Stereocomplementary desymmetrizations of divinylcarbinols by zirconium(IV)- vs. titanium(IV)-mediated asymmetric epoxidations

Article abstract of DOI:10.1002/adsc.200700608

Substituted C_s-symmetric penta-1,4-dien-3-ols ("divinylcarbinols") containing cis- or trans-disubstituted C=C bonds were desymmetrized by asymmetric monoepoxidations. Sharpless conditions gave anti-configured monoepoxides. For cis,cis-divinylca

Full text of DOI:10.1002/adsc.200700608

FULL PAPERS
DOI: 10.1002/adsc.200700608
Stereocomplementary Desymmetrizations of Divinylcarbinols
by Zirconium(IV)- vs. Titanium(IV)-Mediated Asymmetric
Epoxidations
Rainer Kramer,a Thilo Berkenbusch,b and Reinhard Brücknera,*
a
Institut für Organische Chemie und Biochemie, Albert-Ludwigs-Universität, Albertstraße 21, 79104 Freiburg, Germany
Fax : (+49)-761-203-6100; e-mail: reinhard.brueckner@organik.chemie.uni-freiburg.de
Boehringer-Ingelheim Pharma GmbH & Co. KG, Process Development, 55216 Ingelheim am Rhein, Germany
b
Received: December 21, 2007; Published online: April 23, 2008
Dedicated to Professor Andreas Pfaltz on the occasion of his 60th birthday
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Substituted Cs-symmetric penta-1,4-dien-3- conditions the ee of monoepoxide syn-12 increased
ols (“divinylcarbinols”) containing cis- or trans-di- with time from 75 to 99% ee. The reason is the pref-
substituted C=C bonds were desymmetrized by erential overoxidation of its minor enantiomer. The
asymmetric monoepoxidations. Sharpless conditions ease of preparation and multitude of functional
gave anti-configured monoepoxides. For cis,cis-di- groups make epoxides 12–14 worthwhile building
vinylcarbinols this was unprecedented. Oxidation blocks for the synthesis of non-racemic structures.
with tert-butyl hydroperoxide (t-BuOOH) in the
presence of zirconium tetraisopropoxide [Zr(O-i- Keywords: diastereoselectivity; enantioselectivity;
G
Pr)4] and a dialkyl tartrate led to the corresponding epoxy alcohols; kinetic resolution; Sharpless asym-
syn-configured monoepoxides. Under the reaction metric epoxidation
Introduction
The conversion of allyl alcohols into non-racemic
epoxy alcohols by Sharplessꢀ asymmetric epoxida-
tion[1] (SAE) is an almost universally possible trans-
formation.[2] Accordingly, it has gained tremendous
importance in synthesis.[3]
Most SAEs affect achiral primary allyl and racemic
secondary allyl alcohols. In the latter case SAEs per-
form a kinetic resolution of the substrate. To this end,
such an SAE is stopped after slightly more than 50%
conversion. This is because at this moment the more
reactive enantiomer 1 of the substrate has been epoxi-
dized completely while the less reactive enantiomer
ent-1 has hardly begun to react (cf. Scheme 1). As a
consequence, the remaining allyl alcohol (ent-1) is
enantiomerically purer than the major diastereomer
(anti-2/ent-anti-2) of the epoxy alcohol mixture (anti-
2/ent-anti-2+syn-2). It should be noted that the enan-
tio- and the diastereoselectivity, with which epoxy al-
cohol anti-2 emerges from the SAE of Scheme 1 are
highest at the onset of the reaction and decrease with
increasing conversion.
Scheme 1. Sharplessꢀ asymmetric epoxidation (SAE) of
chiral secondary allyl alcohols (“vinylcarbinols”).[1] DET=
diethyl tartrate, DiPT=diisopropyl tartrate.
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1131

Rainer Kramer et al.
FULL PAPERS
The opposite dependence of selectivity from time sembles the reactive (left) enantiomer 1; likewise, the
characterizes the SAEs shown in Scheme 2. They unreactive (right) moiety of 3 resembles the unreac-
were pioneered by the groups of Takano[4] and Jäger[5] tive (right) enantiomer ent-1. Thus, it is plausible to
studying SAEs of penta-1,4-dien-3-ol (3, R1 =R2 =H). assume that the major monoepoxide formed from 3
SAEs of symmetrically substituted penta-1,4-dien-3- must possess stereostructure anti-4 and that the minor
ols 3 (R1 and/or R2 ¼H, “divinyl carbinols”) proceed monoepoxides should be ent-anti-4 and ent-syn-4.
similarly.[6–10] The essence of these reactions is three- Since monoepoxide anti-4 is devoid of the reactive
fold: They provide monoepoxy alcohols 4 in prefer- substructure 1, it undergoes overepoxidation only
ence to bisepoxides 5; they proceed with unusually slowly. In contrast, the minor monoepoxides ent-anti-4
high diastereoselectivities; they exhibit extraordinary and ent-syn-4 still contain substructure 1. Therefore,
enantiopurities (up to 99% ee). Scheme 2 reveals monoepoxide ent-anti-4 should be overepoxidized rel-
these features only implicitly. Explicitly, it explains atively fast, delivering bisepoxide S,S,S,S-5, and mono-
howthey arise. It does so by drawing from analogy to
epoxide ent-syn-4 should proceed to bisepoxide meso-
the reactions depicted in Scheme 1. Thereby the out- 5 about as readily. The ensemble of the rate constants
come of the SAEs of Scheme 2 can be understood as of these overepoxidation steps causes the initially ob-
follows. Step 1 of the epoxidation of divinylcarbinols tained monoepoxide mixture to change composition
3 reflects the reactivity pattern of the epoxidation of upon continued reaction with the oxidant: In essence,
the racemic carbinol 1/ent-1 of Scheme 1. This is be- monoepoxide anti-4 enjoys longevity whereas mono-
cause structurally the reactive (left) moiety of 3 re- epoxides ent-anti-4 and ent-syn-4 are annihilated. Dif-
ferently expressed, the more the SAEs of Scheme 2
progress, the higher enantio- and diastereoselectivity
of monoepoxide formation. The mathematical analy-
sis of these relationships[6a,b,11] let Schreiber et al.
point out that “the ratio of enantiomers can become
arbitrarily large as the reaction goes to comple-
tion”.[6a]
Scheme 3 (upper part) compiles all divinylcarbinol
monoepoxides or their enantiomers, which, to the
best of our knowledge, have ermerged from SAEs to
date. Penta-1,4-dien-3-ol was epoxidized asymmetri-
cally by the groups of Takano,[4] Jäger,[5] and Schreib-
er,[6] yielding monoepoxide anti-6. The ee of anti-6 in-
creased dramatically with time [84% ee (after 3 h) to
>97% ee (after 140 h)].[6a] 2,4-Dimethylpenta-1,4-
dien-3-ol led to monoepoxide anti-7 with an ee, which
again improved considerably with increasing conver-
sion [88% ee (after 0.5 h)!ꢀ99.3% ee (after
1.5 h)].[6a,b] E,E-3,5-Dimethylhepta-2,5-dien-4-ol deliv-
ered monoepoxide anti-8 (>95% ee).[10] In addition,
we are aware of three trans,trans-disubstituted divinyl-
carbinols which were subjected to desymmetrizing
SAEs (!anti-9,[7] anti-10,[8] anti-11[6a,b,9]). For monoep-
oxide anti-11 once more, an inverse proportionality
between ee and time was published [93% ee (after
1 h)!ꢀ97% ee (after 44 h)].[6a]
The bottom half of Scheme 3 shows the monoepoxy
alcohols (anti- and syn-12-anti- and syn-14) obtained
from divinylcarbinols by the desymmetrizations inves-
tigated here.[12] Employing t-BuOOH/dialkyl d- or l-
tartrate/Ti(O-i-Pr)4 mixtures as the oxidant, non-race-
A
mic monoepoxy alcohols anti-12–14 were obtained.
Bis(cis-configured) divinylcarbinols (!anti-12 and
anti-13) had not been involved in this transformation
before. Using t-BuOOH and a dialkyl d- or l-tartrate
as before but nowadding Zr
i-Pr)4, the divinylcarbinols of the present study were
desymmetrized by conversion into the non-racemic
A
G
Scheme 2. SAE of achiral secondary allyl alcohols (“divinyl-
carbinols”).[4–10] DET=diethyl tartrate, DiPT=diisopropyl
tartrate.
1132
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148

Stereocomplementary Desymmetrizations of Divinylcarbinols
FULL PAPERS
Results and Discussion
Preparation of Disubstituted Cs-Symmetrical
Divinylcarbinols
The required divinylcarbinols 20–22 were obtained in
three steps and isomerically pure from propargyl alco-
hol (Scheme 4). Step 1 comprised etherifications of
propargyl alcohol giving (para-methoxybenzyl) prop-
argyl ether 16 (95%) and (tert-butyldimethylsilyl)
propargyl ether 17 (83%), respectively. Conversion
into the corresponding lithium acetylides and reaction
Scheme 4. Synthesis of bis(cis-configured) divinylcarbinols
20 and 21 and bis(trans-configured) divinylcarbinol 22. Re-
agents and conditions: a) NaH (1.2 equiv.), DMF, 08C; addi-
tion of PMB-Cl (1.1 equiv.), 08C, 30 min, !room tempera-
ture, 14 h; 93%. b) TBS-Cl (1.03 equiv.), imidazole
(2.8 equiv.), DMF, 08C, 14 h; 83%. c) 16 (2.3 equiv.), n-BuLi
(2.1 equiv.), THF, À788C, 70 min, addition of HCO2Et; !
À358C, 20 h; 85%. d) 17 (2.03 equiv.), n-BuLi (2.03 equiv.),
THF, À788C, 40 min, addition of HCO2Et, !À308C, 15 h;
Scheme 3. Non-racemic epoxy alcohols from desymmetriz-
ing epoxidations of divinylcarbinols reported earlier[4–10] or
studied here.[12] TBS=t-BuMe2Si, Bn=CH2Ph, PMB=para-
methoxybenzyl.
diastereomeric monoepoxy alcohols syn-12–syn-14.
Accordingly, changing the metal counterpart of the
isopropoxide additive made it feasible to realize
asymmetric epoxidations with complementary favorite
diastereoselectivities. In this regard, our Zr(IV)-medi-
ated desymmetrizations represent a valuable exten-
sion of Sharplessꢀ methodology.
94%. e) Zn (20 equiv.), Cu(OAc)2·H2O (1.0 equiv.), H2O,
N
room temperature, 10 min, addition of AgNO3 (1.0 equiv.),
1 h, filtration, transfer of reductant into MeOH, addition of
18, 408C, 14 h; 82%. f) Zn (15 equiv.), Cu(OAc)2·H2O
G
(0.75 equiv.), H2O, room temperature, 10 min, addition of
AgNO3 (0.75 equiv.), 1 h; filtration, transfer of reductant
into MeOH/H2O (1:1), addition of 19, room temperature,
14 h; 78%. g) Red-Alꢂ (10 equiv.), THF, À408C, 29 h; 57%.
TBS=t-BuMe2Si, PMB=para-methoxybenzyl.
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1133

