Synthesis of a family of fine-tunable new chiral ligands for catalytic asymmetric synthesis. Ligand optimization through the enantioselective addition of diethylzinc to aldehydes

Article abstract of DOI:10.1021/jo9701445

A family of enantiomerically pure (1R,2R)-1-(dialkylamino)-1-phenyl-3- (R-oxy)-2-propanols (4) has been synthesized from (2S,3S)-2,3-epoxy-3- phenylprepanol (1a), arising from the Sharpless epoxydation of cinnamyl alcohol, by two alternative sequences involving either the regioselective ring opening of the epoxide by a secondary amine (C-3 attack) and subsequent chemoselective protection of the primary hydroxy group or the reverse of these operations. A total of 19 different derivatives 4 have been prepared in an iterative process aimed at the optimization of their catalytic properties in the enantioselective addition of diethylzinc to benzaldehyde. In doing this, the steric bulk of the R-oxy group and the choice of the dialkylamino substituent as a nitrogen-containing six-membered ring have been identified as the key structural parameters for high catalytic activity and enantioselectivity in 4. Two optimized ligands fulfilling these structural requirements, 4d-Tr (R-oxy = trityloxy, dialkylamino = piperidino) and 4i-Tr (R-oxy = trityloxy, dialkylamino = 4-methylpiperazin-1-yl), depict a convenient activity and selectivity profile in the addition of Et₂-Zn to a structurally diverse family of aldehydes. These results show how chiral ligands based on non-natural starting materials can accommodate subtle variations of the steric/electronic characteristics key to the fine tuning of catalytic properties and thus represent a convenient alternative to ligands based on natural products.

Full text of DOI:10.1021/jo9701445

4970
J . Org. Chem. 1997, 62, 4970-4982
Syn th esis of a F a m ily of F in e-Tu n a ble New Ch ir a l Liga n d s for
Ca ta lytic Asym m etr ic Syn th esis. Liga n d Op tim iza tion th r ou gh
th e En a n tioselective Ad d ition of Dieth ylzin c to Ald eh yd es
Anton Vidal-Ferran, Albert Moyano, Miquel A. Perica`s,* and Antoni Riera
Unitat de Recerca en Sı´ntesi Asime`trica (URSA), Departament de Quı´mica Orga`nica, Universitat de
Barcelona, Martı´ i Franque`s 1-11, 08028 Barcelona, Spain
Received J anuary 27, 1997X
A family of enantiomerically pure (1R,2R)-1-(dialkylamino)-1-phenyl-3-(R-oxy)-2-propanols (4) has
been synthesized from (2S,3S)-2,3-epoxy-3-phenylpropanol (1a ), arising from the Sharpless
epoxydation of cinnamyl alcohol, by two alternative sequences involving either the regioselective
ring opening of the epoxide by a secondary amine (C-3 attack) and subsequent chemoselective
protection of the primary hydroxy group or the reverse of these operations. A total of 19 different
derivatives 4 have been prepared in an iterative process aimed at the optimization of their catalytic
properties in the enantioselective addition of diethylzinc to benzaldehyde. In doing this, the steric
bulk of the R-oxy group and the choice of the dialkylamino substituent as a nitrogen-containing
six-membered ring have been identified as the key structural parameters for high catalytic activity
and enantioselectivity in 4. Two optimized ligands fulfilling these structural requirements, 4d -Tr
(R-oxy ) trityloxy, dialkylamino ) piperidino) and 4i-Tr (R-oxy ) trityloxy, dialkylamino )
4-methylpiperazin-1-yl), depict a convenient activity and selectivity profile in the addition of Et2-
Zn to a structurally diverse family of aldehydes. These results show how chiral ligands based on
non-natural starting materials can accommodate subtle variations of the steric/electronic charac-
teristics key to the fine tuning of catalytic properties and thus represent a convenient alternative
to ligands based on natural products.
In tr od u ction
ditional primary hydroxy group that allows further
modification and can provide an additional binding site.
Chiral ligands for asymmetric catalysis1,2 have been
generally derived from a small number of readily avail-
able natural products. They thus belong to just a few
structural types and do not allow in many cases signifi-
cant optimization of catalytic properties through struc-
tural modification.3 Synthetic substances of high enan-
tiomeric purity arising from well-established processes
can represent an advantageous alternative to ligands of
natural origin and can allow the circumvention of the
aforementioned limitation.
We have recently reported the highly enantioselective
synthesis of amino acids of many different structural
types from epoxy alcohols (1), through the intermediacy
of 3-amino 1,2-diols (2).4 These substances contain all
the structural characteristics of 1,2-amino alcohols (3),
a well-established class of chiral ligands, plus an ad-
A large number of epoxy alcohols (1) are readily
available in enantiomerically pure form in both enantio-
meric series via the asymmetric Sharpless epoxidation.5
These alcohols can, in principle, be converted into amino
diols with full control of all structural and stereochemical
parameters (Figure 1). This encouraged us to consider
that properly substituted 3-amino 1,2-diols (4) could
constitute an interesting new class of chiral ligands, and
the work described here details our initial efforts in this
area.
X Abstract published in Advance ACS Abstracts, J uly 1, 1997.
(1) (a) Asymmetric Catalysis in Organic Synthesis; Noyori, R., Ed.;
J ohn Wiley & Sons, Inc.: New York, 1994. (b)Catalytic Asymmetric
Synthesis; Ojima, I., Ed.; VCH: New York, 1993.
(2) Berrisford, D. J .; Bolm, C.; Sharpless, K. B. Angew. Chem., Int.
Ed. Engl. 1995, 34, 1050-1070.
(3) For compilations of natural and non-natural enantioselective
catalysts and auxiliaries, see, for example: (a) Blaser, H. U. Chem.
Rev. 1992, 92, 935-952. (b) Chiral Auxiliaries and Ligands in
Asymmetric Synthesis; Seyden-Penne, J ., Ed.; J ohn Wiley & Sons,
Inc.: New York, 1995.
(4) (a) Canas, M.; Poch, M.; Verdaguer, X.; Moyano, A.; Perica`s, M.
A.; Riera, A. Tetrahedron Lett. 1991, 32, 6931-6934. (b) Poch, M.;
Verdaguer, X.; Moyano, A.; Perica`s, M. A.; Riera, A. Tetrahedron Lett.
1991, 32, 6935-6938. (c) Alco´n, M.; Canas, M.; Poch, M.; Moyano, A.;
Perica`s, M. A.; Riera, A. Tetrahedron Lett. 1994, 35, 1589-1592. (d)
Poch, M.; Alco´n, M.; Moyano, A.; Perica`s, M. A.; Riera, A. Tetrahedron
Lett. 1993, 34, 7781-7784. (e) Castejo´n, P.; Pasto´, M.; Moyano, A.;
Perica`s, M. A.; Riera, A. Tetrahedron Lett. 1995, 36, 3019-3022. (f)
Pasto´, M.; Moyano, A.; Perica`s, M. A.; Riera, A. Tetrahedron: Asym-
metry 1995, 6, 2329-2342. (g) Pasto´, M.; Moyano, A.; Perica`s, M. A.;
Riera, A. J . Org. Chem. 1996, 61, 6033-6037. (h) Castejo´n, P.; Moyano,
A.; Perica`s, M. A.; Riera, A. Chem. Eur. J . 1996, 2, 1001-1006. (i)
Castejo´n, P.; Moyano, A.; Perica`s, M. A.; Riera, A. Tetrahedron 1996,
52, 7063-7086.
Resu lts a n d Discu ssion
Gen er a l Str a tegies for th e Syn th esis of P r otected
3-Am in o 1,2-Diols 4. Two different strategies can be
envisaged for the conversion of epoxy alcohols 1 into
protected 3-amino 1,2-diols 4, as shown in Scheme 1. In
the first (route A), the epoxy alcohol is first submitted to
a regioselective ring-opening by a secondary amine, under
the conditions developed by Caron and Sharpless,6 and
the resulting amino diol (5) is then protected by means
of an appropriate R3-X reagent. In the second (route B),
(5) (a) Gao, Y.; Hanson, R. M.; Klunder, J . M.; Ko, S. Y.; Masamune,
H.; Sharpless, K. B. J . Am. Chem. Soc. 1987, 109, 5765-5780. (b)
Reference 1b, pp 101-158. (c) Katsuki, T.; Martin, V. S. In Organic
Reactions; Paquette, L. A., Ed.; J ohn Wiley & Sons, Inc.: New York,
1996; Vol. 48, pp 1-299.
(6) Caron, M.; Sharpless, K. B. J . Org. Chem. 1985, 50, 1557-1560.
S0022-3263(97)00144-8 CCC: $14.00 © 1997 American Chemical Society

Enantioselective Addition of Diethylzinc to Aldehydes
J . Org. Chem., Vol. 62, No. 15, 1997 4971
Ta ble 1. Regioselective Rin g Op en in ga of Ep oxy Alcoh ol
1a b by Secon d a r y Am in es in th e P r esen ce of Ti(OiP r )4
F igu r e 1.
Sch em e 1
a
All reactions were performed in CH2Cl2 solution, in the
presence of a 50% molar excess of both secondary amine and
b
Ti(OiPr)4. Results are for the 2S, 3S enantiomer. c Yields are
referred to pure isolated material and are not optimized.
Sch em e 2
Sch em e 3
the primary hydroxy group in the starting epoxy alcohol
is initially protected, and the resulting epoxy ether (6)
is subjected to regioselective ring opening by a secondary
amine under the conditions developed by Crotti and co-
workers.7
In the present work, we have used both routes for the
preparation of protected 3-amino 1,2-diols 4, starting in
all cases from the 2S,3S enantiomer of the epoxide of
cinnamyl alcohol (1a ).8
Rou te A to P r otected 3-Am in o 1,2-Diols: Regio-
selective Rin g-Op en in g p lu s P r otection . When ep-
oxy alcohol 1a was treated with 1.5 equiv of a secondary
amine in the presence of 1.5 equiv of titanium tetraiso-
propoxide at room temperature in dichloromethane solu-
tion (Scheme 2), a completely regioselective reaction took
place,6 leading to 3-amino 1,2-diols 5 (Table 1).
The reactions proceeded smoothly at room tempera-
ture, with the amino diols 5 being formed in generally
high yield after short reaction periods. Only when the
rather bulky diisopropylamine was used (entry g), the
process turned out to be considerably slower.
The selective protection of the primary hydroxy group
of 5a -h could be conveniently achieved by reaction with
tert-butyldimethylsilyl chloride and imidazole in DMF
solution at 65 °C (Scheme 3 and Table 2),4e,9a the target
amino alcohols 4a -h being obtained in good yield. In
the case of the very polar derivative 5h the protection
could not be driven to completion, and the resulting
mixture (4h /5h ≈ 8/1) could not be separated by standard
chromatographic procedures. As a result, 4h was not
tested as a ligand in catalytic reactions.
For amino diol 5d , several alternative protections of
the primary hydroxy group were also studied (Scheme
4). Thus, treatment of 5d with tert-butyldiphenylsilyl
chloride and imidazole in DMF at 65 °C9b afforded the
corresponding silyl ether 4d -Bp in essentially quantita-
tive yield, whereas reaction with trityl chloride in pyri-
(7) (a) Chini, M.; Crotti, P.; Macchia, F. J . Org. Chem. 1991, 56,
5939-5942. (b) Chini, M.; Crotti, P.; Flippin, L. A.; Macchia, F. J . Org.
Chem. 1991, 56, 7043-7048. (c) For the regioselective ring opening of
some racemic epoxyethers with different nucleophiles, see: Chini, M.;
Crotti, P.; Flippin, L. A.; Gardelli, C.; Giovani, E.; Macchia, F.; Pineschi,
M. J . Org. Chem. 1993, 58, 1221-1227. (d) Calvani, F.; Crotti, P.;
Gardelli, C.; Pineschi, M. Tetrahedron 1994, 50, 12999-13022. (e)
Colombini, M.; Crotti, P.; Di Bussolo, V.; Favero, L.; Gardelli, C.;
Macchia, F.; Pineschi, M. Tetrahedron 1995, 51, 8089-8112. (f) Azzena,
F.; Calvani, F.; Crotti, P.; Gardelli, C.; Macchia, F.; Pineschi, M.
Tetrahedron 1995, 51, 10601-10626.
(8) Similar ligands derived from other epoxides are currently studied
in our laboratories. The results of these studies will be reported in
due time.
(9) (a) Corey, E. J .; Venkateswarlu, A. J . Am. Chem. Soc. 1972, 94,
6190-6191. (b) Hanessian, S.; Lavallee, P. Canad. J . Chem. 1975, 53,
2975-2977.

