 1H, J ) 16.1 Hz), 3.49 (d, 1H, J )
16.2 Hz), 4.15 (t, 1H, J ) 6.6 Hz), 4.38 (d, 1H, J ) 2.6 Hz),
4.41 (d, 1H, J ) 3.4 Hz), 4.60 (d, 2H, J ) 5.6 Hz), 4.65 (m,
1H), 5.26 (m, 3H), 5.88 (m, 1H), 7.12-7.77 (m, 13H); 13C NMR
δ 36.9, 46.8, 47.3, 60.9, 66.1, 67.0, 118.9, 120.0, 125.0, 127.2,
128.8, 129.3, 131.5, 135.9, 141.4, 143.7, 155.8, 166.3, 201.1.
4(S)-ter t-Bu toxycar bon ylam in o-3-oxoh eptan edioic Acid
1-Allyl Ester 7-Meth yl Ester (2g). Compound 2g (3.00 g,
87%) was obtained as a colorless oil after purification on silica
gel eluting with 90:10 hexanes/EtOAc: [R]20D -8.4 (c 1.00, CH2-
Cl2); FTIR (neat) 1751, 1726 cm-1; 1H NMR δ 1.44 (s, 9H), 1.86
(m, 1H), 2.18 (s, 1H), 2.41 (m, 2H), 3.65 (d, 2H, J ) 10.1 Hz),
3.67 (s, 3H), 4.40 (m, 1H), 4.64 (d, 2H, J ) 5.8 Hz), 5.29 (m,
3H), 5.90 (m, 1H); 13C NMR δ 25.9, 28.3, 29.8, 46.1, 51.9, 59.1,
66.1, 80.4, 119.0, 131.6, 155.5, 166.6, 173.4, 201.5.
(1S-Ben zyl-2-oxod ecyl)ca r ba m ic a cid ben zyl ester (1a )
was prepared by method A. Alkylation of 993 mg (2.60 mmol)
of 2a with 1.85 g (2.86 mmol, 1.1 equiv) of heptyl triflate
afforded 770 mg (75%) of 1a as a colorless solid after purifica-
tion on silica gel eluting with 90:10 hexanes/EtOAc. Pure
material could be obtained by recrystallization from hexanes:
mp 52 °C; [R]21D +4.5 (c 1.00, CH2Cl2); FTIR (KBr) 1716, 1692
1
cm-1; H NMR δ 0.88 (t, 3H, J ) 6.5 Hz), 1.23 (m, 10H), 1.49
(m, 2H), 2.36 (m, 2H), 3.03 (t, 2H, J ) 6.8 Hz), 4.61 (q, 1H, J
) 7.2 Hz), 5.07 (s, 2H), 5.42 (d, 1H, J ) 7.9 Hz), 7.22-7.31
(m, 10H); 13C NMR δ 14.1, 22.7, 23.3, 29.1, 29.3, 31.8, 38.0,
40.9, 60.4, 66.9, 76.4, 127.1, 128.1, 128.2, 128.6, 128.7, 129.3,
135.9, 155.7, 208.8. Anal. Calcd for C25H33NO3: C, 75.91; H,
8.41; N, 3.54. Found: C, 75.95; H, 8.35; N, 3.64.
(1S-Ben zyl-2-oxo-4-p h en ylb u t yl)ca r b a m ic a cid 2,2-
d im eth ylp r op yl ester (1b) was prepared by method B.
Alkylation of 903 mg (2.60 mmol) of 2b with 489 mg (2.86
mmol, 1.1 equiv) of benzyl bromide afforded 703 mg (76%) of
1b as a colorless solid after purification on silica gel eluting
with 90:10 hexanes/EtOAc. Pure material could be obtained
5-Ben zyloxy-4S-ter t-bu toxyca r bon yla m in o-3-oxop en -
ta n oic Acid Allyl Ester (2h ). Compound 2h (3.56 g, 87%)
was obtained as a colorless oil after purification on silica gel
by recrystallization from hexanes: mp 84 °C; [R]23 +3.4 (c
D
1.00, CH2Cl2); FTIR (KBr) 1688, 1516 cm-1; H NMR δ 1.40
1
(s, 9H), 2.69 (m, 2H), 2.84 (m, 2H), 2.96 (t, 2H, J ) 6.5 Hz),
4.51 (m, 1H), 5.08 (d, 1H, J ) 7.3 Hz), 7.09-7.27 (m, 10H);
13C NMR δ 28.3, 29.4, 37.9, 42.5, 60.2, 80.0, 126.3, 127.0, 128.4,
128.6, 128.7, 129.3, 136.2, 140.8, 155.2, 208.3. Anal. Calcd for
eluting with 85:15 hexanes/EtOAc: [R]21 +6.1 (c 1.00, CH2-
D
1
Cl2); FTIR (neat) 1755, 1735, 1516 cm-1; H NMR δ 3.58 (s,
2H), 3.66 (dd, 1H, 4.1, 9.9 Hz), 3.91 (dd, 1H, 3.1, 9.6 Hz), 4.47
(s, 2H), 4.57 (d, 2H, J ) 5.9 Hz), 4.58 (m, 1H), 5.10 (s, 2H),
5.25 (m, 2H), 5.83 (m, 2H); 13C NMR δ 46.6, 60.3, 66.0, 67.2,
69.2, 73.5, 118.8, 127.7, 128.1, 128.2, 128.5, 131.6, 136.2, 137.1,
155.9, 166.2, 200.0.
C
22H27NO3: C, 75.76; H, 7.70; N, 3.96. Found: C, 75.44; H,
7.70; N, 3.99.
(1S-1-Meth yl-2-oxo-5-p h en ylp en tyl)ca r ba m ic a cid 2,2-
d im eth ylp r op yl ester (1c) was prepared by method A.
Alkylation of 705 mg (2.60 mmol) of 2c with 727 mg (2.86
mmol, 1.1 equiv) of phenethyl triflate afforded 553 mg (73%)
of 1c as a colorless solid after purification on silica gel eluting
with 90:10 hexanes/EtOAc. Pure material could be obtained
Gen er a l P r oced u r e for th e P r ep a r a tion of th e Ca r -
ba m ic N-P r otected r-Am in o Keton es (1). Meth od A. With
a n Alk yl Tr ifla te a s a n Alk yla tin g Agen t. A solution of
carbamic N-protected R-amino-â-keto allyl ester 2 (2.6 mmol)
in dry THF (10 mL) was added dropwise to a stirred suspen-
sion of NaH (125 mg of 60% in oil, 3.1 mmol, 1.2 equiv) in dry
THF (10 mL) at -20 °C under nitrogen. The mixture was
stirred for 20 min. Then the alkyl triflate (2.86 mmol, 1.1
equiv) in 5 mL of THF was added. The resulting solution was
allowed to warm to room temperature and stirred for 3 h. After
being quenched with 10 mL 10% citric acid, the reaction
mixture was then extracted with ethyl acetate (3 × 50 mL).
The organic extracts were combined, washed with saturated
bicarbonate and brine, dried (MgSO4), passed through a short
pad of silica gel, and concentrated to provide a pale yellow oil.
Without further purification, this oil was dissolved in THF (20
mL) and added dropwise to a stirred solution of palladium
acetate (14.3 mg, 0.064 mmol), PPh3 (33.8 mg, 0.13 mmol),
formic acid (0.2 mL, 5 mmol), and Et3N (0.9 mL, 6.5 mmol) in
dry THF at room temperature under nitrogen. The mixture
was stirred overnight. After the filtrate was concentrated, the
residue was chromatographed on SiO2 with hexanes-ethyl
acetate (90:10) to afford the amino ketone in 65-78% yield.
