Synthesis and evaluation of new 1,7-dioxaspiro[5.5]undecane ligands: Implications for the use of diols in the desymmetrization of meso epoxides

Article abstract of DOI:10.1016/S0040-4020(99)01044-3

The synthesis and preliminary evaluation of two new chiral 1,7- dioxaspiro[5.5]undecane ligands is described. The compounds of interest are readily available in enantiomerically pure form. Chiral organotitanium complexes prepared from these ligands cataly

Full text of DOI:10.1016/S0040-4020(99)01044-3

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 507–513
Synthesis and Evaluation of New 1,7-Dioxaspiro[5.5]undecane
Ligands: Implications for the Use of Diols in the
Desymmetrization of meso Epoxides
*
Debasis Patra, Lihua Yang and Nancy I. Totah
Department of Chemistry, University of Iowa, Iowa City, IA 52242, USA
Received 6 October 1999; revised 29 November 1999; accepted 1 December 1999
Abstract—The synthesis and preliminary evaluation of two new chiral 1,7-dioxaspiro[5.5]undecane ligands is described. The compounds of
interest are readily available in enantiomerically pure form. Chiral organotitanium complexes prepared from these ligands catalyze the ring
opening reaction of meso epoxides by TMSN3. A key observation suggests that silylation of the ligand competes with catalyst turnover in this
application. Implications for the use of diol ligands are discussed. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Introduction
We envisaged that this modification would allow an
expedient entry to the spiroketal ligand 2 that maintains
the basic structural features of the parent system. Herein,
we describe an expedient approach to the synthesis of
1,7-dioxaspiro[5.5]undecane ligands of this type, and report
our findings on the mechanistic implications of using diols
as chiral modifying agents for the enantioselective ring
opening reaction of meso epoxides.3
Despite recent advances in asymmetric catalysis, the
efficient intermolecular transfer of asymmetry remains
elusive in a variety of applications. Though in certain
cases, designed catalyst systems provide promising results
and a basis for further development, more frequently the
nature of the specific catalyst is not easily defined, and
thus the factors that govern enantioselectivity are difficult
to identify. As a result, progress in the development of
asymmetric processes will continue to benefit from the
identification of new ligands and catalyst systems.
Results and Discussion
Our initial objective was to develop a more efficient method
for the synthesis of 1,7-dioxaspiro[5.5]undecane based
ligands. Toward this end, we envisaged the enantioselective
synthesis of ligand 2 via sequential homologation of acetone
dimethylhydrazone,4 with subsequent oxidation and spiro-
cyclization of the resulting linear intermediate. As such, the
synthesis of ligand 2 was achieved as depicted below
(Scheme 1). Thus, lithiation of acetone dimethylhydrazone
3 with nBuLi followed by treatment with 4-bromo-1-butene
provided the dimethylhydrazone 4 which was isolated, but
used without further purification. Subsequent lithiation of
this intermediate (4) and aldol reaction with 3-(t-butyl-
dimethylsilyloxy)propanal (5)5 gave, upon hydrolysis, the
b-hydroxyketone 6. From here, asymmetric dihydroxyl-
ation6 gave the crude trihydroxyketone 7 which was treated
Along these lines, we recently reported a novel chiral
organotitanium species for the ring opening reactions of
meso epoxides.1 The backbone of this reagent was the
4S,6S,8S,9R-spiroketal ligand 1 (Fig. 1) which is prepared
in Ͼ99% ee from propargyl alcohol in 13 steps and 12%
overall yield. Though use of the corresponding ligand
1–Ti(OiPr)4 complex showed initial promise as a catalyst
in the desymmetrization of meso epoxides,2 we considered
further development of this system to be limited by the
length of the ligand 1 synthesis. Since that time, we have
been working on modified ligand systems with an eye
toward minimizing the length of the synthetic sequence
while at the same time maintaining the basic ligand frame-
work. Though the specific nature of the ligand 1–Ti(iOPr)4
catalyst system was unknown, we anticipated that the
methyl group at C9 would have little effect on the efficient
intermolecular transfer of asymmetry in light of its distance
from the chelating functionality of the ligand 1 complex.
Keywords: asymmetric induction; catalysis; spiro compounds; asymmetric
synthesis.
Figure 1.
* Corresponding author. E-mail: nancy.totah@uiowa.edu
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(99)01044-3

508
D. Patra et al. / Tetrahedron 56 (2000) 507–513
Scheme 1.
directly with camphorsulfonic acid to provide an approxi-
mately 1:1 mixture of non-racemic axial (2) and equatorial
(8) alcohols that could be separated by careful flash column
chromatography. By this process, ligand 2 can be prepared
in five steps and 29% overall yield.
then provides, cleanly, the corresponding trans-2-azido
alcohol 13.
We first investigated the ability of complex 11 to catalyze
the reaction of cyclohexene oxide (12a) with trimethylsilyl
azide using equimolar amounts of catalyst and substrate
(Table 1, entries 1–5). In these experiments we found that
enantioselectivity varied with reaction temperature and the
aging period of the chiral titanium complex with TMSN3.