4972 J . Org. Chem., Vol. 62, No. 15, 1997
Vidal-Ferran et al.
Ta ble 3. P r otection of 1a w ith R3X Rea gen ts
Ta ble 2. Selective P r otection a of th e P r im a r y Hyd r oxy
Gr ou p in 5a -h a s a ter t-Bu tyld im eth ylsilyl Eth er
a
All reactions were performed at 65 °C in DMF solution, in the
presence of a 10% molar excess of tert-butyldimethylsilyl chloride
and a 120% molar excess of imidazole. b Yields are referred to pure
isolated material and are not optimized. c Product was obtained
in admixture with a 7.5% yield of starting material.
Sch em e 4
opening polymerization. We were pleased to find that
formation of the sodium alkoxide of 1a with a slight
excess of NaH in DMF at -20 to 0 °C was free of any of
the conceivable side reactions.
Reaction of the so-generated alkoxide with methyl
iodide allowed the isolation of the expected methyl ether
6-Me in 91% yield.12 The same reaction conditions were
employed for the preparation of the corresponding benzyl
(6-Bn ), benzhydryl (6-Bzh ), and trityl (6-Tr ) ethers.
Whereas the preparation of the benzyl ether was fast
even at -20 °C, higher temperatures and longer reaction
times were required as the steric bulk of the alkylating
agent increased. The preparation of 6-Bzh was therefore
best performed at 0 °C, while that of 6-Tr only proceeded
to 13% (15% conversion) after 3 days at 0 °C. In order
to improve the yield of 6-Tr while preventing the occur-
rence of secondary reactions, less basic reaction condi-
tions were investigated. The yield of 6-Tr was increased
to 45% by performing the reaction in pyridine solution,
at 90 °C for 1.5 h in the presence of a 10% molar amount
of 4-DMAP.13 The best reaction conditions, however,
(66% isolated yield) involved the use of 1.1 equiv of
N-(triphenylmethyl)pyridinium tetrafluoroborate in ac-
etonitrile at room temperature.14 Finally, the tert-
butyldimethylsilyl ether 6-Tbs was also prepared in 79%
yield under the standard conditions.
Sch em e 5
dine at 90 °C10 allowed the isolation of the trityl ether
4d -Tr in 73% yield.
Rou te B to P r otected 3-Am in o 1,2-Diols: P r otec-
tion p lu s Regioselective Rin g Op en in g. For the first
part of the sequence (Scheme 5), epoxy alcohol 1a was
subjected to a variety of protection schemes, which have
been summarized in Table 3.
As a first goal, we sought to develop reaction conditions
that minimized possible side reactions of the starting
epoxy alcohol 1a , such as Payne rearrangement11 or ring-
For the regioselective and stereospecific ring opening
of epoxy ethers 6 with secondary amines, the procedure
(11) Behrens, C. H.; Ko, S. Y.; Sharpless, K. B.; Walker, F. J . J .
Org. Chem. 1985, 50, 5687-5696.
(12) Fukuzawa, A.; Sato, H.; Masamune, T. Tetrahedron Lett. 1987,
28, 4303-4306.
(13) (a) Chaudhary, S. C.; Hernandez, O. Tetrahedron Lett. 1979,
20, 99-102. (b) Protecting Groups; Kocienski, P. J ., Ed.; Georg Thieme
Verlag: Stuttgart 1994; pp 58.
(10) (a) Garc´ıa, M. L.; Pascual, J .; Borra`s, L.; Andreu, J . A.; Fos,
E.; Mauleon, D.; Carganico, G.; Arcamone, F.; Tetrahedron 1991, 47,
10023-10034. (b) Ireland, R. E.; Anderson, R. C.; Badoud, R.; Fitzsim-
mons, B. J .; McGarvey, G. J .; Thaisrivongs, S.; Wilcox, C. S. J . Am.
Chem. Soc. 1983, 105, 1988-2006.
(14) (a) Hanessian, S.; Murray, P. J .; Sahoo, S. P. Tetrahedron Lett.
1985, 26, 5623-5626. (b) Hanessian, S.; Murray, P. J .; Sahoo, S. P.
Tetrahedron Lett. 1985, 26, 5627-5630. (c) Hanessian, S.; Sahoo, S.
P.; Murray, P. J Tetrahedron Lett. 1985, 26, 5631-5634. (d) Hanessian,
S.; Staub, A. P. A. Tetrahedron Lett. 1973, 14, 3555-3558.

Enantioselective Addition of Diethylzinc to Aldehydes
J . Org. Chem., Vol. 62, No. 15, 1997 4973
Ta ble 4. Lith iu m P er ch lor a te-In d u ced Regioselective
Rin g Op en in g of Ep oxy Eth er s 6 by Secon d a r y Am in esa
F igu r e 2.
Sch em e 7
Whereas route B is unique in some instances; i.e., when
a nonbulky R3 group has to be chemoselectively intro-
duced, it offers a more subtle advantage in other situa-
tions. It is noteworthy that route A is characterized by
a highly polar intermediate 5, while in route B the
intermediate epoxy ether 6 exhibits much lower polarity.
When an amino alcohol 4 incorporating more than one
nitrogen atom in the amino moiety is prepared according
to route A (i.e., 4h , 4h -Tr , 4i-Bzh , 4i-Tr ), the chromato-
graphic purification of the corresponding precursors 5,
which are extremely polar substances, can be hardly
performed. On the other hand, if the preparation of the
same substance is planned according to route B, the only
polar species in the sequence is the final product 4. Since
4 is obtained from 5 in very high yield, purification at
this stage is a more convenient alternative.
a
All reactions were performed by treating the starting epoxy
The nonproblematic example 4d -Tr has been prepared
according to both routes. Overall yield is slightly higher
in route A (66 vs 61%), but the ease of purification of the
intermediate and final product could favor route B even
in this case.
ether 6 with a 10-fold excess of the corresponding amine in a 6.2
b
M solution of LiClO4 in acetonitrile at 55 °C for 24 h. Yields are
referred to pure isolated material. c Product is a 7:1 mixture of
regioisomers arising from C-3 (major) and C-2 (minor) attacks.
Sch em e 6
In order to understand the factors controlling the
catalytic activity of the various ligands under investiga-
tion, it seemed interesting to modify, at least in one case,
the anti arrangement of the substituents at C-2 and C-3.
Among the different strategies suitable to achieve this
(Figure 2), the Mitsunobu inversion15 of a preformed anti
amino alcohol 4 appeared the most promising,16 since it
is known that the Sharpless epoxidation of Z allyl
alcohols is less efficient than that of the corresponding
E isomers5a and the subsequent ring opening of the
intermediate cis 1,2-disubstituted epoxides takes place
with poor regioselectivity.7
Amino alcohol 4d was selected as the starting material
for the C-2 inversion. A Mitsunobu reaction with p-
nitrobenzoic acid as the nucleophile15c,16 (Scheme 7)
followed by reduction with DIBALH gave syn-4d in
diastereomerically pure form.
of Crotti,7 which involves the use of 5-10 M LiClO4 in
acetonitrile, was selected (Scheme 6).
In this way, the target amino alcohols 4 were obtained
in very high yield as shown in Table 4. It is worth noting
that, except for the reaction of 6-Tbs with N,N,N′-
trimethylethylenediamine, the processes were completely
regioselective and, in all cases, stereospecific.
Com p a r ison of Rou tes A a n d B to Am in o Alcoh ols
4 a n d F u r th er Mod ifica tion of Th ese Su bsta n ces.
The synthetic sequences represented by routes A and B
(Scheme 1) clearly complement to one another in many
different aspects. Route A is well suited for the prepara-
tion of families of amino alcohols 4 containing a bulky
R3 group, while route B can allow the easy preparation
of families of ligands 4 incorporating a common amine
residue while varying at will the nature of the protecting
group R3.
Ligan d Optim ization th r ou gh th e Catalytic En an -
tioselective Ad d ition of Dieth ylzin c to Ben za ld e-
h yd e. The enantioselective addition of diethylzinc and
(15) (a) Mitsunobu, O. Synthesis 1981, 1-28. (b) Robinson, P. L.;
Barry, C. N.; Bass, S. W.; J arvis, S. E.; Evans, S. A. J . Org. Chem.
1983, 48, 5396-5398. (c) Coleman, R. S.; Grant, E. B. Tetrahedron
Lett. 1994, 36, 8341-8344.
(16) For related examples see ref 4f,g.

4974 J . Org. Chem., Vol. 62, No. 15, 1997
Vidal-Ferran et al.
Ta ble 5. Ca ta lytic En a n tioselective Ad d ition to Et2Zn to
Ben za ld eh yd e:a In itia l Scr een in g
reaction
condns
conversion
(%)
selectivity
(%)
ee (8)
(%)
ligand
4g
hexanes, 20 h
toluene, 4 h
toluene, 3 h
hexanes, 20 h
hexanes, 3 h
hexanes, 4 h
>99
>99
>99
>99
>99
>99
98
97
98
99
98
99
32
84
86
79
89
77
4h -Tr
4j-Tr
4a
4d
4f
F igu r e 3.
a
All reactions were performed at room temperature, using a
Et2Zn/C6H5CHO/ligand molar ratio of 2.5/1/0.06.
its congeners to aldehydes mediated by chiral nonracemic
amino alcohols represents one of the most studied
examples of ligand-accelerated catalysis.2 From a practi-
cal perspective, it is placed among the most powerful
methodologies for the production of the ubiquitous sec-
ondary alcohols in enantiomerically pure form.17 Several
characteristics of the reaction, such as the very important
nonlinear effects,17-19 autocatalysis phenomena,20 and the
derived amplification of enantiomeric excess, make it
attractive from an industrial perspective, provided that
sufficiently active catalytic ligands are developed.
Following the pioneering work by Oguni,21 many
different â-amino alcohols have been tested with success
as catalytic ligands: They are mostly derived from
natural products such as isoborneol, proline, and norephe-
drine.17 From the results of these studies, a consistent
empirical correlation between the absolute configuration
of ligand and product has been established (Figure 3),17b
and the mechanism of the process has been studied by
Noyori from both experimental17a,b,19a and theoretical22
points of view.
As mentioned earlier, the functionalized amino alcohols
4 offer the possibility of systematic variations of steric
and electronic characteristics in order to achieve an
improved catalytic behavior. As the initial point in our
study, we synthesized two limited sets of ligands 4
containing as a common feature a bulky (tert-butyldi-
methylsilyl or trityl) R3 group (as defined in Figure 1).
The first of these sets, characterized by the presence of
an acyclic -NR22 group, comprises compounds 4g, 4j-Tr ,
and 4h -Tr , while the second one, incorporating a cyclic
dialkylamino residue, comprises 4a , 4d , and 4f.
Sch em e 8
Sch em e 9
ligands.23 On the other hand, 4h -Tr incorporates the
N,N,N′-trimethylethylenediamino moiety, a structural
motif introduced by Corey in ephedrine-related ligands.24
All three ligands were studied in the addition of Et2Zn
to benzaldehyde (7a ) (Scheme 8), employing a 6% molar
amount of ligand25 and performing the reactions at room
temperature in hexanes or toluene, as indicated in Table
5. In all instances, the predominantly obtained enanti-
omer of 1-phenylpropanol (8a ) had the S configuration,
in full agreement with the empirical model discussed
above.
While the result with 4g was mediocre, indicating that
steric congestion around the nitrogen atom is negative
to catalytic activity and enantioselectivity, much better
results were recorded with 4h -Tr and 4j-Tr , the later
being the most active ligand in this group. Guided by
the results reported by Corey,24 we also explored the use
as a catalyst of the lithium salt of 4h -Tr . Gratifyingly,
the ee of 1-phenylpropanol (8a ) increased to 92% when
a 10% molar amount of this lithium salt was used
(Scheme 9).
The components of the acyclic set were selected as
follows: 4g and 4j-Tr represent two extreme situations
in terms of steric bulk around the nitrogen atom;
moreover, the di-n-butylamino substituent has been
successfully employed by Soai in norephedrine-derived
(17) For reviews on the enantioselective addition of organometallic
reagents to carbonyl compounds see: (a) Reference 1a, Chapter 5, pp
255-297. (b) Noyori, R.; Kitamura, M. Angew. Chem., Int. Ed. Engl.
1991, 30, 49-69. (c) Soai, K. Chem. Rev 1992, 92, 833-856.
(18) Bolm, C.; Schlingloff, G.; Harms, K. Chem. Ber. 1992, 125,
1191-1203.
(19) (a) Kitamura, M.; Okada, S.; Suga, S.; Noyori, R. J . Am. Chem.
Soc. 1989, 111, 4028-4036. (b) Oguni, N.; Matsuda, Y.; Kaneko, T. J .
Am. Chem. Soc. 1988, 110, 7877-7878. (c) Noyori, R. Science 1990,
248, 1194-1199. (d) Kitamura M.; Suga, S.; Niwa, M.; Noyori, R. J .
Am. Chem. Soc. 1995, 117, 4832-4842.
In the study of the set of ligands containing a cyclic
amino residue (4a , 4d , 4f), the amount of catalyst and
the reaction conditions employed were the same as in the
preceding set; these parameters have been kept constant
throughout the present study. It is interesting to note
that a piperidino substituent (in 4d ) is optimal for
(20) (a) Soai, K.; Niwa, S.; Hori, H. J . Chem. Soc., Chem. Commun.
1990, 982-983. (b) Soai, K.; Hayase, T.; Shimada. C.; Isobe, K.
Tetrahedron: Asymmetry 1994, 5, 789-792. (c) Soai, K.; Hayase, T.;
Takai, K. Tetrahedron: Asymmetry 1995, 6, 637-638. (d) Shibata, T.;
Choji, K.; Morioka, H.; Hayase, T.; Soai, K. J . Chem. Soc., Chem.
Commun. 1996, 751-752. (e) Shibata, T.; Morioka, H.; Hayase, T.;
Choji, K.; Soai, K. J . Am. Chem. Soc. 1996, 118, 471-472. (e) Shibata,
T.; Choji, K.; Hayase, T.; Aizu, Y.; Soai, K. J . Chem. Soc., Chem.
Commun. 1996, 1235-1236. (f) Bolm, C.; Bienewald, F.; Seger, A.
Angew. Chem., Int. Ed. Engl. 1996, 35, 1657-1659.
(23) See, for example: Soai, K.; Yokoyama, S.; Hayasaka, T. J . Org.
Chem. 1991, 56, 4264-4268 and references cited therein.
(24) Corey, E. J .; Hannon, F. J . Tetrahedron Lett. 1987, 28, 5233-
5236.
(25) The optimal amount of ligand was determined for 4d as
follows: Using a 3% molar amount of ligand, and performing the
reaction in hexane, conversion was 96% after 21 h and the enantiomeric
excess of the resulting alcohol was 84%. With a 6% molar amount of
ligand, conversion was >99% after 3 h, and ee was 89%. No further
improvement in enantioselectivity was recorded by increasing the
amount of ligand employed.
(21) Oguni, N.; Omi, T. Tetrahedron Lett. 1984, 25, 2823-2824.
(22) Yamakawa, M.; Noyori, R. J . Am. Chem. Soc. 1995, 117, 6327-
6335.

