Discrimination of enantiotopic iodine atoms by an iodine/magnesium exchange reaction

Article abstract of DOI:10.1002/(SICI)1521-3765(19990104)5:1<337::AID-CHEM337>3.0.CO;2-P

An enantioselective iodine/magnesium exchange reaction between the diiodoalkane 4 and a chiral Grignard reagent 12 has been realized at -78°C in THF. The resulting α-iodoalkylmagnesium reagents 13 are configurationally stable under these conditions and during trapping by a benzaldehyde/dimethylaluminium chloride system to furnish the iodohydrins 19 and epoxides 9.

Full text of DOI:10.1002/(SICI)1521-3765(19990104)5:1<337::AID-CHEM337>3.0.CO;2-P

FULL PAPER
Chiral Organometallic Reagents: Part XXII[1]
Discrimination of Enantiotopic Iodine Atoms by an Iodine/Magnesium
Exchange Reaction
Volker Schulze and Reinhard W. Hoffmann*[a]
Abstract: An enantioselective iodine/magnesium exchange reaction between the
diiodoalkane 4 and a chiral Grignard reagent 12 has been realized at 788C in THF.
The resulting a-iodoalkylmagnesium reagents 13 are configurationally stable under
these conditions and during trapping by a benzaldehyde/dimethylaluminium chloride
system to furnish the iodohydrins 19 and epoxides 9.
Keywords: carbenoids
reagents
´ Grignard
us to consider an iodine/magnesium exchange reaction, since
a-heterosubstituted alkylmagnesium species have higher
thermal[4] and probably also configurational stability[5] than
the corresponding lithium compounds.
Introduction
At present, stereoselective synthesis relies almost exclusively
on the differentiation of enantiotopic (or diastereotopic) faces
of planar sp2-hybridized reaction centers.[2] However, the
differentiation of enantiotopic groups is increasingly gaining
importance, as evidenced by the spectacular enantioselective
deprotonation of alkylcarbamates 1 developed by Hoppe
(Scheme 1).[3]
Results and Discussion
We initiated a study of the iodine/magnesium exchange
reaction[6] of the 1,1-diiodoalkane 4, which could be effected
with isopropylmagnesium chloride in THF over 2 h at 788C.
The resulting a-iodoalkylmagnesium compound 5 could be
quenched with D2O to furnish 6 in 98% yield (Scheme 2).
X
H
H
H
nBuLi
O
R
R
R
R
Li
N(iPr)2
Sparteine
X =
O
X
1
X
H
X
X
R-Met
I
I
4
I
I
D2O
ClMg
chiral
ligand
X = Halogen
Met
H
Ph
Ph
Ph
Ph
- 78 oC, THF
D
I
MgCl
2
3
5
6
Scheme 1. Enantioselective deprotonation of alkylcarbamates 1 and the
halogen/metal exchange reaction of dihaloalkanes 2.
- 50 to
PhCHO
Ph
20 o
C
I
PhCHO
Ph
Ph
Ph
It occurred to us, that an enantioselective halogen/metal
exchange reaction on 1,1-dihaloalkanes 2 should have a
similar potential to generate chiral organometallic com-
pounds of the type 3 by a differentiation of enantiotopic
halogen atoms. Provided the compounds 3 have sufficient
thermal and configurational stability, they would be of
considerable interest for stereoselective synthesis. This led
O
OMgCl
Ph
O
7
8
9
Scheme 2. Iodine/magnesium exchange reaction of
4
to give
5
and
subsequent reactions to give 6 ± 9.
Trapping of 5 with benzaldehyde generated the magnesium
alkoxide 8 which cyclized to form the epoxide 9 (55%) with a
cis/trans selectivity of >98%. This
[a] Prof. Dr. Reinhard W. Hoffmann, Dr. V. Schulze
Fachbereich Chemie, Philipps-Universität Marburg
Hans-Meerwein-Strasse, D ± 35032 Marburg (Germany)
Fax: (‡49)6421-28-89-17
encouraged us to use a chiral Grignard
reagent to differentiate between the
enantiotopic iodine atoms of 4. When
using[7] the Grignard reagent 10 de-
MgCl
H
10
E-Mail: rwho@ps1515.chemie.uni-marburg.de
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0337 $ 17.50+.50/0
337

R. W. Hoffmann and V. Schulze
FULL PAPER
rived from ( )-menthyl chloride,[8] the resulting epoxide 9
(41%) showed little (2%) if any enantiomeric enrichment,
according to the 1H NMR spectra recorded in the presence of
tris(3-heptafluorobutyryl-d-camphorato)europium.
dehyde, produced 56% of the product 9 (97% cis) with an ee
value of 49% (scheme 4).
Encouraged by these results we started to investigate this
reaction in greater detail. The key question is whether a
uniform reagent 12 can be generated. We therefore monitored
the reaction between 15 and diisopropylmagnesium (16) by
The low enantiomeric enrichment in 9 could arise either
because the organomagnesium species 5 is not configuration-
ally stable on the time scale of the experiment, or, because the
enantiotopic differentiation between the two iodine atoms in
4 is too low. We assumedÐand later provedÐthat the a-
iodoalkylmagnesium compounds 5 are configurationally sta-
ble at 508C. Therefore the stereodifferentiating step needed
closer attention. The halogen metal exchange reaction is
assumed to proceed in a two-step manner,[9] involving halogen
ate complexes[10] as intermediates. For the case at hand, it
remains open, whether the first or the second step of the
following reaction sequence determines the configuration of
the final a-iodoalkyl Grignard reagent 5.
I
I
I
N
N
Ph
13a
Ph
14a
Ph
N
OMg
4
I
O
Mg
Mg
*
Ph
N
I
PhCHO
*
*
Mg
N
+
I
N
N
N
Ph
13b
Ph
14b
Ph
N
OMg
O
*
12a
O
N
9
Ph
Ph
Scheme 4.
Provided the first step in the reaction between 4 and a
Grignard reagent is irreversible, the formation of the ate
complexes 11a and 11b[11] would be stereodefining
(Scheme 3). In this case some level of enantioselection was
13C NMR spectroscopy. Addition of one equivalent of 15a to a
solution of 16 in THF at 08C was found to result in a 1:2:1
statistical mixture of products 16, 12a, and 17a (Scheme 5).
Evidently, such a reagent mixture is not desirable for
enantioselective halogen/metal exchange reactions, since the
remaining diisopropylmagnesium (16) will react with the
diiodo compound 4 to generate a racemic a-iodoalkylisopro-
pylmagnesium species.[13] Hence, the results obtained above
suggest, that pure 12 should show a substantially higher
enantioselection on reaction with the diiodoalkane 4. We
therefore intensified our efforts to generate uniform 12. On
addition of two equivalents of 15a to diisopropylmagnesium
(16) all of the material could be converted into 17a. Addition
of further diisopropylmagnesium (16) at this stage, however,
did not lead to a comproportionation to give 12a.
-
+MgX
k2
I
I
Ph
-
Ph
I
MgX
R
_
11a
5a
I
R
I
+
R
MgX
Ph
I
4
-
+MgX
I
I
k2
Ph
-
Ph
I
MgX
R
It appears that 12a has a basicity similar to that of
diisopropylmagnesium (16) in deprotonating 15a. In search
of a ligand, the anion of which is less basic than that of 15a, we
turned to the arylthio-substituted ligands 15b and 15c. Using
those, 13C NMR spectroscopy indicated the selective gener-
ation of the desired reagents 12b and 12c. For further ligands
tested see reference [14]. Armed with this knowledge, the
chiral magnesium reagent 12c was treated with the diiodo
compound 4 to give the a-iodoalkylmagnesium compounds 13
which could be trapped by the addition of D2O to give a 94%
yield of 6.
11b
_
5b
I
R
Scheme 3. Course of the reaction of 4 to give 5 via the ate complexes 11.
expected on reaction of the menthyl Grignard reagent.