Thus, when the epoxide opening reaction is run at Ϫ10ЊC,
the product azido alcohol 13a is obtained in 42% ee after a
1 h aging time (entry 1), and in 59% ee when the aging time
is doubled (entry 2). Presumably, this dependence on aging
reflects the formation of an intermediate titanium azide
which is thought to be the active catalytic species.10 Further
extension of the aging period (entries 3 and 4) appears to
have little impact on selectivity. However, the enantio-
selectivity of the stoichiometric process can be moderately
enhanced by decreasing the reaction temperature to Ϫ20ЊC
(entry 5). Under these conditions the 1R,2R11 azidohydrin
13a is produced in 66% ee. This level of enantioselectvity is
comparable to that obtained using the ligand 1–Ti(OiPr)4
complex (64% ee at Ϫ10ЊC after a 1 h aging period).1
Additional studies demonstrated that the ligand 2 complex
promotes the enantioselective ring opening reaction of
cyclopentene oxide 12b (entry 6) and cis-butene oxide
12c (entry 7) to give the corresponding trans azido alcohols
13b and 13c in 41 and 30% ee, respectively.
Though, in this sequence, formation of the equatorial
alcohol 8 reduces the overall yield of the desired ligand 2,
this side-product can be cleanly converted to the desired
axial isomer via selective silylation of the primary alcohol
(9), Swern oxidation of the hydroxyl group at C4 (10),
followed by reduction with Li(s-Bu)3BH, and deprotection
(Scheme 2).7 In this way, the overall yield of the desired
ligand 2 is increased to 47% starting from acetone dimethyl-
hydrazone.
By the process described above, ligand 2 is obtained in 60%
ee as determined by gas chromatography of monosilylated
ligand 9 (OH axial)8 using a chiral Cyclodex B column
(J & W Scientific). However, compound 2 can be obtained
in Ͼ99% ee after a single recrystallization from hexanes to
remove the highly crystalline racemate.
With the desired spiroketal ligand 2 in hand, our next objec-
tive was to evaluate the efficacy of this system as a chiral
modifying agent for the ring opening reaction of meso
epoxides with trimethylsilyl azide. Thus, the ligand
2–Ti(OiPr)4 complex 11 was generated in situ by combining
equimolar amounts of the ligand 2 and Ti(OiPr)4 in CH2Cl2
at room temperature (Scheme 3).9 After 12 h, trimethylsilyl
azide was added, and the resulting mixture aged at 0ЊC for a
specific time period. Subsequent addition of the epoxide 12
As complexes of type 11 will ideally be utilized in sub-
stoichiometric amounts, we next explored the effect of
decreasing the amount of ligand 2–Ti(OiPr)4 complex
Scheme 2.
Scheme 3.

D. Patra et al. / Tetrahedron 56 (2000) 507–513
509
Table 1. Enantioselective ring opening of cyclohexene oxide using ligand 2–Ti(OiPr)4 complex
Entry
Epoxide
Catalyst:
substrate
ratio
Aging
time
(TMSN3)
(h)
Reaction
temperature
(ЊC)
Conversiona
(%)
ee
Configurationc
(%)b
1
2
3
4
5
6
7
8
12a
12a
12a
12a
12a
12b
12c
12a
12a
12a
12a
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:4
1:4
1:4
1:8
1
2
6
9
9
9
9
3
6
9
6
Ϫ10
Ϫ10
Ϫ10
Ϫ10
Ϫ20
Ϫ20
Ϫ20
Ϫ20
Ϫ20
Ϫ20
Ϫ20
46
75
92
91
51
51
57
61
60
61
55
42
59
58
60
66
41
30
52
50
55
35
1R,2R
1R,2R
1R,2R
1R,2R
1R,2R
1R,2R
1R,2R
1R,2R
1R,2R
1R,2R
1R,2R
9
10
11
a Conversions were determined by gas chromatography using tetradecane as an internal quantitative standard. Routinely, the reaction was allowed to proceed
for 48 h after addition of the epoxide.
b Enatioselectivities were determined by capillary GC of the corresponding 1-azido-2-trimethylsilyloxy derivatives using a chiral cyclodex B column.
c Absolute configurations are assigned by comparison to those reported by Jacobsen.2b
(entries 8–11). In these cases, as the ratio of catalyst to
substrate is decreased, the enantioselectivity of the ring
opening process erodes such that at a 1:4 ratio of catalyst
to substrate (entries 8–10), the product azido alcohol is
obtained in only 55% ee at Ϫ20ЊC, whereas at a 1:8 catalyst
to substrate ratio (entry 11) only 35% ee is obtained. Here
too, the impact of catalyst aging with TMSN3 appears to
be minimal beyond the initial incubation period of
approximately 2 h.
use of multidentate ligands or those in which electronic
effects decrease the nucleophilicty of the ligating functions.
This information, combined with the results of the present
studies suggests that an efficient ligand is one that is not
readily silylated. As such, we anticipated that further
improvement in the enantioselectivity of the epoxide open-
ing reaction using 1,7-dioxaspiro-[5.5]undecane based
ligands would necessitate further modification of our ligand
system.
The results of these studies clearly show that the amount of
catalyst utilized has a significant impact on the enantio-
selectivity of the desymmetrization reaction. These results
are consistent with the modification of the catalyst system as
the epoxide opening reaction proceeds. In this regard, one
possibility is silylation of the primary alkoxy function of the
ligand at a rate that is competitive with catalyst turnover
(e.g. silylation of the product).12 This process may result in
the formation of a new catalytic species that is able to
promote epoxide opening, albeit at a slower rate and with
a lower level of selectivity. Support for such a hypothesis is
provided by the observation that significant amounts (ca.