Rainer Kramer et al.
FULL PAPERS
of two equivalents thereof with ethyl formate fur- from syn-12 was difficult. Repeated passages through
nished dialkynylcarbinols 18 (85%) and 19 (94%), re- a flash chromatography column filled with silica gel[22]
spectively, in step 2.
Initially, we reduced the C C bonds of dialkynyl- and pure ent-anti-12 (22%), respectively.
carbinols 18 and 19 cis-selectively with Rieke zinc[13]
Next, we subjected the PMB-containing divinylcar-
prepared from ZnCl2 and molten potassium in reflux- binol 20 to an otherwise identical epoxidation proto-
were required before we obtained pure anti-12 (27%)
ꢁ
–
ing THF[14] – in methanol. This furnished divinylcarbi- col in which we replaced Ti
U
G
20 equiv.) activated by Cu(OAc)2·H2O (0.75– product(s) as before would be obtained, yet possibly
T
1.0 equiv.) and AgNO3 (0.75–1.0 equiv),[15] again in with a higher ee than in the presence of Ti(O-i-Pr)4.[25]
N
methanol.[16] This was also a beneficial change for the Actually we observed a significant acceleration (3 d!
yields, which were augmented to 82% 20 and 78% 21, 4 h) compared to the Ti(O-i-Pr)4-mediated process
A
respectively. The cis-configuration of these com- and a complete reversal of the diastereocontrol: i.e.,
pounds followed from the modest size of their olefinic when we epoxidized divinylcarbinol 20 in the pres-
H,H coupling constants (11.1 Hz both in 20 and 21). ence of Zr(O-i-Pr)4·i-PrOH and l-(+)-DiPT, we iso-
A
The trans-reduction of the C C bonds of dialkynylcar- lated epoxy alcohol syn-12 exclusively (dsꢀ98:2;[21]
binol 18 was achieved at À408C with Red-Alꢂ in 74% yield). Likewise, epoxidation in the presence of
ꢁ
THF.[17] It afforded divinylcarbinol 22 in 57% yield as Zr
(O-i-Pr)4·i-PrOH and d-(À)-DiPT afforded the en-
C=C bonds was established by the magnitude of the reomer (dsꢀ98:2;[21] 64% yield).
olefinic H,H coupling constant (15.6 Hz), too.
Additionally, the last-mentioned epoxidations pro-
vided a bisepoxide in yields of 25% and 32%, respec-
tively. It was sterically homogeneous and turned out
to be the meso-compound syn,syn-23. The formation
of this stereoisomer from divinylcarbinol 20 at the ex-
pense of the two conceivable diastereomers means
Asymmetric Epoxidations of the Bis(cis-configured)
Divinylcarbinol 20
Usually, SAEs of primary allyl alcohols are less enan- that each C=C bond is epoxidized with high syn-dia-
tioselective when the C=C bond is cis- rather than stereoselectivity. As implied (vide infra) by the inves-
trans-configured.[2a,d] SAEs of monoethers of cis-2- tigation of Figure 1 the Zr
(O-i-Pr)4-mediated forma-
G
butene-1,4-diol are no exception: Its PMB monoether tion of bisepoxide syn,syn-23 proceeds mostly via
is epoxidized with 85–88% ee[19] and the correspond- epoxy alcohol enantiomer ent-syn-12 if l-(+)-DiPT is
ing TBS monoether with 84–85% ee.[20] These com- present. Hence, it must stem mostly from syn-12 in
pounds are partial structures of our divinylcarbinols the presence of d-(À)-DiPT.
20 and 21, respectively. Therefore, enantiocontrol of
Due to the much faster Zr(IV)- than Ti(IV)-medi-
the monoepoxidation step of substrates 20 and 21 was ated epoxidation of divinylcarbinol 20 we found it
expected to be less efficient, and a selective overepox- worthwhile testing whether catalytic rather than stoi-
idation of the minor enantiomer seemed crucial for chiometric amounts of Zr
improving the initial ee value. DiPT suffice to effect the reaction (Table 1). Howev-
First, we subjected the PMB-containing divinylcar- er, reducing the amount of Zr(O-i-Pr)·i-PrOH from
A
AHCTREUNG
binol 20 to standard SAEs (Scheme 5, top). We 1.0 to 0.1 equiv. did not result in any conversion even
worked at À208C to À258Cin order to foster stereo- after 24 h (entry 1). Inertness persisted when the
selectivity – and we added 4 Š molecular sieves to amount of Zr(O-i-Pr)·i-PrOH was doubled to
A
the reaction mixtures and stoichiometric amounts of 0.2 equiv. (entry 2). Increasing additionally the con-
Ti(O-i-Pr)4 and diisopropyl tartrate (“DiPT”) in order centration of the substrate from 0.033M to 0.1M was
ACHTREUNG
to increase the turnover. Employing l-(+)- and d- no remedy either (entry 3). The only successful ex-
(À)-DiPT as an additive, epoxy alcohols anti-12 and periment of Table 1 was run at 08C instead of À208C
ent-anti-12, respectively, resulted with moderate anti- (entry 4): After 20 h one half of the divinylcarbinol 20
selectivities (dsꢂ75:25[21] for the crude product and had reacted.[26] Separation by flash chromatograph[22]
ꢂ80:20[21] after flash-chromatography on silica gel[22]). furnished 37% of a 95:5 mixture[21] of monoepoxy al-
Their enantiomeric excesses were 95–97%.[23] It was cohol syn-12 (ee=32%[27]) and its anti-isomer. This
thereby shown for the first time that a bis(cis-config- result could not compete with the stoichiometric pro-
ured) divinylcarbinol can be desymmetrized efficient- cess.
ly under SAE conditions. Unfortunately, separating
The weakness of the Zr(O-i-Pr)4-mediated epoxida-
A
the respective major from the respective minor diaste- tions of divinylcarbinol 20 when allowed to proceed
reomer, i.e., anti-12 from ent-syn-12 and ent-anti-12 for only 4h as displayed in Scheme 5 was a lack of
1134
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148

Stereocomplementary Desymmetrizations of Divinylcarbinols
FULL PAPERS
Scheme 5. Desymmetrizations of divinylcarbinol 20 by classical SAE (top) and by its Zr-analogue (bottom). Reagents and
conditions: a) Ti
(O-i-Pr)4 (1.05 equiv), l-(+)-DiPT (1.1 equiv.), 4 Š molecular sieves, CH2Cl2, À208C, 30 min, addition of t-
BuOOH (2.0 equiv.), 1 h, addition of 20, À208C, 72 h; 69–72% mixture of diastereomers, ds>80:20, anti-12: 95–97% ee, ent-
syn-12: 26–32% ee. b) Ti
(O-i-Pr)4 (1.0 equiv.), d-(À)-DiPT (1.1 equiv.), 4 Š molecular sieves, CH2Cl2, À258C, 1 h, addition of
20, 1 h, addition of t-BuOOH (1.4 equiv.), À258C, 55 h, addition of t-BuOOH (0.7 equiv.), 47 h; 69% mixture of diastereo-
mers, ds<81:19, anti-12: 95% ee, syn-12: 39% ee. c) Zr
(O-i-Pr)4·i-PrOH (1.0 equiv.), 4 Š molecular sieves, CH2Cl2, À208C;
addition of l-(+)-DiPT (1.1 equiv.), t-BuOOH (2.0 equiv.), 30 min, addition of 20, À208C, 4 h; 74% syn-12 (dsꢀ98:2, 82%
ee), 25% syn,syn-23. d) Zr
ACHTREUNG
AHCTREUNG
AHCTREUNG
(O-i-Pr)4·i-PrOH (1.0 equiv.), 4 Š molecular sieves, CH2Cl2, À208C, addition of d-(À)-DiPT
(1.1 equiv.), t-BuOOH (2.0 equiv.), 30 min, addition of 20, À208C, 4 h; 64% ent-syn-12 (dsꢀ98:2, 85% ee), 32% syn,syn-23.
DiPT=diisopropyl tartrate; PMB=para-methoxybenzyl.
enantiocontrol. After the mentioned time, syn-12 pos-
sessed 82% ee[27] and ent-syn-12 85% ee.[27] As dis-
cussed in the context of Scheme 2, an extended reac-
Table 1. Attempts to perform catalytic desymmetrizations of
tion time should increase the enantioselectivity. This
divinylcarbinol 20 in the presence of Zr(O-i-Pr)4·i-PrOH.
G
effect was put to evidence and the underlying mecha-
nism corroborated by monitoring the progress of the
epoxidation 20!syn-12 from the start until 32 h later
(Figure 1). To this end, we dissolved the substrate
(1.56 mmol), l-(+)-DiPT (1.1 equiv.), and Zr(O-i-
U
Pr)4·i-PrOH (1.0 equiv.) in CH2Cl2 (48 mL), added
molecular sieves, cooled to À208C, and started the
epoxidation at t=t0 by the addition of excess t-
BuOOH (2.0 equiv.). At appropriately spaced inter-
vals (namely at tsampling =1, 2, 4, 8, 16, and 32 h), 8 mL
aliquots were removed from the reaction mixture and
worked up extractively as described in the Experi-
mental Section. In the resulting crude product we as-
certained the relative amounts of substrate (20), mon-
oepoxide (syn-12), and bisepoxide (syn,syn-23) by
Entry
x
y
T [8C] t [h] syn-12
Yield [%] syn:anti ee [%]
[a]
1
10 11 À20
20 22 À20
20 22 À20
24
24
24
20
—
—
—
—
[a]
[a]
2
3[b]
4
—
—
20 22
0
37
95:5[18]
32[27]
[a]
No conversion of 20 occurred according to TLC analysis.
The concentration of 20 was three times higher than in
entries 1, 2, and 4.
[b]
1
300 MHz H NMR spectroscopy.[26] Subsequently, we
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1135

Rainer Kramer et al.
FULL PAPERS
Figure 1. Time-resolved analysis of the desymmetrization of divinylcarbinol 20 by a Zr
A
syn-12+ent-syn-12) and overepoxidation (!syn,syn-23) in the presence of l-(+)-DiPT. Elucidating the composition of 20/
(syn-12+ent-syn-12)/syn,syn-23 mixtures[26] sampled at progressive points of the experiment and determining – after separa-
tion from the other components – the enantiopurity of the syn-12/ent-syn-12 mixture established the relative amounts of all
mentioned species as a function of the reaction time. These values – expressed as partial molar fractions”100 – are listed in
the table and plotted in the diagram (full lines). The diagram also traces the evolution of the ee value with the reaction time
(dashed line). Sections from representative HPLC chromatograms,[27] from which the syn-12/ent-syn-12 ratios and ee values
were obtained are inserted.
1136
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148