Enantioselective Addition of Diethylzinc to Aldehydes
J . Org. Chem., Vol. 62, No. 15, 1997 4975
Ta ble 6. Ca ta lytic En a n tioselective Ad d ition of Et2Zn to
Ben za ld eh yd e: Effect of th e R3 P r otectin g Gr ou p
Sch em e 10
reaction
condnsb
conversion selectivity eec
ligand
5d
4d -Me CH3
4d
4d -Bp SiPh2But rt, 3 h, hex.
R3
(%)
(%)
(%)
H
rt, 21 h, tol
rt, 5 h, hex.
93
>99
>99
>99
>99
>99
>99
>99
98
72
99
98
98
99
99
99
99
98
45
69
89
90
66
88
91
92
95
SiMe2But rt, 3 h, hex.
4d -Bn CH2Ph
4d -Bzh CHPh2
rt, 5 h, hex.
rt, 3 h, tol
rt, 3 h, tol
0 °C, 7 h, tol
-18 °C, 24 h, tol
4d -Tr
4d -Tr
4d -Tr
CPh3
CPh3
CPh3
a
A 6% molar amount of ligand was used in all experiments.
tol ) toluene, hex. ) hexanes. c (S)-1-Phenyl-1-propanol was
predominantly obtained in all instances.
b
this aim, a family of ligands containing the piperidino
fragment and different R3 groups were evaluated. The
results are summarized in Table 6.
Sch em e 11
Some important conclusions can be drawn from these
results. Firstly, in order to achieve practical levels of
catalytic activity and enantioselectivity, it is important
that the primary hydroxy group be protected, as clearly
indicated by the results obtained with 5d . Secondly, the
best enantioselectivities are recorded for the bulkiest
protecting groups (4d -Bp > 4d ; 4d -Tr > 4d -Bzh > 4d -
Bn > 4d -Me) without appreciable decrease in catalytic
activity.
For the best ligand in this series, 4d -Tr , we have
investigated the effect of temperature on enantioselec-
tivity. Interestingly, as the reaction temperature is
lowered from room temperature to -18 °C, the enantio-
meric excess of the (S)-1-phenylpropanol formed increases
from 91 to 95% with the turnover of the system being
maintained within reasonable limits.
As the final parameter in this optimization, we have
tested the effect on enantioselectivity of small variations
on the amino substituent keeping constant, however, the
presence of a nitrogen-containing six-membered ring. We
have summarized in Table 7 the results obtained with
ligands 4e (morpholino substituent), 4i-Bzh (4-meth-
ylpiperazin-1-yl substituent), and 4i-Tr (4-methylpiper-
azin-1-yl substituent). For comparison, the results ob-
tained with 4d and 4d -Tr have also been included in
Table 7.
In agreement with the fact that we are already in an
optimal region of the response hypersurface, all ligands
in this group exhibit a high catalytic activity and induce
a consistently high enantioselectivity in the addition of
Et2Zn to benzaldehyde. Whereas ligand 4e does not
represent any improvement over those containing the
piperidino fragment, the presence of a 4-methylpiperazin-
1-yl moiety (a cyclic analogue of the N,N,N′-trimethyl-
ethylenediamino one) seems to impart interesting char-
acteristics to ligands 4. In comparing 4i-Bzh with 4i-
Tr , it is confirmed once again that the bulkiest protecting
groups R3 within an otherwise identical family of ligands
are the most convenient for high enantioselectivity. In
fact, 4i-Tr induces the highest enantioselectivity we have
been able to observe at room temperature with ligands
4.
enantioselectivity and catalytic activity, while a pyrro-
lidin-1-yl substituent (in 4a ) provokes a significant
decrease in catalytic activity.
Within the same structural pattern, we were also
interested in testing the effect of additional chiral centers,
located at the amino substituent, on the catalytic ef-
ficiency of the ligands. Due to their ready availability,
the enantiomeric forms of O-methylprolinol were incor-
porated into ligands 4b and 4c. When the activity of
these ligands was studied (Scheme 10) under our stan-
dard conditions, in no case were we able to detect any
improvement in the enantioselectivity of the reaction (a
79% ee is recorded with 4a , which contains an unsub-
stituted pyrrolidin-1-yl group). In light of these results,
we concentrated our efforts on the structural optimization
of ligands containing a six-membered-ring amino sub-
stituent.
Although it has been established by Noyori that anti
â-amino alcohols (which lead to cis disubstituted zinc
chelates) are in general better ligands than the corre-
sponding syn stereoisomers,17b we wanted to confirm that
the additional alkoxy group present in our ligands did
not modify this behavior. To this end, we compared the
catalytic activities of 4d and syn-4d (Scheme 11) in the
standard addition of diethylzinc to benzaldehyde.
In full agreement with the accepted mechanistic path-
way of the reaction, the syn ligand is less efficient than
the anti analogue both in terms of turnover and enantio-
selectivity. Also in agreement with previous experience,
the absolute configuration at C-2 in the ligand determines
the sense of the enantioselectivity.
We have carried out two additional studies devoted to
the improvement of the enantioselectivity in the reactions
with 4i-Tr . First, we have analyzed the effect of tem-
perature on enantioselectivity by performing reactions
at 0 and -18 °C. In this case (cf. with 4d -Tr , above),
the enantioselectivity of the reaction is completely inde-
With this new parameter in the structure of our target
ligands already fixed, we turned our attention to the
determination of the optimal protecting group R3. With

4976 J . Org. Chem., Vol. 62, No. 15, 1997
Vidal-Ferran et al.
Ta ble 7. Ca ta lytica En a n tioselective Ad d ition of Et2Zn to Ben za ld eh yd e:b Effect of th e Six-Mem ber ed Rin g Am in o
Su bstitu en t
a
b
A 6% molar amount of ligand was used in all cases. All reactions were performed at room temperature and were complete (>99%
conversion) after 3 h.
Ta ble 8. Ca ta lytic En a n tioselective Ad d ition of Et2Zn to
Ald eh yd esa (7b-h ) Lea d in g to Alcoh ols 8b-h ,b Med ia ted
by Liga n d s 4d -Tr a n d 4i-Tr
Sch em e 12
pendent of temperature in the range studied, and this
fact can be of practical importance for the potential
applications of this ligand.
On the other hand, and keeping in mind the analogy
between the 4-methylpiperazin-1-yl and the N,N,N′-
trimethylethylenediamino moieties, we have also tested
the use as a catalyst of the lithium salt of 4i-Tr . Thus,
when a 10% molar amount of this salt was used and the
reaction was performed at 0 °C in 1:1 diethyl ether:
toluene, the enantioselectivity did not significantly im-
prove (93 vs 92%). Accordingly, it appears that the most
practical conditions for the use of 4i-Tr involve working
with the neutral ligand at room temperature.
En a n tioselective Ad d ition of Dieth ylzin c to Se-
lect ed Ald eh yd es Ca t a lyzed b y t h e Op t im ized
Liga n d s 4d -Tr a n d 4i-Tr . As a confirmation of the
validity of the fine-tuning process performed on ligands
4, the optimized ligands, 4d -Tr and 4i-Tr , have been
used in the addition of diethylzinc to a family of alde-
hydes (7b-h ), including both para- and ortho-substituted
aromatic specimens, as well as the generally more
problematic aliphatic analogues.17
In order to facilitate comparison, all experiments have
been performed under identical conditions (Scheme 12).
Results on catalyst efficiency (conversion and selectivity)
and enantiomeric excess of the resulting 1-substituted
1-propanols (8b-h ) have been collected in Table 8.
Interestingly, both ligands exhibit a good profile of
enantioselectivity over the range of studied aldehydes.
Not surprisingly, the best results are obtained with
p-substituted benzaldehydes (7b-d ). However, in con-
trast with many previously reported ligands that are
catalytically active in this particular process, the ee’s
recorded with ortho-substituted benzaldehydes (7e,f) or,
even, with aliphatic ones (7g,h ) remain high.
a
b
All reactions were performed as indicated in Scheme 12. The
S enantiomer is preferentially obtained in all cases.
ligand 4i-Tr turns out to be the best choice for p-
substituted benzaldehydes, whereas 4d -Tr would be
preferable in all other situations.
As a general trend, 4d -Tr behaves as a slightly more
active ligand than 4i-Tr , the most important differences
occurring with aldehydes 7d and 7f. When catalytic
activity and enantioselectivity are jointly considered,
Con clu sion s
In summary, we have developed a family of new chiral
ligands, (1R,2R)-1-(dialkylamino)-1-phenyl-3-(R-oxy)-2-