However, if the first step were reversible, the second step,
the rate- and stereodetermining conversion of the ate com-
plexes 11 into 5 would involve attack of the magnesium ion at
the diiodo-substituted carbon atom of 11. Attack at this site
would be remote from the chirality of the menthyl group due
to the interspersed iodine atom and, hence, the attendant
enantioselectivity is likely to be small. Enantiodifferentiation
in such a situation would rather require a chiral magnesium
cation. Combination of the magnesium dication with a chiral
anion would bind the chiral information tightly to the
magnesium ion by Coulombic
Trapping of the derived a-iodoalkylmagnesium compounds
5 or 13 by benzaldehyde needed also improvement. As
discussed above, trapping of 5 with benzaldehyde furnished
the epoxide 9 in yields, which did not exceed 55%, but were
attraction. For that reason, we
O
O
O
O
O
treated one equivalent of diiso-
propylmagnesium with one
O
O
N
N
N
N
N
R´
Mg
Mg
R
+
R
+
H
+
Mg
R
Mg
R
R´
equivalent of the bisoxazolidine
15a[12] to generate what was
presumed to be 12a. Reaction
of the resulting solution with
the diiodoalkane 4 at 788C
followed by trapping of the
intermediates 13 with benzal-
N
N
N
2
2
O
15
16
16
12
17
a
b
1 equiv
1 equiv
1 equiv
1 equiv
0.5 equiv
0.95 equiv
0.25 equiv
trace
0.25 equiv
trace
Scheme 5. Reaction of 15 with 16. a: R ˆ H, b: R ˆ S-C6H5, c: S-C6H4Cl. R' ˆ CH(CH3)2.
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999 0947-6539/99/0501-0338 $ 17.50+.50/0 Chem. Eur. J. 1999, 5, No. 1
338

Chiral Organometallic Reagents
337±344
frequently lower. This is a consequence of a side reaction, in
which the magnesium alkoxides 8 were oxidized by excess of
benzaldehyde to the a-iodoketone 7. Aside from the fact that
this lowers the yield of 9, oxidation of 8 by benzaldehyde
could be diastereomer-discriminating and could therefore
influence the cis/trans ratio of the product 9. In the case of 14
any differentiation in the oxidation of the diastereomers 14
and 14b would alter the enantiomeric purity of the product 9
ultimately obtained. It was therefore mandatory to develop a
high-yield trapping procedure for 13. This goal was achieved
by using a combination of benzaldehyde and dimethylalumi-
num chloride as trapping reagent. In this way, the a-iodo
alcohols 19 plus some of the epoxide 9 could be obtained in
good yield (Scheme 6).
To address this possibility, 22 was generated from the
bisoxazolidine 15c and trimethylaluminum. When 22 was
added to racemic 5 at
788C followed by addition of
benzaldehyde the reaction produced a 28% yield of the
epoxide 9 (cis only) and a 22% yield of the iodohydrins 19
(syn/anti ˆ 7:3). The much lower enantiomer enrichment of 9
(7%), syn-19 (14%) and of anti-19 (1%) formed in this
experiment suggests that the formation of 5 is not a serious
complication in the trapping of the chiral Grignard reagent 13
with dimethylaluminum chloride and benzaldehyde, a reac-
tion which led to products with a much higher ee value. In fact,
addition of dimethylaluminum chloride to 13 first, followed
by benzaldehyde, or, trapping of 13 with benzaldehyde
precomplexed to dimethylaluminum chloride made little
difference in the course of the reaction. However, the
enantiomeric enrichment of the resulting iodohydrins 19 was
somehow variable. This was eventually traced to a kinetic
resolution during the reaction of the diastereomeric a-
iodoalkyl Grignard reagents 13 with benzaldehyde. The set
of experiments shown in Scheme 7, in which the relative
amount of benzaldehyde was varied, illustrates this point.
AlMe2Cl
4
I
O
O
I
Ph
N
I
Ph
N
N
Ph
Mg
R
14
Mg
*
N
12c
13
O
H2O
S-C6H4Cl (p)
Bu3SnH
R
=
I
O
Ph
Ph
Ph
Ph
+
AIBN, 80 o
benzene
C
Ph
Ph
OH
OH
18
19
9
4
I
KOH
I
O
O
73 %. d.s. 93 %
53 % ee
9 % 50 % ee
Ph
2 eq Me2AlCl
EtOH
Cl
N
Me
Me
Ph
N
Mg
N
I
Al
Mg
SAr
- 78 o
THF
C
- 78 o
THF
C
N
Scheme 6. Reaction of 12c with 4.
23
12c
The enantiomeric excess of the iodohydrins 19 was
determined to be 53% after conversion into the epoxide 9.
For the determination of the absolute configuration of the
iodohydrin 19, a sample that gave epoxide 9 of 29% ee was
reduced with tributyltin hydride to give the dextrorotatory
alcohol 18 (25% optical purity) which is known[15] to have the
(R)-configuration. While the introduction of dimethylalumi-
num chloride into the system led to acceptable yields in
trapping of 13, it created at the same time potential
complications, that required further control experiments.
For example, is there a transmetalation from magnesium in
13 to the aluminum compound 20?
Ar = C6H4Cl(p)
2.5 equiv[a]
I
PhCHO
H2O
Ph
Ph
96 %[a] 39 % ee
d.r. 96 : 4
THF, - 78 oC
0.59 equiv[a]
OH
19
51 %[a] 65 % ee
17 %[a] 68 % ee
d.r. 96 : 4
d.r. 97 : 3
PhCHO
0.23 equiv[a]
PhCHO
[a] referring to the amount of 23 generated
Scheme 7. Influence of the relative amount of benzaldehyde on the ee
value and diastereomeric ratio.
I
I
I
Me2AlCl
Me2AlCl
N
N
Ph
Ph
Ph
AlMe2
Mg
Li
*
In addition, these observations prove the important fact
that any diastereoisomerization of the magnesium compounds
13 or 23 is slower than their rate of reaction with the
complexed benzaldehyde. That the a-iodoalkylmagnesium
compounds 13 do not racemize in the presence of dimethy-
laluminum chloride more rapidly than they are trapped by
aldehydes has also been assessed by another test[16] based on
kinetic resolution (Scheme 8). To this end, a mixture of the
Grignard reagents 5 with dimethylaluminum chloride is added
to 2-benzyloxypropionaldehyde 24 to give the two diastereo-
meric adducts 25 and 26. These products should have the same
configuration at C-2 and C-3 and be epimeric at C-4. In line
with this notion, reduction of a mixture of 25 and 26 with
tributyltin hydride furnished a single isomer of 28. Treatment
of the isomers 25 and 26 separately with potassium hydroxide
led specifically to a cis-epoxide 29 and a trans-epoxide 30
(Scheme 9).
?
20
13
21
In order to assess the reactivity of 20, we generated it from
the lithium compound 21 at 1058C. Adding benzaldehyde
and warming the mixture to 788C led to only negligible
amounts of 19 or 9. Apparently, the reactivity of 20 is much
lower than that of 13, and, hence, the transmetalation of 13 to
give 20 is not considered to interfere in the reactions and
trapping of 13. A second complication could be a ligand
exchange reaction of 13 to give 5.
I
I
Me2AlCl
N
N
N
N
Ph
Me2Al
Ph
+
*
MgCl
Mg
*
?
22
5
13
Me3Al
15
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0339 $ 17.50+.50/0
339

R. W. Hoffmann and V. Schulze
FULL PAPER
I
Since we had established above that the chiral Grignard
reagent 12c is able to differentiate the enantiotopic iodine
atoms of 4, it should also be in a position to discriminate the
enantiomers of the racemic bromoiodo compound 31, and,
hence, should allow a kinetic resolution of 31. When racemic
31 was treated with 0.5 equivalents of the chiral Grignard
reagent 12c, followed by trapping with dimethylaluminum
chloride/benzaldehyde, the resulting bromohydrin 34 was
obtained in 33% yield and 18% ee (Scheme 10).
This indicated[18] that the remaining bromoiodoalkane 31a,
reisolated in 62% yield, should have an ee of about 8%. This
material was then subjected to an iodine ± magnesium ex-
change reaction with achiral isopropylmagnesium chloride.
After two hours at 788C, the resulting 32a was trapped with
dimethylaluminum chloride and benzaldehyde to furnish
72% of the bromohydrin ent-34 of about 5% ee (Scheme 11).