Stereocomplementary Desymmetrizations of Divinylcarbinols
FULL PAPERS
isolated the monoepoxy alcohol syn-12 by flash chro- cohol (24, 34, 45, 56, and 64%), which both was due
matography on silica gel.[22] Finally, we determined to the interference of overepoxidation. According to
the enantiopurity of syn-12 by HPLC.[27]
the data tabulated in Figure 1 overepoxidation affect-
In the preceding paragraph, “syn-12” is a short- ed the minor monepoxy alcohol enantiomer ent-syn-
hand designation for a “mixture composed mainly of 12 disproportionately more than the major monepoxy
the pure enantiomer syn-12 and to a lesser extent of alcohol enantiomer syn-12. Therefore, the yield ratio
the pure enantiomer ent-syn-12”. Bearing this in ent-syn-12/syn-12 shrank the more the overepoxida-
mind, our study revealed at each cumulative reaction tion interfered: from its largest value 0.014 (=10%/
time tsampling the relative amounts of 20, syn-12+ent- 70%) at tsampling =1 h via distinctly smaller values like
syn-12 combined, and syn,syn-23 in the reaction mix- 0.026 (=1.38%/53.6%) at tsampling =8 h to a minimum
1
ture (from H NMR analysis) and the proportion of value of only 0.0050 (=0.18%/35.8%) at tsampling =32 h.
the pure enantiomers syn-12 and ent-syn-12 (from the Correspondingly, the ee value of syn-12 went up
ee value of “syn-12”). This allowed plotting the rela- (Figure 1, dashed curve): from 75% ee after 1 h via
tive amounts of the starting material 20 and its three 95% ee after 8 h to 99% ee after 32 h. At the latter
oxidation products syn-12, ent-syn-12, and syn,syn-23 point in time we stopped the epoxidation experiment
as a function of the reaction time (Figure 1, plain of Figure 1 for good because the enantiopurity, which
curves). When the reaction of Figure 1 had proceeded we had reached, was superb. Admittedly, this ee im-
for 1 h, only 4% of the divinylcarbinol 20 were left provement cost its (intrinsic) price: the loss of a little
and after 2 h just 2% 20, while after 4 h 20 had been more than half of the initially detected monoepoxy al-
completely consumed. As a consequence, at tsampling
=
cohol between tsampling =1 h and 32 h.
1 h already we detected the largest proportion of
monoepoxy alcohols (80%; 70% syn-12+10% ent-
syn-12) of the whole experiment; concomitantly, we Asymmetric Epoxidations of the Bis(cis-configured)
found a smaller proportion of the bisepoxide (16%) Divinylcarbinol 21
than ever again. By the subsequent analyses (at
tsampling =2, 4, 8, 16, and 32 h) we detected decreasing Scheme 6 shows the outcome of asymmetric epoxida-
amounts of the monoepoxy alcohols (74, 66, 55, 44, tions of the TBS-containing bis(cis-configured) di-
and 36%) and increasing amounts of the bisepoxy al- vinylcarbinol 21 under similar conditions as previous-
Scheme 6. Desymmetrizations of divinylcarbinol 21 by classical SAE (top) and by its Zr-analogue (bottom). Reagents and
conditions: a) Ti
1 h, addition of t-BuOOH (1.4 equiv.), 5 d, addition of t-BuOOH (0.65 equiv.), À258C, 17 h; 77% mixture of diastereomers,
ds=62:38, anti-13: 85% ee, ent-syn-13: 27% ee, 17% reisolated 21. b) Ti
A
(O-i-Pr)4 (1.0 equiv.), d-(À)-DiPT (1.1 equiv.), 4 Š
molecular sieves, CH2Cl2, À258C, 50 min, addition of 21, 1 h, addition of t-BuOOH (1.4 equiv.), 5 d, addition of t-BuOOH
(0.65 equiv.), À258C, 17 h; 74% mixture of diastereomers, ds=62:38, ent-anti-13: 84% ee, syn-13: 27% ee, 14% reisolated 21.
c) Zr
A
(2.0 equiv.), 30 min, addition of 21, À208C, 2 h; 66% syn-13 (dsꢀ98:2, 69% ee). d) Zr
(O-i-Pr)4·i-PrOH (1.0 equiv.), 4 Š mo-
lecular sieves, CH2Cl2, À208C; addition of d-(À)-DiPT (1.1 equiv.), t-BuOOH (2.0 equiv.), 30 min, addition of 21, À208C,
8 h; 49% ent-syn-13 (dsꢀ98:2, 91% ee). DiPT=diisopropyl tartrate; TBS=t-BuMe2Si.
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1137

Rainer Kramer et al.
FULL PAPERS
ly applied to the (para-methoxybenzyl)-containing di-
While the SAEs of the bis(cis-configured) sub-
vinylcarbinol 20 (Scheme 5, Figure 1). A remarkable strates 20 (Scheme 5) and 21 (Scheme 6) required sto-
difference between substrates 21 and 20 was the ichiometric amounts of both the isopropoxide and the
lower reactivity of the former. The SAEs of com- tartrate, the presence of 7 mol% of Ti
pound 21 in the presence of stoichiometric amounts mol% either of l-(+)- or d-(À)-DiPT sufficed to let
of Ti
(O-i-Pr)4 and l-(+)-DiPT (!anti-13) or d-(À)- the bis(trans-configured) substrate 22 react to comple-
A
ACHTREUNG
DiPT (!ent-anti-13) proceeded particularly sluggishly tion within 17–18 h (Scheme 7, top). This furnished
(Scheme 6, upper part). Even after 6 d unreacted 21 82% of the monoepoxy alcohol anti-14 highly diaste-
was recovered in yields of 17% and 14%, respectively. reoselectively (ds=97:3[31]) and highly enantioselec-
A more disturbing shortcoming of the two reactions tively (98% ee[32]) and its enantiomer ent-anti-14, re-
were their low levels of diastereocontrol (!anti- spectively, with comparably good values (78% yield,
13:ent-syn-13=62:38 and ent-anti-13:syn-13=62:38, ds=97:3,[31] 97% ee[32]). This is in full agreement with
respectively[28]) and enantiocontrol (eeanti-isomer =85 and the quality of the Hatakeyama and Schreiber synthe-
84%, respectively;[29] eesyn-isomer =27% in both ses, by the same desymmetrization technique
cases[30]). Work-up and purification by “routine” flash (Scheme 1), of the analogous bisbenzyl-containing
chromatography on silica gel[22] provided the monoe- monoepoxy alcohol anti-11 (70–80% yield, dsꢀ94ꢃ6,
poxy alcohols anti-13/ent-syn-13 (77% yield) and ent- 97% ee).[6a,8]
anti-13/syn-13 (74% yield), respectively. Careful re-
The Zr(O-i-Pr)4-mediated monoepoxidations of the
A
newed purification by flash chromatography provided bis(trans-configured) substrate 22 were less syn-selec-
access to 39% pure diastereomer anti-13 in one optical tive (dsꢂ80:20[31]; Scheme 7, bottom) than their pred-
series and to pure diastereomer 19% ent-anti-13 in the ecessors in the cis series (ds ꢀ 98:2, Scheme 5,
other; no pure syn-isomers could be obtained, though. Scheme 6). Epoxidizing for 2.5 h in the presence of l-
The Zr-mediated epoxidations of divinylcarbinol 21 (+)-DiPT, we isolated 19% unchanged divinylcarbinol
with l-(+)-DiPT proceeded much faster than the cor- 22, 40% of a 81:18 mixture[31] of monoepoxy alcohol
responding Ti(IV)-based SAE (Scheme 6, lower left). syn-14 and the isomer ent-anti-14, and 17% of a mix-
After 2 h 21 was already completely consumed (ac- ture of the three conceivable diastereomeric bisepox-
cording to TLC analysis). Monoepoxy alcohol syn-13 ides 24.[33] Monoepoxy alcohol diastereomer syn-14
resulted in 66% yield with almost perfect diastereose- was enriched as a 97:3 mixture[31] with the mentioned
lectivity (dsꢀ98:2[28]). It exhibited 69% ee,[30] which is anti-isomer by repeated flash chromatography on
somewhat inferior to the enantiocontrol of the Zr- silica gel.[22] Its ee was 70%.[34] As expected, the Zr-
mediated epoxidation of divinylcarbinol 20: 74% ee mediated epoxidation of substrate 22 under the influ-
after 1 h (Figure 1).
ence of the enantiomeric auxiliary d-(À)-DiPT deliv-
The Zr-mediated epoxidation of divinylcarbinol 21 ered a closely related result [25% recovered 22, 38%
in the presence of d-(À)-DiPT furnished the enantio- of a 79:21 mixture[31] of monoepoxy alcohol ent-syn-
meric monoepoxy alcohol ent-syn-13 (Scheme 6, 14 (72% ee[34]) and isomer anti-14 (28% ee[32]), 19% of
lower right). In this experiment we prolonged the re- the mixture of three diastereomeric bisepoxides
action time from 2 to 8 h, hoping for an increased 24].[33] Allowing the reactions to proceed for 12 rather
level of enantioselectivity due to the greater suscepti- than 2.5 h, we isolated the monoepoxy alcohols syn-14
bility of one monoepoxy alcohol enantiomer than to (ds=78:22[31]) and ent-syn-14 (ds=79:21[31]) in yields
overepoxidation. Indeed these conditions furnished a of 31% and 34% yield, respectively. Their enantiopur-
diastereomerically pure[28] specimen of ent-syn-13 with ities had increased to 86% and 82% ee,[34] respective-
an improved ee of 91%[30] in a somewhat reduced ly. However, in both reactions the major oxidation
yield (49%). The expected bisepoxy alcohol(s) product was a ternary mixture of bisepoxides (57%
evaded our attention, though.
and 55% yield, respectively).
Asymmetric Epoxidations of the Bis(trans-
configured) Divinylcarbinol 22
Stereochemical Assignment of Monoepoxy Alcohols
12–14
So far, the bis(cis-configured) divinylcarbinols 20 and The relative configurations of the epoxy alcohols anti-
21 had been desymmetrized by asymmetric epoxida- and syn-13 containing the tert-butyldimethylsilyl
tions such that the Ti
A
G
1
mediated processes exhibited complementary diaste- ferences of their H NMR signals resemble the corre-
reoselectivities: The former epoxidation was anti- and sponding H NMR shift differences of the para-me-
1
the latter syn-selective. The same bias was displayed thoxybenzyl-containing epoxy alcohols anti- and syn-
by analogous desymmetrizations of the bis(trans-con- 12. The respective chemical shift values of Table 2
figured) divinylcarbinol 22 (Scheme 7).
render credence to the correctness of this attribution.
1138
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148

Stereocomplementary Desymmetrizations of Divinylcarbinols
FULL PAPERS
Scheme 7. Desymmetrizations of divinylcarbinol 22 by classical SAE (top) and by its Zr-analogue (bottom). Reagents and
conditions: a) Ti
(O-i-Pr)4 (7 mol%), l-(+)-DiPT (9 mol%), 4 Š molecular sieves, CH2Cl2, À258C, 30 min, addition of t-
BuOOH (2.0 equiv.), 45 min, addition of 22, À258C, 18 h; 82% anti-14 (ds=97:3, 98% ee). b) Ti(O-i-Pr)4 (7 mol%), d-(À)-
DiPT (9 mol%), 4 Š molecular sieves, CH2Cl2, À258C, 30 min, addition of t-BuOOH (2.0 equiv.), 45 min, addition of 22,
ACHTREUNG
AHCTREUNG
À258C, 17 h; 78% ent-anti-14 (ds=97:3, 97% ee). c) Zr(O-i-Pr)4·i-PrOH (1.0 equiv.), 4 Š molecular sieves, CH2Cl2, À208C;
G
addition of l-(+)-DiPT (1.1 equiv.), t-BuOOH (2.0 equiv.), 1 h, addition of 22, À208C, 2.5 h; 40% 82:18 mixture syn-14:ent-
anti-14, syn-14: 70% ee, 17% mixture of 3 diastereomeric bisepoxy alcohols;[33] second flash chromatography 25% syn-14
(ds=97:3, 70% ee). d) Zr
A
(1.1 equiv.), t-BuOOH (2.0 equiv.), 1 h, addition of 22, À208C, 12 h; 31% 78:22 mixture syn-14:ent-anti-14, syn-14: 86% ee,
57% 67:27:7 mixture syn,syn-24:anti,syn-24/ent-anti,syn-24:anti,anti-24 (the section of the H NMR spectrum, which corrobo-
rates this assignment, is shown in Figure 2). e) Zr
tion of d-(À)-DiPT (1.1 equiv.), t-BuOOH (2.0 equiv.), 1 h, addition of 22, À208C, 2.5 h; 38% 79:21 mixture ent-syn-14:anti-
14, ent-syn-14: 72% ee, anti-14: 28% ee, 19% mixture of 3 diastereomeric bisepoxy alcohols.[33] f) Zr
(O-i-Pr)4·i-PrOH
(O-i-Pr)4·i-PrOH (1.0 equiv.), 4 Š molecular sieves, CH2Cl2, À208C, addi-
AHCTREUNG
(1.0 equiv.), 4 Š molecular sieves, CH2Cl2, À208C, addition of d-(À)-DiPT (1.1 equiv.), t-BuOOH (2.0 equiv.), 1 h, addition
of 22, À208C, 12 h; 34% 79:21 mixture ent-syn-14:anti-14, ent-syn-14: 82% ee, 55% mixture of 3 diastereomeric bisepoxy al-
cohols.[33] DiPT=diisopropyl tartrate; PMB=para-methoxybenzyl.
The relative configuration of the stereocenters in stereomer anti-12 via this reaction could not possibly
the epoxy alcohol anti-12 follows from the distinct- be an exception.[35]
1
ness of its H and 13C NMR spectra from the corre-
The syn-configuration within epoxy alcohol syn-12
sponding spectra of epoxy alcohol syn-12. This deduc- was established by a chemical correlation beginning
tion implies, of course, that the epoxidation, which with a tandem reduction by Red-Alꢂ (Scheme 8, top
gave rise to anti- and syn-12, rendered cis-substituted right). At À308C, a nucleophilic opening of the epox-
oxiranes. Since this kind of stereoselectivity has not ide ring occurred with some considerable regioselec-
been violated in SAE chemistry,[1–3] our access to dia- tivity. Heating to +608C accomplished another reduc-
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1139