Meth od B. With a n Alk yl Br om id e a s a n Alk yla tin g
Agen t. A solution of carbamic N-protected R-amino-â-keto allyl
ester 2 (2.60 mmol) in dry THF (10 mL) was added dropwise
to a stirred suspension of NaH (125 mg of 60% in oil, 3.10
mmol, 1.2 equiv) in dry THF (10 mL) at -20 °C under nitrogen.
The mixture was stirred for 20 min, and then the alkyl bromide
(2.60 mmol) in 5 mL of THF and 10% NaI were added. The
resulting solution was allowed to warm to room temperature
and refluxed for 2 h in order to complete the reaction. After
by recrystallization from hexanes: mp 46 °C; [R]23 +1.1 (c
D
1.00, CH2Cl2); FTIR (KBr) 3332, 2984, 1710, 1684, 1528, 1452,
1
1368, 1252 cm-1; H NMR δ 1.28 (d, 3H, J ) 7.1 Hz), 1.43 (s,
9H), 1.94 (q, 2H, J ) 7.4 Hz), 2.50 (m, 2H), 2.62 (t, 2H, J )
7.2 Hz), 4.27 (q, 1H, J ) 7.1 Hz), 5.25 (broad, 1H), 7.14-7.27
(m, 5H); 13C NMR δ 17.9, 25.0, 28.4, 35.1, 38.3, 55.2, 79.8,
126.1, 126.4, 141.4, 155.2, 209.5. Anal. Calcd for C17H25NO3:
C, 70.07; H, 8.65; N, 4.81. Found: C, 70.22; H, 8.53; N, 4.88.
(5S-ter t-Bu t oxyca r b on yla m in o-6-oxo-8-p h en yloct yl)-
ca r ba m ic a cid ben zyl ester (1d ) was prepared by method
B. Alkylation of 1000 mg (2.10 mmol) of 2d with 360 mg (2.1
mmol) of benzyl bromide afforded 718 mg (73%) of 1d as a
colorless solid after purification on silica gel eluting with 80:20
hexanes/EtOAc: mp 76 °C; [R]20D +13.4 (c 1.00, CH2Cl2); FTIR
(KBr) 1726, 1696, cm-1; 1H NMR δ 1.42 (broad, 13H), 1.72 (m,
2H), 2.85 (m, 4H), 3.13 (m, 2H), 4.25 (m, 1H), 4.80 (broad, 1H),
5.09 (s, 2H), 5.22 (d, 1H, J ) 7.3 Hz), 7.19-7.34 (m, 10H); 13
C
NMR δ 22.1, 28.3, 29.5, 30.9, 40.5, 41.2, 59.2, 66.6, 79.8, 126.2,
128.0, 128.3, 128.5, 136.7, 140.8, 155.6, 156.5, 208.4. Anal.
Calcd for C27H36N2O5: C, 69.21; H, 7.74; N, 5.98. Found: C,
69.44; H, 7.60; N, 5.99.
(1S-Isobu tyl-2-oxo-5-p h en ylp en tyl-ca r ba m ic a cid ben -
zyl ester (1e) was prepared by method A. Alkylation of 972
mg (2.80 mmol) of 2e with 711 mg (2.80 mmol) of phenethyl
triflate afforded 700 mg (68%) of 1e as a colorless oil after
purification on silica gel eluting with 90:10 hexanes/EtOAc.
[R]22D +9.2 (c 1.00, CH2Cl2); FTIR (CHCl3) 1721 cm-1; 1H NMR
δ 0.89 (d, 3H, J ) 6.6 Hz), 0.95 (d, 3H, J ) 6.2 Hz), 1.34 (m,

Syntheses of syn- and anti-1,2-Amino Alcohols
J . Org. Chem., Vol. 67, No. 4, 2002 1053
2H), 1.67 (m, 1H), 1.92 (m, 2H), 2.51 (m, 2H), 2.61 (t, 2H, J )
7.4 Hz), 4.38 (m, 1H), 5.09 (s, 2H), 5.28 (d, 1H, J ) 8.0 Hz),
7.18-7.34 (m, 10H); 13C NMR δ 21.7, 23.3, 24.9, 35.0, 38.9,
40.7, 58.3, 66.9, 126.0, 128.1, 128.5, 129.0, 141.4, 156.1, 209.6.
(1S-Ben zyl-2-oxod ecyl)ca r ba m ic a cid 9H-flu or en -9-yl-
m eth yl ester (1f) was prepared by method A. Alkylation of
1173 mg (2.50 mmol) of 2f with 620 g (2.50 mmol) of heptyl
triflate afforded 785 mg (65%) of 1f as a colorless solid after
purification on silica gel eluting with 85:15 hexanes/EtOAc.
Pure material could be obtained by recrystallization from
hexanes: mp 104 °C; [R]20D +35.0 (c 1.00, CH2Cl2); FTIR (KBr)
method A afforded 98 mg (98%) of crude anti-3a as a colorless
solid that could be purified by recrystallization from hexanes
with 90% recovery: mp 142 °C; [R]23 -21.4 (c 1.00, CH2Cl2);
D
FTIR (KBr) 3324, 1692 cm-1; H NMR δ 0.86 (t, 3H, J ) 6.5
1
Hz), 1.27 (m, 11H), 1.49 (m, 3H), 2.00 (s, 1H), 2.75 (dd, 1H, J
) 4.6, 14.1 Hz), 2.92 (dd, 1H, J ) 9.4, 14.1 Hz), 3.70 (m, 1H),
3.90 (m, 1H), 4.85 (d, 1H, J ) 6.2 Hz), 5.00 (s, 1H), 7.20-7.31
(m, 10H); 13C NMR δ 14.4, 23.0, 26.4, 29.6, 29.9, 30.0, 32.2,
33.9, 35.7, 57.4, 67.1, 74.3, 126.8, 128.3, 128.8, 128.9, 129.6,
136.8, 138.4, 156.9. Anal. Calcd for C25H35NO3: C, 75.53; H,
8.87; N, 3.52. Found: C, 75.65; H, 8.80; N, 3.58.
1725, 1691 cm-1
;
1H NMR δ 0.88 (t, 3H, J ) 6.6 Hz), 1.23
(1S-Ben zyl-2R-h yd r oxy-4-p h en ylbu tyl)ca r ba m ic Acid
2,2-Dim eth ylp r op yl Ester (a n ti-3b). Reduction of 1b (90
mg, 0.25 mmol) using method A afforded 84 (93%) mg of crude
anti-3b as a colorless solid that could be purified by recrys-
(broad, 10H), 1.53 (m, 2H), 2.37 (m, 2H), 3.04 (m, 2H), 4.18
(m, 1H), 4.38 (m, 2H), 4.62 (m, 1H), 5.43 (d, 1H, J ) 7.6 Hz),
7.12-7.77 (m, 13H); 13C NMR δ 14.1, 22.7, 23.4, 29.1, 29.3,
31.7, 38.0, 40.9, 47.3, 60.4, 67.0, 120.1, 125.1, 127.1, 127.9,
128.7, 129.3, 136.0, 141.4, 143.9, 155.7, 208.8. Anal. Calcd for
tallization from ethyl acetate with 86% recovery: mp 163 °C;
1
[R]23 -1.4 (c 1.00, CH2Cl2); FTIR (KBr) 3360, 1688 cm-1; H
D
C
32H37NO3: C, 79.47; H, 7.71; N, 2.90. Found: C, 79.60; H,
NMR δ 1.35 (s, 9H), 1.83 (m, 2H), 2.68 (m, 2H), 2.91 (m, 2H),
3.72 (m, 1H), 3.84 (m, 1H), 4.60 (d, 1H, J ) 8.1 Hz), 7.15-
7.30 (m, 10H); 13C NMR δ 28.3, 32.4, 35.2, 36.9, 57.1, 73.5,
79.8, 126.0, 126.4, 128.6, 129.3, 138.1, 142.0, 156.4. Anal. Calcd
for C22H29NO3: C, 74.33; H, 8.22; N, 3.94. Found: C, 74.19;
H, 8.03; N, 3.99.