Enantioselective Addition of Diethylzinc to Aldehydes
J . Org. Chem., Vol. 62, No. 15, 1997 4977
Regioselective Rin g Op en in g of (2S,3S)-2,3-Ep oxy-3-
p h en ylp r op a n ol by Secon d a r y Am in es in th e P r esen ce
of Ti(OiP r )4. (2R,3R)-3-P h en yl-3-(p yr r olid in -1-yl)-1,2-
p r op a n ed iol (5a ). To a solution of 1a (250 mg, 1.66 mmol)
in CH2Cl2 (12 mL) were added pyrrolidine (210 µL, 2.52 mmol)
and Ti(OiPr)4 (740 µL, 2.51 mmol) under N2 at room temper-
ature. After 5 h of stirring at room temperature, a 10%
solution of NaOH in brine (10 mL) was added, and vigorous
stirring continued for another 24 h. The mixture was filtered
through Celite and the residue washed with CH2Cl2 (10 mL).
The aqueous solution was extracted with CH2Cl2 (2 × 25 mL).
The combined organic extracts were dried and concentrated
in vacuo. The residual oil was recrystallized from Et2O to give
313 mg (85%) of 5a as white crystals: mp 95 °C; [R]23D ) -65.8
propanols, based on non-natural starting materials. The
molecular architecture of these ligands, built from widely
varying fragments grafted in a controlled manner on a
chiral skeleton that, in turn, arises from the enantiose-
lective epoxidation of an allyl alcohol, allows a gradual
variation of steric/electronic characteristics that is key
to the fine tuning of catalytic properties.
We have employed a well-established enantioselective
process, i.e., the amino alcohol-promoted addition of
diethylzinc to benzaldehyde, to iteratively optimize the
structural characteristics of the ligands for this particular
process. As a result of this study, we have identified the
steric bulk of the R-oxy group and the specification of
the dialkylamino substituent as a nitrogen-containing
six-membered ring as the key parameters for high
catalytic activity and enantioselectivity. Whereas the
steric bulk of the R-oxy group could be important in
shifting the dimer/monomer equilibrium of the ethylzinc
alkoxide derived from the ligand toward the reactive
monomeric form, the results of theoretical (AM1) calcula-
tions on the commonly admitted transition states of the
reaction suggest that interactions between the methylene
groups R to nitrogen (in ligands bearing a six-membered
nitrogen-containing ring) with the ethyl group being
transferred from zinc to carbonyl are responsible for the
high enantiofacial selectivity observed in these cases.26
Two optimized ligands arising from this analysis, 4d -Tr
and 4i-Tr , offer particular interest, allowing us to per-
form the enantioselective addition of Et2Zn to benzalde-
hyde with ee’s in the range 91-95%. Significantly, in
both cases the addition reaction can be performed at room
temperature without significant decrease, if any, in the
enantioselectivity. Closely related to 4i-Tr , the acyclic
ligand 4h -Tr also possesses practical applicability, as its
lithium salt efficiently induces the reference addition to
proceed with 92% ee. In addition, ligands 4d -Tr and 4i-
Tr have been shown to possess general applicability in
the enantioselective addition of Et2Zn to a variety of
aldehydes covering p-substituted benzaldehydes (91-96%
ee), o-substituted benzaldehydes (82-87% ee), and ali-
phatic benzaldehydes (80-86% ee).
1
(c ) 1.0 in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.26-7.33
(m, 5H), 4.23 (d × d × d, 1H, J ) 6.6, 5.7, 5.1 Hz), 3.48 (d ×
d, 1H, J ) 11.9, 5.1 Hz), 3.43 (d, 1H, J ) 5.7 Hz), 3.38 (d × d,
1H, J ) 11.9, 6.6 Hz), 2.93 (br s, 2H, OH), 2.46-2.59 (m, 4H),
1.70-1.75 (m, 4H); 13C-NMR (75 MHz, CDCl3) δ 136.4 (C),
129.3 (CH), 128.2 (CH), 127.8 (CH), 71.9 (CH), 70.7 (CH), 66.0
(CH2), 51.4 (CH2), 22.9 (CH2); IR (KBr) 3506,3267, 3089, 3060,
3027, 2973, 2952, 2937, 2881, 2786, 2732, 1142, 1075, 1024,
+
772, 696 cm-1; MS (EI) m/ z 221 (M+, 0), 160 (M - C2H5O2
,
100). Anal. Calcd for C13H19NO2: C, 70.56; H, 8.65; N, 6.33.
Found: C, 70.51; H, 8.77; N, 6.29.
(2R,3R)-3-[(2S)-2-(Met h oxym et h yl)p yr r olid in -1-yl]-3-
p h en yl-1,2-p r op a n ed iol (5b). Compound 1a (250 mg, 1.66
mmol), CH2Cl2 (12 mL), (2S)-2-(methoxymethyl)pyrrolidine
(290 µL, 2.52 mmol), and Ti(OiPr)4 (735 µL, 2.49 mmol) were
treated as described for 5a during 5 h. The reaction mixture
was quenched with a 10% solution of NaOH in brine (10 mL)
as described for 5a to give 355 mg (80%) of 5b as an oil after
chromatography using hexane:EtOAc (40:60) as eluent: [R]23
D
) -61.8 (c ) 1.0 in CHCl3); 1H-NMR (300 MHz, CDCl3) δ 7.24-
7.39 (m, 5H), 4.24 (d × d × d, 1H, J ) 9.3, 5.4, 4.2 Hz), 4.08
(d, 1H, J ) 9.3 Hz), 3.85 (d × d, 1H, J ) 11.1, 4.2 Hz), 3.69 (d
× d, 1H, J ) 11.1, 5.4 Hz), 3.42 (s, 3H), 3.40-3.50 (m, 2H),
2.81-3.01 (m, 2H), 2.22-2.31 (m, 1H), 2.10 (br s, 2H, OH),
1.24-1.67 (m, 4H); 13C-NMR (75 MHz, CDCl3) δ 134.2 (C),
129.9 (CH), 128.3 (CH), 127.8 (CH), 76.4 (CH2), 69.7 (CH), 67.9
(CH), 66.4 (CH2), 59.1 (CH), 58.7 (CH3), 48.1 (CH2), 27.2 (CH2),
23.0 (CH2); IR (film) 3396, 3106, 3087, 3062, 3029, 2927, 2875
2832, 1452, 1096, 1198, 708, 758 cm-1; MS (EI) m/ z 265 (M+,
0), 204 (M - C2H5O2+, 100).
(2R,3R)-3-[(2R)-2-(Met h oxym et h yl)p yr r olid in -1-yl]-3-
p h en yl-1,2-p r op a n ed iol (5c). Compound 1a (250 mg, 1.66
mmol), CH2Cl2 (12 mL), (2R)-2-(methoxymethyl)pyrrolidine
(290 µL, 2.52 mmol). and Ti(OiPr)4 (735 µL, 2.49 mmol) were
treated as described for 5a during 8 h. The reaction mixture
was quenched with a 10% solution of NaOH in brine (10 mL)
as described for 5a to give 350 mg (79%) of 5c as an oil after
chromatography using hexane:EtOAc (20:80) as eluent. The
residual oil was recrystallized from hexane:EtOAc (80:20) to
give 318 mg (72%) of 5c as pale brown crystals: mp 61 °C;
1
[R]23 ) -9.7 (c ) 1.0 in CHCl3); H-NMR (300 MHz, CDCl3)
D
δ 7.29-7.40 (m, 5H), 4.26 (d × d × d, 1H, J ) 6.6, 5.7, 4.1
Hz), 3.67 (d, 1H, J ) 5.7 Hz), 3.48 (d × d, 1H, J ) 11.1, 5.1
Hz), 3.41 (d × d, 1H, J ) 11.1, 6.6 Hz), 3.18 (s, 3H), 3.04-3.17
(m, 4H), 2.70-2.78 (m, 1H), 1.62-1.78 (m, 4H); 13C-NMR (75
MHz, CDCl3) δ 136.4 (C), 130.0 (CH), 128.3 (CH), 128.1 (CH),
75.3 (CH2), 71.6 (CH), 70.6 (CH), 65.3 (CH2), 59.1 (CH), 58.7
(CH3), 54.1 (CH2), 28.2 (CH2), 23.4 (CH2); IR (KBr) 3417, 3352,
3062, 3033, 2975, 2941, 2927, 2912, 2883, 2846, 2829, 2813,
2734, 1117, 1102, 1070, 1040, 743, 714 cm-1; MS (EI) m/ z 265
(M+, 0), 204 (M - C2H5O2+, 100). Anal. Calcd for C15H23NO3:
C, 67.90; H, 8.74; N, 5.28. Found: C, 67.86; H, 8.78; N, 5.31.
(2R,3R)-3-P h en yl-3-p ip er id in o-1,2-p r op a n ed iol (5d ).
Compound 1a (1.0 g, 6.66 mmol), CH2Cl2 (40 mL), piperidine
(985 µL, 9.96 mmol), and Ti(OiPr)4 (2.95 mL, 9.96 mmol) were
treated as described for 5a during 4 h. The reaction mixture
was quenched with a 10% solution of NaOH in brine (40 mL)
as described for 5a to give 1.42 g (91%) of 5d as a white solid
after recrystallization from hexane:Et2O (2:1): mp 96.5 °C;
We are currently applying the iterative process of
ligand design/fine tuning of catalytic properties to related
reactions catalyzed by amino alcohols. The correspond-
ing results will be reported in due course.
Exp er im en ta l Section
Gen er a l Meth od s. General experimental aspects have
been published elsewhere.4a (2S,3S)-2,3-Epoxy-3-phenylpro-
panol, 1a , was prepared according to the procedure described
by Sharpless; et al.;5a (2S)-2-(methoxymethyl)pyrrolidine and
(2R)-2-(methoxymethyl)pyrrolidine were prepared according
to the procedure described by Enders et al.27
(26) Vidal-Ferran, A.; Moyano, A.; Perica`s, M. A.; Riera, A. To be
published.
(27) Enders, D.; Fey, P.; Kipphardt, H. Org. Synth. 1987, 65, 173-
182.
[R]23 ) +11.7 (c ) 1.3 in CHCl3); 1H-NMR (300 MHz, CDCl3)
D
δ 7.22-7.40 (m, 5H), 5.35 (br s, 1H, OH), 4.34 (d × d × d, 1H,
J ) 9.3, 7.5, 4.8 Hz), 3.84 (d × d, 1H, J ) 10.8, 4.8 Hz), 3.73

4978 J . Org. Chem., Vol. 62, No. 15, 1997
Vidal-Ferran et al.
(d × d, 1H, J ) 10.8, 7.5 Hz), 3.59 (d, 1H, J ) 9.3 Hz), 2.51-
2.53 (br s, 2H), 2.30 (br s, 2H), 1.70 (br s, 1H, OH), 1.51-1.60
(m, 4H), 1.29-1.34 (m, 2H); 13C-NMR (75 MHz, CDCl3) δ 133.2
(C), 129.7 (CH), 128.4 (CH), 128.2 (CH), 75.7 (CH), 67.4 (CH2),
67.1 (CH), 51.6 (CH2), 25.9 (CH2), 23.9 (CH2); IR (KBr) 3544,
3286, 3064, 3029, 2993, 2933, 2852, 2792, 2751, 1451, 1123,
1088, 1063, 1026, 754, 708, 698 cm-1; MS (EI) m/ z 235 (M+,
0.1), 174 (M - C2H5O2+, 100). Anal. Calcd for C14H21NO2: C,
71.46; H, 8.99; N, 5.95. Found: C, 71.48; H, 9.09; N, 5.93.
(2R,3R)-3-Mor p h olin o-3-p h en yl-1,2-p r op a n ed iol (5e).
Compound 1a (250 mg, 1.66 mmol), CH2Cl2 (10 mL), morpho-
line (220 µL, 2.52 mmol), and Ti(OiPr)4 (735 µL, 2.49 mmol)
were treated as described for 5a during 5 h. The reaction
mixture was quenched with a 10% solution of NaOH in brine
(10 mL) as described for 5a to give 345 mg (87%) of 5e as a
(d × d × d, 1H, J ) 9.6, 1.8, 1.8 Hz), 4.20 (br s, 2H, OH), 4.07
(d × d, 1H, J ) 11.7, 1.8 Hz), 3.83 (d, 1H, J ) 9.6 Hz), 3.67 (d
× d, 1H, J ) 11.7, 1.8 Hz), 2.99-3.11 (m, 2H), 2.33 (s, 6H),
2.14-2.45 (m, 4H), 2.05 (s, 3H); 13C-NMR (75 MHz, CDCl3) δ
133.8 (C), 129.7 (CH), 128.1 (CH), 127.5 (CH), 70.8 (CH), 64.2
(CH2), 63.4 (CH), 55.5 (CH2), 50.4 (CH2), 44.2 (CH3), 38.4 (CH3);
IR (film) 3359, 3106, 3087, 3062, 3029, 2942, 2860, 2821, 2794,
1452, 1366, 1102, 1072, 1042, 752, 704 cm-1; MS (EI) m/ z 252
(M+, 0), 191 (M - C2H5O2+, 10), 194 (M - C3H8N+, 100).
Selective p r otection of th e p r im a r y h yd r oxy gr ou p in
5a -h a s a silyl eth er .
(2R,3R)-1-[(ter t-Bu t yld im et h ylsilyl)oxy]-3-p h en yl-3-
(p yr r olid in -1-yl)-2-p r op a n ol (4a ). A solution of 5a (440 mg,
1.99 mmol), tert-butyldimethylsilyl chloride (330 mg, 2.19
mmol), and imidazole (299 mg, 4.39 mmol) in DMF (18 mL)
was heated at 65 °C for 24 h under N2. The reaction mixture
was cooled to room temperature, and Et2O (18 mL) and brine
(18 mL) were added. The aqueous layer was extracted with
Et2O (2 × 25 mL). The combined organic extracts were dried
and concentrated in vacuo. The residual oil was chromato-
graphed using hexane:EtOAc (90:10-80:20) as eluent to give
white solid after recrystallization from hexane:EtOAc (2:1):
1
mp 103 °C; [R]23 ) -23.8 (c ) 1.1 in CHCl3); H-NMR (300
D
MHz, CDCl3) δ 7.24-7.42 (m, 5H), 4.34 (m, 1H), 3.66-3.70
(m, 5H), 3.63 (d × d, 1H, J ) 10.5, 5.1 Hz), 3.49 (d, 1H, J )
7.5 Hz), 2.43-2.55 (m, 4H), 2.02 (br s, 1H, OH), 1.59 (br s,
1H, OH); 13C-NMR (75 MHz, CDCl3) δ 133.6 (C), 129.6 (CH),
128.5 (CH), 128.4 (CH), 73.7 (CH), 68.2 (CH), 66.7 (CH2), 66.2
529 mg (79%) of 4a as a colorless oil: [R]23 ) -19.7 (c ) 1.0
D
1
(CH2), 50.9 (CH2); IR (KBr) 3442, 2815, 1497, 1449, 1117, 872,
in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.27-7.34 (m, 5H),
+
777, 719 cm-1; MS (EI) m/ z 237 (M+, 0.1), 176 (M - C2H5O2
,
4.15 (d × d × d, 1H, J ) 6.6, 6.3, 3.9 Hz), 3.38 (d × d, 1H, J
) 9.9, 6.3 Hz), 3.33 (d, 1H, J ) 3.9 Hz), 3.20 (d × d, 1H, J )
9.9, 6.6 Hz), 3.14 (br s, 1H, OH), 2.60-2.63 (m, 2H), 2.42-
2.46 (m, 2H), 1.74-1.78 (m, 4H), 0.88 (s, 9H), 0.00 (s, 3H),
-0.02 (s, 3H); 13C-NMR (75 MHz, CDCl3) δ 138.0 (C), 129.3
(CH), 127.8 (CH), 127.3 (CH), 71.6 (CH), 71.2 (CH), 64.3 (CH2),
52.4 (CH2), 25.9 (CH3), 23.2 (CH2), 18.2 (C), -5.5 (CH3); IR
100). Anal. Calcd for C13H19NO3‚1/3H2O: C, 64.18; H, 8.15;
N, 5.76. Found: C, 64.12; H, 8.39; N, 5.80.
(2R,3R)-3-(Azep a n -1-yl)-3-p h en yl-1,2-p r op a n ed iol (5f).
Compound 1a (750 mg, 4.99 mmol), CH2Cl2 (25 mL), hexa-
methyleneimine (840 µL, 7.45 mmol), and Ti(OiPr)4 (2.20 mL,
7.46 mmol) were treated as described for 5a during 8 h. The
reaction mixture was quenched with a 10% solution of NaOH
in brine (30 mL) as described for 5a to give 1.13 g (91%) of 5f
as a white solid after recrystallization from Et2O: mp 94 °C;
(film) 3454, 3064, 3031, 2956, 2858, 2804, 1115, 757, 702 cm-1
;
MS (CI, NH3) m/ z 336 (C19H33NO2Si‚H+, 100).
(2R,3R)-1-[(ter t-Bu tyldim eth ylsilyl)oxy]-3-[(2S)-2-(m eth -
oxym eth yl)p yr r olid in -1-yl]-3-p h en yl-2-p r op a n ol (4b). A
solution of 5b (567 mg, 2.14 mmol), tert-butyldimethylsilyl
chloride (354 mg, 2.35 mmol), and imidazole (320 mg, 4.71
mmol) in DMF (20 mL) was heated at 65 °C for 24 h under
N2. A workup identical to the one described for 4a followed
by chromatography using hexane:EtOAc (90:10) as eluent
[R]23 ) +17.5 (c ) 1.0 in CHCl3); 1H-NMR (300 MHz, CDCl3)
D
δ 7.26-7.38 (m, 5H), 4.27 (d × d × d, 1H, J ) 9.0, 6.6, 5.1
Hz), 3.83 (d × d, 1H, J ) 10.9, 5.1 Hz), 3.75 (d × d, 1H, J )
10.9, 6.6 Hz), 3.74 (d, 1H, J ) 9.0 Hz), 2.73 (d × t, 2H, J )
12.7, 6.4 Hz), 2.52 (d × t, 2H, J ) 12.7, 6.4 Hz), 1.80 (br s, 2H,
OH), 1.50-1.61 (m, 8H); 13C-NMR (75 MHz, CDCl3) δ 135.0
(C), 129.3 (CH), 128.3 (CH), 127.9 (CH), 75.2 (CH), 68.0 (CH),
67.5 (CH2), 53.3 (CH2), 28.5 (CH2), 26.3 (CH2); IR (KBr) 3384,
3226, 3087, 3060, 3025, 3002, 2966, 2931, 2885, 2854, 1451,
1059, 1022, 762, 702 cm-1; MS (EI) m/ z 249 (M+, 0), 188 (M
- C2H5O2+, 100). Anal. Calcd for C15H23NO2: C, 72.25; H,
9.30; N, 5.62. Found: C, 72.33; H, 9.41; N, 5.62.
yielded 615 mg (76%) of 4b as a colorless oil: [R]23 ) -42.4
D
1
(c ) 1.1 in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.33-7.35
(m, 5H), 4.19 (d × d × d, 1H, J ) 7.2, 6.3, 6.0 Hz), 3.87 (d, 1H,
J ) 7.2 Hz), 3.53-3.63 (m, 2H), 3.46 (d × d, 1H, J ) 9.3, 6.3
Hz), 3.40 (s, 3H), 3.28 (d × d, 1H, J ) 9.3, 6.0 Hz), 2.76-3.09
(m, 2H), 2.32-2.40 (m, 1H), 1.45-1.76 (m, 4H), 0.91 (s, 9H),
0.05, 0.07 (s, 3H+3H); 13C-NMR (75 MHz, CDCl3) δ 136.3 (C),
130.0 (CH), 127.7 (CH), 127.2 (CH), 76.8 (CH2), 71.0 (CH), 67.0
(CH), 65.0 (CH2), 59.0 (CH + CH3), 49.8 (CH2), 28.0 (CH2), 25.9
(CH3), 23.3 (CH2), 18.3 (C), -5.4 (CH3); IR (film) 3458, 3029,
2929, 2858, 1117, 778, 704 cm-1; MS (CI, NH3) m/ z 380
(C21H37NO3Si‚H+, 100).
(2R,3R)-3-(Diisop r op yla m in o)-3-p h en yl-1,2-p r op a n e-
d iol (5g). Compound 1a (650 mg, 4.33 mmol), CH2Cl2 (20 mL),
diisopropylamine (910 µL, 6.49 mmol), and Ti(OiPr)4 (1.91 mL,
6.47 mmol) were treated as described for 5a during 7 h. The
reaction mixture was quenched with a 10% solution of NaOH
in brine (20 mL) as described for 5a to give 5g as an oil after
chromatography using hexane:EtOAc (20:80-40:60) as eluent.
The oil was recrystallized from hexane:Et2O (10:1) to give 155
(2R,3R)-1-[(ter t-Bu tyldim eth ylsilyl)oxy]-3-[(2R)-2-(m eth -
oxym eth yl)p yr r olid in -1-yl]-3-p h en yl-2-p r op a n ol (4c). A
solution of 5c (425 mg, 1.60 mmol), tert-butyldimethylsilyl
chloride (265 mg, 1.76 mmol), and imidazole (241 mg, 3.54
mmol) in DMF (15 mL) was heated at 65 °C for 24 h under
N2. A workup identical to the one described for 4a followed
by chromatography using hexane:EtOAc (80:20) as eluent
mg (14%) of 5g as a white solid: mp 114 °C; [R]23 ) -51.9 (c
D
1
) 1.0 in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.26-7.38 (m,
5H), 4.16 (d × d × d, 1H, J ) 9.6, 5.7, 5.4 Hz), 3.90 (d, 1H, J
) 9.6 Hz), 3.83 (d × d, 1H, J ) 10.9, 5.4 Hz), 3.74 (d × d, 1H,
J ) 10.9, 5.7 Hz), 3.33 (h, 2H, J ) 6.6 Hz), 1.75 (br s, 2H,
OH), 1.14 (d, 6H, J ) 6.6 Hz), 0.80 (d, 6H, J ) 6.6 Hz); 13C-
NMR (75 MHz, CDCl3) δ 139.7 (C), 129.9 (CH), 128.5 (CH),
127.5 (CH), 70.7 (CH), 65.9 (CH2), 62.6 (CH), 46.4 (CH), 21.4
(CH3), 23.6 (CH3); IR (KBr) 3282, 3195, 3031, 2981, 2966, 2914,
yielded 520 mg (85%) of 4c as a colorless oil: [R]23 ) +9.7 (c
D
1
) 1.1 in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.27-7.39 (m,
5H), 4.18 (d × d × d, 1H, J ) 6.3, 4.2, 3.6 Hz), 3.66 (d, 1H, J
) 4.2 Hz), 3.44 (d × d, 1H, J ) 10.0, 3.6 Hz), 3.27 (d × d, 1H,
J ) 10.0, 6.3 Hz), 3.05-3.18 (m, 2H), 3.12 (s, 3H), 2.95-3.05
(m, 2H), 2.64-2.73 (m, 1H), 1.63-1.74 (m, 4H), 0.86 (s, 9H),
0.02 (s, 3H), -0.05 (s, 3H); 13C-NMR (75 MHz, CDCl3) δ 137.8
(C), 130.1 (CH), 127.8 (CH), 127.5 (CH), 75.9 (CH2), 72.0 (CH),
69.6 (CH), 64.5 (CH2), 59.0 (CH), 58.6 (CH3), 53.6 (CH2), 28.2
(CH2), 25.8 (CH3), 23.5 (CH2), 18.2 (C), -5.5 (CH3); IR (film)
3440, 3029, 2956, 2929, 2858, 1115, 778, 704 cm-1; MS (CI,
NH3) m/ z 380 (C21H37NO3Si‚H+, 100).
2869, 1452, 1395, 1383, 1102, 1082, 1061, 1036, 722, 698 cm-1
;
MS (EI) m/ z 251 (M+, 0), 190 (M - C2H5O2+, 100). Anal.
Calcd for C15H25NO2: C, 71.67; H, 10.25; N, 5.34. Found: C,
71.67; H, 10.02; N, 5.57.
(2R,3R)-3-[[[2-(Dim eth yla m in o)eth yl]m eth yl]a m in o]-3-
p h en yl-1,2-p r op a n ed iol (5h ). Compound 1a (1.0 g, 6.66
mmol), CH2Cl2 (40 mL), N,N,N′-trimethylethylenediamine
(1.27 mL, 9.99 mmol), and Ti(OiPr)4 (2.95 mL, 9.99 mmol) were
treated as described for 5a during 8 h. The reaction mixture
was quenched with a 10% solution of NaOH in brine (40 mL)
as described for 5a to give 1.172 mg (70%) of 5h as an oil after
chromatography using hexane:EtOAc (40:60), EtOAc and
(2R,3R)-1-[(ter t-Bu tyld im eth ylsilyl)oxy]-3-p h en yl-3-p i-
p er id in o-2-p r op a n ol (4d ). A solution of 5d (783 mg, 3.33
mmol), tert-butyldimethylsilyl chloride (551 mg, 3.65 mmol),
and imidazole (500 mg, 7.34 mmol) in DMF (30 mL) was
heated at 65 °C for 14 h under N2. A workup identical to the
one described for 4a followed by chromatography using hexane:
EtOAc:EtOH (90:10) as eluents: [R]23 ) -46.2 (c ) 1.1 in
D
CHCl3); 1H-NMR (300 MHz, CDCl3) δ 7.22-7.37 (m, 5H), 4.22