The latter had the opposite configuration to the bromohydrin
34 obtained above. The fact that the resulting ent-34 has an ee
value similar to that of the starting material demonstrates that
the generation of 32a from 31a and trapping of 32a by
dimethylaluminum chloride/benzaldehyde is not subject to
significant racemization.
OBn
Ph
I
OBn
I
OBn
MgCl . Me2AlCl
Ph
Ph
3
3
4
2
4
2
+
THF, - 78 to 20 o
C
O
OH
OH
25
26
71 %
63 %
91 : 9
rac-24
(S)-24
67 : 33
I
5
Ph
MgCl
OBn
AlMe2Cl
rac-27
(S)-27
THF, - 78 to 20 oC
O
71 %
58 %
91 : 9
64 : 36
Scheme 8. Formation of 25 and 26.
I
OBn
OBn
I
OBn
Bu3SnH
Ph
3
Ph
Ph
3
4
2
+
4
2
AIBN, 80 oC
benzene
OH
25
OH
OH
28
26
Ph
OBn
OBn
Ph
O
O
Therefore, the a-bromoalkylmagnesium compound 32 and,
by inference, the a-iodoalkylmagnesium compound 5 are
30
29
Scheme 9. Reactions of 25 and 26.
0.5 equiv
The differences in the prod-
uct ratio of 25 and 26 obtained
on reaction of 5 with either
racemic or enantiomerically
pure aldehyde 24 indicate that
the Grignard reagent 5 is trap-
ped more rapidly than it race-
mizes. Further experiments de-
lineated that the same result
applies irrespective of whether
dimethylaluminum chloride is
first added to the Grignard
O
O
N
Mg
N
PhCHO
H2O
1.5 equiv Me2AlCl
Br
SAr
N
N
Ph
33
- 78 o
THF
C
- 78 o
THF
C
Mg
*
Br
12c
+
Ph
31
Ar = C6H4Cl (p)
- 78 oC
THF
I
Br
Br
Ph
Ph
Ph
31a
34
I
OH
reisolated in 62 %
ee = ca. 8 %
33 %
ee = 18 %
Scheme 10. Kinetic resolution of 31.
reagent and then benzaldehyde, or whether the aldehyde is
precomplexed with dimethylaluminum chloride first, that is
by using 27 as trapping reagent.
Br
Br
Ph
Ph
Ph
31a
I
OH
ee = ca. 8 %
ee = ca. 5 %
ent-34
This statement of configurational stability of 5 relates to the
time scale of trapping of 5 by the dimethylaluminum chloride/
benzaldehyde system. But this leaves the possibility that the
two diastereomeric complexes 13 have equilibrated prior to
the addition of the trapping reagent. To exclude this
possibility the configurational stability of 5 and of the
diastereomeric complexes 13 had to be established on a
macroscopic time scale at 788C.
For this we envisaged the a-bromoalkyl Grignard reagent
32 to be generated from an enantiomerically enriched
iodobromo compound 31 by a chemoselective iodine/magne-
sium exchange reaction (cf. the preferential exchange of
iodine over bromine for lithium.[17]
72 %
(CH3)2CHMgCl
- 78 oC, THF
1.5 equiv Me2AlCl PhCHO
H2O
Br
- 78 oC
THF
Ph
- 78 o
THF
C
MgCl
32a
Scheme 11. Formation of ent-34.
configurationally stable on a macroscopic time scale at
788C. With this information in hand, the following state-
ments can be made: The chiral Grignard reagent 12 is in a
position to differentiate the enantiotopic iodine atoms of 4
and generates in a kinetically controlled iodine/magnesium
exchange reaction a mixture of two diastereomeric complexes
13a and 13b. These may be trapped in good yield by a reagent
Br
Br
?
+
Ph
Ph
32
ClMg
- 78 oC, THF
I
MgCl
31
340
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0340 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Chiral Organometallic Reagents
337±344
1-Bromo-1-iodo-2-phenylethane (31): A solution of 1,1-dibromo-2-phenyl-
ethane (5.28 g, 20 mmol) in THF (6 mL) was added dropwise at 788C to a
mixture of n-butyllithium in hexane (1.55m, 24.0 mmol), THF (40 mL),
and diethyl ether (20 mL). After stirring for 0.5 h at 1058C a solution of
iodine (7.61 g, 30.0 mmol) in THF (20 mL) was continuously added over a
period of 1.5 h . After stirring for 1 h at 1058C the solution was allowed to
warm to 708C over 3 h. Saturated aqueous NH4Cl solution (30 mL), 20%
Na2S2O3 solution (20 mL), and petroleum ether (30 mL) were added. The
phases were separated and the aqueous phase was extracted with
petroleum ether (3 Â 30 mL). The combined organic phases were washed
with brine (20 mL), dried (Na2SO4), and concentrated. The residue was
chromatographed over silica gel with petroleum ether to furnish 31 as a
colorless oil (4.71 g, 76%). For analysis a sample was purified further by
low-temperature crystallization from petroleum ether/dichloromethane at
788C; 1H NMR (300 MHz, CDCl3): d ˆ 3.61 (dd, J ˆ 14.5 and 7.2 Hz,
1H), 3.70 (dd, J ˆ 14.6 and 7.0 Hz, 1H), 5.47 (t, J ˆ 7.1 Hz, 1H), 7.12 ± 7.30
(m, 5H); 13C NMR (75 MHz, CDCl3): d ˆ 11.6, 53.2, 127.7, 128.6, 129.2,
138.2; elemental analysis calcd for C8H8BrI (311.0): C 30.90, H 2.59; found
C 30.97, H 2.52.
combination of dimethylaluminum chloride and benzalde-
hyde to give the iodohydrins 19. The enantiomeric purity of
these iodohydrins is influenced by a kinetic resolution during
trapping of the diastereomeric complexes 13a and 13b.
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft (SFB 260) and the Fonds
der Chemischen Industrie for support of this study.
Experimental Section
All temperatures quoted are not corrected. Low temperatures were
determined with a GTH 215 precision digital thermometer from Grei-
singer, Regenstauf, Germany. Reactions with organometallic compounds
were carried out in dried solvents under nitrogen or argon. 1H NMR,
13C NMR: Bruker AC 300, AMX 500. Boiling range of petroleum ether:
40 ± 608C. pH7 buffer: NaH2PO4 ´ 2H2O (56.2 g) ‡ Na2HPO4 ´ 2H2O
(213.2 g) in water (1.0 L). Column chromatography: silica gel 60: 63 ±
200 mm or aluminium oxide 90, neutral, activity I (63 ± 200 mm), E. Merck
AG, Darmstadt. Flash chromatography: silica gel 60 (40 ± 63 mm), E. Merck
AG, Darmstadt. MPLC: siliga gel 60 (15 ± 25 mm), E. Merck AG,
Darmstadt.
Phenylthio-[2-((4S)-4,5-dihydro-4-isopropyl-oxazolyl)]-[2-((4S)-4,5-dihy-
dro-4-isopropyl-oxazolidinylidene)]methane (15b): n-Butyllithium in hex-
ane (1.44m, 22.0 mmol) and TMEDA (3.30 mL, 22 mmol) were added to a
solution of bisoxazolidine 15a[12] (4.77 g, 20.0 mmol) in THF (75 mL) at
788C. After stirring for 1.5 h at 788C and 0.5 h at 08C the solution was
cooled to 788C again. A solution of diphenyldisulfide (10.92 g, 50 mmol)
in THF (20 mL) was added dropwise. After stirring for 1.5 h at 788C and
2 h at 08C water (50 mL) and tert-butyl methyl ether (50 mL) were added.