7.63; N, 3.04.
4S-ter t-Bu toxycar bon ylam in o-9-m eth yl-5-oxodec-8-en o-
ic a cid m eth yl ester (1g) was prepared by method B.
Alkylation of 1063 mg (3.1 mmol) of 2g with 462 mg (3.1 mmol)
of prenyl bromide afforded 720 mg (71%) of 1g as a colorless
oil after purification on silica gel eluting with 90:10 hexanes/
EtOAc: [R]20D +29.7 (c 1.00, CH2Cl2); FTIR (CHCl3) 3435, 3026,
2986, 2927, 1741, 1710, 1506, 1451, 1381, 1182 cm-1; 1H NMR
δ 1.43 (s, 9H), 1.61 (s, 3H), 1.67 (s, 3H), 1.73 (m, 2H), 2.29 (m,
2H), 2.31 (m, 2H), 2.54 (m, 2H), 3.68 (s, 3H), 4.35 (m, 1H),
5.04 (t, 1H, J ) 7.1 Hz), 5.25 (d, 1H, J ) 8.0 Hz); 13C NMR δ
17.7, 22.3, 25.6, 26.8, 28.3, 29.7, 39.8, 51.7, 58.6, 79.9, 122.4,
133.1, 155.5, 173.3, 208.5.
(2R-Hydr oxy-1S-m eth yl-5-ph en ylpen tyl)car bam ic Acid
2,2-Dim eth ylp r op yl Ester (a n ti-3c). Reduction of 1c (73 mg,
0.25 mmol) using method A afforded 71 mg (97%) of crude anti-
3c as a colorless solid that could be purified by recrystallization
from hexanes with 92% recovery: mp 79 °C; [R]23D -8.2 (c 1.00,
CH2Cl2); FTIR (KBr) 3356, 1684 cm-1; 1H NMR δ 1.05 (d, 3H,
J ) 6.9 Hz), 1.43 (s, 9H), 1.64 (m, 2H), 1.82 (m, 2H), 2.25 (s,
1H), 2.64 (t, 2H, J ) 7.6 Hz), 3.64 (m, 2H), 4.77 (broad, 1H),
7.15-7.27 (m, 5H); 13C NMR δ 14.5, 27.8, 28.5, 32.8, 35.8, 50.8,
74.4, 79.6, 125.8, 128.4, 128.5, 142.2, 155.9. Anal. Calcd for
(1S-Ben zyloxym eth yl-2-oxod ecyl)ca r ba m ic a cid ben -
zyl ester (1h ) was prepared by method A. Alkylation of 1070
mg (2.60 mmol) of 2h with 1.85 g (2.86 mmol, 1.1 equiv) of
heptyl triflate afforded 818 mg (74%) of 1h as a colorless oil
after purification on silica gel eluting with 90:10 hexanes/
C
17H27NO3: C, 69.59; H, 9.28; N, 4.77. Found: C, 69.72; H,
9.15; N, 4.86.
(5S-ter t-Bu toxyca r bon yla m in o-6R-h yd r oxy-8-p h en yl-
octyl)ca r ba m ic Acid Ben zyl Ester (a n ti-3d ). Reduction of
1d (130 mg, 0.28 mmol) using method A afforded 127 mg (97%)
of crude anti-3d as a colorless solid that could be purified by
recrystallization from EtOH with 85% recovery: mp 113 °C;
[R]20D -15.8 (c 1.00, CH2Cl2); FTIR (KBr) 3366, 1696 cm-1; 1H
NMR δ 1.42 (broad, 15H), 1.70 (m, 2H), 2.48 (broad, 1H), 2.69
(m, 2H), 2.84 (m, 1H), 3.16 (m, 2H), 3.61 (m, 2H), 4.76 (broad,
1H), 4.89 (broad, 1H), 5.09 (s, 2H), 7.21-7.34 (m, 10H); 13C
NMR δ 23.1, 28.4, 29.1, 29.8, 32.4, 35.0, 40.7, 55.4, 66.7, 74.1,
79.7, 118.6, 125.9, 127.6, 128.0, 128.4, 136.6, 142.0, 156.6.
(2R-Hydr oxy-1S-isobu tyl-5-ph en ylpen tyl)car bam ic Acid
Ben zyl Ester (a n ti-3e). Reduction of 1e (200 mg, 0.56 mmol)
using method A afforded 159 mg (97%) of crude anti-3e as a
colorless solid that could be purified by recrystallization from
EtOAc: [R]21 +5.01 (c 1.00, CH2Cl2); FTIR (KBr) 1785, 1730,
D
1666 cm-1; 1H NMR δ 0.88 (t, 3H, J ) 6.0 Hz), 1.24 (m, 10H),
1.53 (m, 2H), 2.45 (m, 2H), 3.69 (dd, 1H, 3.3, 9.8 Hz), 3.90 (dd,
1H, 4.0, 9.9 Hz), 4.45 (m, 3H), 5.10 (s, 2H), 5.80 (d, 1H, J )
7.9 Hz), 7.25-7.31 (m, 10H); 13C NMR δ 14.1, 22.7, 23.4, 29.2,
29.5, 29.7, 31.8, 39.8, 60.3, 67.0, 69.7, 73.3, 127.4, 127.6, 128.1,
128.2, 128.3, 128.5, 128.6, 137.4, 156.0, 207.1. These data are
consistent with literature data.12b
Gen er a l P r oced u r e for th e P r ep a r a tion of th e Ca r -
bam ic N-P r otected a n ti-r-Am in o Alcoh ols (3a-h ). Meth od
A. LiAlH(O-t-Bu)3 (127 mg, 0.50 mmol) was dissolved in EtOH
(3 mL) at -78 °C under nitrogen, and then a solution of the
ketone 1a -h (0.25 mmol) in EtOH (4 mL) was added dropwise.
After 2 h, the solution was quenched with 10% citric acid (2
mL), extracted with ethyl acetate (2 × 50 mL), washed with
H2O (20 mL) and brine (20 mL), dried (MgSO4), and concen-
trated to provide a white solid that could be recrystallized from
an appropriate solvent.
Met h od B. NaBH4 (19 mg, 0.50 mmol) was dissolved in
EtOH (3 mL) at -78 °C under nitrogen, and a solution of the
ketone 1a (0.25 mmol) in EtOH (4 mL) was added dropwise.