15%) of monosilylated ligand 2 (e.g. 14) are formed over
the course of the reaction13 (Fig. 2).
On the basis of these principles, we sought to reduce
the proclivity of diol ligands of type 2 to undergo
silylation. As such, we redesigned our system such
that the hydroxymethyl side chain (1Њ alkoxy function) of
ligand 2 was replaced by a more highly substituted (3Њ)
ligating unit, and turned our attention to the synthesis and
evaluation of the 1,7-dioxaspiro[5.5]undecane ligand 15
(Fig. 3).
In practice, the synthesis of ligand 15 proceeded by a
sequence of reactions analogous to that described for ligand
2 (Scheme 4). Thus, lithiated acetone dimethylhydrazone 3
was treated with 5-bromo-2-methyl-2-pentene, and the
resulting intermediate subjected to aldol condensation
with the aldehyde 5 to give the corresponding b-hydroxy-
ketone 17 in 60% yield upon hydrolysis. From here,
asymmetric dihydroxylation and direct cyclization of the
corresponding trihydroxyketone provided the spiroketal as
a mixture of axial (15) and equatorial (18) alcohols that
could not be separated. This mixture could be converted
cleanly to the desired axial isomer 15 by a two step
oxidation/reduction protocol as indicated below. In this
way, the desired ligand 15 is prepared in seven steps and
28% overall yield from acetone dimethylhydrazone 3. The
enantiomeric purity of this ligand was Ͼ99% ee as
These results are reminiscent of the difficulties with catalyst
turnover that are observed with the tartrate based diol
ligands used by Sinou12 and Oguni.2c Though in these
cases, the observation of a silylated ligand by-product is
not reported, in general, the silylation of diol ligands is
perhaps not unexpected in light of the relative nucleo-
philicity of the ligating function (alkoxide) as compared to
that of more successful catalyst systems in the desymmetri-
zation of meso epoxides.14 Indeed, for this application, high
levels of enantioselectivity are generally observed with the
Figure 2.
Figure 3.

510
D. Patra et al. / Tetrahedron 56 (2000) 507–513
Scheme 4.
determined by capillary GC of the ketone 19 using a chiral
cyclodex B column.
alkoxide function shut down the ligand silylation pathway,
ligand 15 provides significantly lower enantioselectivity
in the desymmetrization process, producing the 1S,2S
azido alcohol 13a with selectivities of up to only 31%
ee under a variety of reaction conditions. A correspond-
ing decrease in reaction rate was also observed.17
Presumably, this loss of selectivity originates with the
increased steric requirements of both catalyst formation
and subsequent reaction.
The 5-bromo-2-methyl-2-pentene used in the initial alkyla-
tion step, though commercially available, is expensive;
therefore, we routinely prepare this bromide in two steps
from g-butyrolactone15 as a ca. 3:1 mixture of double
bond isomers (e.g. 5-bromo-2-methyl-2-pentene, and its
isomer 20, Fig. 4). Though throughout the homologation
sequence the two isomers are inseparable by silica gel
chromatography, the presence of the undesired terminal
olefin (e.g. 21) is of little consequence because of the differ-
ential rates of dihydroxylation of the double bond isomers.16
Indeed, upon asymmetric dihydroxylation and spirocycliza-
tion, the diols 15 and 18 are isolated cleanly from the
reaction mixture along with unreacted hydroxyketone 21.
None of the spiroketal (e.g. 22) resulting from the oxidation
and cyclization of the terminal olefin 21 is identified.
In conclusion, we have developed an efficient, enantio-
selective approach to 1,7-dioxaspiro[5.5]undecane ligands
2 and 15 from readily available starting materials. By this
route, ligands 2 and 15 are available in 29 and 28% overall
yields, respectively, a substantial improvement over our
previous route to the related ligand 1. As the enantio-
selectivity of the ligand synthesis is derived from the Sharp-
less asymmetric dihydroxylation protocol, either
enantiomeric form of these and related ligands should be
readily accessible. In addition, we have demonstrated the
ability of the ligand 2–Ti(OiPr)4 complex to promote the
ring opening reaction of meso epoxides with trimethylsilyl
azide. Though the enantioselectivities observed in these
examples vary from modest to good (up to 66% ee in
stoichiometric applications; 55% ee at 25% catalyst loading
using ligand 2), a key observation in this study is the
presence of the monosilylated ligand 14 that is formed
over the course of the epoxide opening reaction. This result
suggests that ligand silylation competes with catalyst turn-
over, the possibility of which will have broader implications
With the successful preparation of ligand 15, we were ready
to evaluate the effect of increasing ligand substitution on the
stereoselectivity of the desymmetrization reaction. More
specifically, we were interested in seeing whether the incor-
poration of a tertiary alkoxy function would effectively
address the issue of ligand silylation, and hence, catalyst
turnover. Toward this end, the corresponding ligand
15–Ti(OiPr)4 complex was prepared as described
previously and utilized in the epoxide opening reaction of
cyclohexene oxide 12a with TMSN3 (Scheme 5). Unfortu-
nately, while incorporation of the more highly substituted
Figure 4.