Rainer Kramer et al.
FULL PAPERS
1
Table 2. Selected H NMR data (500 MHz, CDCl3, Me4Si as
internal standard) of the anti- and syn-diastereomers of the
bis(cis-configured) epoxy alcohols 12 and 13. PMB=para-
methoxybenzyl, TBS=t-BuMe2Si.
d
(1 H) [ppm] (multiplicity)
G
Proton anti-12
syn-12
anti-13
syn-13
1-HA
1-HB
2-H
3.63 ([a])
3.50 ([a]) 3.83 ([a])
3.75 ([a])
3.82 ([b])
3.17 (ddd)
3.79 ([b])
3.65 ([b]) 4.02 ([b])
3.25 (mC) 3.25
(ddd)
3.19 (mC)
3-H
4-H
3.00 (dd)
4.25 (br.
dd)
3.07 (dd) 3.02 (dd)
4.27 (dd) 4.31–4.37
(m)[c]
3.07 (dd)
4.29–4.32
(m)[d]
[a]
A-part of ABM signal.
B-part of ABM signal.
Not interpretable due to overlap with the AB signal
caused by 7-H2.
Not interpretable due to overlap with the AB signal
caused by 7-H2
[b]
[c]
[d]
Scheme 8. Elucidation of the relative and absolute configu-
rations in selected epoxy alcohol isomers. Reagents and con-
ditions: a) Red-Alꢂ (10 equiv.), toluene, À308C!608C,
tion, which consisted of an SNꢀ displacement of the al-
lylic para-methoxybenzyloxy group by the reduc-
tant.[36] An inseparable 65:35 mixture of the 1,3-diol 2.5 h; 92%. b) Red-Alꢂ (10 equiv.), toluene, À308C!608C,
(2S,4R)-25 and an isomeric 1,2-diol resulted.[37]
4.5 h; NaIO4 (1.0 equiv.), THF/H2O (1:1), room tempera-
ture, 2 h; 51%. c) Red-Alꢂ (4.0 equiv.), THF, À158C!608C,
Sodium periodate cleaved the 1,2-diol portion in this
mixture but let the 1,3-diol intact. This made the
latter readily isolable (51% yield). In order to clarify
2 h; 95%. d) Red-Alꢂ (10 equiv.), toluene, À308C!608C,
4.5 h; 64%. PMB=para-methoxybenzyl.
whether diol (2S,4R)-25 is anti- or syn-configured we
prepared a diastereomer thereof by subjecting epoxy pound already at the occasion of its preparation from
alcohol anti-12 to the same two-step reduction by epoxy alcohol syn-12 (vide supra). The signs of the
Red-Alꢂ (Scheme 9), which provided 1,3-diol (2S,4S)- specific rotation of the two samples of the diol
25. The latter was converted into acetonide cis-26 and
the cis-configuration of this compound established by
NMR spectroscopy.[12a] This proves that the underly-
ing diol (2S,4S)-25 is syn-configured, which in turn im-
plies that the diol (2S,4R)-25 discussed above is anti-
configured.[38,39]
The epoxy alcohols resulting from the bis(trans-
configured) divinylcarbinol 22 by the Ti
mediated (!anti-14 and ent-anti-14, respectively) vs.
Zr
(O-i-Pr)4-
A
syn-14 and ent-syn-14, respectively) were diastereo-
mers since their H NMR spectra differed from one
another (cf. Table 3).
Scheme 9. Determination of the relative configuration of
stereocenters C-2 and C-4 in epoxy alcohol anti-12. Reagents
The anti-configuration of the stereocenters in epoxy
alcohol ent-anti-14 was proved by reduction with Red-
Alꢂ (Scheme 8, top left). It provided the 1,3-diol
(2S,4R)-25 in 92% yield as a pure regioisomer. We
and conditions: a) Red-Alꢂ
(4 equiv.), toluene, 608C, 2 h. b)
A
2,2-Dimethoxypropane (6 equiv.), pyridinium para-toluene-
sulfonate (4 mol%), acetone, 108C, 16 h. PMB=para-me-
had established the anti-configuration of this com- thoxybenzyl.
1140
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148

Stereocomplementary Desymmetrizations of Divinylcarbinols
FULL PAPERS
(2S,4R)-25, which we had prepared, were identical configurations. We felt safe about attributing the ab-
and the absolute values of their specific rotations solute configuration to the stereocenters of epoxy al-
were almost the same (taking into account the differ- cohol anti-14, which we prepared by a SAE in the
ent enantiopurities of their precursors syn-12 and ent- presence of l-(+)-DiPT (Scheme 7, top left). This was
anti-14, namely 71% ee[40] and 97% ee, respectively). because the same epoxidation conditions had provid-
This means that the stereocenters C-2 and C-4 of ed the closely related epoxy alcohol anti-11 (cf.
epoxy alcohols syn-12 and ent-anti-14 have the same, Scheme 3).[6a,b,9b–e] The 3D-structure of the latter was
albeit not yet known absolute configurations.
proved unambiguously by conversions into the natural
The syn-configured 1,3-diol (2R,4R)-25, first pre- products 3-deoxy-d-manno-2-octulosonic acid [(+)-
pared from epoxy alcohol ent-anti-12 and Red-Alꢂ KDO],[6b] (11R,12S,13S,9Z,15Z)-9,12,13-trihydroxyoc-
(Scheme 8, bottom left), was also accessible by the tadeca-9,15-dienoic acid,[9b] prelactone C,[9c] (+)-aspili-
Red-Alꢂ reduction of epoxy alcohol syn-14 cin,[9d] and dysiherbaine.[9e] Knowing the absolute con-
(Scheme 8, bottom right). It provided 64% of the figuration of the enantiomeric epoxy alcohol ent-anti-
mentioned 1,3-diol and 18% of a mixture with the re- 14 then, too, the stereochemical correlation depicted
gioisomeric 1,2-diol[41]. The identity of the 1,3-diols in the upper moiety of Scheme 8 clarified the absolute
was established by NMR spectroscopy and the close configuration of epoxy alcohol syn-12.
match of their specific rotational values. The same
The absolute configuration of the stereocenters in
kind of reasoning as above revealed the relative con- epoxy alcohol anti-12 could not be elucidated earlier
figuration of the stereocenters in epoxy alcohol syn- than after a multistep transformation – undertaken
14. By the same token the stereocenters C-2 and C-4 for a synthetic objective – which led to the para-bro-
of epoxy alcohols syn-14 and ent-anti-12 possess iden- mobenzoate 27. Its stereostructure was determined by
tical but still unknown absolute configurations.
anomalous X-ray diffraction.[42] The mirror-inverted
The last reduction of an epoxy alcohol with Red- configuration was assigned to the enantiomeric epoxy
Alꢂ shown in Scheme 8 (bottom left) affected epoxy alcohol ent-anti-12 (Scheme 10). Thereupon, the ste-
alcohol ent-anti-12. It afforded the same syn-1,3-diol reochemical correlation depicted in the bottom sec-
(2R,4R)-25 (95% yield), which was previously ob- tion of Scheme 8 revealed the absolute configuration
tained from the Red-Alꢂ reduction of epoxy alcohol of epoxy alcohol syn-14.
syn-14. The near identity of the specific rotations of
Finally, it may be assumed that the absolute config-
the two samples implies that their epoxy alcohol pre- urations of the TBS-containing epoxy alcohols syn-
cursors ent-anti-12 and syn-14 are identically config- and ent-syn-13 are the same as in the analogous
ured in the absolute sense at the stereocenters la- PMB-containing epoxy alcohols syn- and ent-syn-12,
belled 2 and 4.[39]
respectively.
The ultimate step of the stereochemical assign-
ments of our epoxy alcohols concerned their absolute
Stereochemical Assignment of Bisepoxy Alcohols 23
and 24
1
Table 3. Selected H NMR shifts (500 MHz, CDCl3, Me4Si as
internal standard) of epoxy alcohols anti- and syn-14.
PMB=para-methoxybenzyl.
Overepoxidation of the non-racemic monoepoxy alco-
hol syn-12 in the presence of Zr(O-i-Pr)4 and l-(+)-
A
DiPT had led to the formation of an isomerically
pure bisepoxy alcohol 23 (Scheme 6, Figure 1). The
following set of data is in accordance with a meso-
d
(1 H) [ppm]
A
Proton
anti-14
syn-14
1-HA
1-HB
2-H
3-H
4-H
3.45
3.74
3.24 (mC)
3.02 (dd)
4.35 (m)
3.46
3.72
3.19 (ddd)
2.99 (dd)
4.09 (m)
Scheme 10. Proof of the absolute configuration of epoxy al-
cohol anti-12 by anomalous X-ray diffraction of the derived
para-bromobenzoate 27.[42] PMB=para-methoxybenzyl.
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1141

Rainer Kramer et al.
FULL PAPERS
The inseparable mixture of bisepoxides 24 obtained
by overepoxidation of the bis(trans-configured) di-
vinylcarbinol 22 consisted of all three conceivable bi-
sepoxide diastereomers according to the the 400 MHz
1H NMR spectrum. That indeed each of the com-
pounds compounds syn,syn-24, anti,syn-24 (or ent-an-
ti,syn-24), and anti,anti-24 was present was concluded
structure, i.e., with stereoformulas syn,syn-23 or anti,- from howmany doublets of doublets protons the pro-
anti-23. (1) The specific rotation of 23 was Æ0. (2) tons 3-H and 5-H caused to arise (cf. Figure 2). The
1
The 500 MHz H NMR spectrum of 23 displayed a best-resolved spectrum[33] displayed these protons as
single set of resonances for the two halves of the mol- four doublets of doublets. Two of them were superim-
ecule. (3) The 126 MHz 13C NMR spectrum of 23 ex- posed and two were completely separated. The major
hibited identical subspectra for both moieties of the constituent (67%) and the minor constituent (6%)
molecule. However, none of these data gives a clue as gave rise to a single doublet of doublet, each [d3-H and
to whether stereostructure syn,syn-23 or stereostruc- 5-H =3.07 (dd) and 3.029 Hz (dd), respectively], while
ture anti,anti-23 is correct. The time-dependent inves- the 27% component of the mixtures caused two such
tigation of Figure 1 allows us to make this distinction, signals [d3-H and 5-H =3.031 (dd) and 3.10 Hz (dd)]. Ac-
though, because it demonstated unequivocally from cordingly, the latter component is chiral – whether it
which precursor(s) the bisepoxide 23 forms: from equals enantiomer anti,syn-24 or its antipode ent-anti,-
both enantiomers of the initially obtained monoepoxy syn-24 is unknown – and the former components are
alcohol. Since the latter arose as a pure syn-diastereo- meso-compounds. Because of the ca. 80:20 bias for
mer, the overepoxidation product needed to retain a syn- vs. anti-selectivity in the Zr(O-i-Pr)4-mediated
G
syn-configured epoxy alcohol substructure. This re- monoepoxidation step of Scheme 7, it is likely (but
quirement is only met when the overepoxidation not proved) that the major meso-bisepoxide possesses
product possesses stereostructure syn,syn-23.
stereoformula syn,syn-24 and the minor meso-bis-
AHCTREUNG
1
Figure 2. Section from the H NMR spectrum (400 MHz, CDCl3, Me4Si as internal standard) of the ternary mixture of bis-
epoxy alcohols syn,syn-24, anti,syn- or ent-anti,syn-24, and anti,anti-24. PMB=para-methoxybenzyl.
1142
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148