After 2 h, the solution was quenched with 10% citric acid (2
mL), extracted with ethyl acetate (2 × 50 mL), washed with
H2O (20 mL) and brine (20 mL), dried (MgSO4), and concen-
trated to provide a white solid, which could be recrystallized
from an appropriate solvent. Other borohydride reagents were
used in place of sodium borohydride to generate the data in
Table 1.
Meth od C. A Selectride solution in THF (1 M, 0.5 mL) was
cooled at -78 °C under nitrogen, and then a solution of the
ketone 1a (100 mg, 0.25 mmol) in THF was added dropwise.
After 2 h, the solution was quenched with 10% citric acid (2
mL), extracted with ethyl acetate (2 × 50 mL), washed with
H2O (20 mL) and brine (20 mL), dried (MgSO4), passed through
a short pad of silica gel, and concentrated to provide the crude
product, which coul be recrystallized from an appropriate
solvent.
hexanes with 87% recovery: mp 108 °C; [R]20 -35.6 (c 1.00,
D
CH2Cl2); FTIR (KBr) 3325, 1696 cm-1; 1H NMR δ 0.90 (d, 6H,
J ) 6.4 Hz), 1.23 (m, 2H), 1.40 (m, 2H), 1.61 (m, 2H), 1.84 (m,
1H), 2.25 (broad, 1H), 2.63 (t, 2H, J ) 7.7 Hz), 3.67 (m, 2H),
4.86 (d, 1H, J ) 8.2 Hz), 5.08 (s, 2H), 7.14-7.33 (m, 10H); 13
C
NMR δ 21.6, 23.7, 24.7, 27.8, 32.5, 35.8, 38.1, 54.0, 66.9, 74.6,
125.8, 128.1, 128,4, 128.5, 136.5, 142.2, 156.8. Anal. Calcd for
C
23H31NO3: C, 74.76; H, 8.46; N, 3.79. Found: C, 74.61; H,
8.35; N, 3.89.
(1S-Ben zyl-2R-h yd r oxyd ecyl)ca r ba m ic Acid 9H-F lu o-
r en -9-ylm eth yl Ester (a n ti-3f). Reduction of 1f (100 mg, 0.20
mmol) using method A afforded 95 mg (95%) of crude anti-3f
as a colorless solid that could be purified by recrystallization
from EtOH with 85% recovery: mp 172 °C; [R]20D -20.8 (c 1.00,
1
CH2Cl2); FTIR (KBr) 3335, 1696 cm-1; H NMR (DMSO-d6) δ
1.23 (t, 3H, J ) 6.2 Hz), 1.62 (broad, 12H), 1.85 (m, 2H), 2.90
(m, 2H), 2.96 (m 1H), 3.42 (dd, 1H, J ) 13.5, 2.3 Hz), 3.83 (m,
1H), 4.51 (m, 3H), 5.09 (d, 1H, J ) 6.3 Hz), 7.55-8.30 (m, 13H);
13C NMR (DMSO-d6) δ 14.1, 22.2, 25.7, 29.3, 31.4, 32.1, 33.5,
35.9, 57.0, 57.6, 73.1, 73.5, 109.9, 120.2, 121.6, 125.9, 127.5,
128.2, 129.2, 135.5, 140.5, 157.1. Anal. Calcd for C32H39NO3:
C, 79.14; H, 8.09; N, 2.88. Found: C, 78.97; H, 7.95; N, 2.92.
4S-ter t-Bu toxycar bon ylam in o-5R-h ydr oxy-9-m eth yldec-
8-en oic Acid Meth yl Ester (a n ti-3g). Reduction of 1g (100
mg, 0.30 mmol) using method A afforded 96 mg (96%) of crude
(1S-Ben zyl-2R-h yd r oxyd ecyl)ca r ba m ic Acid Ben zyl
Ester (a n ti-3a ). Reduction of 1a (100 mg, 0.25 mmol) using

1054 J . Org. Chem., Vol. 67, No. 4, 2002
Hoffman et al.
anti-3g as a colorless solid that could be purified by recrys-
purification on silica gel eluting with 95:5 hexanes/EtOAc.
Pure material could be obtained by recrystallization from
tallization from EtOH with 85% recovery: mp 68 °C; [R]20
D
-17.7 (c 1.00, CH2Cl2); FTIR (KBr) 3366, 1756, 1691 cm-1; 1H
NMR δ 1.44 (broad, 11H), 1.62 (s, 3H), 1.69 (s, 3H), 1.94 (m,
2H), 2.19 (m, 2H), 2.41 (t, 2H, J ) 7.3 Hz), 3.60 (m, 1H), 3.68
(s, 3H), 4.80 (d, 1H, J ) 8.8 Hz), 5.12 (t, 1H, J ) 7.1 Hz); 13C
NMR δ 17.8, 24.6, 25.7, 28.4, 30.9, 33.5, 51.7, 54.9, 74.4, 79.6,
82.0, 123.8, 132.6, 156.2, 174.1.
hexanes with 84% recovery: mp 109 °C; [R]21 +47.2 (c 1.00,
D
CH2Cl2); FTIR (KBr) 1730, 1606 cm-1; 1H NMR δ 2.49 (d, 1H,
J ) 16.3 Hz), 2.61 (d, 1H, J ) 16.3 Hz), 2.87 (t, 2H, J ) 6.6
Hz), 3.05 (broad, 1H), 3.71 (m, 1H), 4.48 (d, 2H, J ) 5.7 Hz),
5.21 (m, 2H), 5.79 (m, 1H), 7.10-7.40 (m, 20H); 13C NMR δ
40.6, 46.8, 63.9, 65.4, 71.2, 117.4, 126.6, 126.9, 128.0, 128.9,
130.0, 131.7, 136.9, 146.0, 166.4, 205.6. Anal. Calcd for C33H31
-
(1S-Ben zyloxym eth yl-2R-h yd r oxyd ecyl)ca r ba m ic Acid
Ben zyl Ester (a n ti-3h ). Reduction of (106 mg, 0.25 mmol)
of 1h using method A afforded 106 mg (99%) of crude anti-3h
as a colorless solid that could be purified by recrystallization
NO3: C, 80.95; H, 6.38; N, 2.86. Found: C, 81.03; H, 6.15; N,
2.95.These data are consistent with the literature data for
8a .12e
from ethyl aceate with 86% recovery: mp 142 °C; [R]23 -21.4
3-Oxo-4S-(tr ityla m in o)p en ta n oic Acid Allyl Ester (8c).
Alanine (0.89 g, 10 mmol) was treated with Me3SiCl (1.086 g,
10 mmol), trityl chloride (2.78 g, 10 mmol), and Et3N (2.79
mL, 20.0 mmol) to give N-tritylalanine. After workup (general
procedure), the crude product was sequentially converted with
CDI (2.79 mL, 20.0 mmol) and then with 2 equiv of lithium
enolate of allyl acetate to afford 3.05 g (74% overall yield) of
8c as a colorless oil after purification on silica gel eluting with
D
(c 1.00, CH2Cl2); FTIR (KBr) 3335, 1700 cm-1; 1H NMR δ 0.88
(t, 3H, J ) 6.8 Hz), 1.26 (broad, 12H), 1.44 (m, 2H), 2.70 (d,
1H, J ) 8.4 Hz), 3.67 (m, 3H), 3.79 (m, 1H), 4.46 (d, 1H, 11.1
Hz), 4.52 (d, 1H, 12.3 Hz), 5.10 (s, 2H), 5.63 (d, 1H, J ) 8.7
Hz), 7.25-7.35 (m, 10H); 13C NMR δ 14.2, 22.7, 26.0, 29.3, 29.6,
29.7, 31.9, 34.7, 54.1, 66.9, 70.1, 73.7, 73.9, 127,9, 128.1, 128.2,
128.3, 128.4, 128.6, 136.6, 137.4, 156.2. Anal. Calcd for C26H37
-
90:10 hexanes/EtOAc: [R]21 +29.3 (c 1.00, CH2Cl2); FTIR
D
NO4: C, 73.03; H, 8.72; N, 3.28. Found: C, 72.89; H, 8.59; N,
3.27.