Scheme 5.

D. Patra et al. / Tetrahedron 56 (2000) 507–513
511
for the use of diol ligands in the present application. Though
silylation is not observed when the corresponding ligand
15–Ti(OiPr)4 complex is used, the increased steric
demands of this catalyst system detract from the efficiency
and selectivity in the epoxide opening reaction. While these
results suggest that diol ligands such as these will be less
useful for the present application, we have demonstrated
the feasibility of using a simplified 1,7-dioxaspiro[5.5]-
undecane skeleton as a chiral modifying agent for the inter-
molecular transfer of asymmetry. As such, these new
ligands may well find application in a variety of other
processes, particularly those in which silylating agents are
not utilized.
n-butyllithium (2.1 mL of a 2.5 M solution in hexane) at
Ϫ78ЊC. After 4 h, a solution of 3-(t-butyldimethylsilyloxy)-
propanal 5 (0.99 g, 5.25 mmol) in 17 mL THF was added.
After an additional 5 h, the reaction mixture was warmed
slowly to room temperature overnight. The reaction was
quenched by the addition of saturated NH4Cl (aq), then
transferred to a separatory funnel and extracted with ether.
The combined organics were washed with brine and dried
over Na2SO4. Subsequent removal of the solvent in vacuo
afforded 1.53 g of crude 1-(t-butyldimethylsilyloxy)-3-
hydroxy-9-decen-5-one dimethylhydrazone as yellow oil.
This material was dissolved in 20 mL THF and added drop-
wise to a stirred solution of Cu(OAc)2·H2O (2.51 g,
12.6 mmol) in 59 mL water and 39 mL THF at room
temperature. Upon complete addition, the reaction mixture
was stirred for 20 h, over which period a rust colored pre-
cipitate deposited. The reaction mixture was concentrated in
vacuo, then partitioned between ethyl acetate and saturated
NH4Cl (aq) (adjusted to pH 8–9 using NH3·H2O). The layers
were separated, the aqueous layer back extracted with ethyl
acetate, and the combined organics washed sequentially
with saturated NH4Cl (aq) (pH 8–9), water and brine, then
dried over Na2SO4 and concentrated in vacuo. Purification
of the residue by flash column chromatography (SiO2;
hexane/ethyl acetate, 9:1) afforded 1.00 g (67%, three
steps) of 1-(t-butyldimethylsilyloxy)-3-hydroxy-9-decen-5-
one 6 as a light-yellow oil. 1H NMR (300 MHz, CDCl3): d
5.77 (1H, m), 5.06–4.96 (2H, m), 4.27 (1H, m), 3.83 (2H,
m), 3.59 (1H, d, Jˆ2.7 Hz, OH), 2.63 (1H, dd, Jˆ16.8,
7.7 Hz), 2.55 (1H, dd, Jˆ16.8, 4.6 Hz), 2.46 (2H, t,
Jˆ7.3 Hz), 2.07 (2H, m), 1.74–1.63 (4H, m), 0.90 (9H, s),
0.08 (6H, s). 13C NMR (90 MHz, CDCl3): d 211.0, 137.9,
115.3, 67.2, 61.5, 49.6, 42.8, 38.3, 33.0, 25.9 (3C), 22.5,
Experimental
General methods
All air sensitive reactions were performed in oven-dried
glassware under an atmosphere of argon. Reaction solvents
were dried over CaH2 (dichloromethane) or sodium/benzo-
phenone ketyl (tetrahydrofuran) and were distilled just prior
to use. Ti(OiPr)4 was distilled under high vacuum prior to
use. All other reagents were reagent grade and purified
1
where necessary. H NMR and 13C spectra were recorded
on Bruker AC-300 or Bruker WM 360 spectrometers.
Chemical shifts are reported in ppm, downfield from tetra-
methylsilane using residual CHCl3 (d 7.27) as the internal
standard. High resolution mass spectra were obtained by the
University of Iowa Mass Spectrometry Laboratory. Elemen-
tal Analyses were performed by Atlantic Microlab, Inc.
(Norcross, GA). Gas chromatographic (GC) analyses were
carried out on a Varian 3400 CX instrument equipped with
FID detectors and 30 M AT-WAX (Altech) and chiral
Cyclodex B (30 mm id×0.25 m film; J & W Scientific)
columns.
18.2, Ϫ5.5 (2C). IR (film): 3499, 2930, 2858, 1710 cmϪ1
.
HRMS (FAB): Calcd for C16H33O3Si: 301.2199 ([MϩH]ϩ),
found: 301.2194.