The phases were separated and the aqueous phase was extracted with tert-
butyl methyl ether (3 Â 50 mL). The combined organic phases were washed
with semisaturated aqueous NaHCO3 solution (50 mL), water (30 mL), and
brine (30 mL), dried (Na2SO4) and concentrated. The residue was
chromatographed over basic alumina (500 g) with petroleum ether/ethyl
acetate ˆ 9:1 to 3:1 to give 15b (4.60 g, 66%) as a colorless solid of m.p. 84 ±
868C. For analysis a sample was recrystallized from petroleumether/tert-
butyl methyl etherˆ 1:1. [a]2D5 ˆ ‡36.9 (c ˆ 1.105, CH2Cl2); 1H NMR
(300 MHz, CDCl3): d ˆ 0.89 (d, J ˆ 6.7 Hz, 6H), 0.99 (d, J ˆ 6.7 Hz, 6H),
1.71 (pseudo-octet, J ˆ 6.7 Hz, 2H), 3.86 (dt, J ˆ 8.6 and 7.1 Hz, 2H), 4.02
(dd, J ˆ 8.4 and 7.3 Hz, 2H), 4.36 (dd, J ˆ 8.6 and 8.6 Hz, 2H), 6.93 ± 7.21
(m, 5H), 9.00 ± 10.50 (br, 1H); 13C NMR (75 MHz, CDCl3): d ˆ 18.3, 18.6,
32.9, 57.9, 67.2, 71.2, 124.0, 124.3, 128.3, 141.1, 168.5; elemental analysis
calcd for C19H26N2O2S (346.5): C 65.86, H 7.56, N 8.09; found C 65.83, H
7.52, N 7.97.
Preparation of starting materials
1,1-Dibromo-2-phenylethane: To a mixture of THF (80 mL), diethyl ether
(80 mL), and diisopropylamine (17.5 g, 162 mmol) was added at 308C a
solution of n-butyllithium in hexane (1.55m, 162 mmol). After stirring for
0.5 h at 308C and 0.5 h at 08C the solution was cooled to 1058C. A
solution of dibromomethane (28.16 g, 162 mmol) in THF (50 mL) was
added dropwise. After stirring for 0.5 h at 1058C a solution of benzyl
bromide (27.71 g, 162 mmol) in THF (30 mL) was added dropwise. Stirring
was continued for 0.5 h at 1058C. After the mixture had been allowed to
warm to room temperature overnight, saturated aqueous NH4Cl solution
(100 mL), aqueous 20% Na2S2O3 solution (20 mL) and petroleum ether
(50 mL) were added. The phases were separated and the aqueous phase
was extracted with petroleum ether (3 Â 50 mL). The combined organic
phases were washed with brine (30 mL), dried (Na2SO4), and concentrated.
Distillation at 858C/10 2 Torr furnished the product (17.50 g, 41%) as a
p-Chlorophenylthio-[2-((4S)-4,5-dihydro-4-isopropyloxazolyl)]-[2-((4S)-
4,5-dihydro-4-isopropyl-oxazolidinylidene)]methane (15c): The bis-oxazo-
lidine 15a[12] (7.15 g, 30.0 mmol) and 4,4'-dichlorodiphenyldisulfide (12.9 g,
44.9 mmol) were allowed to react as described for 15b. The crude product
was purified by chromatography over silica gel with petroleum ether/ethyl
acetate ˆ 10:1 to tert-butyl methyl ether/petroleum etherˆ 1:1 to tert-butyl
methyl ether/petroleum etherˆ 3:1 to give 15c (9.59 g, 84%) as a colorless
solid of m.p. 98 ± 1008C. For analysis a sample was recrystallized from tert-
butyl methyl ether/petroleum etherˆ 1:1. [a]2D5 ˆ ‡ 27.1 (c ˆ 1.045,
CH2Cl2); 1H NMR (300 MHz, CDCl3): d ˆ 0.89 (d, J ˆ 6.7 Hz, 6H), 0.98
(d, J ˆ 6.7 Hz, 6H), 1.71 (pseudo-octet, J ˆ 6.7 Hz, 2H), 3.87 (dt, J ˆ 8.6 and
7.1 Hz, 2H), 4.03 (dd, J ˆ 8.5 and 7.3 Hz, 2H), 4.36 (dd, J ˆ 8.6 and 8.6 Hz,
2H), 7.00 ± 7.08 (m, 2H), 7.10 ± 7.18 (m, 2H); 13C NMR (75 MHz, CDCl3):
d ˆ 18.4, 18.7, 33.0, 57.8, 67.3, 71.3, 125.7, 128.5, 129.7, 139.9, 168.4; elemental
analysis calcd for C19H25N2O2SCl (380.9): C 59.91, H 6.62, N 7.35; found C
59.95, H 6.54, N 7.25.
1
colorless liquid. H NMR (300MHz, CDCl3): d ˆ 3.62 (d, J ˆ 6.8 Hz, 2H),
5.67 (t, J ˆ 6.8 Hz, 1H), 7.15 ± 7.30 (m, 5 H); 13C NMR (75MHz, CDCl3):
d ˆ 45.1, 51.4, 127.7, 128.6, 129.4, 136.7; elemental analysis calcd for C8H8Br2
(264.0): C 36.40, H 3.05; found C 36.61, H 3.00.
1,1-Diiodo-2-phenylethane (4): n-Butyllithium in hexane (1.84m,
190 mmol) was added dropwise to a solution of hexamethyldisilazane
(50 g, 0.31 mol) in THF (50 mL) at 308C. The solution was stirred for
0.5 h at 308C and 0.5 h at 08C. A solution of diiodomethane (53.6 g,
0.2 mol) in THF (100 mL) was cooled to 1058C. The freshly prepared
solution of lithium h examethyldisilazide was added dropwise over 1 h.[19]
Stirring was continued for 0.5 h at 1058C. A solution of benzyl bromide
(42.7 g, 250 mmol) in THF (30 mL) was added dropwise. After stirring for
1.5 h at 1058C the solution was allowed warm to room temperature.
Saturated aqueous NH4Cl solution (90 mL), 20% Na2S2O3 solution
(20 mL) and petroleum ether (50 mL) were added. The phases were
separated and the aqueous phase was extracted with petroleum ether (3 Â
50 mL). The combined organic phases were washed with brine (30 mL),
dried (Na2SO4) and concentrated. The residue was taken up in a mixture of
petroleum ether (68 mL) and dichloromethane (7 mL), and the solution
was stored at 258C. The next day the crystalline product (32 g) could be
collected. The mother liquor was concentrated and the residue was
crystallized from petroleum ether/dichloromethane to furnish further 5.7 g
of 4. Total yield: 37.7 g (55%). For analysis a sample was recrystallized as
above to give crystals (with a pink touch) of m.p. 378C. 1H NMR (300MHz,
CDCl3): d ˆ 3.74 (d, J ˆ 7.4 Hz, 2H), 5.08 (t, J ˆ 7.4 Hz, 1H), 7.20 ± 7.24 (m,
2H), 7.29 ± 7.36 (m, 3H); 13C NMR (125MHz, CDCl3): d ˆ 25.8, 54.4,
127.6, 128.7, 128.9, 139.8; elemental analysis calcd for C8H8I2 (358.0): C
26.84, H 2.25; found C 26.66, H 2.28.
Generation of the chiral magnesium reagent 12: THF (0.1 mL), C6D6
(0.07 mL), and a solution of diisopropylmagnesium (16) in THF (0.59m,
0.20 mmol) were placed in a dry NMR tube which was closed with a
septum. A solution of the bisoxazolidine (15a) in THF (1.00m, 0.200 mmol)
was added to the NMR tube at 08C and 13C NMR (125 MHz) spectra were
recorded at 258C and at 788C. Another 0.2 mmol of 15a were added and
spectra were recorded again. Finally a solution of diisopropylmagnesium
(16) in THF (0.59m, 0.20 mmol) was added, and a third 13C NMR spectrum
was recorded. The following chracteristic signals were observed at 788C:
12a: d ˆ 9.01, 14.7, 19.0, 25.9, 26.0, 33.1, 35.2, 66.2, 67.8, 172.4; 17a: 16.1,
19.2, 33.8, 54.1, 67.6, 68.3, 173.0. In a similar manner the data for 12b
(
308C): d ˆ 9.56, 15.5, 15.6, 19.5, 19.7, 26.3, 26.4, 33.7, 33.8, 55.1, 67.3, 67.5,
69.6, 69.8, 123.5, 124.1, 128.9, 145.7, 174.3, 174.5 and 12c (258C): d ˆ 9.0,
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0341 $ 17.50+.50/0
341

R. W. Hoffmann and V. Schulze
FULL PAPER
15.3, 18.9, 25.3, 25.5, 32.9, 55.4, 67.1, 69.4, 125.5, 125.6, 128.6, 144.3, 174.0
were recorded.