Stereocomplementary Desymmetrizations of Divinylcarbinols
FULL PAPERS
Conclusions
We have realized the first asymmetric monoepoxida-
tions of secondary divinylcarbinols – namely com-
pounds 20 (cis-configured, PMBO-containing), 21
(cis-configured, TBSO-containing), and 22 (trans-con-
figured, PMBO-containing) – by t-BuOOH in the
presence of an enantiomerically pure diisopropyl tar-
trate (DiPT) and Zr(O-i-Pr)4. In each case, the enan-
N
tiopurity of the major monoepoxy alcohol product in-
creased with epoxidation time. For substrate 20, w e
established that this effect is due to a kinetic resolu-
tion of an initially less enantiomerically pure mono-
epoxy alcohol by the preferential overepoxidation of
the minor enantiomer. Based on this enrichment pro-
cess we obtained the monoepoxy alcohol syn-12 with
up to 99% ee.
The mentioned asymmetric monoepoxidations of
the secondary divinylcarbinols 20–22 in the presence
of Zr
ACHTREUNG
counterparts in the presence of Ti
AHCTREUNG
investigated as well. Interestingly, the two procedures
exhibit opposite diastereoselectivities, i.e., they are
stereocomplementary: Leaning on Ti
alcohols syn-12–syn-14 were obtained, and leaning on
Zr(O-i-Pr)4, epoxy alcohols anti-12–anti-14
(Scheme 11, bottom part). The controllability of this
bias turns the Zr(O-i-Pr)4-mediated desymmetriza-
(O-i-Pr)4, epoxy
ACHTREUNG
AHCTREUNG
tions into a valuable extension of the (now) classical
Sharpless methodology.
It is constructive to compare the stereochemical
outcomes of Zr(IV)- vs. Ti(IV)-mediated asymmetric
epoxidations of the secondary divinylcarbinols 20–22
(cf. Scheme 11, bottom part). 1) In the presence of a
given DiPT enantiomer the Zr(IV)- vs. Ti(IV)-mediat-
ed monoepoxidation provides mainly an epoxy alco-
hol, which exhibits identically configured stereocen-
ters in the epoxide ring and a differently configured
stereocenter in the alcohol moiety. Differently ex-
pressed, a given divinylcarbinol and a given DiPT
auxiliary furnish epoxy alcohol epimers when mono-
epoxidized with t-BuOOH in the presence of a partic-
Scheme 11. The Ti(IV)- or Zr(IV)-mediated desymmetriza-
tion of the tertiary divinylcarbinol 28 realized by Spivey
et al.[24] vs. the Ti(IV)- or Zr(IV)-mediated desymmetriza-
tions of the secondary divinylcarbinols 20–22 described here.
DiPT=diisopropyl tartrate; PMB=para-methoxybenzyl;
TBS=t-BuMe2Si.
ular DiPT enantiomer and Zr
N
N
respectively. 2) A syn-configured monoepoxy alcohol
with the R-configuration in the alcohol moiety is ob-
tained from the secondary divinylcarbinols 20–22 by
treatment with t-BuOOH, l-(+)-DiPT, and Zr(O-i-
U
Pr)4, while the anti-configured monoepoxy alcohol
with the identical R-configuration in the alcohol
Finally, Scheme 11 juxtaposes the l-(+)-DiPT-medi-
moiety is obtained from the same substrates and t- ateded desymmetrizations of the current study – af-
(O-i-Pr)4. Similarly, the fecting secondary divinylcarbinols – and those report-
BuOOH, d-(À)-DiPT, and Ti
with the S-configuration in the alcohol moiety are ob- 28.[24] With respect to the latter substrate turning from
ACHTREUNG
syn-configured (anti-configured) monoepoxy alcohols ed by Spivey et al. for the tertiary divinylcarbinol
tained from the secondary divinylcarbinols 20–22, t- Ti
BuOOH, d-(À)-DiPT [l-(+)-DiPT], and Zr(O-i-Pr)4 to producing epoxy alcohol ent-syn-29 rather than
[Ti(O-i-Pr)4].
(O-i-Pr)4- to Zr
(O-i-Pr)4-assistance was tantamount
AHCTREUNG
A
epoxy alcohol syn-29. Accordingly, switching the
metal inverted the sense and the extent of enantio-
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1143

Rainer Kramer et al.
FULL PAPERS
1H NMR (400.13 MHz, CDCl3/TMS): d=2.52 [br. d, 1H of
this isomer, JOH,4 =3.9 Hz, 4-OH (syn-12)], 2.78 [br. d, 1H of
this isomer, JOH,4 =2.8 Hz, 4-OH (anti-12)], 3.00 [dd, 1H of
this isomer, J3,4 =7.7 Hz, J3,2 =4.3 Hz, 3-H (anti-12)], 3.07
[dd, 1H of this isomer, J3,4 =7.3 Hz, J3,2 =4.7 Hz, 3-H (syn-
control of the latter epoxidations[43] while it main-
tained the sense and increased the extent of diaste-
reocontrol. With respect to our substrates (20–22) the
main effect of replacing Ti(IV) by Zr(IV) was turning
the epoxidation from anti- to syn-diastereoselective.
Last but not least, it is pointed out that divinylcar-
binol monoepoxides anti-6,[4,6b,c] 7,[6a,b] anti-10[8], and
anti-11[6b,9b–e,10] served as polyfunctional building
blocks in variety of natural product syntheses. Hence,
applications of divinylcarbinol monoepoxides 12–14
in asymmetric synthesis are conceivable, too.[42]
12)], 3.25 [ddd, 2H of both isomers, J2,1-H(A) =J2,1-H(B)
=
5.3 Hz, J2,3 =4.6 Hz, 2-H (anti- and syn-12)], AB signal [dA =
3.51, dB =3.61–3.67*, 2H of this isomer, JAB =11.6 Hz, in ad-
dition split by JA,2 =6.4 Hz, 1-H2 (syn-12); [* low-field por-
tion superimposed by highfield portion of the following AB
signal], AB signal [dA =3.63, dB =3.79, 2H of this isomer,
JAB =11.1 Hz, in addition split by JA,2 =5.4 Hz, JB,2 =6.0 Hz,
1-H2 (anti-12)], low-field portion superimposed by 3.79 [s,
OMe, 3H of both isomers (anti- and syn-12)], 4.00–4.15 {m,
2H of both isomers, 7-H2 (anti- and syn-12), therein inter-
pretable: AB signal [dA =4.05, dB =4.12, JAB =12.4 Hz, in
Experimental Section
4
addition split by JA,6 =5.5 Hz, JB,5 =1.3 Hz and JB,6 =7.1 Hz,
General remarks and a complete set of procedures and char-
acterization data are found in the Supporting Informations.
4JB,5 =1 Hz.4, 7-H2 (anti-12)], the respective AB signal for
the minor diastereomer syn-12 is not interpretable}, 4.25 [br.
ddd, 1H of this isomer, J4,3 =J4,5 =8.2 Hz, J4,OH =3.0 Hz, 4-H
(anti-12)] superimposed by 4.27 [br. dd, 1H of this isomer,
cis-(2R,3S,4S)-2,3-Epoxy-1,7-bis[(4-methoxybenzyl)-
oxy]hept-5-en-4-ol (anti-12)
J
4,5 =J4,3 =7.8 Hz, 4-H (syn-12)], 4.40–4.56 {m, 4H of both
isomers, 1’-H2, 1’’-H2 (anti- and syn-12); interpretable: d=
4.43 [s, 1’’-H2* (anti-12)], AB signal [dA =4.54, dB =4.46,
J
AB =11.4 Hz, 1’-H2* (anti-12)]; {*assignment interchangea-
ble}, 5.62–5.84 {m, 2H of both isomers, 5-H, 6-H (anti- and
syn-12), interpretable: d=5.70 [dddd, J5,6 =11.4 Hz, J5,4
7.5 Hz, 5-H (anti-12)] and d=5.82 [ddd, 6,5 =11.5 Hz,
6,7-H(A) =J6,7-H(B) =5.7 Hz, 6-H (anti-12)]}, 4 superimposed
AA’BB’-signals at d=6.85–6.91 and 7.22–7.29 [2”4H of
=
J
J
both isomers, 2-HAr-1, 3-HAr-1, 5-HAr-1, 6-HAr-1, 2-HAr-2, 3-HAr-2
,
5-HAr-2, 6-HAr-2 (anti- and syn-12)]; for analytical data of the
At À208C l-(+)-DiPT (11.9 mL, 13.3 g, 56.8 mmol,
pure major diastereomer anti-12 see ref.[12a]; ee (anti-12)=
1.1 equiv.) was added to
a
suspension of Ti(O-i-Pr)4
A
96% (by HPLC)[23]; ee (ent-syn-12)=27% (by HPLC)[27]
.
(14.7 mL, 15.4 g, 54.1 mmol, 1.05 equiv.) and powdered 4 Š
molecular sieves (1.5 g) in CH2Cl2 (700 mL). After 30 min t-
BuOOH (4.4M in CH2Cl2, 23.2 mL, 102 mmol, 2.0 equiv.)
was added and the mixture was stirred for 1 h. Then a solu-
tion of divinylcarbinol 20 (19.8 g, 51.6 mmol) in CH2Cl2
(70 mL) was added dropwise within 20 min. The reaction
mixture was stirred for 3 d at À208C and the reaction was
terminated by adding a solution of FeSO4 (222 g) and citric
acid (81 g) in H2O (750 mL). With vigorous stirring the mix-
ture was allowed to reach room temperature and the phases
were separated. The aqueous phase was extracted with t-
BuOMe (4”150 mL) and the combined organic phases were
evaporated under vacuum to a volume of approximately
250 mL. The remaining solution was treated with a solution
(20 mL) prepared from NaOH (30 g), NaCl (5 g) and H2O
(90 mL) and the resulting mixture was vigorously stirred at
room temperature for 1 h. Thereafter the phases were sepa-
rated and the organic phase was dried over Na2SO4. The sol-
vents were evaporated under vacuum and the residue was
submitted to flash chromatography (12 cm, cyclohexane/
EtOAc, 2:1) to afford a 84:16 mixture[21],* of anti-12 and
ent-syn-12 (fractions 30–63, 14.9 g, 72%). [*During flash-
chromatography only fractions which according to TLC con-
trol contained the major diastereomer anti-12 were collect-
ed]. All following fractions which contained only the minor
diastereomer syn-12 or mixtures of diastereomeric bisepoxy
alcohols were discarded. The ratio of anti-12 and ent-syn-12
before flash chromatography was determined as ꢂ75:25[21]
by 1H NMR spectroscopic analysis of the crude material.
cis-(2R,3S,4R)-2,3-Epoxy-1,7-bis[(4-methoxybenzyl)-
oxy]hept-5-en-4-ol (syn-12)
At À208C l-(+)-DiPT (51 mL, 57 mg, 0.24 mmol, 1.1 equiv.)
and a solution of t-BuOOH (4.43M in CH2Cl2, 100 mL,
0.44 mmol, 2.0 equiv.) were added to a suspension of Zr(O-
A
i-Pr)4·i-PrOH (84 mg, 0.22 mmol, 1.0 equiv.) and powdered
4 Š molecular sieves (140 mg) in CH2Cl2 (4.5 mL). The mix-
ture was stirred for 1 h at À208C before a solution of di-
vinylcarbinol 20 (85 mg, 0.22 mmol) in CH2Cl2 (2.0 mL) was
slowly added. After stirring for 4 h at À208C, the reaction
was terminated by adding a solution (1.0 mL) prepared
from NaOH (30 g), NaCl (5 g) and H2O (90 mL). The cool-
ing bath was removed and the mixture was stirred for 2 h at
room temperature. After addition of t-BuOMe (10 mL),
H2O was removed by addition of Na2SO4 and the resulting
mixture was filtered through a pad of Celite. The filter cake
was washed with t-BuOMe. Excess t-BuOOH was removed
by azeotropic distillation with toluene (4”5 mL). The resi-
due was purified by flash chromatography (2.0 cm, cyclohex-
1144
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148