(Ϫ)-(4S,6S,8R)-8-(Hydroxymethyl)-1,7-dioxaspiro-[5.5]-
undecan-4-ol (2). AD-mix-b (2.1 g) was dissolved in 15 mL
t-BuOH/H2O (1:1) at room temperature, then cooled to 0ЊC
and an orange solid precipitated. 1-(t-Butyldimethylsilyl-
Experimental procedures
6-Hepten-2-one N,N-dimethylhydrazone (4). To a stirred
solution of acetone dimethylhydrazone
oxy)-3-hydroxy-9-decen-5-one
6 (0.430 g, 1.43 mmol)
3
(4.30 g,
was then added in one portion (1 mL t-BuOH was used to
complete the transfer). After 6 h, the reaction was quenched
by addition of solid Na2SO3 (2.3 g), then transferred to a
separatory funnel and extracted with ethyl acetate. The
combined extracts were dried over Na2SO4, filtered and
concentrated in vacuo. The resulting residue was dissolved
in 15 mL CH2Cl2/CH3OH (95:5). Camphorsulfonic acid
(0.17 g, 0.72 mmol) was added, and the resulting solution
stirred for 3 h at room temperature. The reaction mixture
was then diluted with ethyl acetate, and washed with satu-
rated Na2CO3 (aq) and water, respectively. The aqueous
portion was back extracted with ethyl acetate, and the
combined organics dried over Na2SO4, filtered, and concen-
trated in vacuo to afford a ca. 1:1 mixture of axial (2) and
equatorial (8) alcohols. Purification by flash column
chromatography (SiO2; hexane/ethyl acetate, 1:1 to 1:2 to
0:1) gave 0.102 g (35%) of the axial isomer 2 (colorless oil;
[a]2D1ˆϪ88.8Њ, CHCl3, cˆ2.6; after a single recrystalliza-
tion from hexanes), 0.032 g (11%) of the equatorial isomer 8
(colorless oil), and 0.118 g (41%) an isomeric mixture of
alcohols (2 and 8) (total 0.252 g, 87% combined yield).
42.5 mmol) in THF (128 mL) was added n-butyllithium
(17.0 mL of a 2.5 M solution in hexane) at Ϫ78ЊC. Stirring
was continued for 2.5 h over which time a white precipitate
formed. A solution of 4-bromo-1-butene (4.4 mL, 43 mmol)
in 20 mL THF was then added dropwise via cannula. After
an additional 2 h at Ϫ78ЊC, the reaction mixture was
allowed to warm slowly to room temperature overnight.
The resulting clear yellow solution was quenched by the
addition of saturated NH4Cl (aq). The mixture was then
extracted with ether, and the combined extracts washed
with brine, then dried over Na2SO4. Removal of solvent in
vacuo afforded 5.70 g (87%) of crude 6-hepten-2-one
dimethylhydrazone 4 as a 4:1 mixture of isomers that was
used without further purification (yellow oil). 1H NMR
(300 MHz, CDCl3): (major isomer) d 5.80 (1H, m), 5.07–
4.93 (2H, m), 2.42 (6H, s), 2.20 (2H, m), 2.08 (2H, m), 1.94
(3H, s), 1.60 (2H, m).
1-(t-Butyldimethylsilyloxy)-3-hydroxy-9-decen-5-one
(6). To a stirred solution of crude 6-hepten-2-one dimethyl-
hydrazone 5 (0.77 g, 5.0 mmol) in 17.5 mL THF was added

512
D. Patra et al. / Tetrahedron 56 (2000) 507–513
(Ϫ)-(4S,6S,8R)-8-(hydroxymethyl)-1,7-dioxaspiro[5.5]-
undecan-4-ol (Ϫ)-(2): H NMR (360 MHz, CDCl3): d 4.21
afforded 0.470 g (60% over two steps) of 1-(t-butyl-
dimethylsilyloxy)-3-hydroxy-10-methyl-9-undecen-5-one
17 as a pale yellow oil. 1H NMR (300 MHz, CDCl3): d 5.04
(1H, m), 4.22 (1H, m), 3.79 (2H, m), 3.59 (1H, d,
Jˆ2.7 Hz), 2.56 (2H, m), 2.39 (2H, t, Jˆ7.4 Hz), 1.95
(2H, q, Jˆ7.2 Hz), 1.65 (3H, d, Jˆ0.9 Hz), 1.60 (4H,
m), 1.55 (3H, s), 0.86 (9H, s), 0.04 (6H, s). 13C NMR
(75 MHz, CDCl3): d 211.2, 132.2, 123.6, 66.9, 61.2,
49.4, 43.0, 38.3, 27.2, 25.8 (3C), 25.6, 23.7, 18.1,
17.6, Ϫ5.6 (2C). IR (film): 3576, 2932, 2859,
1710 cmϪ1. HRMS (FAB): Calcd for C18H36O3SiNa
351.2331 ([MϩNa]ϩ), found 351.2337.
1
(1H, d, Jˆ9.8 Hz), 4.10–4.00 (2H, m), 3.81 (1H, m), 3.64–
3.34 (3H, m), 2.72 (1H, t, Jˆ6.1 Hz), 1.94–1.25 (10H, m).
13C NMR (90 MHz, CDCl3): d 97.8, 71.1, 65.7, 64.5, 55.1,
40.3, 34.8, 31.9, 26.0, 17.8. IR (film): 3429, 3336, 2945,
1061 cmϪ1
. HRMS (FAB): Calcd for C10H18O4Na:
225.1103 ([MϩNa]ϩ), found 225.1099. Anal Calcd for
C10H18O4 C 59.39, H 8.97; found C 59.34, H 8.94.
(4R,6S,8R)-8-(hydroxy-methyl)-1,7-dioxaspiro[5.5]-
undecan-4-ol (8): 1H NMR (360 MHz, CDCl3): d 4.13 (1H,
m), 3.76–3.45 (5H, m), 2.05 (1H, ddd, Jˆ12.4, 4.8, 2.0 Hz),
1.89 (3H, m), 1.73–1.25 (6H, m).