of the diiodoalkane (0.97m in THF, 0.70 mmol) was added, resulting in the
slow formation of an intensive yellow color.[11] The color had faded after the
mixture had been stirred for 1.5 h at 788C. A solution of benzaldehyde
(0.18 mL, 1.75 mmol) in THF (5 mL at 788C) was prepared to which was
Reactions of magnesium carbenoids
Generation and trapping of 1-iodo-2-phenyl-ethylmagnesium chloride (5):
THF (4 mL), diethyl ether (2 mL), petroleum ether (2 mL), and a solution
of isopropylmagnesium chloride in diethyl ether (1.34m, 1.60 mmol) were
cooled to 788C. A solution of 1,1-diiodo-2-phenylethane (4) (2.88m,
1.00 mmol) in diethyl ether was added dropwise resulting in the formation
of an intensive yellow color.[11] After stirring for 2 h at 788C the color had
faded and water (5 mL) was added dropwise. After the reaction mixture
had reached room temperature, saturated aqueous NH4Cl solution
(20 mL), 20% Na2S2O3 solution (3 mL) and tert-butyl methyl ether
(20 mL) were added. The phases were separated and the aqueous phase
was extracted with tert-butyl methyl ether (3 Â 20 mL). The combined
organic extracts were washed with brine (10 mL), dried (Na2SO4) and
concentrated. Flash chromatography with petroleum ether/ethyl acetate ˆ
20:1 furnished 1-iodo-2-phenylethane (227 mg, 98%) as a colorless oil;
1H NMR (300 MHz, CDCl3): d ˆ 3.19 (t, J ˆ 7.6 Hz, 2H), 3.36 (dt, J ˆ 7.8
and 0.9 Hz, 2H), 7.18 ± 7.37 (m, 5H); 13C NMR (75 MHz, CDCl3): d ˆ 5.45,
40.3, 126.8, 128.3, 128.6, 140.6; elemental analysis calcd for C8H9I (232.1): C
41.41, H 3.91; found C 41.33, H 3.84. The reaction proceeded in the same
manner in pure THF solvent. Quenching with D2O resulted in the
formation of 6.
added
a solution of dimethylaluminum chloride (1.00m in hexane,
1.4 mmol) at 788C. The latter solution was transferred by canula into
the former carbenoid solution at 788C. After stirring for 2 h, the solution
was allowed to warm to room temperature overnight. pH7 Buffer solution
(20 mL) and tert-butyl methyl ether (20 mL) were added, and the
suspension was sonicated for 15 min in a cleaning bath. The phases were
separated and the aqueous phase was extracted with tert-butyl methyl ether
(3 Â 20 mL). The combined organic phases were washed with brine
(10 mL), dried (Na2SO4), and concentrated. Flash chromatography of the
residue with petroleum ether/ethyl acetate ˆ 11:1 to 7:1 tert-butyl methyl
ether/petroleum etherˆ 1:1 furnished 9 (14 mg, 9%), 19 (172 mg, 73%)
and residual ligand 15c (285 mg, 62%). The diastereomeric purity of 9 was
determined from the 1H NMR spectrum to be >97% cis. 19: 1H NMR
(300 MHz, CDCl3): d ˆ 2.51 (d, J ˆ 6.4 Hz, 1H), 3.26 (dd, J ˆ 14.4 and
8.3 Hz, 1 H), 3.30 (dd, J ˆ 14.4 and 7.1 Hz, 1H), 4.32 (dd, J ˆ 5.4 and 5.4 Hz,
1H), 4.55 (ddd, J ˆ 8.2, 7.2, and 4.6 Hz, 1H), 7.17 ± 7.40 (m, 10H); 13C NMR
(75 MHz, CDCl3): d ˆ 43.9, 48.9, 75.1, 125.8, 126.9, 128.0, 128.4, 128.6,
128.9, 139.5, 141.6; for comparison, the data for the anti-diastereomer are:
1H NMR (300 MHz, CDCl3): d ˆ 2.60 (d, J ˆ 1.4 Hz, 1H), 3.02 (dd, J ˆ 15.0
and 9.6 Hz, 1H), 3.09 (dd, J ˆ 15.1 and 4.6 Hz, 1H), 4.59 (ddd, J ˆ 8.9, 4.8,
and 4.8 Hz, 1H), 5.10 (d, J ˆ 3.8 Hz, 1H), 7.03 ± 7.48 (m, 10H); 13C NMR
(75 MHz, CDCl3): d ˆ 39.2, 46.3, 78.0, 126.5, 126.7, 128.1, 128.3, 128.5, 128.9,
139.6, 140.2; elemental analysis calcd for C15H15IO (338.2): C 53.27, H 4.47;
found C 53.41, H 4.22.
2-Benzyl-3-phenyloxirane (9): The carbenoid 5 was generated as described
above from 4 (358 mg, 1.00 mmol) and isopropylmagnesium chloride
(1.50m in diethyl ether, 1.3 mmol) in THF solution. The carbenoid 5 was
trapped by addition of benzaldehyde (0.25 mL, 2.50 mmol). After workup
as described for 5, 13C NMR analysis of the crude product revealed the
presence of 9 and 19 in a 5:1 ratio. The crude product was purified by flash
chromatography with petroleum ether/ethyl acetate ˆ 30:1 changing to
15:1 to give cis-2-benzyl-3-phenyloxirane (9) (116 mg, 55%) as a colorless
oil; 1H NMR (300 MHz, CDCl3): d ˆ 2.50 (dd, J ˆ 14.6 and 6.6 Hz, 1H),
2.72 (dd, J ˆ 14.6 and 5.9 Hz, 1H), 3.38 (ddd, J ˆ 6.3, 6.3, and 4.2 Hz, 1H),
4.10 (d, J ˆ 4.1 Hz, 1H), 6.98 ± 7.01 (m, 2H), 7.10 ± 7.38 (m, 8H); cf. the data
in ref. [20]; 13C NMR (75 MHz, CDCl3): d ˆ 32.2, 57.5, 59.6, 126.5, 126.6,
127.7, 128.1, 128.5, 128.8, 135.4, 137.4.
The iodohydrin 19 (110 mg, 0.32 mmol, 50% ee) was dissolved in ethanol
(15 mL) to which was added KOH (1.80m in ethanol, 0.59 mmol) at 08C.
After stirring the mixture for 2 h at 08C and overnight at room temper-
ature, NH4Cl (ca. 150 mg) was added. The solution was concentrated in
vacuo and the residue was partitioned between saturated aqueous NH4Cl
solution (20 mL) and tert-butyl methyl ether (20 mL). The phases were
separated and the aqueous phase was extracted with tert-butyl methyl ether
(3 Â 20 mL). The combined organic phases were washed with brine
(10 mL), dried (Na2SO4), and concentrated. Flash chromatography with
petroleum ether/ethyl acetate ˆ 12:1 resulted in 9 (51 mg, 75%) as a
colorless oil. The 1H NMR spectra showed a cis/trans ratio of 93:7. The
enantiomeric purity was determined as for 9 to be 53%.
The enantiomeric purity of 9 was determined using tris(3-heptafluorobu-
tyryl-d-camphorato)europium as a shift reagent in CDCl3. Especially the
signal at d ˆ 4.10 was shifted downfield to, for example, d ˆ 5.1 and 5.3. It
was shown that the lowfield signal at d ˆ 5.3 corresponds to (3R)-(‡)-9.
The experiments have also been carried out in the manner that a solution of
dimethylaluminum chloride in hexane was added first immediately
followed by the addition of benzaldehyde. This involved for instance
addition of 2.5, 0.59, or 0.23 equivalents of benzaldehyde resulting in
different enantiomeric enrichments of the iodohydrin 19 obtained.