(neat) 1760, 1725, 1661, 1606 cm-1; H NMR δ 1.25 (d, 3H, J
1
) 7 Hz), 2.66 (d, 1H, J ) 16 Hz), 2.86 (broad, 1H), 2.96 (d, 1H,
J ) 16 Hz), 3.52 (q, 1H, J ) 7.1 Hz), 4.51 (d, 2H, J ) 4.4 Hz),
5.22 (m, 2H), 5.80 (m, 1H), 7.21-7.47 (m, 15H); 13C NMR δ
20.1, 45.8, 58.3, 65.7, 71.5, 118.6, 128.0, 128.9, 131.8, 146.3,
206.3. These data match the literature data for 8c.13e
Gen er a l P r oced u r e for th e P r ep a r a tion of Tr ityl
N-P r otected r-Am in o Acid (6). This procedure is slightly
modified from the literature method.30 Chlorotrimethylsilane
(1.27 mL, 10.0 mmol) was added at room temperature to a
stirred suspension of an L-amino acid (10.0 mmol) in 18 mL of
CHCl3/MeCN (5:1). The reaction mixture was refluxed for 2 h
and then cooled to 0 °C. Dropwise addition of triethylamine
(2.79 mL, 20.0 mmol) was followed by a solution of trityl
chloride (2.79 g, 10.0 mmol) in chloroform (10 mL). The
resulting mixture was stirred for 1 h, and then methanol (2
mL) was added. After concentration, the pale yellow residue
was partitioned between diethyl ether and water. The aqueous
layer was extracted twice with diethyl ether (20 mL). The
combined organic layers were dried (MgSO4) and concentrated
to give the crude N-trityl R-amino acid, which was used for
the next step without further purification.
Gen er a l P r oced u r e for th e P r ep a r a tion of Tr ityl
N-Tr ityla cylim id a zoles (7) fr om N-Tr ityla m in o Acid s.
CDI (carbonyl-1,1′-diimidazole) (1.70 g, 10.5 mmol) was added
to a stirred solution of a trityl N-protected R-amino acid (10
mmol) in dry THF (20 mL) under a N2 atmosphere at room
temperature. The resulting solution was stirred overnight.
After evaporation of the solvent, the residue was partitioned
between water and methylene chloride. The aqueous layer was
extracted twice with methylene chloride. The combined organic
layers were dried (MgSO4) and concentrated to give the crude
N-tritylacylimidazole, which was used for the next step
without further purification.
Gen er a l P r oced u r e for th e P r ep a r a tion of N-Tr ityl
γ-Am in o-â-k eto Ester (8). A solution of lithium enolate of
allyl acetate was made from BuLi (2.5 M, 8 mL, 20 mmol),
diisopropylamine (2.8 mL, 20 mmol), and allyl acetate (2.2 mL,
20 mmol). The crude imidazole was dissolved in THF (20 mL)
and added dropwise to this pale yellow solution of lithium
enolate at -40 °C under N2 atmosphere. The resulting mixture
was stirred for 3 h at the same temperature and then quenched
with water. The reaction mixture was extracted with ethyl
acetate (3 × 50 mL). The organic extracts were combined,
washed with water and brine (50 mL), dried (MgSO4), and
concentrated to provide the crude product, which could be
purified by chromatography. Overall yield 74-78%.
3-Oxo-5-p h en yl-4S-(tr ityla m in o)p en ta n oic Acid Allyl
Ester (8a ). Phenylalanine (1.65 g, 10 mmol) was treated with
Me3SiCl (1.086 g, 10 mmol), trityl chloride (2.78 g, 10 mmol),
and Et3N (2.79 mL, 20.0 mmol) to give the N-tritylamino acid.
After workup (general procedure), the crude product was
sequentially converted with CDI (2.79 mL, 20.0 mmol) and
then with 2 equiv of lithium enolate of allyl acetate to afford
3.6 g (76% overall yield) of 8a as a colorless solid after
3-Oxo-4S-(tr ityla m in o)h ep ta n ed ioic Acid 1-Allyl Ester
7-Meth yl Ester (8g). Glu(OMe)OH (1.61 g, 10 mmol) was
treated with Me3SiCl (1.086 g, 10 mmol), trityl chloride (2.78
g, 10 mmol), and Et3N (2.79 mL, 20.0 mmol) to give the N-trityl
glutamate. After workup (general procedure), the crude prod-
uct was sequentially converted with CDI (2.79 mL, 20.0 mmol)
and then with 2 equiv of lithium enolate of allyl acetate to
afford 3.7 g (78% overall yield) of 8g as a colorless oil after
purification on silica gel eluting with 90:10 hexanes/EtOAc:
[R]21 +34.5 (c 1.00, CH2Cl2); FTIR (neat) 1755, 1730, 1661
D
1
cm-1; H NMR δ 1.88 (m, 1H), 2.25 (m, 2H), 2.62 (d, 1H, J )
16.1 Hz), 2.64 (m, 1H) 3.05 (d, 1H, J ) 16.1 Hz), 3.09 (d, 1H,
J ) 9.2 Hz), 3.67 (s, 3H), 4.49 (d, 2H, J ) 5.7 Hz), 5.23 (m,
2H), 5.80 (m, 1H), 7.20-7.40 (m, 15H); 13C NMR δ 27.8, 28.9,
46.0, 51.6, 60.9, 65.7, 71.4, 118.5, 126.5, 126.7, 127.8, 128.0,
128.8, 131.6, 146.0, 166.1, 173.5, 204.3.
5-Ben zyloxy-3-oxo-4S-(tr ityla m in o)p en ta n oic Acid Al-
lyl Ester (8h ). Ser(OBn)OH (1.95 g, 10 mmol) was treated
with Me3SiCl (1.086 g, 10 mmol), trityl chloride (2.78 g, 10
mmol), and Et3N (2.79 mL, 20.0 mmol) to give the N-
tritylserine derivative. After workup (general procedure), the
crude product was sequentially converted with CDI (2.79 mL,
20.0 mmol) and then with 2 equiv of lithium enolate of allyl
acetate to afford 3.8 g (74% overall yield) of 8h as a colorless
solid after purification on silica gel eluting with 95:5 hexanes/
EtOAc. Pure material could be obtained by recrystallization
from hexanes: mp 125 °C; [R]21D +34.5 (c 1.00, CH2Cl2); FTIR
1
(KBr) 1755, 1730 cm-1; H NMR δ 2.75 (d, 1H, J ) 16.6 Hz),
3.11 (t, 1H, J ) 8.3 Hz), 3.20 (broad, 1H), 3.43 (d, 1H, J )
16.6 Hz), 3.70 (m, 2H), 4.38 (s, 2H), 4.52 (d, 2H, J ) 5.6 Hz),
5.23 (m, 2H), 5.82 (m, 1H), 7.10-7.40 (m, 20H); 13C NMR δ
47.9, 61.9, 65.5, 71.2, 72.6, 73.6, 118.3, 126.8, 127.7, 128.1,
128.9, 132.0, 137.6, 146.1, 166.6, 206.2. Anal. Calcd for C34H33
-
NO4: C, 78.59; H, 6.40; N, 2.70. Found: C, 78.37; H, 6.29; N,
2.73.