(6S, 8R)-8-(2-Hydroxy-2-propyl)-1,7-dioxaspiro[5.5]un-
decan-4-ol (15 and 18). AD-mix-b (2.1 g) and methane-
sulfonamide (0.145 g, 1.52 mmol) was dissolved in 15 mL
t-BuOH/H2O (1:1) at room temperature, then cooled to
0ЊC and an orange solid precipitated. Neat 1-(t-butyl-
dimethylsilyloxy)-3-hydroxy-10-methyl-9-undecen-5-one
17 (0.500 g, 1.52 mmol) was then added, and 1 mL t-BuOH
was used to complete the transfer. After 24 h at 0ЊC, the
reaction was quenched by addition of 2.3 g Na2SO3 and
stirred at room temperature for 30 min. The resulting hetero-
geneous mixture was then extracted with ethyl acetate, and
the combined extracts dried over MgSO4 and concentrated
in vacuo. The residue was dissolved in 15 mL CH2Cl2/
MeOH (95:5); camphorsulfonic acid (0.18 g, 0.76 mmol)
was added and the resulting solution stirred for 4 h.
The reaction mixture was diluted with ethyl acetate,
washed with saturated aqueous Na2CO3 and water, and
the aqueous washings back extracted with ethyl acetate.
The combined organics were dried over MgSO4,
concentrated in vacuo, and the residue purified by
flash chromatography (SiO2; hexanes/ethyl acetate, 1:2)
to afford 0.280 g (81%) of (6S,8R)-8-(2-hydroxy-2-
propyl)-1,7-dioxaspiro[5.5]undecan-4-ol as a mixture of
axial 15 and equatorial 18 alcohols that could not be
separated (clear viscous oil). 1H NMR (300 MHz,
CDCl3): d 4.13–3.52 (4.5H, m), 3.39 (0.5H, dd, Jˆ
11.7, 2.2 Hz), 2.26 (1H, br s), 2.03 (1H, m), 1.92–1.75
(3H, m), 1.70–1.58 (4H, m), 1.56–1.30 (2H, m), 1.24
(1.5H, s), 1.20 (1.5H, s), 1.19 (1.5H, s), 1.13 (1.5H, s).
7-Methyl-6-octen-2-one N,N-dimethylhydrazone (16). To
a stirred solution of acetone dimethyl-hydrazone 3 (0.300 g,
3.00 mmol) in 15 mL THF was added n-butyllithium
(1.2 mL of a 2.5 M solution in hexane) at Ϫ78ЊC. Stirring
was continued for 2 h at Ϫ78ЊC, and resulted in the gradual
formation of a white precipitate. A solution of 5-bromo-2-
methyl-2-pentene (0.490 g, 3.00 mmol) in 5 mL THF was
then added dropwise. The reaction mixture was stirred for
an additional 2 h, after which time it was allowed to warm
slowly to room temperature. The resulting clear yellow
solution was quenched by addition of saturated, aqueous
NH4Cl, the aqueous portion extracted with ether, and the
combined organics washed with brine and dried over
MgSO4. Removal of solvent in vacuo provided 0.470 g
(86%) of crude 7-methyl-6-octen-2-one N,N-dimethyl-
hydrazone 16 as a yellow oil that was used without further
1
purification (ca. 5:1 mixture of hydrazone isomers). H
NMR (300 MHz, CDCl3): (major isomer) d 5.11 (1H, m),
2.43 (6H, s), 2.19 (2H, m), 1.99 (2H, m), 1.94 (3H, s),
1.68 (3H, d, Jˆ1.0 Hz), 1.58 (3H, d, Jˆ0.3 Hz), 1.55
(2H, m).