Stereocomplementary Desymmetrizations of Divinylcarbinols
FULL PAPERS
ane/EtOAc, 5:2) to afford monoepoxy alcohol syn-12 (frac-
tions 35–46, 56 mg, 63%) as pure diastereomer and a 68:32
mixture of bisepoxide syn,syn-23 and monoepoxy alcohol
syn-12 (fractions 47–70, 33 mg; ratio pure syn-12: 12 mg,
11% and 23 mg bisepoxide syn,syn-23, 25%). Both samples
were colorless liquids. The total yield of diastereomerically
pure monoepoxy alcohol syn-12 was 74% and its ee value
14 (signals not listed)]: d=2.09 (br. s, 1H, OH), 3.02 (dd,
1H, J3,4 =3.2 Hz, J3,2 =2.3 Hz, 3-H), 3.24 (ddd, 1H, J2,1-H(A)
=
5.5 Hz, J2,3 =J2,1-H(B) =2.6 Hz, 2-H), AB signal (dA =3.45,
dB =3.74, 2H, JAB =11.7 Hz, in addition split by JA,2 =5.7 Hz,
J
B,2 =2.8 Hz, 1-H2), 3.790 and 3.794 (2 ” s, 2”3H, 2”O-
CH3), 4.01 (dmC, 2H, J7,6 ꢂ5.5 Hz, 7-H2), 4.35 (mC, 1H, 4-
1
H), 4.45 (s, 2H, Ar-CH2 ), AB signal (dA =4.47, dB =4.51,
2
was 82% (by HPLC)[27]; [a]D20: À8.7 (c 0.8, CHCl3); [a]23065
:
2H, JAB =11.5 Hz, Ar-CH2 ), AB signal (dA =5.74, dB =5.92,
4
À33.1 (c 0.8, CHCl3) {this sample stems from the time-re-
2H, JAB =15.7 Hz, in addition split by JA,4 =6.5 Hz, JA,7
=
4
solved experiment (Figure 1) after 16 h, 44% yield and 98%
1.5 Hz, JB,7 =5.8 Hz, JB,4 =1.2 Hz, A: 5-H, B: 6-H), overlap-
ping AA’BB’ signals centered at d=6.866 or 6.869 and d=
7.251 or 7.255 respectively [2”4H, 2”C6H4; contains sol-
vent peak at d=7.26 (CHCl3)]; 13C NMR (125.68 MHz,
CDCl3/CDCl3): d=53.72 (C-2), 55.33 (2”O-CH3), 57.37 (C-
3), 69.24 (C-4), 69.29 (C-1), 69.60 (C-7), 72.12 and 72.99 (C-
1’, C-1’’), 113.88 and 113.90 (C-2Ar-1, C-2Ar-2, C-6Ar-1, C-6Ar-2),
129.44 and 129.47 (C-3Ar-1, C-3Ar-2, C-5Ar-1, C-5Ar-2), 129.64
(C-5), 129.93 and 130.23 (C-1Ar-1, C-1Ar-2), 130.42 (C-6),
159.32 and 159.38 (C-4Ar-1, C-4Ar-2); IR (CDCl3): n=3605,
3005, 2955, 2935, 2860, 1615, 1585, 1515, 1465, 1445, 1365,
1300, 1250, 1175, 1100, 1035, 975, 915, 895, 825 cmÀ1; ele-
mental analysis calcd. (%) for C23H28O6 (400.5): C 68.98, H
7.05; found: C 68.89, H 7.22.
1
ee (by HPLC)[27]}; H NMR (499.87 MHz, CDCl3/TMS): d=
2.67 (br s, 1H, OH), 3.07 (dd, 1H, J3,4 =7.4 Hz, J3,2 =4.5 Hz,
3-H), 3.25 (ddd, 1H, J2,1-H(A) =6.5 Hz, J2,3 =J2,1-H(B) =4.2 Hz,
2-H), AB signal (dA =3.50, dB =3.65, 2H, JAB =11.4 Hz, in
addition split by JA,2 =6.6 Hz, JB,2 =3.7 Hz, 1-H2), 3.79 (s,
3H, 2”O-CH3), AB signal (dA =4.02, dB =4.06, 2H, JAB
=
=
=
4
12.8 Hz, in addition split by JA,6 =5.9 Hz, JA,5 =1.2 Hz, JB,6
6.2 Hz, 4JB,5 =1.5 Hz, 7-H2), 4.27 (br. dd, 1H, J4,5 =J4,3
7.8 Hz, 4-H), 4.39–4.51 (m, 4H, 1’-H2, 1’’-H2), AB signal
(dA =5.66, dB =5.78, 2H, JAB =11.3 Hz, in addition split by
JA,4 =8.2 Hz, JB,7-H(A) =JB,7-H(B) =5.9 Hz, A: 5-H, B: 6-H),
overlapping AA’BB’ signals centered at d=6.87 and d=7.24
[2”4H, 2”C6H4; contains solvent peak at d=7.26
(CHCl3)]; 13C NMR (125.69 MHz, CDCl3/CDCl3): d=55.33
(2”O-CH3), 56.00 (C-2), 58.93 (C-3), 65.83 (C-7), 66.79 (C-
4), 67.79 (C-1), 72.41 and 73.02 (2”C-1’, C-1’’), 113.92 and
113.96 (C-2Ar-1, C-2Ar-2, C-6Ar-1, C-6Ar-2), 129.52 and 129.56
(3Ar-1, C-3Ar-2, C-5Ar-1, C-5Ar-2), 129.78 (C-1Ar-1, C-1Ar-2), 130.24
(C-5), 130.38 (C-6), 159.43 and 159.44 (C-4Ar-1, C-4Ar-2); IR
(CDCl3): n=3590, 3420, 3005, 2960, 2935, 2840, 1615, 1515,
1465, 1305, 1250, 1175, 1085, 1035, 905, 850, 825 cmÀ1; ele-
mental analysis calcd. (%) for C23H28O6 (400.5): C 68.98, H
7.05; found: C 69.16, H 7.16. For analyctical data of a pure
trans-(2S,3S,4R)-2,3-Epoxy-1,7-bis[(4-methoxybenz-
yl)oxy]hept-5-en-4-ol (syn-14)
sample of bisepoxy alcohol syn,syn-23 see ref.[12b]
.
At À208C l-(+)-DiPT (240 mL, 268 mg, 1.14 mmol,
1.1 equiv.) and a solution of t-BuOOH (4.43M in CH2Cl2,
472 mL, 2.08 mmol, 2.0 equiv.) were added to a suspension of
trans-(2R,3R,4R)-2,3-Epoxy-1,7-bis[(4-methoxybenz-
yl)oxy]hept-5-en-4-ol (ent-anti-14)
Zr(O-i-Pr)4·i-PrOH (404 mg, 1.04 mmol, 1.0 equiv.) and
A
powdered 4 Š molecular sieves (210 mg) in CH2Cl2 (26 mL).
The mixture was stirred for 1 h at À208C before a solution
of divinylcarbinol 22 (400 mg, 1.04 mmol) in CH2Cl2
(4.0 mL) was slowly added. After stirring for 2.5 h at
À208C, the reaction was terminated by addition of a cold
solution (1.0 mL) prepared from NaOH (30 g), NaCl (5 g)
and H2O (90 mL). The cooling bath was removed and the
mixture was stirred for 2 h at room temperature. After addi-
tion of t-BuOMe (10 mL), H2O was removed by addition of
Na2SO4 and the resulting mixture was filtered through a pad
of Celite. The filter cake was washed with t-BuOMe. Excess
t-BuOOH was removed by azeotropic distillation with tolu-
ene (4”5 mL). The residue was purified by flash chromatog-
raphy (3.0 cm, cyclohexane/EtOAc, 3:1 fraction 160, 2:1).
Fractions 71–91 afforded divinylcarbinol 22 (76 mg, 19%).
Fractions 95–151 furnished a 92:8[31] mixture of the monoe-
poxy alcohols syn-14 and anti-14 (72 mg, 18%) and the fol-
lowing fractions 122–161 afforded a 75:25[31] mixture of the
monoepoxy alcohols syn-14 and anti-14 (93 mg, 22%). Final-
ly in fractions 163–180 a mixture of 3 diastereomeric bise-
poxy alcohols[33] (73 mg, 17%) was obtained. The total yield
of both diastereomeric monoepoxy alcohols syn-14 and ent-
syn-14 was 40% (49% based on recovered starting material)
and the ratio of syn-14 and anti-14 was 82:18.[31] The 92:8
mixture of monoepoxy alcohols syn<b>-14</b> and anti-
ent-anti-14 was prepared from divinylcarbinol 22 (200 mg,
0.52 mmol) by the same procedure as described for anti-12
using d-(À)-DiPT (9.5 mL, 11 mg, 45 mmol, 9 mol%) as
chiral additive (see above). The epoxidation was performed
using Ti(O-i-Pr)4 (10.5 mL, 10.0 mg, 35 mmol, 7 mol%), t-
A
BuOOH (4.43M in CH2Cl2, 226 mL, 1.0 mmol, 2.0 equiv.)
and powdered 4 Š molecular sieves (310 mg) in CH2Cl2
(total volume: 7.5 mL) at À258C. After 17 h the reaction
was terminated by addition of a cold solution (1.0 mL) pre-
pared from NaOH (30 g), NaCl (5 g) and H2O (90 mL). The
residue of the work-up was purified by flash chromatogra-
phy (2.0 cm, cyclohexane/EtOAc, 5:2) to afford the title
compound together with 3 mol% of the diastereomeric
monoepoxy alcohol syn-14 (fractions 44–75, 164 mg, 82%,
ds=97:3[31]) as colorless oil; ee (ent-anti-14)=97% (by
HPLC)[32]; ee (syn-14) not determined; [a]D: À7.7 (c 1.00,
CHCl3); 1H NMR [499.87 MHz, CDCl3/TMS; sample con-
tains 3 mol% of the diastereomeric monoepoxy alcohol syn-
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1145