After a similar experiment the reaction mixture was allowed to warm to
room temperature very slowly over a period of 15 h. GC analysis of the
crude product revealed the presence of 9 (21%), trans-9 (4%), and the
iodoketone (7). After flash chromatography 47% of 7 was obtained as a
colorless oil; 1H NMR (300 MHz, CDCl3): d ˆ 3.40 (dd, J ˆ 14.3 and 6.7 Hz,
1H), 3.68 (dd, J ˆ 14.3 and 8.1 Hz, 1H), 5.53 (dd, J ˆ 8.1 and 6.8 Hz, 1H),
7.20 ± 7.80 (m, 10H); 13C NMR (75 MHz, CDCl3): d ˆ 25.5, 40.9, 127.0,
(R)-(‡)-1,3-Diphenyl-1-propanol (18): The iodohydrin 19 (70 mg,
0.21 mmol) was dissolved in benzene (15 mL) to which was added
tributyltin hydride (0.20 mL, 0.76 mmol) and AIBN (ca. 200 mg). The
mixture was refluxed for 2 h, further AIBN was added, and refluxing was
continued for 2 h. Saturated aqueous NH4Cl solution (15 mL) and tert-
butyl methyl ether (20 mL) were added, the phases were separated, and the
aqueous phase was extracted with tert-butyl methyl ether (3 Â 20 mL). The
combined organic phases were washed with brine (10 mL), dried (Na2SO4),
and concentrated. Flash chromatography of the residue with petroleum
ether/ethyl acetate ˆ 10:1 furnished the alcohol 18 (43 mg, 95%) as a
colorless oil. [a]2D5 ˆ ‡ 3.9 (c ˆ 3.60, ethanol); 1H NMR (300 MHz, CDCl3):
d ˆ 1.90 ± 2.08 (m, 3H), 2.52 ± 2.69 (m, 2H), 4.58 (dd, J ˆ 7.7 and 4.4 Hz,
1H), 7.06 ± 7.31 (m, 10H); 13C NMR (75 MHz, CDCl3): d ˆ 32.0, 40.4, 73.8,
125.8, 125.9, 127.6, 128.3, 128.4, 128.6, 141.8, 144.6, cf. the data in ref. [15].
1
128.5, 128.6, 128.7, 129.1, 133.5, 134.0, 139.1, 194.1; characteristic H NMR
data of trans-9 (300 MHz, CDCl3): d ˆ 2.97 (d, J ˆ 5.7 Hz, 2H), 3.15 (dt, J ˆ
2.0 and 5.5 Hz, 1H), 3.65 (d, J ˆ 2.0 Hz, 1H).
Generation of 13 and trapping with D2O: A solution of 15c (0.82m in THF,
1.19 mmol) was added slowly to a solution of diisopropylmagnesium (1.59m
in diethyl ether, 1.05 mmol) at 08C, and the mixture was stirred for 1 h at
room temperature. After cooling to 788C a solution of the diiodoalkane 4
(1.10m in THF, 0.70 mmol) was added, leading to the slow generation of an
intensive yellow color.[11] The color had faded after the mixture had been
stirred for 1.5 h at 788C. A CH3OD/D2O mixture (2:1, 1.0 mL) was
added. Stirring was continued for 0.5 h at this temperature. After allowing
the mixture to warm to room temperature, saturated aqueous NH4Cl
solution (20 mL) and tert-butyl methyl ether (20 mL) were added. The
phases were separated and the aqueous phase was extracted with tert-butyl
methyl ether (3 Â 20 mL). The combined organic extracts were washed with
brine (10 mL), dried (Na2SO4), and concentrated. Flash chromatography
with petroleum ether/ethyl acetate ˆ 15:1 changing to 1:1 to tert-butyl
methyl ether furnished the iodo compound 6 (153 mg, 94%) as a colorless
oil and the recovered 15c (371 mg, 81%).
Control experiments with (1-iodo-2-phenylethyl)dimethylaluminum (20):
A solution of n-butyllithium (1.5m in hexane, 1.5 mmol), THF (8 mL),
diethyl ether (4 mL), and petroleum ether (4 mL) were mixed at 788C
and cooled to 1058C. A solution of the diiodo compound 4 (1.00m in
THF, 1.00 mmol) was added dropwise resulting in an intense yellow
coloration. The yellow color faded over 15 min at 1058C. A solution of
dimethylaluminum chloride (1.00m in hexane, 2.00 mmol) was added,
resulting in the formation of a second yellowish phase. Benzaldehyde
(0.20 mL, 2.00 mmol) was added and the emulsion was stirred vigorously
for 1 h at 1058C. The mixture was allowed to warm to 788C over 2 h,
was stirred for 1 h at this temperature, and was quenched by addition of
pH7 buffer (20 mL). 20% Na2S2O3 solution (3 mL) and tert-butyl methyl-
1-Hydroxy-2-iodo-1,3-diphenylpropane (19): A solution of 15c (0.79m in
THF, 1.20 mmol) at 08C was added dropwise to a solution of diisopro-
pylmagnesium (16) (0.90m in diethyl ether, 1.05 mmol). After stirring for
1 h at room temperature the solution was cooled to 788C and a solution
342
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0342 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Chiral Organometallic Reagents
337±344
ether (20 mL) were added and the mixture was sonicated for 15 min in a
cleaning bath. The phases were separated and the aqueous phase was
extracted with tert-butyl methyl ether (3 Â 20 mL). The combined organic
phases were washed with brine (10 mL), dried (Na2SO4), and concentrated.
The residue (225 mg) was essentially pure 1-iodo-2-phenylethane.
was allowed to warm to room temperature. A pH7 buffer solution (20 mL)
and tert-butyl methyl ether (20 mL) were added. The resulting mixture was
sonicated for 15 min in a cleaning bath and the phases were separated. The
aqueous phase was extracted with tert-butyl methyl ether (3 Â 20 mL). The
combined organic phases were washed with brine (10 mL), dried (Na2SO4),
and concentrated. The residue consisted of excess aldehyde 24 and the
iodohydrins 25 and 26. In order to facilitate separation, the crude product
was taken up in THF (15 mL) and cooled to 788C. Borane-dimethylsul-
fide complex (152 mg, 2.0 mmol) was added and the mixture was stirred for
2 h at 788C and 5 h at room temperature. Saturated aqueous NH4Cl
solution (20 mL) and tert-butyl methyl ether (20 mL) were added. The
phases were separated and the aqueous phase was extracted with tert-butyl
methyl ether (3 Â 20 mL). The combined organic phases were washed with
brine (10 mL), dried (Na2SO4), and concentrated. Flash chromatography of
the residue with petroleum ether/ethyl acetate ˆ 10:1 to 7:1 furnished a
mixture of 25 and 26 (249 mg, 63%) as a colorless oil. The diastereomer
ratio was determined by 13C NMR spectrosocpy to be 67:33. Pure samples
of 25 and 26 were obtained by MPLC separation with petroleum ether/
ethyl acetate ˆ 7:1. 25: [a]2D5 ˆ ‡ 13.6 (c ˆ 2.80, CH2Cl2); 1H NMR
(300 MHz, CDCl3): d ˆ 0.99 (d, J ˆ 6.2 Hz, 3H), 2.48 (dd, J ˆ 7.5 and ca.
2 Hz, 1H), 2.95 (s, br., 1H), 3.29 (dd, J ˆ 14.0 and 7.5 Hz, 1H), 3.36 (dd, J ˆ
14.0 and 8.5 Hz, 1H), 3.55 (dq, J ˆ 7.4 and 6.2 Hz, 1H), 4.16 (td, J ˆ 7.9 and
2.1 Hz, 1H), 4.51 (d, J ˆ 11.2 Hz, 1H), 4.55 (d, J ˆ 11.2 Hz, 1H), 7.09 ± 7.30
(m, 10H); 13C NMR (75 MHz, CDCl3): d ˆ 14.8, 38.6, 44.4, 71.3, 74.7, 81.1,
126.8, 127.8, 128.4, 128.5, 129.0, 137.9, 139.5; elemental analysis calcd for
C18H21IO2 (396.3): C 54.56, H 5.34; found C 54.74, H 5.50. 26: [a]2D5 ˆ ‡ 25.9
(c ˆ 2.02, CH2Cl2); 1H NMR (300 MHz, CDCl3): d ˆ 1.20 (d, J ˆ 6.3 Hz,
3H), 2.54 (d, J ˆ 9.1 Hz, 1H), 3.04 (dd, J ˆ 14.8 and 9.6 Hz, 1H), 3.43 (dd,
J ˆ 14.8 and 3.5 Hz, 1H), 3.52 (ddd, J ˆ 8.8, 8.8, and 2.3 Hz, 1H), 4.12 (qd,
J ˆ 6.3 and 2.4 Hz, 1H), 4.29 (ddd, J ˆ 9.6, 8.6, and 3.5 Hz, 1H), 4.39 (d, J ˆ
11.4 Hz, 1H), 4.58 (d, J ˆ 11.3 Hz, 1H), 7.10 ± 7.32 (m, 10H); 13C NMR
(75 MHz, CDCl3): d ˆ 17.0, 40.1, 41.6, 71.3, 75.2, 78.6, 126.6, 127.9, 128.0,
128.2, 128.5, 129.4, 138.0, 139.5; elemental anlysis calcd for C18H21IO2
(396.3): C 54.56, H 5.34; found C 54.61, H 5.26.