Enantioselective Addition of Diethylzinc to Aldehydes
J . Org. Chem., Vol. 62, No. 15, 1997 4979
EtOAc (90:10-80:20) as eluent yielded 889 mg (76%) of 4d as
a colorless oil: [R]23D ) -9.3 (c ) 1.0 in CHCl3); 1H-NMR (300
MHz, CDCl3) δ 7.29-7.32 (m, 5H), 4.29 (d × d × d, 1H, J )
6.6, 6.0, 4.8 Hz), 3.62 (d × d, 1H, J ) 9.9, 6.0 Hz), 3.55 (d × d,
1H, J ) 9.9, 4.8 Hz), 3.48 (d, 1H, J ) 6.6 Hz), 3.04 (br s, 1H,
OH), 2.43 (br s, 4H), 1.57-1.59 (m, 4H), 1.34-1.40 (m, 2H),
0.90 (s, 9H), 0.05 (s, 3H), 0.03 (s, 3H); 13C-NMR (75 MHz,
CDCl3) δ 134.5 (C), 129.8 (CH), 127.9 (CH), 127.5 (CH), 71.6
(CH), 69.5 (CH), 64.7 (CH2), 51.9 (CH2), 25.90 (CH2), 25.91
(CH3), 24.3 (CH2), 18.3 (CH2), -5.5, -5.4 (CH3); IR (film) 3448,
3064, 3029, 2932, 2858, 2805, 1252, 1113, 778, 702 cm-1; MS
(CI, NH3) m/ z 350 (C20H35NO2Si‚H+, 100).
(CH), 70.4 (CH), 70.2 (CH2), 65.0 (CH), 57.9 (CH2), 52.2 (CH2),
45.8 (CH3), 40.0 (CH3), 25.9 (CH3), 18.5 (C), -5.4 (CH3); MS
(CI, NH3) m/ z 367 (C20H38N2O2Si‚H+, 100) for 4h , 253
(C14H24N2O2‚H+, 7) for 5h .
(2R,3R)-1-[(ter t-Bu tyld ip h en ylsilyl)oxy]-3-p h en yl-3-p i-
p er id in o-2-p r op a n ol (4d -Bp ). A solution of 5b (200 mg, 0.85
mmol), tert-butyldiphenylsilyl chloride (240 µL, 0.94 mmol),
and imidazole (127 mg, 1.87 mmol) in DMF (10 mL) was
heated at 65 °C for 24 h under N2. A workup identical to the
one described for 4a followed by chromatography using hexane:
EtOAc (95:5-90:10) as eluent yielded 400 mg (99%) of 4d -Bp
as a colorless oil: [R]23 ) -10.3 (c ) 1.0 in CHCl3); 1H-NMR
D
(2R,3R)-1-[(ter t-Bu tyld im eth ylsilyl)oxy]-3-m or p h olin o-
3-p h en yl-2-p r op a n ol (4e). A solution of 5e (1.19 g, 5.01
mmol), tert-butyldimethylsilyl chloride (830 mg, 5.51 mmol),
and imidazole (753 mg, 11.06 mmol) in DMF (45 mL) was
heated at 65 °C for 24 h under N2. A workup identical to the
one described for 4a followed by chromatography using hexane:
EtOAc (90:10-80:20) as eluent yielded 1.31 g (74%) of 4e as a
(300 MHz, CDCl3) δ 7.60-7.70 (m, 4H), 7.22-7.41 (m, 11H),
4.28 (d × d × d, 1H, J ) 7.7, 6.0, 4.7 Hz), 3.78 (d × d, 1H, J
) 10.2, 6.0 Hz), 3.69 (d × d, 1H, J ) 10.2, 4.7 Hz), 3.49 (d, 1H,
J ) 7.7 Hz), 2.70 (br s, 1H, OH), 2.28 (t, 4H, J ) 5.0 Hz), 1.41-
1.47 (m, 4H), 1.26-1.33 (m, 2H), 1.06 (s, 9H); 13C-NMR (75
MHz, CDCl3) δ 135.9 (C), 135.53 (CH), 135.47 (CH), 133.34
(C), 133.28 (C), 129.6 (CH), 129.5 (CH), 127.7 (CH), 127.6 (CH),
127.2 (CH), 71.3 (CH), 69.7 (CH), 65.6 (CH2), 51.5 (CH2), 26.8
(CH3), 26.3 (CH2), 24.5 (CH2), 19.2 (CH); IR (film) 3575, 3450,
colorless oil: [R]23 ) -10.4 (c ) 1.0 in CHCl3); 1H-NMR (300
D
MHz, CDCl3) δ 7.27-7.31 (m, 5H), 4.23 (d × d × d, 1H, J )
6.0, 5.7, 5.4 Hz), 3.68 (t, 4H, J ) 5.1 Hz), 3.51 (d × d, 1H, J )
9.9, 6.0 Hz), 3.43 (d × d, 1H, J ) 9.9, 5.4 Hz), 3.40 (d, 1H, J
) 5.7 Hz), 2.84 (br s, 1H, OH), 2.38-2.48 (m, 4H), 0.86 (s, 9H),
0.00 (s, 3H), -0.01 (s, 3H); 13C-NMR (75 MHz, CDCl3) δ 135.3
(C), 129.7 (CH), 128.1 (CH), 127.8 (CH), 71.4 (CH), 69.2 (CH),
67.0 (CH2), 64.5 (CH2), 51.3 (CH2), 25.9 (CH3), 18.2 (C), -5.4
(CH3), -5.5 (CH3); IR (film) 3460, 3064, 3029, 2956, 2858, 1119,
1005, 778, 704 cm-1; MS (CI, NH3) m/ z 352 (C19H33NO3Si‚H+,
100).
3072, 2933, 2858, 2805, 1472, 1452, 1113, 1067, 741, 702 cm-1
;
MS (CI, NH3) m/ z 474 (C30H39NO2Si‚H+, 100).
Selective P r otection of th e P r im a r y Hyd r oxyl Gr ou p
in 5d a s a Tr ityl Eth er : (1R,2R)-1-P h en yl-1-p ip er id in o-
3-(tr ip h en ylm eth oxy)-2-p r op a n ol (4d -Tr ). A solution of 5d
(300 mg, 1.27 mmol) and triphenylmethyl chloride (426 mg,
1.53 mmol) in pyridine (12 mL) was heated at 90 °C for 12 h
under N2. The solvent was removed in vacuo, and the residual
oil was chromatographed using firstly hexane:Et2O (90:10) as
eluent to separate the excess of tritylating agent and finally
hexane:EtOAc (80:20) to give 443 mg (73%) of 4d -Tr as white
(1R,2R)-1-(Azep a n -1-yl)-3-[(ter t-b u t yld im et h ylsilyl)-
oxy]-1-p h en yl-2-p r op a n ol (4f). A solution of 5f (500 mg,
2.00 mmol), tert-butyldimethylsilyl chloride (332 mg, 2.20
mmol), and imidazole (302 mg, 4.44 mmol) in DMF (20 mL)
was heated at 65 °C for 24 h under N2. A workup identical to
the one described for 4a followed by chromatography using
crystals: mp 150 °C; [R]23 ) +3.1 (c ) 1.1 in CHCl3); 1H-
D
NMR (300 MHz, CDCl3) δ 7.40-7.43 (m, 6H), 7.17-7.29 (m,
14H), 4.33 (d × d × d, 1H, J ) 7.2, 6.3, 4.8 Hz), 3.46 (d, 1H, J
) 7.2 Hz), 3.27 (d × d, 1H, J ) 9.6, 6.3 Hz), 3.04 (d × d, 1H,
J ) 9.6, 4.8 Hz), 2.78 (br s, 1H, OH), 2.23-2.30 (br s, 4H),
1.41-1.48 (m, 4H), 1.28-1.33 (m, 2H); 13C-NMR (75 MHz,
CDCl3) δ 144.1 (C), 136.1 (C), 129.4 (CH), 128.6 (CH), 127.74
(CH), 127.70 (CH), 127.2 (CH), 126.9 (CH), 86.7 (C), 71.7 (CH),
68.8 (CH), 65.6 (CH2), 51.7 (CH2), 26.3 (CH2), 24.5 (CH2); IR
(KBr) 3357, 3056, 3033, 2966, 2929, 2879, 2861, 2807, 1597,
1491, 1449, 1094, 1079, 764, 754, 702 cm-1; MS (CI, NH3) m/ z
478 (C33H35NO2‚H+, 100). Anal. Calcd for C33H35NO2: C,
82.98; H, 7.39; N, 2.93. Found: C, 83.10; H, 7.46; N, 2.97.
P r ot ect ion of t h e Hyd r oxy Gr ou p in 1a w it h R 3X
Rea gen ts. (2S,3S)-2-(m eth oxym eth yl)-3-p h en yloxir a n e
(6-Me). A solution of 1a (1.0 g, 6.66 mmol) in DMF (7 mL)
was added via canula to a suspension of sodium hydride (231
mg, ca. 7.7 mmol) in DMF (8 mL) at -20 °C under N2. The
mixture was stirred for 20 min, and methyl iodide (540 µL,
8.66 mmol) was syringed into the mixture. After being stirred
for 4 h at -20 °C, the mixture was allowed to reach room
temperature and stirred for another hour. MeOH (25 mL) and
brine (25 mL) were added. The aqueous solution was extracted
with CH2Cl2 (3 × 50 mL). The combined organic extracts were
dried and concentrated in vacuo. The residual oil was chro-
matographed using hexane:Et2O (90:10-70:30) as eluent to
hexane:Et2O (97:3) as eluent yielded 518 mg (71%) of 4f as a
1
colorless oil: [R]23 ) -3.5 (c ) 1.2 in CHCl3); H-NMR (300
D
MHz, CDCl3) δ 7.25-7.30 (m, 5H), 4.17 (d × d × d, 1H, J )
7.8, 5.7, 4.8 Hz), 3.64 (d, 1H, J ) 7.8 Hz), 3.63-3.72 (m, 2H),
2.75 (br s, 1H, OH), 2.50-2.70 (m, 4H), 1.54 (br s, 8H), 0.91
(s, 9H), 0.07 (s, 3H), 0.06 (s, 3H); 13C-NMR (75 MHz, CDCl3)
δ 137.3 (C), 129.3 (CH), 127.8 (CH), 127.1 (CH), 70.41 (CH),
70.38 (CH2), 65.2 (CH2), 52.6 (CH2), 29.2 (CH2), 26.8 (CH2),
25.9 (CH3), 18.3 (C), -5.3 (CH3), -5.4 (CH3); IR (film) 3570,
3442, 3087, 3064, 3029, 2929, 2858, 1463, 1452, 1117, 1075,
1061, 778, 702 cm-1; MS (CI, NH3) m/ z 364 (C21H37NO2Si‚H+,
100).
(2R,3R)-1-[(ter t-Bu t yld im et h ylsilyl)oxy]-3-(d iisop r o-
p yla m in o)-3-p h en yl-2-p r op a n ol (4g). A solution of 5g (175
mg, 0.70 mmol), tert-butyldimethylsilyl chloride (115 mg, 0.76
mmol), and imidazole (105 mg, 1.54 mmol) in DMF (7 mL) was
heated at 65 °C for 24 h under N2. A workup identical to the
one described for 4a followed by chromatography using hexane:
Et2O (99:1) as eluent yielded 133 mg (52%) of 4g as a colorless
oil: [R]23 ) -39.8 (c ) 0.9 in CHCl3); 1H-NMR (300 MHz,
D
CDCl3) δ 7.25-7.46 (m, 5H), 4.10 (d × d × d, 1H, J ) 8.5, 6.7,
3.7 Hz), 3.88 (d, 1H, J ) 8.5 Hz), 3.86 (d × d, 1H, J ) 10.0, 3.7
Hz), 3.64 (d × d, 1H, J ) 10.0, 6.7 Hz), 3.31 (h, 2H J ) 6.5
Hz), 2.30 (br s, 1H, OH), 1.08 (d, 6H J ) 6.5 Hz), 0.91 (s, 9H),
0.80 (d, 6H, J ) 6.5 Hz), 0.07 (s, 3H), 0.06 (s, 3H); 13C-NMR
(75 MHz, CDCl3) δ 141.1 (C), 129.8 (CH), 127.9 (CH), 126.7
(CH), 71.8 (CH), 65.9 (CH2), 60.3 (CH), 46.1 (CH), 25.9 (CH3),
23.9 (CH3), 21.5 (CH3), 18.3 (C), -5.3 (CH3), -5.4 (CH3); IR
(film) 3571, 3064, 3031, 2962, 2931, 2860, 1464, 1115, 778, 702
cm-1; MS (CI, NH3) m/ z 366 (C21H39NO2Si‚H+, 100).
give 995 mg (91%) of 6-Me as an oil: [R]23 ) -43.9 (c ) 1.0
D
1
in CHCl3); H-NMR (200 MHz, CDCl3) δ 7.25-7.33 (m, 5H),
3.79 (d, 1H, J ) 2.2 Hz), 3.77 (d × d, 1H, J ) 11.6, 2.8 Hz),
3.53 (d × d, 1H, J ) 11.6, 5.6 Hz), 3.44 (s, 3H), 3.20 (m, 1H);
13C-NMR (50 MHz, CDCl3) δ 136.8 (C), 128.4 (CH), 128.2 (CH),
125.6 (CH), 72.1 (CH2), 60.9 (CH), 59.2 (CH3), 55.7 (CH); IR
(film) 3031, 2989, 2931, 2883, 2829, 1605, 1499, 1463, 1436,
1378, 1138, 1117, 752, 698 cm-1; MS (EI) m/ z 164 (M+, 1),
121 (100).
(2R,3R)-1-[(ter t-Bu tyld im eth ylsilyl)oxy]-3-[[[2-(d im eth -
yla m in o)eth yl]m eth yl]a m in o]-3-p h en yl-2-p r op a n ol (4h ).
A solution of 5h (480 mg, 1.90 mmol), tert-butyldimethylsilyl
chloride (315 mg, 2.09 mmol), and imidazole (286 mg, 4.20
mmol) in DMF (18 mL) was heated at 65 °C for 24 h under
N2. A workup to the one described for 4a followed by
chromatography using hexane:EtOAc (50:50-30:70), EtOAc,
and EtOAc:EtOH (90:10) as eluents yielded 460 mg of a
mixture of 4h and 5h (8:1) as a colorless oil: 13C-NMR for 5h
(50 MHz, CDCl3) δ 136.4 (C), 129.7 (CH), 127.8 (CH), 127.3
P r ep a r a tion of (2S,3S)-2-(Ben zyloxym eth yl)-3-p h en yl-
oxir a n e (6-Bn ) Usin g Ben zyl Br om id e a s P r ot ect in g
Rea gen t. Compound 1a (380 mg, 2.53 mmol) in DMF (5 mL),
sodium hydride (88 mg, ca. 2.9 mmol) in DMF (3 mL), and
benzyl bromide (300 µL, 2.54 mmol) were treated as described
for 6-Me to give 558 mg (91%) of 6-Bn as an oil after
chromatography using hexane:Et2O (90:10) as eluent: [R]23
D
) -36.1 (c ) 0.9 in CHCl3); 1H-NMR (200 MHz, CDCl3) δ 7.28-
7.37 (m, 10H), 3.85 (d × d, 1H, J ) 11.6, 2.8 Hz), 3.78 (d, 1H,