Rainer Kramer et al.
FULL PAPERS
14 was submitted to a second flash chromatography (2.0 cm,
cyclohexane/EtOAc, 3:1) delivering the title compound to-
gether with 3 mol% of the diastereomeric monoepoxy alco-
hol syn-14 (fractions 70–92, 46 mg, ds=97:3[31]) as colorless
oil; ee (anti-14)=70% (by HPLC)[32]; ee (syn-14) not deter-
(300 mL). After drying of the organic phase over Na2SO4
the solvent was evaporated under reduced pressure and the
residue was submitted to flash chromatography (12 cm, cy-
clohexane/EtOAc, 3:2) to afford the title compound (frac-
tions 14–46, 19.8 g, 82%) as a slightly yellowoil. 1H NMR
(499.87 MHz, CDCl3/TMS): d=2.35 (br. d, 1H, JOH,4 =
3.1 Hz, 4-OH), 3.79 (s, 6H, 2”O-CH3), AB signal (dA =4.01,
dB =4.08, 4H, JAB =12.5 Hz, in addition split by JA,2 and
1
mined; [a]2D0: À2.04 (c 1.08, CHCl3); H NMR [499.87 MHz,
CDCl3; sample contains 3 mol% of the diastereomeric mon-
oepoxy alcohol anti-14 (signals not listed)]: d=2.12 (br. s,
1H, OH), 2.99 (dd, 1H, J3,4 =4.6 Hz, J3,2 =2.3 Hz, 3-H), 3.19
(ddd, 1H, J2,1-H(A) =5.4 Hz, J2,3 =J2,1-H(B) =2.6 Hz, 2-H), AB
signal (dA =3.46, dB =3.72, 2H, JAB =11.7 Hz, in addition
JA,6 =5.2 Hz respectively, JB,2 and JB,6 =5.2 Hz respectively, 1-
H2 and 7-H2 respectively), 4.42 (s, 4H, 1’H2, 1’’-H2), 5.15 (td,
1H, J4,3 =J4,5 =6.8 Hz, J4,OH =2.9 Hz, 4-H), 5.62–5.70 {m, 4H,
approximately interpretable as AB signal [dA =5.64, dB =
5.67, JAB =11.1 Hz (ꢁJcis), in addition split by JA,4 =6.6 Hz,
JB,1 and JB,7 =5.5 Hz, respectively; A: 3-H and 5-H respec-
tively, B: 2-H and. 6-H respectively)}, AAꢀBBꢀ signal cen-
tered at d=6.87 and d=7.25 [2”4H, 2”C6H4; contains sol-
vent peak at d=7.26 (CHCl3)]; 13C NMR (125.69 MHz,
CDCl3/CDCl3): d=55.26 (2”OCH3), 64.48 (C-4), 65.61 (C-
1, C-7), 72.15 (C-1’, C-1’’), 113.85 (C-2Ar-1, C-2Ar-2, C-6Ar-1, C-
6Ar-2), 128.04 (C-2, C-6), 129.46 (3Ar-1, C-3Ar-2, C-5Ar-1, C-5Ar-2),
129.97 (C-1Ar-1, C-1Ar-2), 134.22 (C-3, C-5), 159.32 (C-4Ar-1, C-
4Ar-2); IR (film): n=3415, 3010, 2995, 2910, 2855, 2840, 1615,
1585, 1515, 1460, 1385, 1300, 1250, 1175, 1080, 1035, 820,
760, 710 cmÀ1; elemental analysis calcd. (%) for C23H28O5
(384.5): C 71.85%, H 7.34%; found: 71.75%, 7.40%.
split by JA,2 =5.5 Hz, JB,2 =3.0 Hz, 1-H2), 3.791 and 3.795 (2”
4
s, 2”3H, 2”O-CH3), 4.01 (ddd, 2H, J7,6 =5.4 Hz, J7,5
=
5J7,4 =1.2 Hz, 7-H2), 4.09 (mC, 1H, presumably interpretable
as br. ddd, J4,5 ꢂJ4,3 ꢂJ4,OH ꢂ4.6 Hz, 4-H), 4.45 (s, 2H, Ar-
1
CH2 ), AB signal (dA =4.47, dB =4.51, 2H, JAB =11.5 Hz,
2
Ar-CH2 ), AB signal (dA =5.82, dB =5.92, 2H, JAB =15.7 Hz,
in addition split by JA,4 =5.7 Hz, JA,7 =1.4 Hz, JB,7 =5.4 Hz,
4
4JB,4 =1.2 Hz, A: 5-H, B: 6-H), overlapping AA’BB’ signals
centered at d=6.868 or 6.870 and 7.250 or 7.254, respective-
ly [2”4H, 2”C6H4; contains solvent peak at d=7.26
(CHCl3)]; 13C NMR (125.69 MHz, CDCl3/CDCl3): d=54.94
(C-2), 55.33 (2”O-CH3), 58.11 (C-3), 69.20 (C-1), 69.61 (C-
7), 71.03 (C-4), 72.14 and 73.03 (C-1’, C-1’’), 113.88 and
113.90 (C-2Ar-1, C-2Ar-2, C-6Ar-1, C-6Ar-2), 129.44 and 129.47
(C-3Ar-1, C-3Ar-2, C-5Ar-1, C-5Ar-2), 129.49 (C-6), 129.92 and
130.24 (C-1Ar-1, C-1Ar-2), 130.52 (C-5; same intensity as peak
,
at d=129.49), 159.31 and 159.39 (C-4Ar-1 C-4Ar-2); IR
(CDCl3): n=3690, 3605, 2980, 2935, 2900, 2860, 2840, 1615, Acknowledgements
1585, 1515, 1465, 1385, 1365, 1300, 1250, 1175, 1100, 1035,
975, 940, 825 cmÀ1
C23H28O6 (400.5): C 68.98, H 7.05; found: C 68.96, H 7.10.
;
elemental analysis calcd. (%) for
We gratefully acknowledge financial support by the European
Graduate College “Catalysts and Catalytic Reactions for Or-
ganic Synthesis” of the German Research Foundation
(DFG).
A
dien-4-ol (20)
References
[1] Short communications: a) T. Katsuki, K. B. Sharpless, J.
Am. Chem. Soc. 1980, 102, 5974–5976 (using stoichio-
metric or near stoichiometric amounts of titanium iso-
propoxide and the dialkyl tartrate); b) R. M. Hanson,
K. B. Sharpless, J. Org. Chem. 1986, 51, 1922–1925
(using molecular sieves and <10 mol% both of titani-
um isopropoxide and the dialkyl tartrate); full paper:
c) Y. Gao, R. M. Hanson, J. M. Klunder, S. Y. Ko, H.
Masamune, J. Am. Chem. Soc. 1987, 109, 5765–5780.
[2] Reviews: a) A. Pfenniger, Synthesis 1986, 89–116; b) T.
Katsuki, V. S. Martin, Org. React. 1996, 48, 1–299; c) T.
Katsuki, in: Comprehensive Asymmetric Catalysis,
(Eds: A. Pfaltz, E. N. Jacobsen, H. Yamamoto), Spring-
er-Verlag, Berlin, 1999, pp 621–648; d) J. A. Johnson,
K. B. Sharpless, in: Catalytic Asymmetric Synthesis,
(Ed: I. Ojima), 2nd edn., Wiley-VCH, Weinheim, 2000,
pp 231–286.
[3] a) R. M. Hanson, Chem. Rev. 1991, 91, 437–478;
b) P. C. A. Pena, S. M. Roberts, Curr. Org. Chem. 2003,
7, 555–571.
[4] S. Hatakeyama, K. Sakurai, S. Takano, J. Chem. Soc.
Chem. Commun. 1985, 1759–1761.
[5] a) B. Häfele, D. Schrçter, V. Jäger, Angew. Chem. 1986,
98, 89–90: Angew. Chem. Int. Ed. Engl. 1986, 25, 87–
At room temperature Cu(OAc)2·H2O (12.5 g, 62.5 mmol,
A
1.0 equiv.) was added in one portion to a suspension of Zn
dust (86,3 g, 1.25 mol, 20 equiv.) in H2O (500 mL). After stir-
ring at this temperature for 10 min AgNO3 (10.6 g,
62.5 mmol, 1.0 equiv.) was added in small portions within
30 min (Caution! exothermic reaction) and stirred for 1 h.
The suspension was filtered under an inert atmosphere
maintained by placing over the apparatus an inverted funnel
through which N2 was continuously passed. The residue was
successively washed with H2O (200 mL), methanol
(400 mL), acetone (400 mL) and t-BuOMe (400 mL). The
metal powder was suspended in methanol (100 mL) and a
solution of dialkynylcarbinol 18 (23.8 g, 62.5 mmol) in meth-
anol (60 mL) was added at room temperature. The resulting
mixture was vigorously stirred at 408C for 14 h. The reaction
mixture was cooled to room temperature, filtered through a
pad of celite, and the filter cake was washed with t-BuOMe.
The filtrate and washings were washed with saturated aque-
ous NaHCO3 (300 mL) and saturated aqueous NaCl
1146
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148