Control experiment involving dimethylaluminum bisoxazolidine complex
22: A solution of 15c (0.785m in THF, 1.60 mmol) was added at 08C to a
solution of trimethylaluminum (2.00m in hexane, 1.50 mmol) and THF
(5 mL). The mixture was stirred first at 08C, then at room temperature.
This solution was added dropwise at 788C to a solution of the carbenoid 5
generated from the diiodo compound 4 (1.00 mmol) as described for 5.
After the mixture had been stirred for 1 h at 788C, benzaldehyde
(0.35 mL, 3.5 mmol) was added dropwise and stirring was continued for 2 h.
After the mixture had been allowed to warm to room temperature
overnight workup as described for 19 provided 9 (60 mg, 28%) and 19
(75 mg, 22%) as colorless oils in addition to recovered 15c (588 mg). The
diastereomer ratio of 9 was determined by 1H NMR spectroscopy to be
>97: < 3. The enantiomer enrichment of 9 obtained directly, and that of the
epoxide formed by subsequent cyclization of the iodohydrin 19 was
recorded as described for 9.
Kinetic resolution of 1-bromo-1-iodo-2-phenylethane: A solution of 15c
(0.75m in THF, 2.10 mmol) was added dropwise at 08C to a solution of
diisopropylmagnesium (1.46m in diethyl ether, 1.80 mmol) in THF
(15 mL). After stirring for 30 min the mixture was allowed to warm to
room temperature. This solution was added with a motor-driven syringe to
a solution of the bromoiodo compound 31 (3.22m in THF, 3.00 mmol) and
THF (15 mL) over a period of 1.5 h at 788C. Stirring was continued for
another 1.5 h at this temperature. Solutions of dimethylaluminum chloride
(1.00m in hexane, 3.24 mmol) and benzaldehyde (0.33 mL, 3.30 mmol)
were added sequentially. After stirring for 2 h at 788C, the mixture was
allowed to warm to room temperature. A pH7 buffer solution (20 mL) and
tert-butyl methyl ether (20 mL) were added and the suspension was
sonicated for 15 min in a cleaning bath. The phases were separated and the
aqueous phase was extracted with tert-butyl methyl ether (3 Â 20 mL). The
combined organic phases were washed with brine (10 mL), dried (Na2SO4),
and concentrated. Flash chromatography with petroleum ether/ethyl
acetate ˆ 15:1 to 7:1 to tert-butyl methyl ether/petroleum etherˆ 1:1 to
tert-butyl methyl ether furnished the recovered bromoiodo compound 31a
An identical experiment using racemic 2-benzyloxypropionaldehyde (24)
furnished 71% of 25 and 26 in a 91:9 diastereomer ratio. Essentially
identical experiments were also carried out in which the solution of the
carbenoid 5 was introduced by canula into a solution of 2-benzyloxypro-
pionaldehyde precomplexed with dimethylaluminum chloride (cf. 27).
(580 mg, 62%, [a]D25
ˆ
0.652 (c ˆ 10.5, CH2Cl2)) and the bromohydrin 34
(290 mg, 33%) in addition to the recovered ligand 15c (747 mg).
Structure assignment of 25 and 26: A crude product containing a mixture of
25 and 26 (64:36, 230 mg, 0.58 mmol) and excess aldehyde 24 was taken up
in ethanol (15 mL). Carefully powdered NaBH4 (76 mg, 2.0 mmol) was
added. The mixture was stirred for 3 h at room temperature and KOH
(1.8m in ethanol, 0.2 mL) was added. After the mixture had been stirred
overnight, saturated aqueous NH4Cl solution (20 mL), hydrochloric acid
(3m, 1 mL) and tert-butyl methyl ether (20 mL) were added. After the
hydrogen evolution had ceased, the mixture was brought to pH7 by
addition of aqueous NaHCO3 solution. The phases were separated, and the
aqueous phase was extracted with tert-butyl methyl ether (3 Â 20 mL). The
combined organic phases were washed with brine (10 mL) dried (Na2SO4),
and concentrated. Flash chromatography with petroleum ether/ethyl
acetate ˆ 10:1 furnished a mixture of 29 and 30 (150 mg, 96%) as a
colorless oil. The diastereomer ratio was determined from the 13C NMR
spectrum to be 64:36; elemental analysis calcd for C18H20O2 (268.4): C
80.56, H 7.51; found C 80.34, H 7.64.
Bromohydrin 34: syn/anti ˆ 97:3; [a]D25
ˆ
1.27 (c ˆ 6.90, CH2Cl2);
1H NMR (300 MHz, CDCl3): d ˆ 2.74 (d, J ˆ 5.7 Hz, 1H), 3.08
(dd, J ˆ 14.3 and 8.9 Hz, 1H), 3.22 (dd, J ˆ 14.3 and 5.8 Hz, 1H), 4.43
(ddd, J ˆ 8.9, 5.5, and 5.5 Hz, 1H), 4.69 (dd, J ˆ 5.4 and 5.4 Hz, 1H),
7.13 ± 7.40 (m, 10H); 13C NMR (75 MHz, CDCl3): d ˆ 35.2, 64.9, 75.3,
126.2, 126.9, 128.2, 128.48, 128.50, 129.1, 138.0, 140.6; elemental analysis
calcd for C15H15BrO (291.2): C 61.87, H 5.19; found C 61.73, H 5.26.
The bromohydrin 34 obtained was cyclized as described for 19 to (3R)-9,
which showed an enantiomeric excess of 18%. [a]2D5 ˆ ‡ 3.52 (c ˆ 2.73,
CH2Cl2).
Generation of 1-bromo-2-phenylethylmagnesium chloride from 1-bromo-
1-iodo-2-phenylethane (31a): Starting from (S)-( )-31a of about 8% ee
(500 mg, 1.6 mmol) the carbenoid 32a was generated as described for 5 by
using isopropylmagnesium chloride (2.00m in diethyl ether, 2.09 mmol).
The carbenoid was quenched with dimethylaluminum chloride (1m in
hexane, 3.70 mmol) followed by benzaldehyde (0.45 mL, 4.50 mmol).
Workup as described for 19 provided the bromohydrin ent-34 (336 mg,
The diastereomeric epoxides 29 and 30 were separated by MPLC with
petroleum ether/ethyl acetate ˆ 7:1. 29: [a]D25
ˆ
43.6 (c ˆ 2.825, CH2Cl2);
72%) as
a
colorless oil. According to 1H NMR spectroscopy, the
1H NMR (300 MHz, CDCl3): d ˆ 1.17 (d, J ˆ 6.3 Hz, 3H), 2.70 (dd, J ˆ 15.6
and 7.5 Hz, 1H), 2.76 (dd, J ˆ 15.2 and 5.1 Hz, 1H), 2.98 (dd, J ˆ 8.1 and
4.4 Hz, 1H), 3.10 (ddd, J ˆ 7.1, 4.9 and 4.9 Hz, 1H), 3.46 (dq, J ˆ 8.1 and
6.5 Hz, 1H), 4.57 (d, J ˆ 11.8 Hz, 1H), 4.76 (d, J ˆ 11.8 Hz, 1H), 7.11 ± 7.35
(m, 10H); 13C NMR (75 MHz, CDCl3): d ˆ 17.7, 34.7, 54.7, 60.7, 71.2, 73.7,
126.7, 127.5, 127.8, 128.3, 128.6, 137.7, 138.7. 30: [a]2D5 ˆ ‡ 0.84 (c ˆ 0.96,
diastereomer ratio was syn/anti ˆ 94:6. [a]2D5 ˆ ‡ 0.347, (c ˆ 5.62, CH2Cl2).