4980 J . Org. Chem., Vol. 62, No. 15, 1997
Vidal-Ferran et al.
J ) 2.2 Hz), 3.61 (d × d, 1H, J ) 11.6, 5.2 Hz), 3.24 (m, 1H);
13C-NMR (50 MHz, CDCl3) δ 137.8 (C), 136.8 (C), 128.42 (CH),
128.39 (CH), 128.2 (CH), 127.7 (CH), 125.6 (CH), 73.2 (CH2),
69.8 (CH2), 61.1 (CH), 55.8 (CH); IR (film) 3064, 3031, 2860,
1605, 1497, 1455, 1102, 743, 698 cm-1; MS (CI, NH3) m/ z 258
(C16H16O2‚H‚NH3+, 100).
P r ep a r a tion of (2S,3S)-2-(Ben zyloxym eth yl)-3-p h en yl-
oxir a n e (6-Bn ) Usin g Ben zyl Ch lor id e a s P r ot ect in g
Rea gen t. Compound 1a (200 mg, 1.33 mmol) in DMF (2 mL),
sodium hydride (52 mg, ca. 1.7 mmol) in DMF (2 mL) and
benzyl chloride (190 µL, 1.40 mmol) were treated as described,
using benzyl bromide as the protecting reagent and stirring
36 h at -20 °C, to give 262 mg (82%) of 6-Bn . This compound
was spectroscopically identical to the one obtained using benzyl
bromide as reagent.
oil that was recrystallized afterwards from hexane. This
compound was spectroscopically identical to the one obtained
using chlorotriphenylmethane and sodium hydride as re-
agents.
(2S,3S)-3-P h en yl-2-[(ter t-b u t yld im et h ylsilyl)oxy]oxi-
r a n e (6-Tbs). A solution of 1a (500 mg, 3.33 mmol), tert-
butyldimethylsilyl chloride (552 mg, 3.66 mmol), and imidazole
(499 mg, 7.32 mmol) in DMF (30 mL) was stirred at room
temperature for 12 h under N2. Et2O (25 mL) and a saturated
NH4Cl solution (25 mL) were added. The aqueous layer was
extracted with Et2O (2 × 25 mL). The combined organic
extracts were dried and concentrated in vacuo. The residual
oil was chromatographed using hexane:Et2O (99:1) as eluent
to give 695 mg (79%) of 6-Tbs as a colorless oil: [R]23D ) -28.5
(c ) 1.3 in CHCl3) (lit.28 [R]23 ) -26.8 (c ) 1.04 g in MeOH));
D
(2S ,3S )-2-[(Dip h e n ylm e t h oxy)m e t h yl]-3-p h e n yloxi-
r a n e (6-Bzh ). Compound 1a (1.0 g, 6.66 mmol) in DMF (7
mL), sodium hydride (231 mg, ca. 7.7 mmol) in DMF (8 mL),
and benzhydryl bromide (2.06 g, 8.32 mmol) were treated as
described for 6-Me, stirring 22 h at 0 °C, to give 1.440 g (68%)
of 6-Bzh as an oil after chromatography, using hexane:Et2O
(98:2) as eluent. One hundred mg of the oil was recrystallized
from hexane to give analytically pure white crystals: mp 62
spectroscopic data were in agreement with the those already
described for 6-Tbs.28
Lit h iu m P er ch lor a t e-In d u ced R egioselect ive R in g
Op en in g of E p oxy E t h er s 6 by Secon d a r y Am in es.
(2R,3R)-1-Meth oxy-3-ph en yl-3-piper idin opr opan -2-ol (4d-
Me). Piperidine (990 µL, 10 mmol) was syringed into a
mixture of 6-Me (164 mg, 1 mmol) and LiClO4 (1.65 g, 15.5
mmol) in acetonitrile (2.5 mL) at 55 °C under N2. After the
mixture was stirred for 24 h at 55 °C, H2O (20 mL) was added,
and the aqueous layer was extracted with CH2Cl2 (3 × 25 mL).
The combined organic extracts were dried and concentrated
in vacuo. The residual oil was chromatographed using hexane:
EtOAc (90:10-80:20) as eluent to give 234 mg (94%) of 4d -
Me as a colorless oil that was afterwards recrystallized from
°C; [R]23 ) -30.9 (c ) 1.1 in CHCl3); 1H-NMR (300 MHz,
D
CDCl3) δ 7.21-7.40 (m, 15H), 3.82 (d × d, 1H, J ) 11.5, 3.0
Hz), 3.75 (d, 1H, J ) 2.1 Hz), 3.62 (d × d, 1H, J ) 11.5, 5.1
Hz), 3.26 (m, 1H); 13C-NMR (75 MHz, CDCl3) δ 141.73 (C),
141.67 (C), 136.9 (C), 128.4 (CH), 128.2 (CH), 127.6 (CH), 127.5
(CH), 127.0 (CH), 126.9 (CH), 125.7 (CH), 83.9 (CH), 68.6
(CH2), 61.3 (CH), 55.9 (CH); IR (KBr) 3087, 3062, 3029, 3004,
hexane:Et2O to give 224 mg (90%) of 4d -Me as white crys-
2861, 1601, 1586, 1495, 1453, 1096, 1079, 741, 702, 695 cm-1
;
1
tals: mp 69 °C; [R]23 ) -39.1 (c ) 1.0 in CHCl3); H-NMR
D
MS (CI, NH3) m/ z 334 (C22H20O2‚NH4+, 59), 317 (C22H20O2‚H+,
17), 167 (100). Anal. Calcd for C22H20O2: C, 83.51; H, 6.38.
Found: C, 83.64; H, 6.43.
(300 MHz, CDCl3) δ 7.27-7.31 (m, 5H), 4.39 (d × d × d, 1H,
J ) 7.2, 6.3, 3.6 Hz), 3.31 (s, 3H), 3.32 (d, 1H, J ) 6.3 Hz),
3.31 (d × d, 1H, J ) 9.6, 3.6 Hz), 3.16 (d × d, 1H, J ) 9.6, 7.2
Hz), 2.92 (br s, 1H, OH), 2.30-2.40 (m, 4H), 1.50-1.57 (m,
4H), 1.33-1.40 (m, 2H); 13C-NMR (50 MHz, CDCl3) δ 136.4
(C), 129.3 (CH), 127.9 (CH), 127.4 (CH), 75.2 (CH2), 71.8 (CH),
68.4 (CH), 59.0 (CH3), 51.9 (CH2), 26.3 (CH2), 24.5 (CH2); IR
P r epar ation of (2S,3S)-3-P h en yl-2-[(tr iph en ylm eth oxy)-
m et h yl]oxir a n e (6-Tr ) Usin g Ch lor ot r ip h en ylm et h a n e
a n d Sod iu m Hyd r id e a s P r otectin g Rea gen ts. Compound
1a (250 mg, 1.66 mmol) in DMF (3 mL), sodium hydride (60
mg, ca. 2.0 mmol) in DMF (2 mL), and chlorotriphenylmethane
(466 mg, 1.67 mmol) in DMF (4 mL) were treated as described
for 6-Me, stirring 72 h at 0 °C, to give 83 mg (13%) of 6-Tr as
an oil after chromatography, using hexane:Et2O (99:1) as
eluent. This oil was recrystallized from hexane to give
analytically pure white crystals: mp 120 °C; [R]23D ) -32.4 (c
(KBr) 3422, 3064, 3027, 2981, 2937, 2802, 1096, 704, 751 cm-1
;
MS (CI, NH3) m/ z 250 (C15H23NO2‚H+, 1), 174 (100). Anal.
Calcd for C15H23NO2: C, 72.25; H, 9.30; N, 5.62. Found: C,
72.17; H, 9.41; N, 5.56.
(2R,3R)-1-(Ben zyloxy)-3-p h en yl-3-p ip er id in op r op a n -
2-ol (4d -Bn ). Compound 6-Bn (240 mg, 1.0 mmol), LiClO4
(1.65 g, 15.5 mmol), and piperidine (990 µL, 10 mmol) in
acetonitrile (2.5 mL) were treated as described for 4d -Me
during 24 h. The workup was identical to the one described
for 4d -Me to give 318 mg (98%) of 4d -Bn as an oil after
chromatography using hexane:EtOAc (90:10-80:20) as elu-
1
) 1.0 in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.21-7.50 (m,
20H), 3.78 (d, 1H, J ) 1.8 Hz), 3.45 (d × d, 1H J ) 10.2, 2.4
Hz), 3.21-3.29 (m, 2H); 13C-NMR (75 MHz, CDCl3) δ 143.8
(C), 137.1 (C), 128.6 (CH), 128.4 (CH), 128.2 (CH), 127.9 (CH),
127.1 (CH), 125.7 (CH), 86.8 (C), 64.2 (CH2), 61.3 (CH), 56.1
(CH); IR (KBr) 3064, 3023, 2925, 2869, 1595, 1489, 1468, 1447,
1109, 1094, 1078, 762, 751, 700 cm-1; MS (CI, NH3) m/ z 410
(C28H24O2‚H+, 1), 243 (100). Anal. Calcd for C28H24O2: C,
85.68; H, 6.16. Found: C, 85.62; H, 6.16.
ent: [R]23 ) -5.7 (c ) 1.0 in CHCl3); 1H-NMR (300 MHz,
D
CDCl3) δ 7.29-7.35 (m, 10H), 4.43-4.53 (m, 3H), 3.35-3.46
(m, 3H), 2.95 (br s, 1H, OH), 2.32-2.40 (m, 4H), 1.50-1.58
(m, 4H), 1.24-1.41 (m, 2H); 13C-NMR (50 MHz, CDCl3) δ 138.2
(C), 136.3 (C), 129.4 (CH), 128.3 (CH), 127.9 (CH), 127.6 (CH),
127.5 (CH), 127.3 (CH), 73.3 (CH2), 72.7 (CH2), 71.7 (CH), 68.5
(CH), 51.8 (CH2), 26.3 (CH2), 24.5 (CH2); IR (film) 3454, 3087,
3064, 3029, 2933, 2856, 2805, 1113, 1028, 700, 751 cm-1; MS
(CI, NH3) m/ z 326 (C21H27NO2‚H+, 100).
P r ep a r a tion of (2S,3S)-3-P h en yl-2-(tr ip h en ylm eth oxy-
m eth yl)oxir a n e (6-Tr ) Usin g Ch lor otr ip h en ylm eth a n e,
4-(Dim eth yla m in o)p yr id in e, a n d P yr id in e a s Rea gen ts.
Compound 1a (250 mg, 1.66 mmol), chlorotriphenylmethane
(487 mg, 1.75 mmol), and DMAP (20 mg, 0.16 mmol) in
pyridine (10 mL) were stirred for 90 min at 90 °C under N2.
Et2O (25 mL) was added, and the mixture was washed with
HCl 10% (3 × 30 mL) and H2O (2 × 25 mL). The solvent was
removed in vacuo, and the residual oil was chromatographed
using hexane:Et2O (99:1) as eluent to give 293 mg (45%) of
6-Tr as an oil that was recrystallized afterwards from hexane.
This compound was spectroscopically identical to the one
obtained using chlorotriphenylmethane and sodium hydride
as reagents.
P r epar ation of (2S,3S)-3-P h en yl-2-[(tr iph en ylm eth oxy)-
m eth yl]oxir a n e (6-Tr ) Usin g N-(Tr ip h en ylm eth yl)p yr i-
d in iu m Tetr a flu or obor a te. 1a (250 mg, 1.66 mmol) and
N-triphenylmethylpyridinium tetrafluoroborate (750 mg, 1.83
mmol) in acetonitrile (4 mL) were stirred for 24 h at room
temperature under N2. Et2O (15 mL) was added, and the
precipitate was filtered out. The solvent was removed in
vacuo, and the residual oil was chromatographed using hex-
ane:Et2O (99:1) as eluent to give 433 mg (66%) of 6-Tr as an
(2R,3R)-1-(Dip h en ylm et h oxy)-3-p h en yl-3-p ip er id in o-
p r op a n -2-ol (4d -Bzh ). Compound 6-Bzh (250 mg, 0.79
mmol), LiClO4 (1.30 g, 12.2 mmol), and piperidine (780 µL,
7.9 mmol) in acetonitrile (2 mL) were treated as described for
4d -Me during 24 h. The workup was identical to the one
described for 4d -Me to give 305 mg (96%) of 4d -Bzh as an oil
after chromatography using hexane:EtOAc (80:20) as eluent:
1
[R]23 ) -3.8 (c ) 1.0 in CHCl3); H-NMR (300 MHz, CDCl3)
D
δ 7.22-7.32 (m, 15H), 5.31 (s, 1H), 4.55 (m, 1H), 3.40-3.49
(m, 3H), 2.80 (br s, 1H, OH), 2.34-2.35 (m, 4H), 1.50-1.57
(m, 4H), 1.33-1.40 (m, 2H); 13C-NMR (75 MHz, CDCl3) δ 142.2
(C), 142.1 (C), 136.2 (C), 129.5 (CH), 128.31 (CH), 128.30 (CH),
127.9 (CH), 127.4 (CH), 127.3 (CH), 126.9 (CH), 84.1 (CH), 71.6
(CH), 71.3 (CH2), 68.7 (CH), 51.8 (CH2), 26.3 (CH2), 24.5 (CH2);
(28) Maruoka, K.; Ooi, T.; Nagahara, S.; Yamamoto, H. Tetrahedron
1991, 47, 6983-6998.