Gen er a l P r oced u r e for th e P r ep a r a tion of Tr ityl
N-P r otected r-Am in o Keton es (9). Meth od A. With a n
Alk yl Tr ifla te a s a n Alk yla tin g Agen t. A solution of allyl
N-trityl γ-amino-â-keto ester 8 (5.0 mmol) in dry THF (20 mL)
was added dropwise under nitrogen to a stirred suspension of
NaH (240 mg of 60% in oil, 6.0 mmol, 1.2 equiv) in dry THF
(10 mL) at -10 °C and allowed to warm to 0 °C. The mixture
was stirred for 20 min, and then the alkyl triflate (5.5 mmol,
1.1 equiv) in 5 mL of THF was added. The resulting solution
was allowed to warm to room temperature and stirred for 3
h. After being quenched with 10 mL of water, the reaction
mixture was then extracted with ethyl acetate (3 × 50 mL).
The organic extracts were combined, washed with water and
(30) Barlos, K.; Papaioannou, D.; Theodoropoulos, D. J . Org. Chem.
1982, 47, 1324.

Syntheses of syn- and anti-1,2-Amino Alcohols
J . Org. Chem., Vol. 67, No. 4, 2002 1055
brine, dried (MgSO4), and concentrated to provide a pale yellow
oil. Without further purification, this oil was dissolved in THF
(20 mL) and added dropwise to a stirred solution of tetrakis-
(triphenylphosphine)palladium (578 mg, 0.5 mmol) and mor-
pholine (4.3 mL, 50 mmol) in dry THF under nitrogen at room
temperature. The mixture was stirred overnight. After the
filtrate was concentrated the residue was chromatographed
on SiO2 with hexanes-ethyl acetate (95:5) to afford the amino
ketone in 70-76% yield.
Meth od B. With a n Alk yl Br om id e a s a Alk yla tin g
Agen t. A solution of allyl N-trityl γ-amino-â-keto ester 8 (5.0
mmol) in dry THF (10 mL) was added dropwise to a stirred
suspension of NaH (240 mg of 60% in oil, 6.0 mmol, 1.2 equiv)
in dry THF (10 mL) at -20 °C under nitrogen. The mixture
was stirred for 20 min, and then the alkyl bromide (5.5 mmol)
in 5 mL of THF and 10% NaI were added. The resulting
solution was allowed to warm to room temperature and
refluxed for 2 h in order to complete the reaction. After the
reaction was quenched with 10 mL of water, the mixture was
extracted with ethyl acetate (3 × 50 mL). The organic extracts
were combined, washed with water and brine, dried (MgSO4),
and concentrated to provide a pale yellow oil. Without further
purification, this oil was dissolved in THF (20 mL) and added
dropwise to a stirred solution of tetrakis(triphenylphosphine)-
palladium (578 mg, 0.5 mmol) and morpholine (4.3 mL, 50
mmol) in dry THF under nitrogen at room temperature. The
mixture was stirred overnight. After the filtrate was concen-
trated, the residue was chromatographed on SiO2 with hex-
anes-ethyl acetate (95:5) to afford the amino ketone in 68%
yield.
31.8, 41.9, 61.3, 71.1, 73.3, 126.5, 127.8, 128.4, 129.0, 138.0,
146.5, 213.4.
Gen er a l P r oced u r e for th e P r ep a r a tion th e Ca r ba m ic
N-P r otected syn -r-Am in o Alcoh ols. LiAlH(O-t-Bu)3 127 mg
(0.50 mmol) was dissolved in THF (4 mL) at -5 °C under
nitrogen, and then a solution of the N-tritylamino ketone (0.25
mmol) in THF (3 mL) was added dropwise. The reaction was
monitored by TLC (95:5 hexanes/EtOAc). After 8-60 h, the
solution was quenched with water (2 mL), extracted with ethyl
acetate (2 × 50 mL), washed with H2O (20 mL) and brine
(20 mL), dried (MgSO4), and concentrated to provide a pale
yellow oil. Without further purification, the crude product was
used to exchange the trityl with the appropriate carbamic
protecting group. Treatment with aqueous HCl in acetone
removed the trityl group and treatment of the residue with a
carbamoylating agent and triethylamine gave the carbamate.
The final product was then purified by chromatography.
(1S-Ben zyl-2S-h ydr oxydecyl)car bam ic Acid Ben zyl Es-
ter (syn -3a ). Reduction of 9a (100 mg, 0.20 mmol) afforded
69.5 mg (88%) of syn-3a as a colorless solid that could be
purified by recrystallization from hexanes with 87% recov-
ery: mp 141 °C; [R]23D -17.5 (c 1.00, Cl2Cl2); FTIR (KBr) 3328,
1692 cm-1; 1H NMR δ 0.86 (t, 3H, J ) 6.5 Hz), 1.21 (m, 12H),
1.42 (m, 2H), 2.00 (s, 1H), 2.89 (d, 2H, J ) 7.3 Hz), 3.56 (m,
1H), 3.90 (q, 1H, J ) 7.7 Hz), 5.06 (s, 1H), 5.17 (d, 1H, J ) 9.0
Hz), 7.20-7.31 (m, 10H); 13C NMR δ 14.5, 23.1, 26.1, 29.6, 29.8,
29.9, 32.2, 34.9, 39.2, 56.4, 67.1, 71.9, 126,8, 128.3, 128.4, 128.9,
129.6, 137.0, 138.7, 156.9. Anal. Calcd for C25H35NO3: C, 75.53;
H, 8.87; N, 3.52. Found: C, 75.65; H, 8.80; N, 3.58.
(2S-Hydr oxy-1S-m eth yl-5-ph en ylpen tyl)car bam ic Acid
2,2-Dim eth ylp r op yl Ester (syn -3c). Reduction of 9c 108 mg
(0.25 mmol) afforded 67 mg (91%) of a mixture of both
diastereomers syn-3c and anti-3c (4:1). The mixture was not
purified further. Spectral characterization of the major isomer
1-P h en yl-2S-(tr ityla m in o)u n d eca n -3-on e (9a ). Alkyla-
tion of 8a (2.44 g, 5 mmol) with heptyl triflate (1.36 g, 5.50
mmol, 1.1 equiv) using method A afforded 1.83 g (73%) of 9a
as a colorless oil after purification on silica gel eluting with
95:5 hexanes/EtOAc: [R]21D +58.0 (c 1.00, CH2Cl2); FTIR (neat)
was obtained from the mixture: FTIR (KBr) 3365,1691 cm-1
;
1H NMR δ 1.13 (d, 3H, J ) 6.7 Hz), 1.43 (broad, 9H), 1.48 (m,
1
1730, 1606 cm-1; H NMR δ 0.87 (t, 3H, J ) 7 Hz), 1.19 (m,
2H), 1.75 (m, 2H), 2.25 (s, 1H), 2.62 (m, 2H), 3.46 (m, 1H),
12H), 1.48 (m, 2H), 2.79 (dd, 1H, J ) 7.5, 13.3 Hz), 2.94 (dd,
1H, J ) 6.1, 13.3 Hz), 3.16 (broad, 1H), 3.58 (t, 1H, J ) 6.6
Hz), 7.15-7.42 (m, 20H); 13C NMR δ 14.2, 22.5, 28.9, 29.2, 31.9,
41.7, 42.2, 63.6, 71.3, 126.5, 127.9, 128.5, 129.3, 129.8, 138.0,
146.7, 213.7.