The bromohydrin ent-34 was converted into (3S)-9 as described for 19 with
an ee of 5%; [a]2D5
ˆ
0.965, (c ˆ 1.71, CH2Cl2).
Reaction of the carbenoid 5 complexed with dimethylaluminum chloride
with 2-benzyloxypropionaldehyde (24): A solution of 1,1-diiodo-2-phenyl-
ethane (4) (1.00m in THF, 1.00 mmol) was added at 788C to a solution of
isopropylmagnesium chloride (2.00m in diethyl ether, 1.2 mmol) in THF
(15 mL). After the mixture had been stirred for 2.5 h at 788C, the initial
intense yellow color had faded. A solution of dimethylaluminum chloride
(1.00m in hexane, 2.50 mmol) was added and stirring was continued for
5 min. The resulting solution was transferred by canula into a precooled
1
CH2Cl2); H NMR (300 MHz, CDCl3): d ˆ 1.12 (d, J ˆ 6.5 Hz, 3H), 2.70 ±
2.88 (m, 4H), 3.21 (dq, J ˆ 6.5 and 6.5 Hz, 1H), 4.48 (d, J ˆ 11.9 Hz, 1H),
4.65 (d, J ˆ 11.9 Hz, 1H), 7.10 ± 7.29 (m, 10H); 13C NMR (75 MHz, CDCl3):
d ˆ 17.4, 38.2, 54.9, 61.9, 71.2, 75.6, 126.7, 127.5, 127.7, 128.3, 128.6, 128.9,
137.1, 138.6.
A mixture of the iodohydrins 25 and 26 (66 mg, 0.17 mmol) was reduced
with tributyltin hydride as described for 18. Flash chromatography of the
crude product with petroleum ether/ethyl acetate ˆ 8:1 furnished the
(
788C) solution of (2S)-benzyloxypropionaldehyde (24)[21] (493 mg,
3.0 mmol) in THF (15 mL). After stirring for 2 h at 788C the mixture
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0343 $ 17.50+.50/0
343

R. W. Hoffmann and V. Schulze
FULL PAPER
alcohol 28 (33 mg, 73%) as a colorless oil; [a]2D5 ˆ ‡ 12 (c ˆ 0.11, CH2Cl2);
1H NMR (300 MHz, CDCl3): d ˆ 1.09 (d, J ˆ 6.0 Hz, 3H), 1.61 ± 1.72 (m,
2H), 2.1 ± 2.7 (br. s, 1H), 2.60 (ddd, J ˆ 13.7, 9.3, and 7.3 Hz, 1H), 2.77 (ddd,
J ˆ 14.1, 8.9, and 5.6 Hz, 1H), 3.28 ± 3.40 (m, 2H), 4.34 (d, J ˆ 11.5 Hz, 1H),
4.58 (d, J ˆ 11.5 Hz, 1H), 7.01 ± 7.31 (m, 10H); 13C NMR (75 MHz, CDCl3):
d ˆ 15.5, 31.9, 34.8, 71.0, 74.2, 78.4, 125.7, 127.7, 127.8, 128.3, 128.4, 128.5,
138.3, 142.2; elemental analysis calcd for C18H22O2 (270.4): C 79.96, H 8.20;
found C 79.61, H 8.41.
[8] a) D. J. Procter, M. Thornton-Pett, C. M. Rayner, Tetrahedron 1996,
52, 1841 ± 1854; b) M. Tanaka, I. Ogata, Bull. Chem. Soc. Jpn. 1975, 48,
1094.
[9] W. F. Bailey, J. J. Patricia, J. Organomet. Chem. 1988, 352, 1 ± 46.
[10] G. Wittig, U. Schöllkopf, Tetrahedron 1958, 3, 91 ± 93.
[11] V. Schulze, M. Brönstrup, V. P. W. Böhm, P. Schwerdtfeger, M.
Schimeczek, R. W. Hoffmann, Angew. Chem. 1998, 110, 869 ± 871;
Angew. Chem. Int. Ed. Engl. 1998, 37, 824 ± 826.
[12] a) R. E. Lowenthal, A. Abiko, S. Masamune, Tetrahedron Lett. 1990,
31, 6005 ± 6008; b) D. Müller, G. Umbricht, B. Weber, A. Pfaltz, Helv.
Chim. Acta 1991, 74, 232 ± 240.
[13] V. Schulze, R. Löwe, S. Fau, R. W. Hoffmann, J. Chem. Soc. Perkin
Trans. 2 1998, 463 ± 465.
[14] V. Schulze, Dissertation, Universität Marburg, 1997.
[15] J. L. von dem Bussche-Hünnefeld, D. Seebach, Tetrahedron 1992, 48,
5719 ± 5730.
[16] a) R. Hirsch, R. W. Hoffmann, Chem. Ber. 1992, 125, 975 ± 982;
b) R. W. Hoffmann in Organic Synthesis via Organometallics (OSM4)
(Eds.: D. Enders, H.-J. Gais, W. Keim), Vieweg, Braunschweig, 1993,
pp. 79 ± 91.
[17] a) G. Wittig, G. Fuhrmann, Ber. Dtsch. Chem. Ges. 1940, 73, 1197 ±
1218; b) H. C. Stiasny, R. W. Hoffmann, Chem. Eur. J. 1995, 619 ± 624.
[18] H. B. Kagan, J. C. Fiaud, Top. Stereochem. 1988, 18, 249 ± 330.
[19] P. Charreau, M. Julia, J.-N. Verpeaux, Bull. Soc. Chim. Fr. 1990, 127,
275 ± 282.
[20] Y. Tamura, S. M. Bayomi, K. Sumoto, M. Ikeda, Synthesis 1977, 693 ±
695.
[21] P. Varelis, B. L. Johnson, Austr. J. Chem. 1995, 48, 1775 ± 1779.
[1] For Part XXI see: H. C. Stiasny, V. P. W. Böhm, R. W. Hoffmann,
Chem. Ber. 1997, 130, 341 ± 346.
[2] Houben-Weyl, Methods of Organic Chemistry, 4th ed; Stereoselective
Synthesis (Eds.: G.Helmchen, R. W.Hoffmann, J. Mulzer, E. Schau-
mann), Thieme, Stuttgart, E21, 1995.
[3] D. Hoppe, T. Hense, Angew. Chem. 1997, 109, 2376 ± 3410; Angew.
Chem. Int. Ed. Engl. 1997, 36, 2282 ± 2316.
Â
[4] J. Villieras, Organomet. Chem. Rev. A. 1971, 7, 81 ± 94.
[5] a) R. W. Hoffmann, M. Julius, F. Chemla, T. Ruhland, G. Frenzen,
Tetrahedron 1994, 50, 6049 ± 6060; b) W. Klute, M. Krüger, R. W.
Hoffmann, Chem. Ber. 1996, 129, 633 ± 638.
Â
[6] a) J. Villieras, Bull. Soc. Chim. Fr. 1967, 1520 ± 1532; b) D. Seyferth,
R. L. Lambert Jr., E. M. Hanson, J. Organomet. Chem. 1970, 24, 647 ±
661; c) R. C. Hahn, J. Tompkins, Tetrahedron Lett. 1990, 31, 937 ± 940;
d) C. DeLima, M. Julia, J.-N. Verpeaux, Synlett. 1992, 133 ± 134;
e) R. W. Hoffmann, A. Kusche, Chem. Ber. 1994, 127, 1311 ± 1316.
[7] We thank Dr. R. Löwe for carrying out this experiment: R. Löwe,
Dissertation, Universität Marburg, 1997.
Received: May 6, 1998 [F1144]
344
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0344 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1