Enantioselective Addition of Diethylzinc to Aldehydes
J . Org. Chem., Vol. 62, No. 15, 1997 4981
IR (film) 3448, 3087, 3062, 3029, 2931, 2856, 1094, 1073, 1030,
743, 702 cm-1; MS (CI, NH3) m/ z 402 (C27H31NO2‚H+, 100).
(2R,3R)-1-(Dip h en ylm eth oxy)-3-(4-m eth ylp ip er a zin -1-
yl)-3-p h en ylp r op a n -2-ol (4i-Bzh ). Compound 6-Bzh (455
mg, 1.44 mmol), LiClO4 (2.37 g, 22.3 mmol), and 4-methylpi-
perazine (1.6 mL, 14.4 mmol) in acetonitrile (3.6 mL) were
treated as described for 4d -Me during 24 h. The workup was
identical to the one described for 4d -Me to give 565 mg (94%)
of 4i-Bzh after chromatography using hexane:EtOAc (50:50)
The oil was further recrystallized in hexane to give analytically
pure white crystals: mp 81 °C; [R]23 ) -23.7 (c ) 0.9 in
D
CHCl3); 1H-NMR (300 MHz, CDCl3) δ 7.42-7.46 (m, 6H), 7.22-
7.31 (m, 14H), 4.38 (d × d × d, 1H, J ) 7.5, 7.3, 4.2 Hz), 3.65
(d, 1H, J ) 7.5 Hz), 3.34 (d × d, 1H, J ) 9.6, 4.2 Hz), 3.16 (d
× d, 1H, J ) 9.6, 7.3 Hz), 2.62 (br s, 1H, OH), 2.05-2.43 (m,
4H), 0.98-1.30 (m, 8H), 0.80 (t, 6H, J ) 7.2 Hz); 13C-NMR (75
MHz, CDCl3) δ 144.0 (C), 136.2 (C), 129.5 (CH), 128.7 (CH),
127.8 (CH), 127.7 (CH), 127.1 (CH), 126.9 (CH), 86.7 (C), 69.7
(CH), 66.6 (CH2) 65.5 (CH), 49.9 (CH2), 29.8 (CH2), 20.4 (CH2),
14.1 (CH3); IR (KBr) 3587, 3483, 3162, 3087, 3060, 3029, 3002,
2958, 2931, 2871, 2861, 2817, 1491, 1466, 1449, 1074, 773, 762,
746, 702 cm-1; MS (CI, NH3) m/ z 522 (C36H43NO2‚H+, 100).
Anal. Calcd for C36H43NO2: C, 82.88; H, 8.31; N, 2.68.
Found: C, 82.98; H, 8.46; N, 2.70.
as eluent. This material solidified upon standing: [R]23
)
D
1
+3.5 (c ) 0.9 in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.25-
7.30 (m, 15H), 5.26 (s, 1H), 4.41-4.47 (m, 1H), 3.42 (d, J )
6.5 Hz, 1H), 3.31-3.40 (m, 2H), 2.67 (br s, 1H, OH), 2.40-
2.42 (br s, 8H), 2.24 (s, 3H); 13C-NMR (75 MHz, CDCl3) δ 142.0
(C), 141.9 (C), 135.9 (C), 129.5 (CH), 128.3 (CH), 128.0 (CH),
127.6 (CH), 127.4 (CH), 126.9 (CH), 84.1 (CH), 71.1 (CH2), 71.0
(CH), 68.4 (CH), 55.3 (CH2), 50.4 (CH2), 45.8 (CH3); IR (KBr)
3421, 3087, 3062, 3029, 2935, 2875, 2798, 2695, 1452, 1144,
1107, 1072, 758, 742, 702 cm-1; MS (CI, NH3) m/ z 417
(C27H32N2O2‚H+, 100). Anal. Calcd for C27H32N2O2: C, 77.85;
H, 7.74; N, 6.72. Found: C, 77.63; H, 7.84; N, 6.63.
(2R,3R)-1-[(ter t-bu tyld im eth ylsilyl)oxy]-3-[[[2-(d im eth -
yla m in o)eth yl]m eth yl]a m in o]-3-p h en ylp r op a n -2-ol (4h ).
Compound 6-Tbs (312 mg, 1.18 mmol), LiClO4 (1.94 g, 18.3
mmol), and N,N′,N′-trimethylethylenediamine (1.5 mL, 11.8
mmol) in acetonitrile (3.0 mL) were treated as described for
4d -Me during 24 h. The residual oil after removal of the
solvents was distilled in vacuo (Kugelrohr apparatus, 150-
160 °C, 0.3 Torr) to give 424 mg (98%) of a 7:1 mixture of 4h
and its regioisomer.
(1R,2R)-1-P h en yl-1-p ip er id in o-3-(tr ip h en ylm eth oxy)-
2-p r op a n ol (4d -Tr ). Compound 6-Tr (455 mg, 1.16 mmol),
LiClO4 (1.91 g, 17.99 mmol), and piperidine (1.15 mL, 11.62
mmol) in acetonitrile (3 mL) were treated as described for 4d -
Me during 24 h. The workup was identical to the one
described for 4d -Me to give 520 mg (93%) of 4d -Tr as an oil
after chromatography using hexane:Et2O (90:10) as eluent.
This oil was recrystallized in hexane:Et2O to give 481 mg (86%)
of 4d -Tr as analytically pure white crystals. This compound
was spectroscopically identical to the one obtained by selective
tritylation of the primary hydroxyl group in 5d .
Mitsu n obu In ver sion a t C-2 of th e An ti Am in o Alcoh ol
4. 4-Nitr oben zoic Acid (1S,2R)-1-[[(ter t-Bu tyld im eth yl-
silyl)oxy]m eth yl]-2-p h en yl-2-p ip er id in oeth yl Ester .
A
solution of 4d (358 mg, 1.02 mmol), PPh3 (1.32 g, 5.04 mmol),
and 4-nitrobenzoic acid (750 mg, 1.02 mmol) in toluene (9 mL)
and THF (9 mL) under N2 was cooled at -20 °C. DEAD (795
µL, 5.04 mmol) was syringed into the solution. The mixture
was stirred at -20 °C for 3 h, allowed to reach room
temperature, and stirred for another 14 h at room tempera-
ture. The solvents were removed in vacuo, and the residual
oil was chromatographed using hexane:EtOAc (95:5) as eluent
to give 340 mg (67%) of 4-nitrobenzoic acid (1S,2R)-1-[[(tert-
(1R,2R)-1-[[[2-(Dim eth yla m in o)eth yl]m eth yl]a m in o]-1-
p h en yl-3-(tr ip h en ylm eth oxy)p r op a n -2-ol (4h -Tr ). Com-
pound 6-Tr (525 mg, 1.33 mmol), LiClO4 (2.21 g, 20.8 mmol),
and N,N,N′-trimethylethylenediamine (1.7 mL, 13.4 mmol) in
acetonitrile (4.0 mL) were treated as described for 4d -Me
during 24 h. The workup was identical to the one described
for 4d -Me to give 650 mg (98%) of 4h -Tr as a vitreous material
butyldimethylsilyl)oxy]methyl]-2-phenyl-2-piperidinoethyl es-
1
ter as a colorless oil: [R]23 ) -41.7 (c ) 1.0 in CHCl3); H-
D
NMR (300 MHz, CDCl3) δ 8.20-8.31 (m, 4H), 7.26-7.43 (m,
5H), 6.15 (d, 1H, J ) 9.0 Hz), 3.55 (m, 2H), 3.14 (d × d × d,
1H, J ) 6.6, 6.0, 4.8 Hz), 3.04 (br s, 1H, OH), 2.42 (br s, 4H),
1.55-1.58 (m, 4H), 1.33-1.38 (m, 2H), 0.85 (s, 9H), -0.07 (s,
3H), 0.00 (s, 3H); 13C-NMR (75 MHz, CDCl3) δ 163.5 (C), 150.5
(C), 139.5 (C), 136.0 (C), 130.6 (CH), 127.9 (CH), 127.7 (CH),
127.1 (CH), 123.5 (CH), 75.4 (CH), 69.9 (CH), 59.6 (CH2), 51.6
(CH2), 26.7 (CH2), 25.8 (CH3), 24.7 (CH2), 18.0 (C), -5.5 (CH3),
-5.4 (CH3); IR (film) 3035, 2932, 1728, 1609, 1530, 1472, 1270,
1102, 1015, 776, 698 cm-1; MS (CI, NH3) m/ z 495 (C27H38N2O5-
Si‚H+, 100).
after chromatography using EtOAc as eluent: [R]23 ) +0.7
D
1
(c ) 1.0 in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.35-7.39
(m, 6H), 7.16-7.27 (m, 14H), 4.24 (d × d × d, 1H, J ) 6.3, 6.0,
5.7 Hz), 4.13 (br s, 1H, OH), 3.53 (d, 1H, J ) 5.7 Hz), 3.13 (d
× d, 1H, J ) 9.4, 6.3 Hz), 2.88 (d × d, 1H, J ) 9.4, 6.0 Hz),
2.38-2.51 (m, 4H), 2.23 (s, 6H), 2.17 (s, 3H); 13C-NMR (75
MHz, CDCl3) δ 143.9 (C), 136.5 (C), 129.6 (CH), 128.5 (CH),
127.8 (CH), 127.6 (CH), 127.2 (CH), 126.8 (CH), 86.6 (C), 70.7
(CH), 69.5 (CH), 65.5 (CH2), 57.5 (CH2), 51.5 (CH2), 45.5 (CH3),
40.3 (CH3); IR (KBr) 3409, 3087, 3058, 3025, 2939, 2858, 2821,
1447, 1073, 1032, 772, 749, 704 cm-1; MS (CI, NH3) m/ z 495
(C38H38N2O2‚H+, 42), 253 (100).
P r ep a r a t ion of (2S,3R)-1-[(ter t-Bu t yld im et h ylsilyl)-
oxy]-3-p h en yl-3-p ip er id in op r op a n -2-ol (syn -4d ). To a
solution 4-nitrobenzoic acid (1S,2R)-1-[[(tert-butyldimethylsi-
lyl)oxy]methyl]-2-phenyl-2-piperidinoethyl ester (157 mg, 0.31
mmol) in CH2Cl2 (2 mL) at -20 °C under N2 was added
DIBALH 1 M in hexanes (2.5 mL, 2.5 mmol) dropwise. The
mixture was stirred at -20 °C for 19 h. Brine (20 mL) and
CH2Cl2 (30 mL) were added carefully, and the mixture was
stirred vigorously. The organic phase was separated and the
aqueous layer extracted with CH2Cl2 (2 × 30 mL). The
combined organic extracts were dried and concentrated in
vacuo. The residual oil was chromatographed using hexane:
(1R ,2R )-1-(4-Me t h ylp ip e r a zin -1-yl)-1-p h e n yl-3-(t r i-
p h en ylm eth oxy)p r op a n -2-ol (4i-Tr ). Compound 6-Tr (500
mg, 1.27 mmol), LiClO4 (2.10 g, 19.7 mmol), and 4-methylpi-
perazine (1.4 mL, 12.8 mmol) in acetonitrile (3.2 mL) were
treated as described for 4d -Me during 24 h. The workup was
identical to the one described for 4d -Me to give 559 mg (89%)
of 4i-Tr as an amorphous solid after chromatography using
hexane:EtOAc (50:50-0:100) as eluent: [R]23D ) -7.5 (c ) 0.9
1
in CHCl3); H-NMR (300 MHz, CDCl3) δ 7.36-7.40 (m, 6H),
7.14-7.29 (m, 14H), 4.30 (d × d × d, 1H, J ) 6.0, 5.7, 5.1 Hz),
3.44 (d, 1H, J ) 5.7 Hz), 3.15 (d × d, 1H, J ) 9.6, 6.0 Hz), 2.94
(d × d, 1H, J ) 9.6, 5.1 Hz), 2.40-2.43 (br s, 8H), 2.31 (s, 3H);
13C-NMR (75 MHz, CDCl3) δ 143.9 (C), 135.6 (C), 129.3 (CH),
128.6 (CH), 128.0 (CH), 127.7 (CH), 127.5 (CH), 126.9 (CH),
86.8 (C), 80.0 (CH), 68.5 (CH), 65.3 (CH2), 55.2 (CH2), 50.0
(CH2), 45.6 (CH3); IR (KBr) 3448, 3087, 3060, 3031, 2935, 2883,
2811, 1449, 1144, 1115, 1079, 762, 748, 704 cm-1; MS (CI, NH3)
m/ z 493 (C33H36N2O2‚H+, 19), 251 (100).
EtOAc (95:5) as eluent to give 58 mg (53%) of syn-4d : [R]23
D
) -5.6 (c ) 1.1 in CHCl3); 1H-NMR (300 MHz, CDCl3) δ 7.23-
7.38 (m, 5H), 5.04 (br s, 1H), 3.68-3.81 (m, 2H), 2.81-2.87
(m, 1H), 2.67-2.75 (br s, 4H), 1.60 (br s, 4H), 1.41-1.47 (m,
2H), 0.89 (s, 9H), 0.02 (s, 6H); 13C-NMR (75 MHz, CDCl3) δ
142.4 (C), 127.9 (CH), 126.9 (CH), 126.0 (CH), 71.5 (CH), 70.8
(CH), 60.2 (CH2), 52.5 (CH2), 26.3 (CH2), 25.7 (CH3), 24.2 (CH2),
18.0 (C), -5.61 (CH3), -5.63 (CH3); IR (film) 3433, 3064, 3031,
2931, 2858, 1526, 1453, 1096, 778, 700 cm-1; MS (CI, NH3)
m/ z 350 (C20H35NO2Si‚H+, 100).
(1R,2R)-1-(Di-n -bu tylam in o)-1-ph en yl-3-(tr iph en ylm eth -
oxy)p r op a n -2-ol (4j-Tr ). Compound 6-Tr (150 mg, 0.38
mmol), LiClO4 (0.63 g, 5.92 mmol), and di-n-butylamine (640
µL, 3.8 mmol) in acetonitrile (1 mL) were treated as described
for 4d -Me during 24 h. The workup was identical to the one
described for 4d -Me to give 187 mg (94%) of 4j-Tr as an oil
after chromatography using hexane:EtOAc (80:20) as eluent.
Gen er a l P r oced u r e for th e En a n tioselective Am in o
Alcoh ol-Ca ta lyzed Ad d ition of Dieth ylzin c to Ald eh yd es.
To a solution of the chiral catalyst (0.06 mmol, 6 mol %) in
hexane or toluene (2 mL) was added the aldehyde (1 mmol) at
room temperature. The mixture was stirred for 20 min and