Stereocomplementary Desymmetrizations of Divinylcarbinols
FULL PAPERS
89; b) V. Jäger, W. Hümmer, U. Stahl, T. Grazca, Syn-
thesis 1991, 769–775.
[20] a) P. C. Astles, E. J. Thomas, J. Chem. Soc. Perkin
Trans. 1 1997, 845–856 (85% ee); b) H. Shibuya, K. Ka-
washima, N. Narita, M. Ikeda, I. Kitagawa, Chem.
Pharm. Bull. 1992, 40, 1154–1165 (84–85% ee).
[6] a) S. L. Schreiber, T. S. Schreiber, D. B. Smith, J. Am.
Chem. Soc. 1987, 109, 1525–1529; b) D. B. Smith, Z.
Wang, S. L. Schreiber, Tetrahedron 1990, 46, 4793–
4808; c) M. Nakatsuka, J. A. Ragan, T. Sammakia,
D. B. Smith, D. E. Uehling, S. L. Schreiber, J. Am.
Chem. Soc. 1990, 112, 5583–5601.
[7] A. Bayer, J. S. Svendsen, Eur. J. Org. Chem. 2001,
1769–1780.
[8] A. Herunsalee, M. Isobe, S. Pikul, T. Goto, Synlett
1991, 199–201.
[9] a) S. Hatakeyama, K. Satoh, S. Takano, Tetrahedron
Lett. 1993, 34, 7425–7428; b) S. Hatakeyama, K.
Kojima, H. Fukuyama, H. Irie, Chem. Lett. 1995, 763–
764; c) T. Esumi, H. Fukuyama, R. Oribe, K. Kawazoe,
Y. Iwabuchi, H. Irie, S. Hatakeyama, Tetrahedron Lett.
1997, 38, 4823–4826; d) T. Nishioka, Y. Iwabuchi, H.
Irie, S. Hatakeyama, Tetrahedron Lett. 1998, 39, 5597–
5600; e) E. Masaki, J. Maeyama, K. Kamada, T. Esumi,
Y. Iwabuchi, S. Hatakeyama, J. Am. Chem. Soc. 2000,
122, 5216–5217.
[10] S. Hatakeyama, K. Sakurai, H. Numata, N. Ochi, S.
Takano, J. Am. Chem. Soc. 1988, 110, 5201–5203.
[11] A more general analysis of stereocontrol in consecutive
asymmetric functionalizations of compounds, which
contain two constitutionally identical prochiral sites led
to a complementary conclusion: Unless a fast and a
slowreacting enantiomer emerge from the first func-
tionalization, the ee does not change during the second
functionalization: R. Rautenstrauch, Bull. Soc. Chim.
Fr. 1994, 131, 515–524.
[21] The diastereomeric compositions of monoepoxy alco-
hol 12 and its enantiomer were determined from the
1
ratio of the integrals over the following H NMR sig-
nals (300 or 400 MHz, CDCl3, Me4Si as internal stan-
dard): d=3.00 [dd, J3,4 =7.7, J3,2 =4.3 Hz, 3-H (anti-12)]
vs. d=3.07 [dd, J3,4 =7.3, J3,2 =4.7 Hz, 3-H (syn-12)].
“dsꢀ98:2” (cf. Scheme 5) means that the mentioned
1H NMR signal of the minor diastereomer could not be
detected.
[22] W. C. Still, M. Kahn, A. Mitra, J. Org. Chem. 1978, 43,
2923–2925.
[23] The ee of monoepoxy alcohol anti-12 and its enantio-
mer were determined by HPLC (Chiralpak AD
column, n-heptane/i-PrOH, 90:10, UV detector:
238 min, flowrate: 1.0 mLmin À1, 158C isothermal);
tR =35.7 min for anti-12, tR =40.1 min for ent-anti-12.
[24] A. C. Spivey, S. J. Woodhead, M. Weston, B. I. An-
drews, Angew. Chem. 2001, 113, 791–793; Angew.
Chem. Int. Ed. 2001, 40, 769–771; cf. also A. C. Spivey,
M. Weston, S. Woodhead, Chem. Soc. Rev. 2002, 31,
43–59.
[25] This precedent is compared with our results in the con-
cluding section of this paper.
[26] The molar ratio of divinylcarbinol 20, monoepoxy alco-
hol syn-12 (including some enantiomer), and bisepoxy
alcohol syn,syn-23 in the reaction mixture was deter-
mined from the ratio of the integrals over the following
1H NMR signals of the isolated crude product
(300 MHz, CDCl3, Me4Si as internal standard): d=3.00
[dd, J3,4 =7.7 Hz, J3,2 =4.3 Hz, 3-H (syn-12)] vs. d=3.13
[12] Preliminary communications: a) T. Berkenbusch, R.
Brückner, Synlett 2003, 1813–1816; b) R. Kramer, R.
Brückner, Synlett 2006, 33–38.
[dd,
J
3,4 =J5,4 =6.0 Hz,
J3,2 =J5,6 =4.5 Hz, 3-H, 5-H
[13] Method for preparing Rieke zinc: a) R. D. Rieke, S. J .
Uhm, P. M. Hudnall, J. Chem. Soc. Chem. Commun.
1973, 269–270; b) R. D. Rieke, Top. Curr. Chem. 1975,
59, 1–33; c) R. D. Rieke, S. J. Uhm, Synthesis 1975,
452–453; d) R. D. Rieke, Aldrich. Acta, 2000, 33, 52–
60.
[14] Procedure: W.-N. Chou, D. L. Clark, J. B. White, Tetra-
hedron Lett. 1991, 32, 299–302.
[15] Method: W. Boland, N. Schroer, C. Sieler, Helv. Chim.
(syn,syn-23)] vs. d=5.23 [dmc, J4,OH ꢂ7.5, 4-H (20)].
[27] The ee of monoepoxy alcohol syn-12 and its enantio-
mer were determined by HPLC (Chiralpak AD
column, n-heptane/i-PrOH 90:10, UV detector:
238 min, flowrate: 1.0 mLmin À1, 158C isothermal);
tR =48.8 min for syn-12, tR =54.0 min for ent-syn-12.
[28] The diastereomeric compositions of monoepoxy alco-
hol 13 and its enantiomer were determined from the
1
ratio of the integrals over the following H NMR sig-
Acta 1987, 70, 1025–1040.
nals (300 MHz, CDCl3, Me4Si as internal standard): d=
3.02 [dd, J3,4 =7.5 Hz, J3,2 =4.1 Hz, 3-H (anti-13)] vs. d=
3.07 [dd, J3,4 =7.4 Hz, J3,2 =4.5 Hz, 3-H (syn-13)]. “dsꢀ
[16] Procedure: a) R. W. Bates, R. Fernandez-Moro, S. V.
Ley, Tetrahedron Lett. 1991, 32, 2651–2654; b) M.
Avignon-Tropis, J. R. Pougny, Tetrahedron Lett. 1989,
30, 4951–4952.
1
98:2” means that the mentioned H NMR signal of the
minor diastereomer could not be detected.
[17] Procedure: a) S. E. Denmark, T. K. Jones, J .Org.
Chem. 1982, 47, 4595–4597; b) R. W. Bates, D. D.
Martin, W. J. Kerr, J. G. Knight, S. V. Ley, A. Sakellar-
dis, Tetrahedron 1990, 46, 4063–4082.
[18] An almost identical 2-step synthesis of divinylcarbinol
22 (61% overall yield) from ether 16 was described:
B. M. Trost, S. T. Wrobleski, J. D. Chisholm, P. E. Har-
rington, M. Jung, J. Am. Chem. Soc. 2005, 127, 13589–
13597.
[19] a) R. M. Williams, S. B. Rollins, T. C. Judd, Tetrahedron
2000, 56, 521–532 (88% ee); b) T. Yoshino, Y. Nagata,
E. Itoh, M. Hashimoto, T. Katoh, S. Terashima, Tetra-
hedron 1997, 53, 10239-10252 (85% ee).
[29] The ee of monoepoxy alcohol anti-13 and its enantio-
mer were determined by HPLC (Chiralpak OD-H
column, n-heptane/i-PrOH, 99.5:0.5, UV detector:
210 min, flowrate: 1.0 mLmin À1, 208C isothermal);
tR =13.8 min for ent-anti-13, tR =15.1 min for anti-13.
[30] The ee of monoepoxy alcohol syn-13 and its enantio-
mer were determined by HPLC (Chiralpak OD-H
column, n-heptane/i-PrOH, 200:1, UV detector:
210 min, flowrate: 1.0 mLmin À1, 208C isothermal);
tR =24.5 min for syn-13, tR =27.8 min for ent-syn-13.
[31] The diastereomeric compositions of monoepoxy alco-
hol 14 and its enantiomer were determined from the
1
ratio of the integrals over the following H NMR sig-
Adv. Synth. Catal. 2008, 350, 1131 – 1148
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1147

Rainer Kramer et al.
FULL PAPERS
nals (500 MHz, CDCl3, Me4Si as internal standard): d=
2.99 [dd, J3,4 =4.6 Hz, J3,2 =2.3 Hz, 3-H (syn-14)] vs. d=
3.02 [dd, J3,4 =3.2 Hz, J3,2 =2.3 Hz, 3-H (anti-14)].
split by JA,4 =8.8 Hz, JA,2 =3.5 Hz and JB,2 =8.5 Hz,
J
B,4 =2.9 Hz, 3-H2 (2S,4R)-25] vs. AB signal [dA =1.77,
dB =1.87,
JAB =14.7 Hz, in addition split by JA,3
=
[32] The ee of monoepoxy alcohol anti-14 and its enantio-
mer were determined by HPLC (Chiralpak AD
column, n-heptane/i-PrOH, 85:15, UV detector:
227 min, flowrate: 1.0 mLmin À1, 308C isothermal);
tR =52.7 min for anti-14, tR =58.1 min for ent-anti-14.
[33] The molar ratio of bisepoxy alcohols syn,syn-24, anti,-
anti-24, and anti,syn-24/ent-anti,syn-24 could be extact-
ed from a 400 MHz 1H NMR spectrum (CDCl3) as
shown in Figure 2, but not from the routinely registered
5.7 Hz,
J
A,1-H(A) =3.9 Hz, A,1-H(B) =3.5 Hz and JB,3
J
=
J
B,1-H(B) =8.4 Hz, JB,1-H(A) =4.7 Hz, 2-H2 (1,2-diol)] and
d=4.46 [s, benzylPMB-CH2 (1,2-diol)] vs. d=4.49 [s, ben-
zylPMB-CH2 [(2S,4R)-25]}.
[38] An independent assignment of 1,3-diol diastereomers
syn- und anti-25 is possible by the 13C NMR criterion of
R. W. Hoffmann, U. Weidmann, Chem. Ber. 1985, 118,
3980–3992. The found that in a pair of diastereomeric
1,3-diols the sum of the chemical shifts of the oxygen-
bearing carbon nuclei (here: C-2 and C-4) is smaller
for the anti- than the syn-isomer. We observed d=
67.93 and 67.99 (C-2 and C-4; )Sd =135.92) for anti-25
and d=71.08 (C-4) and d=71.16 (C-2) ()Sd =142.24)
for syn-25..
[39] The absolute configurations mentioned in this para-
graph and drawn in Scheme 8 and Scheme 9 were not
yet known at this point of our investigation.
[40] The reduction of monoepoxy alcohol syn-12 was per-
formed before we learnt to prepare this compound as
enantioselectively as documented in Scheme 4 and
Figure 1. The enantiomeric monoepoxy alcohol ent-syn-
12 (85% ee) was transformed into the enantiomeric
1,3-diol 2R,4S-25 {[a]D:+5.1 (c 0.98, CHCl3)} by same
procedure (47% yield).
1
300 MHz H NMR spectra.
[34] The ee of monoepoxy alcohol syn-14 and its enantio-
mer were determined by HPLC (Chiralpak AD-H
column, n-heptane/i-PrOH, 80:20, UV detector:
227 min, flowrate: 1.0 mLmin À1, 308C isothermal);
tR =26.6 min for ent-syn-14, tR =28.8 min for syn-14.
[35] Apart from this, the cis-configuration of the oxirane
rings in the monoepoxy alcohols in question can be de-
duced from the absolute value of the H,H coupling be-
tween the ring protons: It was relatively large (4.2 Hz)
in compounds anti-12 and syn-12 and smaller (2.3 Hz)
in the isomeric epoxy alcohols anti- and syn-14, each of
which contains a trans-configured oxirane ring. Both
values agree well with the rule of thumb, according to
3
3
which Jcis-oxirane is typically 4.5 Hz and Jtrans-oxirane 3.0 Hz
according to: E. Pretsch, H. Seibl, W. Simon, Tabellen
zur Strukturaufklärung organischer Verbindungen mit
spektroskopischen Methoden, 3rd edn., Springer Verlag,
Berlin 1990, H65.
[41] The composition of this mixture could not be deter-
1
mined by H NMR analysis (300 MHz, CDCl3, Me4Si as
internal standard) due to insufficiently separated sig-
nals ()1,3-diol:1,2-diolꢀ79:21).
[36] Procedure (described for the analogous transformation
of monoepoxy alcohol anti-11!homoallyl-1,3-diol in
81% yield): S. Hatakeyama, K. Satoh, S. Takano, Tetra-
hedron Lett. 1993, 34, 7425–7428.
[37] The ratio between 1,3-diol (2S,4R)-25 and the isomeric
1,2-diol was determined by averaging the ratios of the
integrals over the following pairs of 1H NMR signals
(500 MHz, CDCl3, Me4Si as internal standard): AB
signal [dA =1.55, dB =1.65, JAB =14.4 Hz, in addition
[42] a) R. Kramer, Dissertation, Universität Freiburg, 2007;
b) R. Kramer, R. Brückner, manuscript in preparation.
[43] It was no surprise that 28 underwent SAE with as little
as 14% ee because it is well-known that tertiary allyl al-
cohols bind poorly to the tartrate/titanium complex;[2]
cf. also p. 721 in D. C. Dittmer, R. P. Discordia, Y.
Zhang, C. K. Murphy, A. Kumar, A. S. Pepito, Y.
Wang, J. Org. Chem. 1993, 58, 718–731.
1148
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1131 – 1148