3.62 (m, 1H), 4.81 (d, 1H, J ) 8.9 Hz), 7.15-7.27 (m, 5H); 13
C
NMR δ 18.3, 27.4, 28.5, 33.9, 35.9, 50.5, 74.9, 79.4, 125.8, 128.4,
128.5, 142.3, 156.3. Anal. Calcd for C17H27NO3: C, 69.59; H,
9.28; N, 4.77. Found: C, 69.42; H, 9.12; N, 4.76.
(1S-Ben zyloxym eth yl-2S-h yd r oxyd ecyl)ca r ba m ic Acid
Ben zyl Ester (syn -3h ). Reduction of 133 mg (0.25 mmol) of
9h afforded 96 mg (90%) of a mixture of syn-3h and anti-3h
(12:1). Spectral characterization of the major isomer was
obtained from the mixture. The major isomer syn-3h could be
purified by recrystallization from ethyl acetate, 85%: FTIR
6-P h en yl-2S-(tr ityla m in o)h exa n -3-on e (9c). Alkylation
of 8a (2.00 g, 5.00 mmol) with 1.40 g (5.5 mmol, 1.1 equiv) of
phenethyl triflate using method A afforded 1.60 g (76%) of 9c
as a colorless solid after purification on silica gel eluting with
95:5 hexanes/EtOAc. Pure material could be obtained by
recrystallization from hexanes with 86% recovery: mp 113 °C;
[R]21D +42.6 (c 1.00, CH2Cl2); FTIR (KBr) 1725, 1606 cm-1; 1H
NMR δ 1.20 (d, 3H, J ) 6.9 Hz), 1.46 (m, 3H), 2.02 (m, 1H),
2.35 (m, 2H), 3.21 (broad, 1H), 3.38 (m, 1H), 7.15-7.47 (m,
20H); 13C NMR δ 20.8, 24.4, 35.0, 38.8, 57.5, 71.4, 125.9, 126.4,
127.8, 128.1, 128.3, 128.5, 129.0, 141.7, 146.6, 213.3. Anal.
Calcd for C31H31NO: C, 85.87; H, 7.21; N, 3.23. Found: C,
85.91; H, 7.04; N, 3.23.
1
(KBr) 3335, 1700 cm-1; H NMR δ 0.87 (t, 3H, J ) 6.9 Hz),
1.25 (broad, 12H), 1.36 (m, 2H), 3.00 (broad, 1H), 3.71 (m, 3H),
3.89 (1H), 4.49 (s, 2H), 5.10 (s, 2H), 5.50 (d, 1H, J ) 8.5 Hz),
7.22-7.34 (m, 10H); 13C NMR δ 14.1, 22.7, 25.7, 26.0, 29.3,
29.5, 29.6, 31.9, 33.9, 53.6, 66.8, 70.0, 72.6, 73.7, 127,7, 127.8,
128.0, 128.2, 128.3, 128.6, 136.6, 137.5, 156.6. Anal. Calcd for
C
26H37NO4: C, 73.03; H, 8.72; N, 3.28. Found: C, 73.20; H,
8.66; N, 3.28.
9-Meth yl-5-oxo-4S-(tr ityla m in o)d ec-8-en oic Acid Meth -
yl Ester (9g). Alkylation of 8g (2.42 g, 5 mmol) of with 820
mg (5.50 mmol, 1.1 equiv) of prenyl bromide using method B
afforded 1.59 g (68%) of 9g as a colorless oil after purification
on silica gel eluting with 95:5 hexanes/EtOAc: [R]21D +47.9 (c
1.00, CH2Cl2); FTIR (neat) 1750, 1716, 1611 cm-1; 1H NMR δ
1.52 (s, 3H), 1.55 (m, 1H), 1.60 (s, 3H), 1.81 (m, 2H), 2.15 (m,
2H), 2.58 (m, 2H), 3.19 (broad, 1H), 3.54 (m, 1H), 3.67 (s, 3H),
4.82 (t, 1H), 7.15-7.42 (m, 15H); 13C NMR δ 17.6, 21.8, 25.6,
28.4, 29.0, 40.0, 51.6, 60.1, 71.5, 122.9, 126.5, 127.5, 127.9,
129.0, 132.3, 146.4, 173.8, 211.5.
Gen er a l P r oced u r e for th e Cycliza tion of th e Ca r -
ba m ic N-P r otected r-Am in o Alcoh ols 3 to th e Cor r e-
sp on d in g Oxa zolid in on es 4. Meth od A. Aqueous sodium
hydroxide (8 M, 0.5 mL) was added to a stirred solution of the
protected R-amino alcohol 3 (0.25 mmol) in 5 mL of methanol-
THF (1:2). The mixture was stirred for 3-5 h at room
temperature and then evaporated. The residue was dissolved
in 10 mL of water and extracted twice with 10 mL of ethyl
acetate, washed with H2O (5 mL) and brine (5 mL), dried
(MgSO4), and concentrated to provide the crude product, which
could be recrystallized from an appropriate solvent.
Meth od B. A solution of the protected R-amino alcohol 0.25
mmol in 2 mL of DMF was added to a suspension of NaH (18
mg, 0.75 mmol) in 3 mL of DMF. The mixture was stirred for
3-5 h at room temperature. The mixture was dissolved in 10
mL of water, extracted twice with 10 mL of ethyl acetate,
washed with H2O (5 mL) and brine (5 mL), dried (MgSO4),
and concentrated to provide the crude product, which could
be purified by chromatography or by recrystallization from an
appropriate solvent.
1-Ben zyloxy-2S-(tr ityla m in o)u n d eca n -3-on e (9h ). Alky-
lation of 8h (2.67 g, 5 mmol) of with 1.36 g (5.50 mmol, 1.1
equiv) of heptyl triflate using method A afforded 1.90 g (70%)
of 9h as a colorless oil after purification on silica gel eluting
with 95:5 hexanes/EtOAc: [R]21D +13.1 (c 1.00, CH2Cl2); FTIR
1
(neat) 1727, 1611 cm-1; H NMR δ 0.87 (t, 3H, J ) 6.9 Hz),
1.19 (m, 12H), 1.48 (m, 1H), 2.22 (m, 1H), 3.30 (m, 2H), 3.56
(dd, 1H, J ) 4.6, 8.9 Hz), 3.77 (dd, 1H, J ) 4.6, 7.6 Hz), 4.42
(s, 2H), 7.15-7.42 (m, 20H); 13C NMR δ 14.1, 22.7, 29.0, 29.2,

1056 J . Org. Chem., Vol. 67, No. 4, 2002
Hoffman et al.