1-(t-Butyldimethylsilyloxy)-3-hydroxy-10-methyl-9-
undecen-5-one (17). To a stirred solution of 7-methyl-6-
octen-2-one N,N-dimethylhydrazone 16 (0.440 g, 2.4
mmol) in 10 mL THF was added n-butyllithium (0.96 mL
of a 2.5 M solution in hexanes) at Ϫ78ЊC and the resulting
mixture stirred for 4 h. A solution of 3-(t-butyldimethyl-
silyloxy)propanal 5 (0.545 g, 2.90 mmol) in 5 mL THF
was then added. After additional 5 h at Ϫ78ЊC, the reaction
mixture was allowed to warm to room temperature. The
reaction was quenched with saturated, aqueous NH4Cl and
the aqueous portion extracted with ether. The combined
organics were washed with brine, dried over MgSO4 and
concentrated in vacuo to afford 0.876 g of the crude
dimethylhydrazone as a yellow oil. The crude product was
then dissolved in 10 mL THF and added dropwise to a stir-
red solution of Cu(OAc)2·H2O (1.21 g, 6 mmol) in 50 mL of
3:2 water:THF at room temperature. The color of the reac-
tion mixture changed from blue to green and finally a rust
colored precipitate deposited. After 15 h, the reaction
mixture was concentrated in vacuo, then partitioned
between ethyl acetate and saturated aqueous NH4Cl
(adjusted to pH 8–9 with aqueous ammonia). The aqueous
layer was extracted with ethyl acetate, and the combined
extracts washed sequentially with saturated aqueous
NH4Cl (pH 8–9), water and brine, then dried over MgSO4
and concentrated in vacuo. Purification of the residue by
flash chromatography (SiO2; hexane/ethyl acetate, 7:1)
(Ϫ)-(6S, 8R)-8-(2-Hydroxy-2-propyl)-1,7-dioxaspiro[5.5]-
undecan-4-one (19). To a solution of oxalyl chloride
(0.080 mL, 0.92 mmol) in 5 mL CH2Cl2 at Ϫ78ЊC was
slowly added a solution of DMSO (0.13 mL, 1.8 mmol)
in 2 mL CH2Cl2. After 5 min, a solution of (6S,8R)-8-
(2-hydroxy-2-propyl)-1,7-dioxaspiro[5.5]-undecan-4-ol 17
(0.085 g, 0.37 mmol) in CH2Cl2 was added dropwise and
stirred for 1 h at Ϫ78ЊC. Diisopropylethylamine (0.5 mL,
3.0 mmol) was added, the reaction mixture stirred for
15 min, then allowed to warm to room temperature. The
reaction mixture was diluted with water, extracted with
CH2Cl2, and the organic extracts washed sequentially with
1% aqueous HCl, saturated aqueous NaHCO3, and brine,
dried over MgSO4 and concentrated in vacuo. Purification
of the residue by flash chromatography (SiO2; hexanes/ethyl
acetate, 1:1) afforded 0.070 g (83%) of (6S,8R)-8-(2-
hydroxy-2-propyl)-1,7-dioxaspiro[5.5]-undecan-4-one 19
1
as a clear viscous oil. ([a]2D2ˆϪ98.5Њ, CHCl3, cˆ1.7); H
NMR (300 MHz, CDCl3): d 4.00 (1H, m), 3.93 (1H, dt,
Jˆ11.7, 3.1 Hz), 3.43 (1H, dd, Jˆ11.7, 1.9 Hz), 2.59 (1H,

D. Patra et al. / Tetrahedron 56 (2000) 507–513
513
m), 2.46 (2H, s), 2.32 (1H, dd, Jˆ14.4, 1.5 Hz), 2.13 (1H,
Acknowledgements
broad s), 1.93–1.81 (2H, m), 1.73–1.60 (2H, m), 1.48–1.26
(2H, m), 1.16 (3H, s), 1.12 (3H, s). 13C NMR (75 MHz,
CDCl3): d 205.4, 100.2, 76.3, 71.5, 59.0, 52.3, 40.9, 34.4,
We thank the National Science Foundation (CHE-9700141)
and the University of Iowa for support of this work.
25.9, 24.1, 23.8, 18.3. IR (neat): 3468, 2951, 1721 cmϪ1
.
HRMS (FAB): Calcd for C12H20O4Na 251.1259
([MϩNa]ϩ), found 251.1262.
References
1. Eppley, A. W.; Totah, N. I. Tetrahedron 1997, 53, 16545–
16552.
2. (a) Nugent, W. A. J. Am. Chem. Soc. 1992, 114, 2768–2769. (b)
Martinez, L. E.; Leighton, J. L.; Carsten, D. H.; Jacobsen, E. N.
J. Am. Chem. Soc. 1995, 117, 5897–5898. (c) Hayashi, M.;
Kohmura, K.; Oguni, N. Synlett 1991, 774–776.
3. For mechanistic discussions on the desymmetrization of meso
epoxides, see: (a) Hansen, K. B.; Leighton, J. L.; Jacobsen, E. N.
J. Am. Chem. Soc. 1996, 118, 10924–10925. (b) Palucki, M.;
Finney, N. S.; Pospisil, P. J.; Gu¨ler, M. L.; Ishida, T.; Jacobsen,
E. N. J. Am. Chem. Soc. 1998, 120, 948–954. (c) McCleland,
B. W.; Nugent, W. A.; Finn, M. G. J. Org. Chem., 1998, 63,
6656–6666.
4. (a) Schreiber, S. L.; Wang, Z. J. Am. Chem. Soc. 1985, 107,
5303–5305. (b) Enders, D.; Dahmen, W.; Dederichs, E.;
Gatzweiler, W.; Weuster, P. Synthesis 1990, 1013–1019.
5. This aldehyde 5 can be prepared in two steps from 1,3-propane-
diol (1) NaH, THF, TBSCl; (2) (COCl)2, DMSO, CH2Cl2;
iPr2EtN).
(Ϫ)-(4S, 6S, 8R)-8-(2-Hydroxy-2-propyl)-1,7-dioxaspiro-
[5.5]undecan-4-ol (15). To a solution of (6S, 8R)-8-
(2-hydroxy-2-propyl)-1,7-dioxaspiro[5.5]undecan-4-one 19
(0.065 g, 0.28 mmol) in 2 mL THF was added Li(s-Bu)3BH
(0.63 mL of a 1 M solution in THF) dropwise at Ϫ78ЊC.