4982 J . Org. Chem., Vol. 62, No. 15, 1997
Vidal-Ferran et al.
then cooled to the desired temperature if necessary. Dieth-
ylzinc (2.5 mL of a 1 M hexanes solution, 2.5 mmol) was added
dropwise. The mixture was stirred for the corresponding
reaction time under N2. The reaction was quenched by the
addition of a saturated NH4Cl solution (10 mL). The mixture
was then extracted with CH2Cl2 (3 × 10 mL). The combined
organic extracts were dried and concentrated in vacuo. The
enantiomeric excesses were determined from the crude mix-
ture by GC analyses. Conditions of GC analyses: â-DEX or
R-DEX 120 column, 30 m length, 0.25 mm internal diameter,
isotherm temperature program, He as carrier gas (2.4 mL/
min). For 1-phenylpropanol: â-DEX 120 column, 112 °C, tR
R isomer 49.3 min, tR S isomer 52.0 min. For 1-(p-tolyl)-
propanol: â-DEX 120 column, 120 °C, tR R isomer 48.1 min,
tR S isomer 51.9 min. For 1-(4-methoxyphenyl)propanol:
â-DEX 120 column, 135 °C, tR R isomer 65.3 min, tR S isomer
68.1 min. For 1-(4-fluorophenyl)propanol: â-DEX 120 column,
112 °C, tR R isomer 53.1 min, tR S isomer 58.8 min. For 1-(2-
Gen er a l P r oced u r e for th e En a n tioselective Am in o
Alcoh ol-Ca ta lyzed Ad d ition of Dieth ylzin c to Ben za ld e-
h yd e Usin g th e Lith iu m Sa lt of 4h -Tr or 4i-Tr . A solution
of 4h -Tr or 4i-Tr (0.1 mmol) in Et2O (4 mL) at -78 °C was
treated with n-BuLi (65 µL of a 1.6 M solution in Et2O, 0.1
mmol) under N2. The solution was stirred for 10 min, and
then benzaldehyde (100 µL, 1 mmol) and diethylzinc (2.5 mL
of a 1 M hexanes solution, 2.5 mmol) were added. The reaction
mixture was warmed to 0 °C and stirred for 8 h at 0 °C.
Quenching and determination of the ee. of the product was
performed as described above.
Ack n ow led gm en t. Financial support from CIRIT-
CICYT (QFN93-4407), from DGICYT (PB95-0265), and
from CIRIT (GRQ93-1083) is gratefully acknowledged.
A.V.-F. thanks MEC for a contract under the program
Acciones para la Incorporacio´n a Espan˜a de Doctores y
Tecno´logos.
chlorophenyl)propanol: â-DEX 120 column, 135 °C, tR
R
Su p p or tin g In for m a tion Ava ila ble: 13C NMR spectra of
compounds 5b,h , 4a -g, 4h /5h (8:1 mixture), 4d -Bp , 6-Me,
6-Bn , 4d -Bn , 4d -Bzh , 4h -Tr , 4i-Tr , syn -4d (p-nitrobenzoate),
and syn -4d (19 pages). This material is contained in libraries
on microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
isomer 45.8 min, tR S isomer 50.2 min. For 1-(1-naphthyl)-
propanol: â-DEX 120 column, 160 °C, tR S isomer 96.5 min,
tR R isomer 101.1 min. For 1-phenyl-3-pentanol: R-DEX 120
column, 115 °C, tR R isomer 74.4 min, tR S isomer 75.6 min.
For 5-methyl-3-hexanol: R-DEX 120 column, 65 °C, tR R isomer
14.7 min, tR S isomer 15.2 min.
In order to establish the absolute configuration of the final
compounds, the alcohols were purified by bulb-to-bulb distil-
lation of the crude mixtures. The optical rotation was mea-
sured in each case, and its sign was compared with the
reported value ((S)-1-phenylpropanol,29 (S)-1-(p-tolyl)pro-
panol,30 (S)-1-(4-methoxyphenyl)propanol,30 (R)-1-(4-fluoro-
phenyl)propanol,31 (R)-1-(1-naphthyl)propanol,32 (S)-1-phenyl-
3-pentanol,33 and (S)-5-methyl-3-hexanol34).
J O9701445
(30) Capillon, J .; Gue´tte´, J . P. Tetrahedron 1979, 35, 1817-1820.
(31) Kang, J .; Lee, J . W.; Kim, J . I. J . Chem. Soc., Chem. Commun.
1994, 2009-2010.
(32) Niwa, S.; Soai, K. J . Chem. Soc., Perkin Trans. 1 1991, 2717-
2720.
(33) Sato, T.; Gotoh, T.; Wakabayashi, Y.; Fujisawa, T. Tetrahedron
Lett. 1983, 24, 4123-4126.
(34) Soai, K.; Ookawa, A.; Kaba, T.; Ogawa, K. J . Am. Chem. Soc.
1987, 109, 7111-7115.
(29) Kitamura, M.; Suga, S.; Kawai, K.; Noyori, R. J . Am. Chem.
Soc. 1986, 108, 6071-6072.