4S-Ben zyl-5R-octyloxa zolid in -2-on e (syn -4a ) was pre-
pared by method A from 100 mg (0.25 mmol) of alcohol anti-
3a to give syn-4a as a colorless oil that was purified by
chromatography to afford 64 mg (89%): 1H NMR δ 0.89 (t,
3H, J ) 6.8 Hz), 1.28 (m, 11H), 1.60 (m, 2H), 1.83 (m, 1H),
2.66 (dd, 1H, J ) 3.9, 13.5 Hz), 2.86 (dd, 1H, J ) 10.1, 14.3
Hz), 3.94 (m, 1H), 4.65 (m, 1H), 4.90 (s, 1H), 7.17-7.34 (m,
5H); 13C NMR δ 14.1, 22.7, 26.0, 29.0, 29.2, 29.4, 29.7, 32.0,
36.3, 56.9, 80.0, 127.2, 129.0, 129.1, 136.8, 158.7. Irradiation
of the proton signals at 2.75 and 1.83 ppm causes the methine
signals at 4.66 and 3.96 ppm, respectively, to collapse to
doublets with J ) 7.3 Hz.
4S-Ben zyl-5R-p h en eth yloxa zolid in -2-on e (syn -4b) was
prepared by method B from 90 mg (0.25 mmol) of alcohol anti-
3b to give syn-4b as a colorless solid, which was purified by
chromatography to afford 60 mg (86%): 1H NMR δ 1.95 (m,
1H), 2.17 (m, 1H), 2.72 (m, 4H), 3.92 (m, 1H), 4.62 (m, 1H),
5.10 (s, 1H), 7.14-7.32 (m, 10H); 13C NMR δ 31.4, 32.1, 36.4,
56.7, 78.8, 126.3, 127.2, 129.0, 129.1, 136.1, 140.6, 158.5.
Irradiation of the proton signals at 2.21 and 1.83 ppm causes
the methine signals at 4.65 and 3.94 ppm, respectively, to
collapse to doublets with J ) 7.4 Hz.
4S-Meth yl-5R-(3-p h en ylp r op yl)oxa zolid in -2-on e (syn -
4c) was prepared by method B from 73 mg (0.25 mmol) of
alcohol anti-3c to give syn-4c as a colorless solid, which was
purified by chromatography to afford 50 mg (90%). A crystal
was prepared in toluene for X-ray analysis: 1H NMR δ 1.11
(d, 3H, J ) 6.4 Hz), 1.74 (m, 4H), 2.67 (m, 2H), 3.87 (m, 1H),
4.55 (m, 1H), 6.22 (s, 1H), 7.15-7.32 (m, 5H); 13C NMR δ 15.9,
27.6, 28.7, 35.5, 51.1, 80.0, 126.0, 128.4, 141.7, 159.7. Irradia-
tion of the proton signals at 1.65 and 1.11 ppm causes the
methine signals at 4.56 and 3.87 ppm, respectively, to collapse
to doublets with J ) 7.6 Hz.
4S-Ben zyloxym eth yl-5R-octyl-oxa zolid in -2-on e (syn -
4h ) was prepared by method A from 107 mg (0.25 mmol) of
alcohol anti-3h to give syn-4a as a colorless oil that was
purified by chromatography to afford 74 mg (86%): 1H NMR
δ 0.88 (t, 3H, J ) 6.6 Hz), 1.26 (broad, 11H), 1.56 (m, 3H),
3.49 (m, 2H), 3.91 (m, 1H), 4.52 (s, 2H), 4.60 (m, 1H), 5.64 (s,
1H), 7.26-7.36 (m, 5H); 13C NMR δ 14.1, 22.7, 26.1, 29.1, 29.2,
29.3, 29.4, 31.9, 55.0, 68.8, 73.8, 79.9, 127.8, 128.1, 128.7, 137.4,
159.1. Irradiation of the proton signals at 1.56 and 3.47 ppm
causes the methine signals at 4.60 and 3.92 ppm, respectively,
to collapse to doublets with J ) 7.3 Hz.
3H, J ) 6.8 Hz), 1.23 (m, 11H), 1.50 (m, 2H), 1.65 (m, 1H),
2.83 (d, 2H, J ) 7.0 Hz), 3.64 (q, 1H, J ) 6.5 Hz), 4.28 (m,
1H), 5.45 (s, 1H), 7.15-7.34 (m, 5H); 13C NMR δ 14.1, 22.8,
26.0, 29.0, 29.2, 29.4, 29.7, 31.8, 36.3, 57.0, 80.1, 127.2, 129.0,
129.2, 136.8, 158.7. Irradiation of the proton signals at 1.59
and 2.83 ppm causes the methine signals at 4.28 and 3.68 ppm,
respectively, to collapse to doublets with J ) 5.5 Hz.
4S-Meth yl-5S-(3-p h en ylp r op yl)oxa zolid in -2-on e (a n ti-
4c) was prepared by method B from 73 mg (0.25 mmol) of the
alcohol mixture of anti-3c and syn-3c (4:1) to give a mixture
of syn-4c and anti-4c in the same ratio as a colorless oil, which
was purified by chromatography to afford 51 mg (94%).
Spectral characterization of the major isomer was obtained
from the mixture: 1H NMR δ 1.22 (d, 3H, J ) 6.2 Hz), 1.71
(m, 4H), 2.66 (m, 2H), 3.52 (m, 1H), 4.07 (m, 1H), 6.58 (broad,
1H), 7.15-7.32 (m, 5H); 13C NMR δ 20.58, 25.6, 33.6, 35.4,
53.7, 84.1, 126.0, 128.4, 141.6, 159.5. Irradiation of the proton
signals at 1.73 and 1.22 ppm causes the methine signals at
4.07 and 3.65 ppm, respectively, to collapse to doublets with
J ) 6.4 Hz.
4S-Ben zyloxym eth yl-5S-octyloxa zolid in -2-on e (a n ti-
4h ) was prepared by method A from 107 mg (0.25 mmol) of
an alcohol mixture syn-3h and anti-3h (12:1) with NaOH in
THF-methanol to give a mixture of syn-4h and anti-4h in the
same ratio as a colorless oil that was purified by chromatog-
raphy to afford 73 mg (91%). Spectral characterization of the
major isomer was obtained from the mixture: 1H NMR δ 0.88
(t, 3H, J ) 6.8 Hz), 1.26 (broad, 11H), 1.65 (m, 3H), 3.44 (m,
2H), 3.61 (m, 1H), 4.25 (m, 1H), 4.53 (s, 2H), 6.1 (s, 1H), 7.26-
7.36 (m, 5H); 13C NMR δ 14.1, 22.7, 24.6, 29.2, 29.3, 29.4, 31.9,
34.9, 57.3, 71.9, 73.6, 79.6, 127.7, 128.0, 128.6, 137.4, 159.1.
Irradiation of the proton signals at 3.44 and 1.67 ppm causes
the methine signals at 3.65 and 4.25 ppm, respectively, to
collapse to doublets with J ) 5.2 Hz.
Ack n ow led gm en t. This work was supported by a
grant from the National Science Foundation (CHE-
9900402). The authors thank Prof. Phil Hampton of the
University of New Mexico for helpful insights. We also
thank Michael Critz for help with the cover.
Su p p or tin g In for m a tion Ava ila ble: 13C NMR spectra of
1e,g, anti-3d , anti-3g, 8g, and 9a ,g,h , complete X-ray data
for 4c, and space-filling structures of 9a ,g,h . This material is
available free of charge via the Internet at http://pubs.acs.org.
4S-Bezyl-5S-octyl-oxa zolid in -2-on e (a n ti-4a ) was pre-
pared by method A from 100 mg (0.25 mmol) of alcohol syn-
3a to give anti-4a as a colorless oil that was purified by
chromatography to afford 60 mg (83%): 1H NMR δ 0.88 (t,
J O010270F