After stirring for 1 h, the reaction mixture was allowed to
warm to 0ЊC and quenched by the addition of 10% aqueous
NaOH (1.9 mL) and 30% H2O2 (1.3 mL). The resulting
mixture was allowed to warm slowly to room temperature
overnight and then diluted with ethyl acetate. The aqueous
layer was saturated with NaCl and extracted with ethyl
acetate. The combined extracts were washed with saturated,
aqueous Na2S2O3 and brine, then dried over MgSO4 and
concentrated in vacuo. Purification of the residue by flash
chromatography (SiO2; ethyl acetate, 1:2) afforded 0.060 g
(91%) of (4S,6S,8R)-8-(2-hydroxy-2-propyl)-1,7-dioxa-
spiro[5.5]undecan-4-ol 15 as a white crystalline solid.
([a]2D5ˆϪ86.3Њ, CHCl3, cˆ1.8); 1H NMR (300 MHz,
CDCl3): d 4.03–3.92 (3H, m), 3.58 (1H, dd, Jˆ11.7,
4.6 Hz), 3.51 (1H, dd, Jˆ11.7, 2.1 Hz), 2.31 (1H, broad
s), 1.89–1.71 (3H, m), 1.67–1.54 (5H, m), 1.40 (1H, dd,
Jˆ13.8, 4.7 Hz), 1.33 (1H, m), 1.20 (3H, s), 1.15 (3H, s).
13C NMR (75 MHz, CDCl3): d 97.9, 76.6, 71.4, 64.2, 55.2,
40.6, 35.0, 31.9, 26.7, 25.0, 24.4, 18.2. IR (neat): 3403,
2950 cmϪ1. Anal Calcd for C12H22O4 C 62.58, H 9.63;
found C 62.46, H 9.68.
6. Kolb, H. C.; Van Nieuwenhze, M. S.; Sharpless, K. B. Chem.
Rev. 1994, 94, 2483–2547.
7. Mixed fractions from the initial column chromatography of 2
and 8 are most conveniently separated into their component
alcohols by monosilylation and subsequent chromatographic
separation on silica gel. The enantiomeric purity of the mono-
silylated axial isomer can then be evaluated directly by gas
chromatography.
8. Prepared by monosilylation of ligand 2 (TBSCl, Et3N, DMAP,
CH2Cl2).
General procedure for epoxide opening reactions
9. The anticipated mode of complexation is as indicated in
Scheme 3, though the specifics of this structure remain to be
elucidated.
10. (a) Choukroun, R.; Gervais, D. J. Chem. Soc., Dalton Trans.
1980, 1800–1802. (b) Blandy, C.; Choukroun, R.; Gervais, D.
Tetrahedron Lett. 1983, 24, 4189–4192. See also: (c) Caron, M.;
Sharpless, K. B. J. Org. Chem. 1985, 50, 1557–5189.
11. Absolute configurations are assigned by comparison of the
relative retention times of individual enantiomers with those
reported by Jacobsen. See Ref. 2b.
12. Emziane, K. I.; Sutowardoyo, K. I.; Sinou, D. J. Organomet.
Chem. 1988, 346, C7.
13. None of the corresponding bis-silylated ligand is obtained,
presumably due to the increased steric hindrance about the axial
alcohol at C4.
14. Shibasaki has recently reported the use of a binaphthol based
system for the desymmetrization of meso epoxides using non-
silylated nucleophiles. See: Iida, T.; Yamamoto, N.; Sasai, H.;
Shibasaki, M. J. Am. Chem. Soc. 1997, 119, 4783–4784.
15. Sarmah, B. K.; Barua, N. C. Tetrahedron 1993, 49, 2253–
2260.
16. Daqiang, X.; Crispino, G. A.; Sharpless, K. B. J. Am. Chem.
Soc. 1992, 114, 7570–7571.
17. When ligand 15 is used in this application, reaction times
increase substantially. Thus, the reaction requires approximately
six days after addition of the epoxide to reach 50% conversion.
The procedure used for entry 5 in Table 1 is representative:
To a solution of ligand 2 (1.0 mL of a 0.05 M solution in
CH2Cl2) was added titanium (IV) isopropoxide (0.013 mL,
0.045 mmol) and the solution allowed to stir at room
temperature. After 12 h, the reaction mixture was cooled
to 0ЊC, treated with azidotrimethylsilane (0.012 mL,
0.09 mmol), and stirred for an additional 9 h. Tetradecane
(0.012 mL, 0.045 mmol) was added as an internal standard,
followed by cyclohexene oxide 12a (0.0046 mL,
0.045 mmol). The reaction mixture cooled immediately to
Ϫ20ЊC (cryocool). After 48 h, an aliquot was removed from
the reaction mixture, quenched with saturated, aqueous
NaHCO3, extracted with ether, and analyzed directly by
capillary GC using an AT-WAX column (Altech; Tˆ50ЊC
(1 min) to 200ЊC (20Њ/min)) to determine percent conver-
sion. The remaining reaction mixture was treated directly
with chlorotrimethylsilane (0.06 mL, 0.5 mmol) and
triethylamine (0.12 mL, 0.9 mmol). After 2 h, the reaction
mixture was quenched with saturated, aqueous NaHCO3,
extracted with ether and the organic layer filtered through
Celite. Enantioselectivities were determined by analysis of
the ether layer by capillary GC using a Cyclodex-B chiral
column (J & W Scientific; Tˆ120ЊC). Enantiomers of
1-azido-2-(trimethyl-silyloxy)cyclohexane 13a were base-
line resolved.